Enhancement of amino acid detection and quantification by electrospray ionization mass spectrometry

Article abstract of DOI:10.1021/ac0600510

A new strategy for amino acid analysis is reported involving derivatization with an N-hydroxysuccinimide ester of N-alkylnicotinic acid (C _n-NA-NHS) followed by reversed-phase chromatography and electrospray ionization mass spectrometry (RPC-MS). Detection sensitivity increased as the N-alkyl chain length of the nicotinic acid derivatizing agent was increased from 1 to 4. N-Acylation of amino acids with the C_n-NA-NHS reagents in water produced a stable product in roughly 1 min using a 4-fold molar excess of derivatizing agent in 0.1 M sodium borate buffer at pH values ranging from 8.5 to 10. Some O-acylation of tyrosine was also observed, but the product hydrolyzed within a few minutes at pH 10. The cystine product also degraded slowly over the course of a few days from reduction of the disulfide bond to form cysteine. The retention time of C_n-NA derivatized amino acids was lengthened in reversed-phase chromatography to the extent that polar amino acids were retained beyond the solvent peak, particularly in the cases of the C₃-NA and C₄-NA derivatives. Complete resolution of 18 amino acids was achieved in 28 min using the C₄-NA-NHS reagent. Compared to N-acylation with benzoic acid, derivatization with C ₄-NA-NHS increased MS detection sensitivity 6-80-fold. This was attributed to the surfactant properties of the C_n-NA-NHS reagents. The quaternary amine increased the charge on amino acid conjugates while the presence of an adjacent alkyl chain further increased ionization efficiency by apparently enhancing amino acid migration to the surface of electrospray droplets. Further modification of the C_n-NA-NHS reagents with deuterium was used to prepare coded sets of derivatizing agents. These coding agents were used to differentially code samples and after mixing carry out comparative concentration measurements between samples using extracted ion chromatograms to estimate relative peak areas of derivatized amino acids.

Full text of DOI:10.1021/ac0600510

Anal. Chem. 2006, 78, 4702-4708
Enhancement of Amino Acid Detection and
Quantification by Electrospray Ionization Mass
Spectrometry
Wen-Chu Yang, Hamid Mirzaei, Xiuping Liu, and Fred E. Regnier*
Department of Chemistry, Purdue University, West Lafayette, Indiana 47907
method with postcolumn ninhydrin-based detection1 that led to
HPLC, and finally reversed-phase chromatographic (RPC) separa-
tion methods with precolumn derivatization. Derivatization of
amino acids with reagents such as o-phthaladehyde-mercapto-
ethanol,2 dansyl chloride,3 phenyl isothiocyanate,4 and its ana-
logues5,6 fluorenylmethyl chlorofomate,7 and 6-aminoquiline car-
bamate8 vastly increased detection sensitivity.
Now mass spectral methods are playing a role in detection.
Among the many advantages of mass spectrometry (MS) are that
analytes do not have to be chromatographically resolved to allow
detection. Another is that multiple dimensions of structure analysis
are available through MS/MS analysis without adding time to the
analysis. Moreover, extracted ion chromatograms allow multiple
mass-related features of a mixture to be recognized, quantified,
and displayed. This enables the detection of both parent ions and
fragment ions common to multiple species.
But there are problems with the reversed-phase chromatog-
raphy-mass spectrometry (RPC-MS) approach as well. One is
that matrix effects can suppress the ionization of amino acids. This
is particularly true of amino acids that do not ionize well in the
first place. Yet another is that polar amino acids elute from RPC
columns unretained in the column void volume with a large
number of other substances. This problem has been addressed
by derivatization with a more hydrophobic group such as benzoic
acid9 or 2,4-dinitrofluorobennzene (DNFB)10 to increase RPC
retention of amino acids. Unfortunately, derivatization can also
impact ionization efficiency. Although all 19 DNFB derivatized
amino acids examined were well resolved by RPC, only H, R, and
K were observed in the positive ion mode of electrospray ionization
(ESI)-MS.11 This is apparently due to a loss in charge caused by
derivatization.
