ANALYTICAL

BIOCHEMISTRY

Analytical Biochemistry 348 (2006) 139–147

www.elsevier.com/locate/yabio

Effective novel dissociation methods for intact protein: Heat-assistednozzle-skimmer collisionally induced dissociation and infrared

multiphoton dissociation using a Fourier transform ion cyclotronresonance mass spectrometer equipped with a micrometal

electrospray ionization emitter

Naoyuki Yamada *, Ei-ichiro Suzuki, Kazuo Hirayama

Institute of Life Sciences, Ajinomoto Company Inc., Kawasaki-ku, Kawasaki-shi 210-8681, Japan

Received 1 September 2005Available online 2 November 2005

Abstract

Heating of a nano-electrospray ionization (nanoESI) source can improve the dissociation efficiency of collisionally induceddissociation (CID) methods, such as nozzle-skimmer CID (NS–CID) and infrared multiphoton dissociation (IRMPD), for largebiomolecule fragmentation. A metal nanoESI emitter was used due to its resistance to heating above 250 �C. This novel methodfor the dissociation of large biomolecular ions is termed ‘‘heat-assisted NS–CID’’ (HANS–CID) or ‘‘heat-assisted IRMPD’’(HA–IRMPD). Multiple charged nonreduced protein ions (8.6 Da ubiquitin, 14 kDa lysozyme, and 67 kDa bovine serum albumin)were directly dissociated by HANS–CID and HA–IRMPD to effectively yield fragment ions that could be assigned. The fragmentions of ubiquitin by HANS–CID can be analyzed by tandem mass spectrometry (MS/MS) using sustained off-resonance irradiationCID (SORI–CID) and IRMPD. In addition, a native large protein, immunoglobulin G (IgG, 150 kDa), was efficiently dissociatedby HA–IRMPD. The product ions that were obtained reflected the domain structure of IgG. However, these product ions of IgGand lysozyme were not dissociated by MS/MS using the same heating energetic methods such as IRMPD and SORI–CID.� 2005 Elsevier Inc. All rights reserved.

Keywords: Protein; Dissociation; Fragmentation; IgG; FTICR; Mass spectrometer; Heat-assisted IRMPD; Heat-assisted NS–CID; Nonreducedlarge protein

Mass spectrometry (MS)1 has become an importantand useful tool for the study of large molecules of inter-est in biology. In particular, in the field of proteomics(i.e., the comprehensive analysis of the expressed

0003-2697/$ - see front matter � 2005 Elsevier Inc. All rights reserved.

doi:10.1016/j.ab.2005.10.026

* Corresponding author. Fax: +81 44 210 5872.E-mail address: naoyuki_yamada@ajinomoto.com (N. Yamada).

1 Abbreviations used: MS, mass spectrometry; MS/MS tandem MS; FTmultiphoton dissociation; SAD, surface-activated dissociation; BIRD, blacdissociation; SORI–CID, sustained off-resonance irradiation CID; ECD, elecCID; ESI, electrospray ionization; RF, radio frequency; IgG2, immunoglobuinfrared; HPLC, high-performance liquid chromatography; CapExit, capillarIRMPD; SWIFT, stored waveband inverse Fourier transform; LC, liquid ch

proteins in a cell, tissue, or an organ or in body fluids),tandem MS (MS/MS) is a highly suitable method forprotein identification. In general ‘‘bottom-up’’ method-ology, the complex protein mixture from a cell is digest-

ICR, Fourier transform ion cyclotron resonance; IRMPD, infraredk body infrared radioactive dissociation; CID, collisionally inducedtron capture dissociation; NS–CID or hexapole CID, nozzle-skimmerlin G2; BSA, bovine serum albumin; ICR, ion cyclotron resonance; IR,y exit; HANS–CID, heat-assisted NS–CID; HA–IRMPD, heat-assistedromatography.

