Enhancing electrospray ionization efficiency of peptides by derivatization

Article abstract of DOI:10.1021/ac0602266

With the advent of electrospray ionization mass spectrometry, the world was given a new way to look at complex peptide mixtures. Identification of proteins via their signature peptides requires ionization of a representative portion of the peptides derived from proteins by proteolysis. Unfortunately, matrix effects prohibited electrospray ionization of many peptides. This paper describes the development of a new labeling reagent that simultaneously adds a permanent positive charge to peptides and increases their hydrophobicity to enhance their ionization efficiency. The labeling agent is preactivated with N-hydroxysuccinimide to react with primary amines to form a peptide bond. In the most dramatic case, ionization efficiency of the peptide ADRDQYELLCLDNTRKPVDEYK increased 500-fold after derivatization as opposed to other peptides where ionization efficiency was impacted little. Ionization efficiency of peptides was enhanced roughly 10-fold in general by derivatization. Peptides of less than 500 Da experienced the greatest increase in ionization efficiency by derivatization. Poor ionization efficiency of native peptides was found to be due more to their inherent structural properties than the matrix in which ionization occurs.

Full text of DOI:10.1021/ac0602266

Anal. Chem. 2006, 78, 4175-4183
Enhancing Electrospray Ionization Efficiency of
Peptides by Derivatization
Hamid Mirzaei and Fred Regnier*
Department of Chemistry, Purdue University, West Lafayette, Indiana 47907
Passage of peptides from droplets formed during the electro-
spray process into the gas phase is the result of peptide
desolvation.6 Peptide hydrophobicity seems to play a role in
desolvation, probably because hydrophobicity dictates the rate at
which peptides migrate to the surface of droplets.7-13 But for these
gas-phase species to be detected they need to acquire charge.
Obviously basic peptides are more likely to absorb a proton and
ionize. Attaching a quaternary amine to peptides enhances
ionization by providing a permanent positive charge. It has been
noted that when hydrophobicity and good gas-phase proton affinity
are combined in hydrophobic, cationic peptides they ionize more
readily along with suppressing the ionization of other peptides.14,15
It is interesting that addition of tetramethylammonim bromide to
a solution of peptides being electrosprayed suppresses the
ionization of all peptides.16 This means that surfactantlike species
can saturate the droplet surface and push peptides toward the
interior of droplets. It also explains how hydrophobic peptides
derived from more abundant proteins suppress ionization. They
diminish or eliminate droplet surface area for other peptides to
be protonated and pass into the gas phase.
With the advent of electrospray ionization mass spectrom-
etry, the world was given a new way to look at complex
peptide mixtures. Identification of proteins via their
signature peptides requires ionization of a representative
portion of the peptides derived from proteins by proteoly-
sis. Unfortunately, matrix effects prohibited electrospray
ionization of many peptides. This paper describes the
development of a new labeling reagent that simultaneously
adds a permanent positive charge to peptides and in-
creases their hydrophobicity to enhance their ionization
efficiency. The labeling agent is preactivated with N-
hydroxysuccinimide to react with primary amines to form
a peptide bond. In the most dramatic case, ionization
efficiency of the peptide ADRDQYELLCLDNTRKPVDEYK
increased 500-fold after derivatization as opposed to other
peptides where ionization efficiency was impacted little.
Ionization efficiency of peptides was enhanced roughly 10-
fold in general by derivatization. Peptides of less than 500
Da experienced the greatest increase in ionization ef-
ficiency by derivatization. Poor ionization efficiency of
native peptides was found to be due more to their inherent
structural properties than the matrix in which ionization
occurs.
The hypothesis being tested in this paper is that by tagging
peptides with a derivatizing agent that contains both a quaternary
amine and an alkyl group it will be possible to increase peptide
migration to the surface of droplets and concomitantly electrospray
ionization efficiency. The derivatization process used in these
studies was similar to that used in stable isotope coding for relative
quantification.17
Protein identification in proteomics is achieved in several ways.
One is to tryptic digest a proteome and after several dimensions
of chromatographic fractionation peptide cleavage fragments are
identified by electrospray ionization mass spectrometry (ESI-
MS).1,2 Because a tryptic digest of even a simple proteome can
contain several hundred thousand peptides, chromatographic
fractions being introduced into an ESI-MS can contain hundreds
to thousands of components.3,4 This causes a problem in ESI-MS.
As the complexity of samples being introduced into the instrument
increases, many peptides fail to ionize because of a phenomenon
known as matrix suppression of ionization.5
EXPERIMENTAL SECTION
Materials. Synthetic peptides Ac-Gln-Lys-Arg-Pro-Ser-Gln-Arg-
Ser-Lys-Tyr-Leu-OH, H-Asp-Arg-Val-Tyr-Ile-His-Pro-Phe-His-Leu-
OH, H-Ala-Phe-Pro-Leu-Glu-Phe-OH, H-Ser-Tyr-Ser-Met-Glu-His-
Phe-Arg-Trp-Gly-OH, H-Cys-Asp-Pro-Gly-Tyr-Ile-Gly-Ser-Arg-OH,
H-Tyr-Gly-Gly-Phe-Met-Lys-OH, H-Pro-His-Pro-Phe-His-Phe-His-
(6) Ohashi, Y. Kagaku (Kyoto, Japan) 1991, 46, 627-633.
(7) Iribarne, J. V.; Dziedzic, P. J.; Thomson, B. A. Int. J. Mass Spectrom. Ion
Phys. 1983, 50, 331-347.
* Corresponding author. E-mail: fregnier@purdue.edu.
(1) Fenyoe, D.; Eriksson, J.; Beavis, R. C. Informatics Proteomics 2005, 267-
275.
(2) Chen, E. I.; Hewel, J.; Felding-Habermann, B.; Yates, J. R., III. Mol. Cell.
Proteomics 2006, 5, 53-56.
(8) Fenn, J. B. J. Am. Soc. Mass Spectrom. 1993, 4, 524-535.