A new strategy for amino acid analysis is reported involv-
ing derivatization with an N-hydroxysuccinimide ester of
N-alkylnicotinic acid (Cn-NA-NHS) followed by reversed-
phase chromatography and electrospray ionization mass
spectrometry (RPC-MS). Detection sensitivity increased
as the N-alkyl chain length of the nicotinic acid deriva-
tizing agent was increased from 1 to 4. N-Acylation of
amino acids with the Cn-NA-NHS reagents in water
produced a stable product in roughly 1 min using a 4-fold
molar excess of derivatizing agent in 0.1 M sodium borate
buffer at pH values ranging from 8.5 to 10. Some
O-acylation of tyrosine was also observed, but the product
hydrolyzed within a few minutes at pH 10. The cystine
product also degraded slowly over the course of a few days
from reduction of the disulfide bond to form cysteine. The
retention time of Cn-NA derivatized amino acids was
lengthened in reversed-phase chromatography to the
extent that polar amino acids were retained beyond the
solvent peak, particularly in the cases of the C3-NA and
C4-NA derivatives. Complete resolution of 18 amino acids
was achieved in 28 min using the C4-NA-NHS reagent.
Compared to N-acylation with benzoic acid, derivatization
with C4-NA-NHS increased MS detection sensitivity 6-80-
fold. This was attributed to the surfactant properties of
the Cn-NA-NHS reagents. The quaternary amine increased
the charge on amino acid conjugates while the presence
of an adjacent alkyl chain further increased ionization
efficiency by apparently enhancing amino acid migration
to the surface of electrospray droplets. Further modifica-
tion of the Cn-NA-NHS reagents with deuterium was used
to prepare coded sets of derivatizing agents. These coding
agents were used to differentially code samples and after
mixing carry out comparative concentration measure-
ments between samples using extracted ion chromato-
grams to estimate relative peak areas of derivatized amino
acids.
(1) Moore, S.; Stein, W. H. Methods Enzymol. 1958, 6, 819-831.
(2) Hill, D. W.; Walters, F. W.; Wilson, T. D.; Stuart, J. D. Anal. Chem. 1979,
51, 1338-1341.
(3) Tapuhi, Y.; Schmidt, D. E.; Lindner, W.; Karger, B. L. Anal. Biochem. 1981,
115, 123-129.
(4) Cohen, S. A.; Bidlingmeger, B. A.; Tarvin, T. L. Nature 1986, 320, 769-
70.
(5) Woo, K. L.; Lee, S. H. J. Chromatogr., A 1994, 667, 105-111.
(6) Woo, K. L.; Ahan, Y.-K. J. Chromatogr., A 1996, 740, 41-50.
(7) Einarsson, S.; Josefsson, B.; Lagerkvist, S. J. Chromatogr., A 1983, 282,
609-618.
(8) Cohen, S. A.; Michaud, D. P. Anal. Biochem. 1993, 211, 279-287.
(9) Julka, S.; Regnier, F. E. Anal. Chem. 2004, 76, 5799-5806.
(10) Sanger, F. Biochem. J. 1945, 39, 507-15.
The evolution of methods for amino acid analysis has closely
paralleled the development of analytical tools for the life sciences
during the past 50 years. First there were the old paper chroma-
tography methods, then the Moore and Stein ion exchange
* To whom correspondence should be addressed. Phone: (765)494-9390.
Fax: (765)494-0239. E-mail: fregnier@purdue.edu.
(11) Liu, Z. F.; Minkler, P. E.; Lin, D.; Sayre, L. M. Rapid Commun. Mass
Spectrom. 2004, 18, 1059-1065.
4702 Analytical Chemistry, Vol. 78, No. 13, July 1, 2006
10.1021/ac0600510 CCC: $33.50 © 2006 American Chemical Society
Published on Web 05/17/2006

It is becoming clear that a strategy for dealing with poor
ionization of amino acids is to derivatize them with a group that
is easily protonated, such as a tertiary amine.12 Dansyl derivatizes
for example are more easily protonated during ESI because of
the dimethylamino moiety in the naphthyl ring.13 Derivatizing
amino acids with dimethylformamide (DMF)-dimethylacetal also
enhances detection in tandem mass spectrometry, especially for
smaller amino acids such as glycine.14 Derivatization with N,N-
dimethyl-2,4-dinitro-5-fluorobenzylamine is another example in
which a robust mass signal is obtained by introduction of a tertiary
amine.11 Quaternization could increase ionization efficiency even
more as has been shown with steroids and peptides.15-17
The objective of the work reported here was to explore the
use of derivatizing agents to enhance RPC-MS analysis of amino
acids by adding desirable features to analytes. One was to add
sufficient hydrophobicity to cause the retention of amino acids
well beyond the RPC solvent peak. A second was to add a
quaternary amine group that would increase ionization efficiency.