140 Effective novel dissociation methods for intact protein / N. Yamada et al. / Anal. Biochem. 348 (2006) 139–147

ed by proteolysis and the peptides obtained are analyzedby MS and MS/MS. Recently, ‘‘top-down’’ method-ology has emerged as a different strategy for thehigh-throughput identification of proteins and the char-acterization of posttranslational modifications [1,2]. Thetop-down strategy is based on fragmentation of agaseous, large, intact protein ion by MS and the high-resolving power of a Fourier transform ion cyclotronresonance (FTICR) mass spectrometer. These newapproaches have been used recently to directly obtainprotein sequence information by means of the dissocia-tion of protonated intact protein ions without the use ofproteolysis and peptide separation [3]. Other proteindissociation methods that have been reported are infra-red multiphoton dissociation (IRMPD) [4], surface-acti-vated dissociation (SAD) [5], black body infraredradioactive dissociation (BIRD) [6], and collisionallyinduced dissociation (CID) such as sustained off-reso-nance irradiation CID (SORI–CID) [7,8] at an ion trapcell. Electron capture dissociation (ECD) is an alternatedissociation method developed by Zubarev and co-workers [9,10]. On the other hand, there are some in-source dissociation methods such as nozzle-skimmerCID (NS–CID, hexapole CID) [11,12] and multipolestorage-assisted dissociation [13,14]. In general, the effi-ciency of the protein dissociation depends on an increasein the protein�s molecular weight. When different ionactivation methods, such as hexapole CID and ECD,are used, the dissociation conditions provide differentfragmentation efficiencies and dissociation sites. Thesuccessful dissociation of the largest protein, by ECDand NS–CID, has been reported to be the 45-kDa phos-phopantetheinyl cysteine synthetase [15]. Usually, thedissociation of intact proteins less than 67 kDa, suchas serum albumin, is done successfully by slow heatingmethods (e.g., SORI–CID, IRMPD). NS–CID for largeprotonated protein ions has been performed at the high-pressure region in the electrospray ionization (ESI)source, and collision has occurred by increasing internalion energy by means of voltage differences between thenozzle and skimmer [11,12,16–19]. In 1991, Loo andco-workers [20] reported low-resolution NS–CID spec-tra for 67 kDa serum albumin using a triple quadrupolemass spectrometer. Feng and Konishi [21,22] tried theCID of multiple charged 150-kDa antibody ions, andseveral fragment ions were observed. That report sug-gested that the CID of multiple protonated protein ionshas the potential to become a direct fragmentation tech-nique for the structural analysis of large molecular pro-teins, and in 1995 the high-resolution fragment spectraof comparative large proteins (37–67 kDa) wereobserved by use of NS–CID and IRMPD with anFTICR mass spectrometer [23]. Sannes-Lowery andco-workers demonstrated that, under appropriate condi-tions, electrospray-generated ions could be effectivelydissociated by use of extended accumulation intervals

in a radio frequency (RF)-only hexapole in multipolestorage-assisted dissociation [13,24] and by IRMPD inan external ion reservoir [25].

Kelleher and co-workers [26] reported the dissocia-tion of the largest intact protein, the 74-kDa reducedThiC protein, using a combination of NS–CID andIRMPD. In general, the dissociation of large protonatedprotein ions with reduced disulfide bonds resulted inmore fragment ions than did dissociation without disul-fide reduction. In a previous study, the presence of disul-fide bonds in proteins significantly inhibited theproduction of the fragment ions [27]. Zhai and co-work-ers [12] studied the effects of the nozzle-skimmer voltage,distance, and capillary heating current on NS–CID onthe fragmentation of ubiquitin.

In top-down proteomics, the fragmentation of intactprotein ions is a very useful tool not only for proteinidentification but also for analysis of the interaction sur-faces of the protein–ligand complex and of hydrogen–deuterium exchange [28–30]. In this article, we describedissociation techniques for large intact biomolecules,such as antibody immunoglobulin 2 (IgG2), withoutreduced disulfide linkages. Our methods combinedNS–CID or IRMPD with heating above 200 �C at theatmosphere–vacuum nanoESI region by use of a stain-less-steel nanoESI emitter. NS–CID and IRMPD assist-ed by gas heating are the dissociation methods for thefragmentation of lysozyme, serum albumin, and IgG,which have many disulfide bonds, without reduction.By using a stainless-steel nanoESI emitter, our methodshave the advantages of high dissociation efficiency, highsensitivity, and robustness.

Materials and methods

Ubiquitin and bovine serum albumin (BSA) wereobtained from Sigma (St. Louis, MO, USA) and wereused without further purification. Anti-dansyl mouseIgG2 was kindly provided by Ichio Shimada (Universityof Tokyo).

MS experiments were performed on a Fourier trans-form mass spectrometer Apex II (Bruker Daltonics,Billerica, MA, USA) equipped with a 7 T superconduc-ting self-shielded magnet and an external nanoESIsource (Analitica, Branford, CT, USA). All mass spectrawere measured by a dynamic trapping method or a gas-assisted method using argon as the cooling gas.