(9) Kebarle, P.; Tang, L. Anal. Chem. 1993, 65, 972A-986A.
(10) Tang, K.; Smith, R. D. J. Am. Soc. Mass Spectrom. 2001, 12, 343-347.
(11) Cech, N. B.; Enke, C. G. Anal. Chem. 2000, 72, 2717-2723.
(12) Zhou, S.; Cook, K. D. J. Am. Soc. Mass Spectrom. 2001, 12, 206-214.
(13) Cech, N. B.; Krone, J. R.; Enke, C. G. Anal. Chem. 2001, 73, 208-213.
(14) Rundlett, K. L.; Armstrong, D. W. Anal. Chem. 1996, 68, 3493-3497.
(15) Sioma, C. S. MS, Purdue University, West Lafayette, IN, 2003.
(16) Pan, P.; McLuckey, S. A. Anal. Chem. 2003, 75, 5468-5474.
(17) Julka, S.; Regnier, F. E. Briefings Funct. Genomics Proteomics 2005, 4, 158-
177.
(3) Naylor, S.; Adamec, J.; Meys, M. Beyond Genomics, USA. Application: WO,
2004.
(4) Martosella, J.; Zolotarjova, N.; Liu, H.; Nicol, G.; Boyes, B. E. J. Proteome
Res. 2005, 4, 1522-1537.
(5) Regnier, F. E.; Chakraborty, A. B.; Dormady, S. J.; G′Eng, M.; Ji, J.; Riggs,
L. D.; Sioma, C. S.; Wang, S.; Zhang, X. Purdue Research Foundation. USA
Application: WO, 2001.
10.1021/ac0602266 CCC: $33.50 © 2006 American Chemical Society
Published on Web 04/28/2006
Analytical Chemistry, Vol. 78, No. 12, June 15, 2006 4175

Phe-Phe-Val-Tyr-Lys-OH, H-Ala-Leu-Gly-OH, H-Ala-Gly-Gly-OH,
H-Ala-Met-OH, and H-Ala-Ser-OH were purchased from Bachem
Biosience Inc. (King of Prussia, PA). Trifluoroacetic acid (TFA)
was purchased from Pierce Co. (Rockford, IL). Dithiothreitol
(DTT), trypsin, N-R-tosyl-L-lysine chloromethyl ketone (TLCK),
apotransferrin (human), 4-iodobutyric acid, N-hydroxysuccinimide
(NHS), dicyclohexylcarbodiimide (DCC), 4-aminobutanoic acid,
potassium bicarbonate, methyl iodide, chloroform, and N,N-
dimethyl-N-octylamine were obtained from Sigma-Aldrich Chemi-
cal Co. (St. Louis, MO). HPLC-grade acetonitrile (ACN) and urea
were purchased from Mallinckrodt. (St. Louis, MO). 218TP54
reversed-phase C18 columns were purchased from Vydac (W. R.
Grace & Co.-Conn. Columbia, MD). DE44H10426 Zorbax reversed-
phase C18 column (0.5 × 150 mm) and DE45C00085 Zorbax
reversed-phase C8 column (0.3 × 150 mm) were purchased from
Agilent Technologies, Inc. (Palo Alto, CA). Reversed-phase chro-
matography (RPC) analyses were done on a BioCAD 20 Mi-
croanalytical Workstation (PE Biosystems, Framingham, MA).
The LC system used in conjunction with the mass spectrometer
was an Agilent 1100 series instrument. LC/MS mass spectral
analyses were done using a Sciex QSTAR hybrid LC/MS/MS
Quadrupole TOF mass spectrometer (Applied Biosystems, Foster
City, CA). All spectra were obtained in the positive ion mode.
Methods. Synthesis of 4-Iodo(2,5-dioxopyrrolidin-1-yl)
Butyrate. This reagent was synthesized according to the proce-
dure of Taran et al.18 4-Iodobutyric acid (1 g, 467 mmol), NHS
(0.54 mg, 4.69 mmol), and DCC (0.96 g, 4.66 mmol) were
dissolved in 10 mL of AcOEt. The reaction mixture was stirred at
room temperature for 16 h. The dicyclohexyl urea formed was
discarded by filtration over Celite, and the crude material was
concentrated and chromatographed on silica gel (petroleum
ether/AcOEt, 1/1). The resultant yellow powder was recrystalized
from petrolumn ether/EtOH(9/1) yielding activated ester 4-iodo-
(2,5-dioxopyrrolidin-1-yl) butyrate.
Synthesis of [3-(2,5)-Dioxopyrrolidin-1-yloxycarbonyl)-
propyl]dimethyloctylammonium (C8-QAT). This reagent was
synthesized according to the procedure by Taran et al. with slight
modifications.18 N,N-Dimethyl-N-octylamine (280 mg, 1.8 mmol),
activated ester 4-iodo[2,5-dioxopyrrolidin-1-yl] butyrate (550 mg,
1.8 mmol), and silver trifluoromethanesulfonate (460 mg, 1.8
mmol) were dissolved in 2 mL of anhydrous acetone. The solution
was stirred for 16 h at room temperature. The silver iodide formed
was discarded by filtration over Celite, and the solvent was
evaporated. The resultant orange oil was dissolved in 1 mL of
ACN, and the final product was precipitated by addition of 10 mL
of EtOAc.
Figure 1. Structures of the QAT and C8-QAT reagents. The C8-
QAT reagent is more hydrophobic than QAT and will convey greater
hydrophobicity to labeled peptides.
Figure 2. Reversed-phase chromatograms for (1) unlabeled and
(2) C8-QAT-labeled transferrin digest separated on a C18 column. The
digests were separated on a Vydac C18 column using a 60-min
gradient from 99.5% buffer A (0.01% TFA in dI H2O) to 60% buffer B
(95% ACN/0.01% TFA in dI H2O). There is a 15-min delay in retention
time caused by increased hydrophobicity of the labeled peptides. The
labeled digest also shows a higher level of complexity and longer
gradient elution due to retention of very hydrophilic peptides that did
not retain before labeling. The labeling reagent peaks were removed
by subtraction of labeling reagent only chromatogram from the labeled
digest’s.
remaining residue extracted with 400 × 2 mL of anhydrous
acetone, concentrated under reduced pressure, slurried with 200
mL of THF, and filtered to yield the final product.