A third was to introduce stable isotope codes into amino acids to
enhance relative quantification in comparative metabolomics.
Derivatizing agents evaluated in these studies were obtained by
synthesizing a family of N-alkylnicotinic acid derivatizing agents
activated with N-hydroxysuccinimide.
Scheme 1. Synthesis of N-Alkylnicotinic Acid
N-Hydroxysuccinimide Ester (Cn-NA-NHS)
to 5 mL of dry DMF containing 40 mmol (5.0 g) of nicotinic acid
and 50 mmol of N-hydroxysuccinimide (4.6 g). The reaction
mixture was stirred at room temperature for 48 h and then treated
with glacial acetic acid (0.2 mL) to decompose excess DCC. After
allowing the mixture to stir at room temperature for an additional
hour, the temperature was reduced to 3-4 °C and stirring
continued for another 4 h. Precipitated dicyclohexylurea was
removed by filtration and the filtrate evaporated under vacuum
to reduce the volume to ∼10 mL. A pale yellow product was
precipitated by addition of a mixture of hexane and 2-propanol
(6:1 in volume). The nicotinic acid N-hydroxysuccinimide ester
precipitate was dried under vacuum and used to prepare a series
of N-alkylnicotinic acid N-hydroxysuccinimide esters (alkyl ) C1,
C2, C3, C4) by direct quaternization. A 5-mmol aliquot of nicotinic
acid N-hydroxysuccinimide ester was added to 10 mmol of an alkyl
EXPERIMENTAL SECTION
Materials and Reagents. An amino acid standard mixture
was obtained from Sigma-Aldrich (St. Louis, MO). The following
amino acids including ammonium chloride were in the mixture:
L
-alanine,
L
-arginine,
-histidine,
-phenylalanine, -proline,
-valine. Amino acids in this standard mixture were at a concentra-
-cystine at
L
-aspartic acid,
-isoleucine, -leucine,
-serine, -threonine,
L
-cystine,
L
-glutamic acid,
-methionine,
-tyrosine, and
glycine,
L
L
L
L
L
L
-lysine,
L
L
L
L
L
o
iodide in 5 mL of ACN. The mixture was refluxed at 70-75 C
tion of 2.5 µmol/mL in 0.1 N HCl. The exception was
L
under argon for 24-48 h. A yellow product precipitated from the
solution either during the reaction or after removing some of the
solvent, depending on the alkyl groups. After cooling, the Cn-NA-
NHS product was collected by filtration, washed with cold acetone,
and dried to yield yellow crystals that were stored in a desiccator
until used for amino acid derivatization.
1.25 µmol/mL. Anhydrous acetonitrile (ACN), dicyclohexylcar-
bodiimide (DCC), alkyl iodide, N-hydroxysuccinamide, and DMF
were also purchased from Sigma-Aldrich. HPLC grade acetonitrile
and acetone were obtained from Mallinckrodt Baker (Phillipsburg,
NJ). Trifluoroacetic acid (TFA) (Sequanal grade) was obtained
from Pierce (Rockford, IL). 1-Iodobutane-d9 was the product of
Cambridge Isotope Laboratories (Andover, MA). Double-deionized
water was produced by a Milli-Q gradient A10 system (Millipore,
Bedford, MA).
Synthesis of N-Hydroxysuccinimide-Activated N-Alkylni-
cotinic Acid Esters (Cn-NA-NHS). The scheme for synthesizing
N-alkylnicotinic acid N-hydroxysuccinimide ester (Cn-NA-NHS)
derivatizing agents is shown in Scheme 1. The procedure is an
adaptation of a method first described by Srivastava.18 An aliquote
of diclyclohexyl carbodiimide (DCC) (44 mmol, 9.0 g) was added
Derivatization of Amino Acids. A Cn-NA-NHS solution was
prepared at a concentration of 40.0 mg/mL in dry ACN. This
solution is stable for 5 days at ambient temperature or at least
one month at 4 °C. Two types of standard solutions were used in
optimization studies. One type contained 50.0 µmol/mL of an
individual amino acid in 0.1 mol/L hydrochloride. Tryptophan was
an exception in that it was dissolved in deionized water instead
of HCl. The other type of standard solution was made by adding
10.0 µL of all the individual amino acid solutions in HCl and 0.5
µL of 50.0 µmol/mL tryptophan in deionized water to 200 µL of
0.1 mol/L sodium borate buffer at pH 9.0. After mixing, 20 µL of
40.0 mg/mL Cn-NA-NHS (roughly a 4-fold molar excess) was
added. The reaction was allowed to proceed for 1 min before RPC
analysis. Samples of unknown amino acid concentration were
analyzed by adding 10.0 µL of sample and 20 µL of 40.0 mg/mL
Cn-NA-NHS solution to 200 µL of 0.1 mol/L sodium borate buffer
at pH 9.0. After a minimum of 1-min incubation, derivatized amino
acids were analyzed by RPC-MS. Incubation times of 10 min were
used when it was the objective to allow O-acylated tyrosine to
hydrolyze to a single N-acylated product.