All samples were dissolved in a 50/50 (v/v) solution of1% aqueous acetic acid/methanol, and these samplesolutions were measured by a nanoESI FTICR massspectrometer equipped with a 30-lm inner diame-ter · 30-mm stainless-steel nanoESI emitter (MicrometalTip, Eisho Metal, Tokyo, Japan) at a flow rate between50 nl/min and 1 ll/min. A typical ion cyclotron reso-nance (ICR) cell pressure was 2–4 · 10�9 Torr. For the

Effective novel dissociation methods for intact protein / N. Yamada et al. / Anal. Biochem. 348 (2006) 139–147 141

gas-assisted dynamic trapping method, the ICR cellpressure was 1 to 5 · 10�7 during ion trapping usingthe cooling gas. The PV1 trapping voltage was 0.39 V,and the PV2 trapping voltage was 0.43 V. The ions weresubjected to broadband detection with a 512,000-datapoint and 217-kHz sweep width. The typical spray volt-age and hexapole accumulation time were �1700 V and0.5 s, respectively. Drying gas pressure was 5–10 psi. AllIRMPD experiments were performed with a continu-ous-wave, 40-W CO2 laser that provided infrared (IR)irradiation through a BaF2 window into the ICR cell.Typically, the laser power was 8 W and the irradiationtime was 100 to 200 ms; these were controlled by anXmass system (Bruker Daltonics). NS–CID andIRMPD spectra are reported as averages of 1 to 32scans. The ICR data were acquired by the Xmass datasystem and were transformed by the Fourier mass spec-trometer. The most abundant isotopic peak and themonoisotopic peak of the fragment masses wereassigned by use of the computer program THRASH[31] and the Web-based search engine Protein Prospec-tor [32].

Results and discussion

Intact ubiquitin fragmentation by heat-assisted NS–CID

and IRMPD

NS–CID and IRMPD are energetic ion dissociationmethods. These collisionally induced activations areslow heating processes. To improve the dissociationefficiency for large protonated protein ions, NS–CIDand IRMPD were performed under heated conditionsat the nanoESI interface of an FTICR mass spectrom-eter equipped with a stainless-steel nanoESI emitter(Fig. 1). Samples were introduced to the nanoESI

Fig. 1. Schematic illustration of the ESI interface equi

source by using a syringe pump or micro high-perfor-mance liquid chromatography (HPLC) pump, and thespray voltage was set at 900–1700 V. The temperatureof the atmosphere–vacuum interface was heated andcontrolled, at 100–300 �C, by the use of a heated dry-ing gas. For the normal ESI source, the thermal energywas mainly burnt to desolvation due to a high sampleflow rate. Consequently, we used nanoESI to heat theprotein ion efficiently. The heated protonated proteinions were dissociated at the atmosphere–vacuum inter-face by the nozzle-skimmer voltage. The nozzle-skim-mer voltage was controlled by the difference betweenthe capillary exit (CapExit) voltage and the skimmerlens voltage. Commonly used nanoESI emitters aremade from gold- or polymer-coated fused silica capil-lary. However, these coated fused silica capillary nano-ESI emitters were not durable at high-temperature andhigh-voltage conditions. Thus, a highly robust stain-less-steel nanoESI emitter with a 30-lm diameter wasused to measure NS–CID under heated conditions.The NS–CID spectra of ubiquitin (1.2 pmol/ll,8.6 kDa) at several different temperatures are shownin Fig. 2. All such spectra were obtained using a noz-zle-skimmer voltage of 290 V and only one scan. At100 �C, multiple charged protonated ions of ubiquitinwere observed, but the fragment ions were notobserved (Fig. 2A). When temperatures increased from150 to 250 �C, multiple protonated molecules disap-peared and fragment ions were produced (Figs. 2B–D). Fig. 3 shows NS–CID spectra for the temperaturesof 150, 200, 250, and 300 �C (panels A–D, respectively)and a nozzle-skimmer voltage of 90 V. At typical noz-zle-skimmer voltages (90 V), the fragmentation of mul-tiple protonated molecules did not occur at the typicaldrying gas temperature of 150 �C (Fig. 3A). At 200 �C(Fig. 3B), some fragment ions were observed. Anincrease in the detection of fragment ions was depen-

pped with a micrometal emitter for HANS–CID.

500 1000 1500 2000 m/z

Rel

ativ

e In

tens

ity

MH 11+ MH 8+

MH 7+

*

y142+ b41

5+ b414+

b214+ y58

5+

A

B

C

D

Fig. 2. ESI MS spectra of ubiquitin obtained by use of a micrometal ESI emitter at a CapExit voltage of 290 V after a single scan. Drying gastemperatures were 100 �C (A), 150 �C (B), 200 �C (C), and 250 �C (D). *Hen egg lysozyme, for internal calibration.

b526+y58

7+

y588+

MH 9+MH 10+

MH 11+

y244+

y183+

Abs

olut

e In

tens

ity

MH 12+MH 8+ MH 7+

600 800 1000 1200 m/z

A

B

C

D

Fig. 3. ESI MS spectra of ubiquitin, obtained by use of a micrometal ESI emitter at a moderate CapExit voltage (90 V) and a spray voltage of1700 V, after a single scan. Drying gas temperatures were 150 �C (A), 200 �C (B), 250 �C (C), and 300 �C (D).