Synthesis of [3-(2,5)-dioxopyrrolidin-1-yloxycarbonyl)-
propyl]trimethyllammonium (QAT). This reagent was prepared
according to the method of Sioma et al.15 A 13.2-g (0.073 mol)
aliquot of (3-carboxypropyl)trimethylammonium chloride, 8.4
g(0.073 mol) of NHS, and 16.5 g (0.080 mol) of DCC were
dissolved in ACN. Once dissolved, the resulting solution was
refrigerated for 24 h. The formed 1,3 dicyclohexylurea was filtered
off, and the ACN was evaporated under vacuum. The crude
residue was slurried with 200 mL of THF, filtered, and recrystal-
lized from ACN/THF to yield the final product.
Synthesis of (3-Carboxypropyl)trimethylammonium Chlo-
ride. (CH3)N+(Cl)CH2CH2CH2(CdO)OH was prepared according
to the method of Sioma et al.15 A 10.0-g (0.097 mol) aliquot of
4-aminobutanopic acid, 48.0 g (0.497 mol) of KHCO3, and 30 mL
(0.482 mol) of CH3I were stirred in 1 L of MeOH for 24 h. The
MeOH was removed under vacuum and the crude residue slurried
with 200 mL of CHCl3, then filtered, and acidified with 50 mL of
concentrated HCl. The H2O was removed under vacuum and the
Derivatization of Model Peptides with QAT and C8-QAT
Reagent. Model peptides were dissolved in 50 mM Hepes, pH
8.00, at a final concentration of 1 mg/mL. A 50-fold molar excess
of each derivatization reagent was added individually to the peptide
(18) Taran, F.; Renard, P. Y.; Bernard, H.; Mioskowski, C.; Frobert, Y.; Pradelles,
P.; Grassi, J. J. Am. Chem. Soc. 1998, 120, 3332-3339.
4176 Analytical Chemistry, Vol. 78, No. 12, June 15, 2006

Figure 3. Ion chromatograms for (1) QAT-labeled and (2) C8-QAT-labeled transferrin digest separated on a C8 reversed-phase column.
Separation of transferrin digests on a less hydrophobic reversed-phase column such as C8 allows the hydrophobic effect to become more
dominant. The retention time delay as a result of C8-QAT labeling is reduced to 5 min and elution of C8-QAT-labeled peptides at lower organic
percentage enhances the hydrophobic effect. Some QAT-labeled peptides (5%) do not retain on C8 column, but overall there is 1.25-fold increase
in ionization efficiency. The C8-QAT-labeled peptides are also distributed more evenly across the chromatogram. Decrease in local concentration
of peptides in droplets also decreases the ionization suppression.
solutions, and the reaction was allowed to proceed for 2 h at room
temperature. After the reaction was finished, N-hydroxylamine was
added in excess, and the pH was adjusted to 11-12 with sodium
hydroxide to hydrolyze esters. The reaction was allowed to
proceed for 10 min, and then the pH was adjusted back to 7-8
with 1-2 drops of glacial acetic acid.
Proteolysis. To denature, reduce, and alkylate transferrin,
urea and DTT were added to a final concentration of 6 M and 10
mM, respectively. Mixtures were incubated for 1 h at 37 °C,
iodoacetamide was then added to a final concentration of 20 mM,
and the reaction allowed to proceed for an additional 30 min at 4
°C. Cysteine was then added to a final concentration of 10 mM to
quench extra iodoacetamide. Samples were diluted 6-fold with 50
mM HEPES, pH 8.0, 10 mM MgCl2, and 10 mM CaCl2. Sequence
grade trypsin was added (2%), and the reaction mixture was
incubated at 37 °C for at least 8 h. Proteolysis was stopped by
adding TLCK (trypsin/TLCK ratio of 1:1 (w/w)).
The reaction was allowed to proceed for 10 min, and then the pH
was adjusted back to 7-8 with 1-2 drops of glacial acetic acid.
LC/MS Analysis. Derivatized and nondrivatized model pep-
tides as well as transferrin digest were separated on an Agilent
Zorbax C8 (0.3 × 150) and C18 column (0.5 × 150 mm) using an
Agilent 1100 series instrument (Agilent Technologies, Inc.) at 4
µL/min. Solvent A was 0.01% TFA in deionized H2O (dI H2O),
and solvent B was 95% ACN/0.01% TFA in dI H2O. The flow from
the column was directed to the Q-STAR workstation (Applied
Biosystems, Framingham, MA) equipped with an ESI source. Flow
from the HPLC was diverted to waste for 10 min after sample
injection at 100% solvent A to remove salts, remaining derivatizing
reagent APTA, and weakly adsorbed peptides. The Q-STAR was
then reconnected and peptides were separated in a 15-min linear
gradient (from 0 to 60% B). MS spectra were obtained in the
positive ion mode at a sampling rate of one spectrum/s.
LC Separation of Tryptic Peptides. Peptides from the
transferrin tryptic digest were separated on a Vydac C18 column
(4.6 × 250 mm) using an BioCAD 20 Micro-Analytical Workstation
(Applied Biosystems, Framingham, MA) at 1 mL/min. Solvent A
was 0.1% TFA in dI H2O and solvent B was 95% ACN/0.1% TFA in
dI H2O. Peptides were separated in a 60-min linear gradient (from
0 to 60% B).
Derivatization of Model Peptides with QAT and C8-QAT
Reagent. A 50-fold molar excess of each derivatization reagent
was added to the tryptic digest peptide, and the reaction was
allowed to proceed for 2 h at room temperature. After the reaction
was finished, N-hydroxylamine was added in excess and the pH
was adjusted to 11-12 with sodium hydroxide to hydrolyze esters.