(12) Quirke, J. M. E.; Adams, C. L.; Van Berkel, G. J. Anal. Chem. 1994, 66,
1302-1315.
(13) Durden, D. A.; Davis, B. A.; Boulton, A. A. Biomed. Mass Spectrom. 1974,
1, 83-95.
(14) Johnson, D. W. Rapid Commun. Mass Spectrom. 2001, 15, 2198-2205.
(15) Higashi, T.; Yamauchi, A.; Shimada, K.; Koh, E.; Mizokmi, A.; Namiki, M.
Anal. Biochem. 2005, 382, 1035-1043.
(16) Stewart, N. A.; Pham, V. T.; Choma, C. T.; Kaplan, H. Rapid Commun. Mass
Spectrom. 2002, 16, 1448-1453.
(17) Ren, D.; Julka, S.; Inerowicz, H.; Regnier, F. E. Anal. Chem. 2004, 76,
4522-4530.
(18) Tedjamulia, M. L.; Srivastava, P. C.; Knapp, Jr., F. F. J. Med. Chem. 1985,
28, 1574-1580.
Analytical Chemistry, Vol. 78, No. 13, July 1, 2006 4703

Scheme 2. Reactions Involved in the Derivatization of Amino Acids with Cn-NA-NHS and Subsequent
Hydrolysis of the Unused Derivatizing Agenta
a Amino acid derivatization occurs rapidly. Hydrolysis of Cn-NA-NHS in contrast is a much slower reaction.
RPC-ESI-MS Analysis. Optimization of the derivatization
reaction was achieved using RPC on an Integral Micro-Analytical
workstation with a 205-nm-wavelength UV detector (Applied
Biosystems, Framingham, MA). Amino acid derivatives were also
purified with this system for direct MS analysis in some cases.
RPC with ESI mass spectral analysis of amino acid derivatives
was carried out on a Waters Alliance system, equipped with
MassLynx 4.0 software, a Waters 996 photodiode array detector
(PAD), and a Micromass Q-Tof micromass spectrometer. The
PAD was used in the scanning mode to detect analytes between
200 and 400 nm. Positive ion mass spectra were acquired in the
continuous mode. Ionization was achieved with an electrospray
voltage of 3 kV, cone voltage of +30 V, and cone and desolvation
gas flows of 5-8 and 500 L/h, respectively. The source block and
desolvation temperatures were maintained at 150 and 350 °C,
respectively. All separations were performed on a low TFA Vydac
analytical column (C18 Mass Spec, 4.6 mm × 250 mm) using 15
µL of sample and gradient elution with two sets of mobile phases.
Optimization of derivatization was achieved by eluting the RPC
column initially with 5% ACN and 0.1% TFA in water for 5 min
and then proceeding on in a 30-min linear gradient to 5% water
and 0.1% TFA in ACN. General analyses were achieved with a
second elution protocol using two different eluting solvents.
Solvent A was composed of 0.05% TFA in water, and solvent B
contained 5% H2O and 0.05% TFA in ACN. After 4 min of isocratic
elution with solvent A, an 11-min linear gradient to 10% solvent B
was initiated, followed by a 5-min linear gradient to 20% mobile
phase B, and finally a 10-min linear gradient to 50% solvent B.
The flow rate was 1 mL/min in all cases.
declustering potential 2, 20.0. Collision energy was optimized to
obtain maximum fragmentation for tandem MS.