142 Effective novel dissociation methods for intact protein / N. Yamada et al. / Anal. Biochem. 348 (2006) 139–147

dent on gas temperature (Fig. 3C and D). At 250 �C,many fragment ions were observed (Fig. 3C). Mainfragment ions were designated y or b ions (Fig. 3D).The efficient fragmentation occurred at the typical noz-zle-skimmer voltage (90 V) by heating around theatmosphere–vacuum interface. Because such fragmen-tation was not observed at 100 �C and the CapExitvoltage was 290 V (Fig. 2A), the heating by dryinggas efficiently assisted the dissociation of multiple pro-tonated molecules of ubiquitin. In the ESI MS experi-ments of ubiquitin at the CapExit voltage of 0 V,

under varying heating conditions between 100 and250 �C, dissociation of ubiquitin did not appear.Therefore, the high drying gas temperatures did notdirectly affect the ion dissociation at the ESI andatmosphere–vacuum interfaces. These results indicatethat dissociation occurred between the CapExit (noz-zle) and the skimmer lens, by the nozzle-skimmer volt-age, and was assisted by heating at the ESI andatmosphere–vacuum interfaces. Thus, this effective dis-sociation method was termed ‘‘heat-assisted NS–CID’’(HANS-CID).

Effective novel dissociation methods for intact protein / N. Yamada et al. / Anal. Biochem. 348 (2006) 139–147 143

IRMPD is a low energy collision by means of thesame ion heating used in NS–CID. To improve the effi-ciency of the dissociation of multiple charged proteinions, IRMPD was combined with heating at the ESIinterface (i.e., heat-assisted IRMPD [HA–IRMPD]).By IRMPD of ubiquitin, at the drying gas temperatureof 200 �C, significant fragmentation occurred (data notshown). IRMPD spectra were measured at a CapExitvoltage of 90 V, and irradiation by the CO2 laser wasdone at a power of 8 W for 300 ms. The ESI spectraof ubiquitin, at a drying gas temperature of 200 �C,are shown in Fig. 4A, and [M + 12H]12+ ions were iso-lated and trapped in ICR cells by stored wavebandinverse Fourier transform (SWIFT) (Fig. 4B). In addi-tion, MS/MS spectra were obtained by IRMPD(Fig. 4C). Therefore, HA–IRMPD was an effectivemethod for the sequencing of isolated ions by MS/MS.

Intact lysozyme and BSA fragmentation by HANS–CID

and HA–IRMPD

Hen egg lysozyme and BSA have 4 and 17 intermo-lecular disulfide bonds, respectively. In a previous study,the collisionally activated dissociation mass spectraobtained by ESI of serum albumin were reported byLoo and co-workers [20] and Nemeth-Cawley andRouse [33]. Usually, it is difficult to obtain good sig-nal-to-noise dissociation spectra for these large proteinswith disulfide bonds [34]. Hence, disulfide bond cleavageby the addition of dithiothreitol is effective in promotingsignal-to-noise dissociation spectra [35–37]. In the com-prehensive top-down approach, the protein sample wasusually reduced and alkylated before MS analysis. How-

y198+

a344+

y535+

a345+

400 600 800 1000 12

y244+

Rel

ativ

e In

tens

ity

MH12+

Fig. 4. ESI MS and HA–IRMPD spectra of ubiquitin by the MS/MS metubiquitin ions in ICR cells by SWIFT; (C) HA–IRMPD spectra of isolated

ever, the solubility of the protein was decreased byreduction and alkylation; hence, several proteins wereprecipitated or aggregated. a-Lactalbumin (14.2 kDa),which contains four disulfide bonds and has a highstructural homology with lysozyme, resisted CID.Fig. 5 shows NS–CID and HANS–CID spectra of intacthen egg lysozyme obtained with varying CapExit voltag-es and drying gas temperatures. All spectra were accu-mulated eight scans of 7 pmol/ll of lysozyme withoutreduction and denaturation. As shown in Fig. 5A, the[M + 10H]10+, [M + 11H]11+, and [M + 12H]12+ ionsof lysozyme were observed. At a normal temperatureof 150 �C and a high CapExit voltage of 300 V (typicalNS–CID conditions), the dissociation of intact lysozymedid not occur (Fig. 5A). When the drying gas tempera-ture was increased to 200 �C, the fragment ions wereobserved at a CapExit voltage of 300 V (Fig. 5B). Inaddition, at the typical CapExit voltage of 150 V, theeffective dissociation of intact lysozyme occurred byHANS–CID at 250 and 300 �C (Figs. 5C and D, respec-tively). These fragment ions were generated in the atmo-sphere–vacuum interface by the heating and CapExitvoltage, and some fragment ions were assigned tosequences 32 to 41 (or 33–42) and 43 to 52, which areshown in Fig. 5D.