Analytical Chemistry, Vol. 78, No. 12, June 15, 2006 4177

Table 1. List of Model Peptides Used To Study the Effect of QAT and C8-QAT Labeling on Ionization Efficiencya
Ri based ratio
peptide
native
area
QAT
area
203
C8-QAT
area
native:QAT:C8-QAT
1
Ac-Gln-Lys-Arg-
Pro-Ser-Gln-Arg-
Ser-Lys-Tyr-Leu-
OH (HP: -24.5)
(715.89)2+
0
(844.50)2+
(942.11)2+
0
1:12:20
(477.60)3+
(358.95)4+
(1296.69)1+
260
0
2362
(563.34)3+
(422.75)4+
(1423.78)1+
1364
1663
25
(628.41)3+
(471.56)4+
(1521.89)1+
2519
2471
24
2
3
4
H-Asp-Arg-Val-Tyr-
Ile-His-Pro-Phe-
1:0.82:0.21
1:1:0.56
His-Leu-OH (HP: -1.3)
(648.89)2+
(432.89)3+
(723.37)1+
30491
25886
26652
(712.39)2+
(475.26)3+
(850.47)1+
18204
30599
23236
(761.45)2+
(507.49)3+
(948.58)1+
14058
5175
11548
H-Ala-Phe-Pro-
Leu-Glu-Phe-OH
(HP: 6.1)
(362.19)2+
(241.79)3+
(1299.55)1+
724
0
1470
(425.73)2+
(284.16)3+
(1426.65)1+
3190
0
208
(474.79)2+
(316.86)3+
(1524.76)1+
3372
296
111
H-Ser-Tyr-Ser-Met-
Glu-His-Phe-Arg-
Trp-Gly-OH
1:1.91:0.83
(HP: -10.7)
(650.28)2+
(433.85)3+
(967.43)1+
21985
6631
5797
(713.83)2+
(476.22)3+
(1094.53)1+
15123
42773
686
(762.88)2+
(508.59)3+
(1192.64)1+
11110
14162
50
5
6
7
H-Cys-Asp-Pro-
1:0.14:0.34
1:6.67:33.3
1:2:1
Gly-Tyr-Ile-Gly-Ser-
Arg-OH (HP: -5.1)
(484.21)2+
(323.14)3+
(701.32)1+
4160
0
0
(554.27)2+
(369.84)3+
(957.53)1+
893
0
339
(596.82)2+
(398.22)3+
(1153.75)1+
3691
0
28938
H-Tyr-Gly-Gly-Phe-
Met-Lys-OH
(HP: -1.3)
(351.16)2+
(234.78)3+
(1318.67)1+
1656
0
771
(479.27)2+
(319.51)3+
(1445.77)1+
9766
0
53
(576.87)2+
(384.92)3+
(1543.88)1+
28603
0
46
H-Pro-His-Pro-Phe-
His-Phe-His-Phe-
Phe-Val-Tyr-Lys-OH
(HP: -2.6)
(659.84)2+
(440.22)3+
(259.15)1+
9579
12440
3082
(723.39)2+
(482.59)3+
(387.26)1+
4056
6724
2687
(772.44)2+
(515.29)3+
(485.37)1+
5569
17173
47781
8
H-Ala-Leu-Gly-OH
(HP: 5.2)
1:1:16.67
1:16.67
(130.07)2+
(203.09)1+
0
0
(194.14)2+
(331.19)1+
0
3170
(243.18)2+
(429.31)1+
0
2781
9
H-Ala-Gly-Gly-OH
(HP: 1)
(102.16)2+
(220.08)1+
0
355
(166.10)2+
(348.19)1+
0
(215.15)2+
(446.30)1+
10251
43094
10
11
H-Ala-Met-OH
(HP: 3.7)
10989
1:31.25:125
1:20:20
(110.54)2+
(176.07)1+
0
155
(174.60)2+
(304.18)1+
0
2156
(223.652+
(402.29)1+
0
2304
H-Ala-Ser-OH
(HP: 1)
(130.07)2+
0
(152.50)2+
0
(201.65)2+
0
a All peptides were labeled with QAT and C8-QAT. Native, QAT-labeled and C8-QAT-labeled forms of each peptide were mixed in 1:1:1 ratio and
separated on a C18 column. The injection volume was 5 µL in all cases. The area under the peak for all charge states of each peptide were summed
and used as a total ionization measure. The ratios were calculated by dividing the total peak areas for each peptide by the total peak area of the
native peptide.
RESULTS
Derivatizing Agent Architecture. The strategy for enhancing
ure 1) which has been previously described.15 Peptides derivatized
with this quaternary amine tag (QAT) have been shown to have
improved ionization efficiency.19
electrospray ionization of peptides through derivatization was to
use a derivatizing agent that would (1) have both quaternary amine
and alkyl groups, (2) N-acylate primary amines, (3) be preactivated
to form peptide bonds, and (4) acylate peptides in water. The
reagent [3-(2,5)-dioxopyrrolidin-1-yloxycarbonyl)propyl]dimethy-
loctylammonium (Figure 1) was developed to meet these criteria.
Peptides derivatized with this reagent will be designated as having
a C8-quaternary amine tag (C8-QAT). When activated with NHS,
this tagging agent will be referred to as C8-QAT-NHS. The
synthesis of this derivatizing agent is very similar to that of [3-(2,5)-
dioxopyrrolidin-1-yloxycarbonyl)-propyl]-trimethyllammonium (Fig-
Peptide Derivatization. The C8-QAT-NHS labeling agent was
added to peptide mixtures in at least a 50-fold molar excess and
incubated at room temperature overnight. The reaction was
terminated by addition of excess N-hydroxylamine and the pH
adjusted to 11-12 with a few drops of saturated sodium hydroxide
solution. Hydrolysis of C8-QAT-NHS and any O-acylation products
was achieved in 10 min after the addition of base. The pH was
then adjusted back to 7-8 with 1-2 drops of glacial acetic acid.