RESULTS AND DISCUSSION
Amino Acid Derivatization. Amino acid (AA) derivatization
was achieved by amide bond formation through nucleophilic
displacement of NHS from an N-alkylnicotinic acid (NA) by the
R-amino group of the AA (reaction Scheme 2). In the case of
lysine, the ꢀ-amino group is sufficiently nucleophilic at slightly
basic pH that it too is derivatized. Beyond being easily displaced,
a major advantage of NHS activation is that derivatization can be
achieved in water. This is a great benefit because amino acids
are water soluble. A similar protocol for R-amino group derivati-
zation is used with peptides where it has been shown that NHS
activated nicotinic acid is sufficiently stable in water to quantita-
tively acylated primary amine groups before finally hydrolyzing.19
The reagent used in these studies differs from the one used in
peptide derivatization only in the addition of alkyl groups (Cn) to
nicotinic acid (NA) and quaternization of the ring nitrogen.
N-Alkylnicotinic acid reagents activated with NHS will be repre-
sented by the general formula Cn-NA-NHS throughout the rest of
the paper. Hydrolysis of residual Cn-NA-NHS after derivatization
can be accelerated by adjusting to pH 10 with hydroxlamine.20
Hydrolysis is achieved in less than 1 min under these conditions
to yield Cn-NA according to MS analysis. Cn-NA-NHS derivatizing
agents were stored in acetonitrile to preclude premature hydroly-
sis.
Optimization of derivatization was examined using the C3-NA-
NHS reagent and tryptophan (Trp) with RPC-MS analysis of the
reaction products. Peaks eluting at 7.5, 13.1, and 16.8 min were
found to be of m/z 166.1(+1), 204.2(+1), and 352.1(+1), respec-
tively. The number in parentheses behind the m/z value is the
charge state of the ion observed in the MS spectrum. These three
peaks were identified as C3-NA, free Trp, and C3-NA-Trp, according
to their retention times and m/z values. The derivatization reaction
was investigated by carrying the reaction out in 200 µL of 0.1
ESI-MS/MS Analysis. ESI-MS and ESI-MS/MS analysis was
performed on an API QSTAR Pulsar LC/MS/MS System (Applied
Biosystem, Framingham, MA) equipped with an ESI ion source.
Fractions from the Intergral LC were collected, dried, reconsti-
tuted with CH3OH/H2O/acetic acid (50%/49%/1%), and injected
into the ESI source by infusion at 5-10 µL/min. Typical settings
for analysis in the positive mode of ESI were as follows: ion spray
voltage, 5000; curtain gas, 20; ion source gas 1, 20; ion source
gas 2, 0; declustering potential, 45; focusing potential, 220;
(19) Munchbach, M.; Quadroni, M.; Miotto, G.; James, P. Anal. Chem. 2000,
72, 4047-4057.
(20) Wang, S.; Regnier, F. E. J Chromatogr., A 2001, 924, 345-357.
4704 Analytical Chemistry, Vol. 78, No. 13, July 1, 2006

mol/L sodium borate buffer at pH 8.0 to which 10 µL of a 50 µmol/
mL tryptophan solution was added along with 20 µL of derivati-
zation reagent. C3-NA-NHS and Trp were mixed in 1:1 and 3:1
molar ratios, and the reaction was allowed to proceed for 5 min.
Although some Trp was seen at the 1:1 C3-NA-NHS/Trp ratio, at
the 3:1 molar ratio, Trp derivatization was complete in 30 s. A
1-min reaction time and 4:1 molar ratio were used in subsequent
experiments to ensure that derivatization would be complete.
Further optimization of the reaction was carried out with a
standard 17-amino acid mixture from Sigma (see Experimental
Section) using the C3-NA-NHS derivatizing agent. Yields were
studied using 0.1 M sodium borate buffers ranging in pH from
7.5 to 10. Although the amount of product formed increased
slightly between pH 7.5 and 8.5, there was little additional increase
in product above pH 8.5. Similar findings have been reported with
other acylating agents. 9,21 In addition to N-acylation, Tyr showed
O-acylation in the pH range from 7.5 to 8.5. Three peaks containing
Tyr were seen in the RPC tracing. One of the Tyr products was
the expected R-amino group acylation product. Another was from
acylation of the phenolic OH. O-Acylation of tyrosine has also been
reported in the case of peptides where O-acylated products were
hydrolyzed after derivatization by the addition of hydroxylamine.20
The third arose from derivatizing both the phenolic hydroxyl and
the R-amino group. At pH 7.5, the phenolic hydroxyl derivative
was dominant with roughly 95% of the derivatization occurring at
this site. At pH 8.0, greater than 90% of the Try product resulted
from double derivatization. When the reaction was carried out at
pH 9.0, the R-amino derivatized product dominated the mixture.