IRMPD and HA–IRMPD spectra of intact lysozyme,without reduction of disulfide bonds, are shown in Fig. 6.Multiple charged protonated molecules of intact lyso-zyme were trapped in ICR cells irradiated by a CO2 laserfor 100 ms.At anormal temperature of 150 �C, a few frag-ment ions were observed at a high CapExit voltage of300 V (Fig. 6A). At a high temperature of 200 �C and ahigh CapExit voltage of 220 V, multiple charged proton-

00 1400 1600 1800 m/z

A

B

C

hod: (A) ESI MS spectra of ubiquitin; (B) isolation of [M + 12H]12+

[M + 12H]12+ ubiquitin ions.

400 600 800 1000 1200 m/z

Rel

ativ

e In

tens

ity

43-52

32-41 or 33-42

A

B

C

D

Fig. 6. HA–IRMPD spectra of nonreduced intact lysozyme accumu-lated eight scans under the following conditions: CapExit voltage at300 V, hexapole accumulation at 3 s, and drying gas temperature at150 �C (A); CapExit voltage at 220 V, hexapole accumulation at 3 s,and drying gas temperature at 200 �C (B); CapExit voltage at 120 V,hexapole accumulation at 2 s, and drying gas temperature at 250 �C(C); CapExit voltage at 50 V, hexapole accumulation at 2 s, and dryinggas temperature at 300 �C (D).

500 1000 1500 2000 m/z

MH 10+

MH 11+

MH 12+

32-41 or 33-4243-52

Rel

ativ

e In

tens

ity

A

B

C

D

Fig. 5. NS–CID and HANS–CID spectra of nonreduced intact lysozyme accumulated eight scans under the following conditions: CapExit voltage at300 V, hexapole accumulation at 5 s, and drying gas temperature at 150 �C (A); CapExit voltage at 300 V, hexapole accumulation at 3 s, and dryinggas temperature at 200 �C (B); CapExit voltage at 150 V, hexapole accumulation at 2 s, and drying gas temperature at 250 �C (C); CapExit voltage at150 V, hexapole accumulation at 2 s, and drying gas temperature at 300 �C (D).

144 Effective novel dissociation methods for intact protein / N. Yamada et al. / Anal. Biochem. 348 (2006) 139–147

ated molecules of intact lysozyme were dissociated anddisappeared completely (Fig. 6B). At the typical CapExitvoltage of 120 V and a low voltage of 50 V (Figs. 6C andD, respectively), some fragment ions were observed withexcellent signal-to-noise spectra by HA–IRMPD at250 �C. In these HA–IRMPD spectra, the acquired Cap-Exit voltage for dissociation decreased, depending on thedrying gas temperature. For lysozyme, HA–IRMPD wasmore efficient for the dissociation of intact protein ions

than was HANS–CID. Several main fragment ions wereassigned by the use of TRASH and Protein Prospectorsoftware. Their peptide fragmentswere obtainedby disso-ciation at the C terminal of glutamine. In addition, weexamined whether main fragments were isolated and dis-sociated further by IRMPD and SORI–CID, but frag-ment ions were not observed. These results suggestedthat obtained fragment ions had rigid structure given thatIRMPD and HANS–CID were powerful dissociationtechniques. Therefore, it seems that these fragment ionswere not dissociated by the same heating energetic disso-ciation methods.

In a previous pioneering study by Loo and co-work-ers, [20] serum albumin from 10 different species, includ-ing BSA, were dissociated by NS–CID (CapExit voltageof 350 V) and MS/MS. Nearly all of the more abundantfragment ions were generated from the N-terminal pep-tide of BSA. Fig. 7A shows the HANS–CID spectrum ofnonreduced BSA (5 pmol/ll). The accumulation timeand flow rate were 10 s and 200 nl/min, respectively.The amount of sample needed to obtain this very goodsignal-to-noise spectrum was 166 fmol. The dissociationefficiency of the HANS–CID method is enough toobtain the sequence information. From the deconvolut-ed spectrum (Fig. 7B), the amino acid sequence could bedetermined to be L(/I)VL(/I)L(/I)AFSQYL(/I)QQ. Thiscandidate sequence confirmed that the amino acidsequence is LVLIAFSQYLQQ at positions 22 to 33 ofBSA. The assigned sequences were obtained from frag-mentation within the first 30 residues of the N terminusand are the same as results reported previously [20]. Thisassigned N-terminus region is not S–S bound, and thefragmentation efficiency of BSA was improved by gas-phase heating.