(19) Sioma, C. S., 2003.
4178 Analytical Chemistry, Vol. 78, No. 12, June 15, 2006

Figure 4. Extracted ion chromatograms for native, QAT-, and C8-QAT-labeled model peptide Ac-Gln-Lys-Arg-Pro-Ser-Gln-Arg-Ser-Lys-Tyr-
Leu-OH. All the charge states for different forms of the peptide were extracted from the total ion chromatogram. The total ion chromatogram
was obtained by separation of a mixture of native, QAT-, and C8-QAT-labeled peptide mixed in 1:1:1 ratio on a C18 column. The area under the
peak for each charge state of each labeled peptides was calculated and then summed. The summation of peak areas of all charge states of a
peptide form was used as a representative of total ionization for that peptide.
Reactivity of the C8-QAT-NHS was found to be the same as that
of the previously described QAT-NHS reagent (Figure 1).
Effect of C8-QAT Labeling on Reversed-Phase Chromato-
graphic Retention Time. The impact of C8-QAT derivatization
on peptide retention in RPC was studied using a transferrin tryptic
digest. Native peptides and those tagged with C8-QAT were
prepared and fractionated in separate runs on a Vydac C18 column
using a 60-min linear gradient starting from solvent A to 60%
solvent B. Figure 2 shows overlaid chromatograms of the labeled
and unlabeled tryptic peptide digests. C8-QAT labeling induced
roughly a 20-min increase in the retention time of peptides. The
relative increase in retention of small hydrophilic peptides was
greater than that of larger hydrophobic peptides.
Analyte retention in reversed-phase chromatography (RPC) is
generally accepted to be due to a solvophobic effect driven by
the surface tension of a polar mobile phase. According to
solvophobic theory, polar solvents force hydrophobic molecules
to interact with molecules that are similarly hydrophobic to
minimize their contact area with the solvent.20,21 Although peptides
have a large number of polar groups, the side chains of many
amino acids in peptides are hydrophobic. The solvophobicity of
these hydrophobic groups in polar mobile phases drives peptides
to interact with the solvophobic surface of RPC columns. But not
all peptides are retained by RPC columns. Those that are not lack
sufficient hydrophobicity to interact with the stationary phase. As
observed, the hydrophobic C8-QAT would be expected to increase
the retention of a small hydrophilic peptide in RPC more than
that of a large hydrophobic peptide that already interacts strongly
with the stationary phase.
Although increasing the retention of hydrophilic peptides
beyond the solvent front is desirable, a 20-min increase in retention
is greater than needed. The QAT and C8-QAT derivatized trans-
ferrin digests were also examined with an octylsilane (C8 Zorbax)
column (Figure 3). Comparing the chromatograms in Figures 2
and 3, it is seen that retention of peptides tagged with the C8-
QAT reagent is substantially shorter on the octyl (C8) than the
octadecyl (C18) RPC column. The reduced hydrophobicity of the
C8 Zorbax column compensates for the increased hydrophobicity
of peptides derivatized with the C8-QAT reagent. Directing the
nonretained peak from the C8 column into a C18 column showed
that 5% of the peptides captured by the C18 column were not
retained on the C8 column. Peptides not retained by the C8 column
are generally small hydrophilic peptides of minimal value in
protein identification (data not shown).
Effect of C8-QAT Labeling on Electrospray Ionization of
Model Peptides. Tagging peptides with C8-QAT alters them in
two important ways that might impact electrospray ionization. One
is the introduction of a quaternary amine, giving them a permanent
positive charge. The second is to make them more hydrophobic.
The relative contribution of quaternization was studied by deriva-
tizing model peptides with the QAT reagent. As noted in Figure
(20) Molnar, I. Chromatographia 2005, 62, S7-S17.
(21) Vailaya, A.; Horvath, C. Book of Abstracts, 218th ACS National Meeting, New
Orleans, August 22-26 1999; ANYL-029.
Analytical Chemistry, Vol. 78, No. 12, June 15, 2006 4179

Table 2. List of Model Peptides Used to Study the Effect of QAT and C8-QAT Labeling on Ionization Efficiencya
Ri based ratio
peptide
native
area
QAT
area
C8-QAT
area
127
native:QAT:C8-QAT
1
2
Ac-Gln-Lys-Arg-Pro-