These data were interpreted to mean that derivatization of the
phenolic hydroxyl group occurred faster than at the R-amino group
but was also being hydrolyzed. Given enough time at the higher
pH, the exclusive product at pH 9.0 should be from derivatization
on the R-amino group. This was in fact the case. After 10-min
incubation at pH 9.0, the exclusive product was from R-amino
group derivatization. Neither buffer concentration nor reagent-
to-amino acid ratio had much influence on product distribution
and yield under these conditions. As a consequence of the
additional complication added by Tyr, derivatization reactions were
carried out in 0.1 M sodium borate at pH 9.0 for 10 min at a
reactant-to-analyte ratio of no less than 4:1. Although higher pH
also favors hydrolysis of NHS esters, the rate of derivatization
exceeded the rate of hydrolysis in the early phase of the reaction.
Reproducibility of derivatization was studied by analyzing the
standard mixture 10 times according to the procedure described
in the Experimental Section. Concentration was determined using
peak areas derived from absorbance detection. The results of these
analyses are presented in Table 1 as relative standard derivations.
Reproducibility was generally better than 5%.
Table 1. m/z Value, Retention Time, and
Reproducibility of Derivatization (RSD of peak area) for
C4-NA Derivatives of Amino Acids
retention
peak area RSD
amino acid
m/z
time (min)
(%, n ) 10)a
His
317.02 (+1)
267.01 (+1)
237.02 (+1)
295.00 (+1)
336.05 (+1)
309.01 (+1)
277.04 (+1)
281.03 (+1)
251.04 (+1)
343.01 (+1)
343.03 (+1)
469.13 (+1)
235.07 (+2)
279.05 (+1)
562.01 (+1)
281.51 (+2)
311.01 (+1)
504.09 (+1)
252.55 (+2)
293.05 (+1)
293.05 (+1)
327.02 (+1)
366.02 (+1)
11.89
13.25
13.83
14.18
14.64
16.34
16.67
17.16
17.55
19.21
23.03
23.31
ndb
1.3
1.8
4.2
2.3
3.3
4.5
3.8
1.6
nd
Ser
Gly
Asp
Arg
Glu
Pro
Thr
Ala
Tyr-OH
Tyr-NH2
(di-) Lys
3.5
2.4
Val
(di-) Cyn
23.64
23.84
2.6
4.4
Met
Tyr-OH, NH2
24.00
25.15
4.6
nd
Ile
25.63
26.08
26.52
26.88
1.1
1.0
1.3
1.9
Leu
Phe
Trp
a From UV absorbance at 205 nm. b nd, not detected.
elutes between Asp and Arg in RPC. Degradation products for
Met were not identified. Tyr O-acylation is also a problem.
Generally it would be recommended to remove O-acylation by
hydrolysis before either analysis or storage before analysis.
Hydrolysis of the O-acylated Tyr derivative accelerates with
increasing pH. It is for this reason that this product is easily
destroyed by addition of hydroxylamine to achieve a slightly basic
pH.21 Hydroxylamine treatment of samples before analysis is thus
a good idea.
Reversed-Phase Chromatography. Hydrophilic amino acids
do not interact with the hydrophobic stationary phase of an RPC
column, typically eluting in the column void volume. One of the
objectives in this work was to circumvent this problem by
increasing the hydrophobicity and concomitantly the RPC reten-
tion time of amino acids through derivatization. It was for this
reason that alkyl groups were added to nicotinic acid. The
experiments below were based on the rationale that by varying
the aliphatic chain length in Cn-NA derivatizing agents it should
be possible to manipulate the retention times of amino acids widely
in the reversed-phase chromatogram. It will be shown later in the
paper that extending the aliphatic chain length of Cn-NA also
increased detection sensitivity.
It is seen in Figure 1 that the RPC retention time of even the
most hydrophilic amino acids could be increased to greater than
10 min by derivatization. With the C1-NA-NHS reagent, only Ile,
Leu, Phe, and Trp were detected. This reagent did not increase
hydrophobility of the hydrophilic amino acids sufficiently to retain
them beyond the trailing solvent peak. As the alkyl chain length
was increased to 3, all 18 amino acids were adequately retained
to be detected. C4-NA-NHS was used in subsequent reactions
because detection sensitivity was higher with C4-NA derivatization.