600 700 800 900 1000 1200 m/z

2800 3300 3800 m/z

L/IV

A

F SQ Y Q Q

L/IL/I

L/I

Rel

ativ

e In

tens

ityR

elat

ive

Inte

nsity A

B

Fig. 7. HANS–CID spectrum of nonreduced intact BSA (A) and its deconvoluted spectrum (B).

Effective novel dissociation methods for intact protein / N. Yamada et al. / Anal. Biochem. 348 (2006) 139–147 145

Intact IgG2b (150 kDa) fragmentation by HA–CID and

HA–IRMPD

All IgG molecules have a four-chain structure thatconsists of two identical light chains (23 kDa) and twoidentical heavy chains (50–70 kDa). IgG is a larger,150-kDa soluble glycoprotein and consists of a hetero-tetramer in the light and heavy chains. The tertiarystructure for a whole mouse IgG2a molecule was thefirst complete IgG molecule to be sequenced [38,39].The approximate structure of the 23-kDa light chaincontains one intrachain disulfide bond and consists oftwo regions that are VL and CL domains. The heavychain has 3 intrachain and 2 interchain disulfide bonds,

960412273

1210511694

11862

11880

10510

125012186

12291

11

2256

Rel

ativ

e In

tens

ity

1340 1390

11862

1188

1000 1100 1200

10510

Fig. 8. HA–IRMPD spectrum of nonreduced whole IgG2b (A

and the light chain has 1 interchain disulfide bond. Intotal, whole IgG2b has 12 disulfide bonds. The completeamino acid sequence and carbohydrate structure ofmouse anti-dansyl monoclonal antibody IgG2b weredetermined by liquid chromatography (LC) MS/MS[39]. The CID spectrum of multiple charged 150-kDaantibody ions has been reported previously, but fewfragment ions were observed due to very low fragmenta-tion efficiency and low signal-to-noise spectra [21,22].Fig. 8A shows the HANS–CID spectrum of intact non-reduced mouse anti-dansyl monoclonal antibody IgG2b,and Fig. 8B shows the expanded HANS–CID spectrumbetween 1300 and 1490 m/z. This good dissociationspectrum was obtained by use of a drying gas tempera-

1304813066

11862127342 11496 11694

11880

1304813066960412273

11862694

11862

1440 m/z

0

1300 1400 m/z

A

B

) and its deconvoluted spectrum (B) at 1000 to 1500 m/z.

146 Effective novel dissociation methods for intact protein / N. Yamada et al. / Anal. Biochem. 348 (2006) 139–147

ture of 250 �C and a CapExit voltage of 350 V. Themolecular weights of these observed fragment ions werecalculated by use of m/z values and differences in theweights of isotopic ions. The calculated molecularweights of fragment ions are indicated in Fig. 8A andB. Unfortunately, the amino acid sequences of theobserved fragment ions could not be assigned becausethere are many candidate sequences for the calculationof long amino acid sequences of IgG. All major frag-ment ions were approximately 10–13 kDa. A sugar moi-ety of IgG, such as N- or O-linked glycoside, wasprobably dissociated by CID [39]. For that reason, 10-to 13-kDa fragment ions consisted of a peptide moietywithout a carbohydrate. IgG consists of domains suchas CH1 and CH2. All domains have a similar molecularweight of approximately 10 to 13 kDa. Therefore, theobserved IgG fragment ions may be reflected in theIgG domain structure. For the identification and assign-ment of fragment ions from large molecules such as IgG,a further MS/MS analysis is necessary. Additional MS/MS analysis of the observed fragments obtained byHANS–CID and HA–IRMPD would be difficult bymeans of a thermal method, so some fragments ionsand those dehydrated ions were observed in HANS–CID spectrum (Fig. 8B). However, we believe that ourmethods seemed to be able to use the effective first disso-ciation step for large protein. A combination withanother dissociation method, such as ECD or anECD-related method, would probably be effective inthe identification of large protein.

Acknowledgment

The authors thank Ichio Shimada (Tokyo University)for providing the mouse IgG2b.

References

[1] G.E. Reid, S.A. McLuckey, ‘‘Top down’’ protein characterizationvia tandem mass spectrometry, J. Mass Spectrom. 37 (2002) 663–675.

[2] J.A. Loo, J.P. Quinn, S.I. Ryu, K.D. Henry, M.W. Senko,F.W. McLafferty, High-resolution tandem mass spectrometry oflarge biomolecules, Proc. Natl. Acad. Sci. USA 89 (1992) 286–289.