Ser-Gln-Arg-Ser-Lys-
Tyr-Leu-OH
(715.89)2+
0
(844.50)2+
0
(942.11)2+
1:2.85:7.14
(HP: -24.5)
(478.27)3+
(358.95)4+
(1296.69)1+
591
1587
305
(563.34)3+
(422.75)4+
(1423.78)1+
3798
2189
59
(628.41)3+
(471.56)4+
(1521.89)1+
7621
7154
78
H-Asp-Arg-Val-Tyr-Ile-
His-Pro-Phe-His-Leu-
OH (HP: -1.3)
1:1:2
(648.89)2+
(432.89)3+
(723.37)1+
7259
5865
32622
(712.39)2+
(475.26)3+
(850.47)1+
3845
6388
29211
(761.45)2+
(507.49)3+
(948.58)1+
13672
11320
19766
3
4
H-Ala-Phe-Pro-Leu-
1:1:1
Glu-Phe-OH (HP: 6.1)
(362.19)2+
(241.79)3+
(1299.55)1+
0
559
43
(425.73)2+
(284.16)3+
(1426.65)1+
3661
138
20
(474.79)2+
(316.86)3+
(1524.76)1+
13012
1076
229
H-Ser-Tyr-Ser-Met-Glu-
His-Phe-Arg-Trp-Gly-
OH (HP: -10.7)
1:1.46:7.69
(650.28)2+
(433.85)3+
(967.43)1+
2286
2848
2310
(713.83)2+
(476.22)3+
(1094.53)1+
2525
4787
1121
(762.88)2+
(508.59)3+
(1192.64)1+
22748
15638
928
5
H-Cys-Asp-Pro-Gly-
Tyr-Ile-Gly-Ser-Arg-OH
(HP: -5.1)
1:0.2:2
(484.21)2+
(323.14)3+
(701.32)1+
3148
0
0
(554.27)2+
(369.84)3+
(957.53)1+
0
0
668
(596.82)2+
(398.22)3+
(1153.75)1+
9955
0
576
6
7
H-Tyr-Gly-Gly-Phe-Met-
1:6.8:20
Lys-OH (HP: -1.3)
(351.16)2+
(234.78)3+
(1318.67)1+
1316
0
188
(479.27)2+
(319.51)3+
(1445.)1+
8674
00
129
(576.87)2+
(384.92)3+
(1543.88)1+
26563
0
0
H-Pro-His-Pro-Phe-His-
Phe-His-Phe-Phe-Val-
Tyr-Lys-OH (HP: -2.6)
1:0.33:6.67
(659.84)2+
(440.22)3+
(259.15)1+
3958
2437
129
(723.39)2+
(482.59)3+
(387.26)1+
921
1112
2002
(772.44)2+
(515.29)3+
(485.37)1+
9056
33390
50151
8
H-Ala-Leu-Gly-OH
(HP: 5.2)
1:13.3:333
(130.07)2+
(203.09)1+
0
0
(194.14)2+
(331.19)1+
0
0
(243.18)2+
(429.31)1+
0
6598
9
H-Ala-Gly-Gly-OH
(HP: 1)
∞
(102.16)2+
(220.08)1+
0
0
(166.10)2+
(348.19)1+
0
0
(215.15)2+
(446.30)1+
0
10
11
H-Ala-Met-OH
(HP: 3.7)
12507
∞
(110.54)2+
(176.07)1+
0
0
(174.60)2+
(304.18)1+
0
189
(223.652+
(402.29)1+
0
3524
H-Ala-Ser-OH
(HP: 1)
1:20
(89.04)2+
0
(152.50)2+
0
(201.65)2+
0
a All peptides were labeled with QAT and C8-QAT. Native, QAT-labeled, and C8-QAT-labeled forms of each peptide were mixed in 1:1:1 ratio and
separated on a C18 column. The injection volume was 20 µL in all cases (4 times the amount used in previous table). The area under the peak for
all charge states of each peptide were summed and used as a total ionization measure. The ratios were calculated by dividing the total peak areas
for each peptide by the total peak area of the native peptide (Ri). In higher concentration, all C8-QAT-labeled peptides show higher ionization
efficiency than other forms of the same peptide (native and QAT labeled).
1, the QAT and C8-QAT reagents are identical in structure with
the exception of the substitution of an octyl group for a methyl
group on the quaternary amine. It has been previously reported
that some peptides tagged with the (QAT) reagent experience a
moderate increase in electrospray ionization efficiency.19
60% buffer B (95% ACN/0.01% TFA in dI H2O). Relative ionization
(Ri) of peptides was calculated using the formula
Ri ) A+i n
/
A+n
∑ ∑ native
The relative ionization of native, QAT-labeled, and C8-QAT-
labeled peptides was studied using 11 model peptides varying in
size, hydrophilic to hydrophobic amino acid ratio, and number of
cationic amino acids (Table 1). Hydrophobicity (HP) of native
peptides was calculated using relative hydrophobicity of their
amino acids. All the model peptides were labeled separately with
both the QAT and C8-QAT reagents and mixed with native
peptides in a 1:1:1 ratio (See Figure 4). The mixture (5-µL injection
volume) was separated on a Zorbax C18 column using a 60-min
gradient starting from 99.5% buffer A (0.01% TFA in dI H2O) to
+n
where ∑A
is the sum of the area under the curves of all
native
charge states of the native peptide and ∑A+i n is the sum of the
area under the curves for all the charge states of the peptide for
which Ri is being calculated (peptide i). Since Ri is normalized
against a native peptide, the value of Ri for the native peptide is
always 1.
It is seen in Table 1 that for the peptide Ac-Gln-Lys-Arg-Pro-
Ser-Gln-Arg-Ser-Lys-Tyr-Leu-OH that Ri for the C8-QAT peptide
is 20 while Ri for the QAT derivatized form of the peptide is 12.
4180 Analytical Chemistry, Vol. 78, No. 12, June 15, 2006

Table 3. List of Native and C8-QAT-Labeled Transferrin Peptides Identified by LC/MS/MS Analysisa
Ri based ratio
no.