[3] N.L. Kelleher, H.Y. Lin, G.A. Valaskovic, D.J. Aaserud, E.K.Fridriksson, F.W. McLafferty, Top down versus bottom upprotein characterization by tandem high-resolution mass spec-trometry, J. Am. Chem. Soc. 121 (1999) 806–812.

[4] D.P. Little, J.P. Speir, M.W. Senko, P.B. O�Connor, F.W.McLafferty, Infrared multiphoton dissociation of large multiplycharged ions for biomolecule sequencing, Anal. Chem. 66 (1994)2809–2815.

[5] R.A. Chorush, D.P. Little, S.C. Beu, T.D. Wood, F.W. McLaff-erty, Surface induced dissociation of multiply-protonated pro-teins, Anal. Chem. 67 (1995) 1042–1046.

[6] J.R. Johnson, F. Meng, A.J. Forbes, B.J. Cargile, N.L. Kelleher,Fourier-transform mass spectrometry for automated fragmenta-tion and identification of 5–20 kDa proteins in mixtures, Electro-phoresis 23 (2002) 3217–3223.

[7] J.E. Gauthier, T.R. Trautman, D.B. Jacobson, Sustained off-resonance irradiation for collision-activated dissociation involvingFourier-transform mass spectrometry: collision-activated dissoci-ation technique that emulates infrared multiphoton dissociation,Anal. Chim. Acta 246 (1991) 211–225.

[8] M.W. Senko, J.P. Speir, F.W. McLafferty, Collisional activationof large multiply charged ions using Fourier transform-massspectrometry, Anal. Chem. 66 (1994) 2801–2808.

[9] R.A. Zubarev, N.L. Kelleher, F.W. McLafferty, Electron capturedissociation of multiply charged protein cations: a nonergodicprocess, J. Am. Chem. Soc. 120 (1998) 3265–3266.

[10] R.A. Zubarev, Electron-capture dissociation tandem mass spec-trometry, Curr. Opin. Biotechnol. 15 (2004) 12–16.

[11] K. Biemann, H.A. Scoble, Characterization by tandem massspectrometry of structural modifications in proteins, Science 237(1987) 992–998.

[12] K. Zhai, X. Han, K. Breuker, F.W. McLafferty, Consecutive ionactivation for top down mass spectrometry: improved proteinsequencing by nozzle-skimmer dissociation, Anal. Chem. 77(2005) 5777–5784.

[13] K.A. Sannes-Lowery, S.A. Hofstadler, Characterization of mul-tipole storage assisted dissociation: Implications for electrosprayionization mass spectrometry characterization of biomolecules, J.Am. Soc. Mass Spectrom. 11 (2000) 1–9.

[14] C. Pan, R.L. Hettich, Multipole-storage-assisted dissociation forthe characterization of large proteins and simple protein mixturesby ESI–FTICR–MS, Anal. Chem. 77 (2005) 3072–3082.

[15] Y. Ge, B.G. Lawhorn, M. ElNaggar, E. Strauss, J.H. Park, T.P.Begley, F.W. McLafferty, Top down characterization of largerproteins (45 kDa) by electron capture dissociation mass spec-trometry, J. Am. Chem. Soc. 124 (2002) 672–678.

[16] J.A. Loo, H.R. Udseth, R.D. Smith, Collisional effects on thecharge distribution of ions from large molecules, formed byelectrospray ionization mass spectrometry, Rapid Commun. MassSpectrom. 2 (1988) 207–210.

[17] R.D. Smith, J.A. Loo, C.G. Edmonds, C.J. Barinaga, H.R.Udseth, New developments in biochemical mass spectrometry:electrospray ionization, Anal. Chem. 62 (1990) 882–899.

[18] R.D. Smith, J.A. Loo, C.J. Barinaga, C.G. Edmonds, H.R.Udseth, Collisional activation and collision-activated dissociationof large multiply charged polypeptides and proteins produced byelectrospray ionization, J. Am. Soc. Mass Spectrom. 1 (1990)53–65.

[19] J.A. Loo, C.G. Edmonds, R.D. Smith, Primary sequence infor-mation from intact proteins by electrospray ionization tandemmass spectrometry, Science 248 (1990) 201–204.

[20] J.A. Loo, C.G. Edmonds, R.D. Smith, Tandem mass spectrom-etry of very large molecules: Serum albumin sequence informationfrom multiply charged ions formed by electrospray ionization,Anal. Chem. 63 (1991) 2488–2499.

[21] R. Feng, Y. Konishi, Analysis of antibodies and other largeglycoproteins in the mass range of 150,000–200,000 daltons byelectrospray ionization mass spectrometry, Anal. Chem. 64 (1992)2090–2095.