1
native peptide
m/z
RT
13
labeled peptide
m/z
RT
52
native: C8-QAT
AD*RD*QYE*LLCLD
NTRKPVDEYK
CD*E*WSVNSVGK
CLKD*GAGD*VAFVK
CQSFRDHM*K
CQSFRD*HMK
DCHLAQVPSHTVVAR
D*CHLAQVPSHTVVAR*
(917.44)3+
A*DRD*QYE*LLCLDNT
RK*PVDEYK*
*CD*E*WSVNSVGK*
*CLK*DGAGDVAFVK*
*CQSFRD*HM*K*
*CQSFRD*HMK*
*DCHLAQVPSHTVVAR
*D*CHLAQVPSHTVVAR°
(852.48)4+
1:500
2
3
4
5
6
7
(633.87)2+
(683.82)2+
(584.27)2+
(1173.96)2+
(817.01)2+
(839.04)2+
(559.67)3+
(710.01)2+
(1035.52)2+
40
18
22
18
30
40
(860.47)2+
(1001.56)2+
(821.34)2+
(813.46)2+
(929.97)2+
(951.61)2+
(634.75)3+
(936.06)2+
(1261.84)2+
(841.56)3+
(820.04)3+
(1229.56)2+
(863.43)2+
(851.38)2+
(862.54)2+
(1057.02)2+
45
54
44
38
49
54
1:72
1:2
1:8
1:43
1:453
1:19
8
9
D*SGFQM*NQLRGK
EDLIWELLNQAQEHFGK
27
49
*D*SGFQM*NQLRGK*
*EDLIWELLNQAQEH
FGK*
55
70
1:21
1:1.6
10
E*D*PQTFYYAVAVVKK
(901.43)2+
20
*ED*PQTFYYAVAVVK*K*
27
1:29
11
12
13
14
HSTIFENLANK
(637.32)2+
(625.31)2+
(636.43)2+
(539.68)3+
1008
20
27
43
37
*HSTIFENLANK*
51
57
50
57
1:14
1:79
1:37
1:2.2
SASDLTWDNLK
SASDLTWD*NLK
TVRWCAVSEHEATK
*SASDLTWDNLK*
*SASD*LTWDNLK*
*TVRWCAVSE*HE*ATK*
15
16
VPPRMD*AK
RLAVGALLVCAVLGL
LAVPD*KTVR
(468.25)2+
(858.41)3+
29
47
*VPPRMD*AK*
*RLAVGALLVCAVLGL
CLAVPD*K*TVR*
*AIAANEADAVTLDAG
LVYDAYLAPNNLK*PV
VAEFYGSK*
*K*CSTSSLLE*ACTF
RRP°
*WCAVSEHE*ATK*C
QSFR
(694.41)2+
47
69
1:8
1:8
(1008.92)3+
17
AIAANEADAVTLDAG
VYDAYLAPNNLKPVVA
EFYGSK
(989.28)4+
46
(1158.95)4+
70
1:3
18
19
KCSTSSLLE*ACTFRRP
(921.94)2+
(952.50)2+
24
38
(765.76)3+
(1148.13)2+
(786.07)3+
(1178.58)2+
61
48
1:20
1:14
WCAVSEHE*ATKCQSFR
a The average increases in ionization efficiency of these peptides was calculated to be 70-fold. The area under the peak for all charge states of
each peptide were summed and used as a total ionization measure. The area under the peak for all charge states of each peptide were summed
and used as a total ionization measure. The ratios were calculated by dividing the total peak areas for each peptide by the total peak area of the
native peptide. Key: *AA, C8QAT-labeled N-terminal; E*, sodiated glutamic acid; D*, sodiated aspartic acid M*, oxidized methionine to methionine
sulfoxide; K*, C8QAT labeled; AA°, sodiated C-termial.
This peptide has two primary amine groups and was derivatized
twice. A similar behavior is seen in the peptide H-Tyr-Gly-Gly-
Phe-Met-Lys-OH, which is labeled twice. Ionization efficiency of
the C8-QAT derivative is 33 times that of the native peptide and 5
times greater than the QAT derivative. The most dramatic
differences in ionization efficiency were seen with the peptides
H-Ala-Gly-Gly-OH and H-Ala-Met-OH, which were singly labeled.
Ionization of native H-Ala-Gly-Gly-OH was not detectable relative
to the C8-QAT derivative while that of native H-Ala-Met-OH was
125 times lower than that of the C8-QAT tagged peptide. Ionization
efficiency of the C8-QAT derivative was also much larger than that
of the QAT derivative in the case of these peptides. The small
peptide H-Ala-Ser-OH also experienced a 20-fold increase in
ionization efficiency after C8-QAT derivatization. Although the QAT
tag increased the ionization efficiency of these peptides, deriva-
tization with the C8-QAT group clearly had a much more dramatic
impact on ionization efficiency of these peptides.
In contrast, derivatization of the peptides H-Asp-Arg-Val-Tyr-
Ile-His-Pro-Phe-His-Leu-OH, H-Cys-Asp-Pro-Gly-Tyr-Ile-Gly-Ser-Arg-
OH, H-Ser-Tyr-Ser-Met-Glu-His-Phe-Arg-Trp-Gly-OH, and H-Ala-
Phe-Pro-Leu-Glu-Phe-OH with the QAT or C8-QAT labels either
decreased ionization efficiency slightly or had little effect. The
single derivatizable primary amine in these peptides was at the
amino terminus. Derivatization had an equally small impact on
the ionization efficiency of H-Pro-His-Pro-Phe-His-Phe-His-Phe-Phe-
Val-Tyr-Lys-OH, which was singly labeled at the C-terminal lysine.
When the same 11 model peptides along with the QAT and
C8-QAT derivatives were examined at 4 times higher concentra-
tion, a different picture of ionization efficiency emerged (Table
2). Increases in ionization efficiency of the peptides Ac-Gln-Lys-
Arg-Pro-Ser-Gln-Arg-Ser-Lys-Tyr-Leu-OH and H-Tyr-Gly-Gly-Phe-
Met-Lys-OH after C8-QAT derivatization were 7 and 20, respec-
tively, at the higher concentration as opposed to 20 and 33 at the
lower concentration. There was also a 1-7-fold increase in the
ionization efficiency of other peptides greater than 500 Da in the
mixture. Only with peptides of lower than 500 Da was there an
increase in ionization efficiency greater than 300.
Effect of C8-QAT Labeling on Electrospray Ionization of
Tryptic Peptides. The objective in the experiments described
below was to determine the impact of C8-QAT labeling on
electrospray ionization of typical tryptic peptides. This was
achieved using native and C8-QAT-labeled transferrin digests with
an equimolar distribution of peptides at a concentration of 10-6
M. The labeled and unlabeled digests were mixed in a 1 to 1 ratio
and separated on the C18 Zorbax column as described above before
electrospray ionization and introduction into the mass spectrom-
eter. MASCOT was used for peptide identification after new
masses for the tagged amine groups at the N-terminus and on
lysine were included as a variable modification. Peptides identified
by this procedure are listed in Table 3. Ri values were calculated
and used to determined differences in ionization efficiency.
The average increase in ionization was estimated to be 70-
fold. A 500-fold increase in ionization efficiency was observed for
the peptide ADRDQYELLCLDNTRKPVDEYK. This miss-cleaved
peptide was labeled on three different amino acids, i.e., the
N-terminus and both lysine residues. But the enhancement of
Analytical Chemistry, Vol. 78, No. 12, June 15, 2006 4181

Figure 5. MS/MS spectrum for the peptide MYLGYEYVTAIR. All the y ions and seven b ions were found for this peptide and mapped on the
spectrum. The mass of methionine labeled with C8-QAT reagent was calculated using both y and b ions.
ionization seen with this peptide was probably not due to being
derivatized at multiple sites. Ionization of the peptide DCHLAQVP-
SHTVVAR with a single tag was enhanced 453-fold. The number
of sites tagged seemed not to correlate with increases in ionization
efficiency throughout this work. For example, the peptide CLKD-
GAGDVAFVK is triply tagged (at N-terminus and both lysine
residues) but showed only a 2-fold increase in ionization efficiency.
MS/MS Analysis of C8-QAT-Labeled Peptides from Trans-
ferrin. The effect of C8-QAT labeling on CID-induced fragmenta-
tion was examined using peptides from the transferrin digest. The
MS/MS spectrum and the peak map for MYLGYEYVTAIR is
shown in Figure 5, showing that all the y ions and seven b ions
were found. The mass of C8-QAT-labeled methionine at the
N-terminus of the peptide can be calculated using either the b(1)
ion or the difference between the precursor ion mass and the mass
of y(11)fragment ion f (852.45)2+ - 1347.70 y(11) - 1.0079. A
proton is subtracted from the mass of y(11) ion because this
fragment has absorbed a proton to be positively charged following
loss of the quaternary amine tag. All the peptides in Table 3 were
identified from their MS/MS spectra. A total of 25 peptides were
identified using manual sequencing (data not shown).
Discussion. Electrospray ionization of peptides is governed
by two major factors. One is the ability of the peptide to acquire
charge (generally a proton) before or during escape from the
surface of electrospray droplets. The second is their surface
activity, i.e., the ease with which the peptide migrates to the
4182 Analytical Chemistry, Vol. 78, No. 12, June 15, 2006

droplet surface. The rationale in the design of the C8-QAT labeling
agent was that introduction of a permanent positive charge into
peptides and addition of an adjacent aliphatic hydrocarbon tail
would augment both of these processes. Data presented here and
in previous work7 indicate that in most cases quaternization does
in fact improve ionization efficiency. It is also clear that adding
an octyl side chain to the quaternary amine can have a very
substantial impact on ionization efficiency.
Ideally it would be possible to look at the structures of a family
of peptides and predict their relative ionization efficiency before
and after derivatization with a reagent such as the C8-QAT. That
is not possible from the data presented here. Generally the
ionization efficiency of peptides that ionize well in native form was
not improved substantially by derivatization. In contrast, ionization
of many peptides that ionized poorly was improved 1 order of
magnitude or more through derivatization with the C8-QAT. One
of the more predictable observations was that the ionization
efficiency of peptides under 500 Da would improve 1 order of
magnitude or more with derivatization. Actually, a 300-fold increase
in ionization efficiency was seen in one case. Another observation
of major importance was that a relatively minor change in peptide
concentration can change the relative ionization efficiency of
peptides in a mixture. Tagging peptides with C8-QAT moderated
this effect.
The lack of correlation between ionization efficiency and (1)
peptide size above 500 Da, (2) the presence of cationic amino acids
such as histidine and arginine in peptides, (3) number of C8-QAT
tags added to a peptide, and (4) peptide hydrophobicity was
surprising. A dramatic 500-fold increase in ionization efficiency
following C8-QAT tagging was seen in a peptide of almost 1000
Da. In another case a similar peptide showed a 2-fold increase.
The presence of cationic amino acids such as arginine and
histidine is generally thought to be very desirable in peptide
ionization because of the ability to acquire protons during the
electrospray process. Surprisingly, the number of cationic amino
acids in a peptide was not a good indicator of ionization efficiency.
This is a clear indicator that the electrospray ionization process
depends on more than acquiring a protein to become ionized.
Based on the rationale that peptide hydrophobicity could
impact migration to the droplet surface, it was surprising that
neither the number of hydrophobic C8-QAT groups added to
peptides nor the net hydrophobicity of native peptides was useful
in predicting relative changes in ionization efficiency follow
tagging. One contributor to this anomaly could be that peptides
entering the electrospray inlet have just left a RPC column. RPC
effluent can contain up to 60% acetonitrile. This will seriously
diminish solvophobicity toward hydrophobic peptides in electro-
spray droplets. A high content of organic solvent in the droplet
would suppress the surface activity of the hydrophobic tags as
well by decreasing surface tension in the droplet. Peptide
conformation could be another that would contribute to the surface
activity of peptides, particularly with larger peptides It is conceiv-
able that hydrophobic regions of a peptide could be shielded from
a polar solution by polar amino acid side chains. Derivatization
could either enhance or diminish this shielding.
CONCLUSION
Based on data presented in this paper, it is concluded that
addition of a tag containing a quaternary amine with an adjacent
n-octyl chain to primary groups in peptides can increase the
electrospray ionization efficiency of many peptides by 10-fold or
more. This enhancement of ionization efficiency occurs most
consistently with peptides under 500 Da but can have an equally
large impact on larger peptides as well. The most dramatic
increases in ionization efficiency following derivatization are likely
to occur with peptides that are difficult to ionize in the native state.
Smaller degrees of augmentation are likely in peptides that ionized
well in the untagged form. It is further concluded that improve-
ments in ionization efficiency following C8-QAT tagging are most
likely due to increased accumulation of peptides at droplet surfaces
by imparting surfactant properties to peptides.
Finally it is concluded that derivatization of peptides with
reagents such as C8-QAT can improve detection sensitivity of many
peptides while minimizing differences in ionization efficiency. This
may be particularly useful in the analysis of complex peptide
mixtures such as those derived from tryptic digests of a proteome.
ACKNOWLEDGMENT
This work was supported by grant AG13319 from the National
Institute of Health.
Received for review February 2, 2006. Accepted March 31,
2006.
AC0602266
Analytical Chemistry, Vol. 78, No. 12, June 15, 2006 4183