[22] R. Feng, Y. Konishi, Collisionally-activated dissociation ofmultiply charged 150-kDa antibody ions, Anal. Chem. 65 (1993)645–649.

[23] J.P. Speir, M.W. Senko, D.P. Little, J.A. Loo, F.W. McLafferty,High-resolution tandem mass spectra of 37–67 kDa proteins, J.Mass Spectrom. 30 (1995) 39–42.

[24] K.A. Sannes-Lowery, R.H. Griffey, G.H. Kruppa, J.P. Speir, S.A.Hofstadler, Multipole storage assisted dissociation: a novelin-source dissociation technique for electrospray ionization gen-

Effective novel dissociation methods for intact protein / N. Yamada et al. / Anal. Biochem. 348 (2006) 139–147 147

erated ions, Rapid Commun. Mass Spectrom. 12 (1998) 1957–1961.

[25] S.A. Hofstadler, K.A. Sannes-Lowery, R.H. Griffey, Infraredmultiphoton dissociation in an external ion reservoir, Anal. Chem.71 (1999) 2067–2070.

[26] N.L. Kelleher, S.V. Taylor, D. Grannis, C. Kinsland, H.J. Chiu,T.P. Begley, F.W. McLafferty, Efficient sequence analysis of thesix gene products (7–74 kDa) from the Escherichia coli thiaminbiosynthetic operon by tandem high-resolution mass spectrome-try, Protein Sci. 7 (1998) 1796–1801.

[27] J.L. Stephenson Jr., B.J. Cargile, S.A. McLuckey, Ion trapcollisional activation of disulfide linkage intact and reducedmultiply protonated polypeptides, Rapid Commun. Mass Spec-trom. 13 (1999) 2040–2048.

[28] N. Yamada, E. Suzuki, K. Hirayama, Identification of theinterface of a large protein–protein complex using H/D exchangeand Fourier transform ion cyclotron resonance mass spectrome-try, Rapid Commun. Mass Spectrom. 16 (2002) 293–299.

[29] S. Akashi, Y. Naito, K. Takio, Observation of hydrogen–deuterium exchange of ubiquitin by direct analysis of electrospraycapillary–skimmer dissociation with Fourier transform ion cyclo-tron resonance mass spectrometry, Anal. Chem. 71 (1999) 4974–4980.

[30] S. Akashi, K. Takio, Characterization of the interface structure ofenzyme–inhibitor complex by using hydrogen–deuteriumexchange and electrospray ionization Fourier transform ioncyclotron resonance mass spectrometry, Protein Sci. 9 (2000)2497–2505.

[31] D.M. Horn, R.A. Zubarev, F.W. McLafferty, Automated reduc-tion and interpretation of high resolution electrospray mass

spectra of large molecules, J. Am. Soc. Mass Spectrom. 11 (2000)320–332.

[32] Protein Prospector, version 4.0.5, http://prospector.ucsf.edu/(Retrieved 25 January 2005).

[33] J.F. Nemeth-Cawley, J.C. Rouse, Identification and sequencinganalysis of intact proteins via collision-induced dissociation andquadrupole time-of-flight mass spectrometry, J. Mass Spectrom.37 (2002) 270–282.

[34] J.J. Lennon, K.A. Walsh, Direct sequence analysis of proteins byin-source fragmentation during delayed ion extraction, ProteinSci. 6 (1997) 2446–2453.

[35] M. Thevis, R.R. Loo, J.A. Loo, Mass spectrometric character-ization of transferrins and their fragments derived by reductionof disulfide bonds, J. Am. Soc. Mass Spectrom. 14 (2003) 635–647.

[36] J.A. Loo, C.G. Edmonds, H.R. Udseth, R.D. Smith, Effect ofreducing disulfide-containing proteins on electrospray ionizationmass spectra, Anal. Chem. 62 (1990) 693–698.

[37] J.M. Ginter, F. Zhou, M.V. Johnston, Generating proteinsequence tags by combining cone and conventional collisioninduced dissociation in a quadrupole time-of-flight mass spec-trometer, J. Am. Soc. Mass Spectrom. 15 (2004) 1478–1486.

[38] L.J. Harris, S.B. Larson, K.W. Hasel, J. Day, A. Greenwood, A.McPherson, The three-dimensional structure of an intact mono-clonal antibody for canine lymphoma, Nature 360 (1992) 369–372.

[39] K. Hirayama, R. Yuji, N. Yamada, K. Kato, Y. Arata, I.Shimada, Complete and rapid peptide and glycopeptide mappingof mouse monoclonal antibody by LC/MS/MS using ion trapmass spectrometry, Anal. Chem. 70 (1998) 2718–2725.