The RPC separation of 18 C4-NA derivatized amino acids within
27 min is seen in Figure 2. The derviatization time in this case
Derivative stability was examined by measuring UV peak area
in reversed-phase chromatograms at 205 nm over one week.
Amino acids acylated on primary amine groups with C3-NA were
found to be stable for one week at 4 °C or less. But there were
additional complications with Met and Cyn. Peak areas for these
amino acids could decrease up to 20% in one week of storage at
ambient temperature. Cyn degradation was the result of disulfide
bond cleavage with the formation of Cys (m/z 308.1(+1)). Cys
(21) Rajagopalan, R.; Kuntz, R. R.; Sharma, U.; Volkert, W. A.; Pandurangi R. S.
J. Org. Chem. 2002, 67, 6748-6757.
Analytical Chemistry, Vol. 78, No. 13, July 1, 2006 4705

Figure 1. Effects of alkyl chain length on retention time of
derivatized amino acids. Buffer A, 100% H2O + 0.1% TFA; buffer B,
95% ACN + 0.1% TFA. Elution protocol: 0-5 min, 100% A; 5-30
min, 100% A-100% B.
Figure 3. Extracted total ion chromatogram from 18 amino acids
derivatized with C4-NA-NHS using 10-min incubation time. Detection
was achieved by monitoring the m/z 190-600 range. Separation
conditions and peak identification are the same as in Figure 2.
In addition to N-acylation, O-acylation of tyrosine (peak 10)
and bistyrosine (peak 16) was observed with a short derivatization
time of 1 min (Figure 2). As the derivatization time was increased
from 1 to 10 min, it is clearly seen in Figure 3 that these peaks
are no longer present. Although phenolic hydroxyl groups
derivatize rapidly, they are not stable in water and hydrolyze
during a 10-min incubation.
One of the advantages of Cn-NA derivatization is that it also
facilitates detection by UV absorbance through the addition of
the aromatic ring in nicotinic acid. It is seen from Figure 2 that
the UV response of nonaromatic amino acids was very similar
with the exception of Lys and cystine (Cyn). These amino acids
were doubly labeled by Cn-NA-NHS and thus have one more
nicotinic acid residue than the other nonaromatic amino acids.
Nicotinic acid residues show weak absorbance at 265 nm and
absorb strongly below 210 nm.
Enhancement of Ionization. The requisite acylation of amino
groups in amino acids to achieve derivatization reduces their
charge. Concomitantly, this is expected to decrease ionization
efficiency in positive ionization mode ESI-MS. The impact of
charge on ionization efficiency of derivatized amino acids was
examined using nicotinoate and benzoate derivatized Glu, Try,
and Lys. 13C6 Isotopically coded benzoyl derivatives (BA-13C6) of
Glu, Trp, and Lys were prepared using methods described in the
literature.9 Use of BA-13C6 was necessary because benzoate and
nicotinoate vary by only 1 atomic mass unit. Equal amounts of
these three amino acids were derivatized individually with 13C6-
BA-NHS and NA-NHS, respectively, and examined individually by
RPC. Peaks of analyte were collected, dried, and redissolved in
50% methanol + 49% H2O + 1% acetic acid (v/v/v). Equal aliquots
of the six derivatives were mixed with the exception of BA-13C6-
Glu, which was added at five times the volume of the others. An
ESI-MS analysis of the mixture was performed as described in
the Experimental Section. Relative differences in detection sen-
sitivity are summarized in Table 2. Derivatization with nicotinic
acid was seen to enhance detection sensitivity with Glu and Try,
but not Lys. This is attributed to the ionization of nitrogen in the
pyridine ring.
Figure 2. Reversed-phase chromatogram from 18 amino acids
derivatized with C4-NA-NHS using 1-min incubation time. The upper
trace was obtained with photodiode array detection from 200 to 400
nm. The lower trace was obtained by extraction ion monitoring from
m/z 190 to 600. Elution of the RPC columns was achieved with a
two-solvent gradient. Buffer A was 0.05% TFA in water while buffer
B contained 0.05% TFA in ACN. The gradient protocol was as
follows: 0-4 min, 100% A; 4-15 min, 0% B-10% B; 15-20 min,
10% B-20% B; and 20-30 min, 20% B-50% B. Peak identification:
1, His; 2, Ser; 3, Gly; 4, Asp; 5, Arg; 6, Glu; 7, Pro; 8, Thr; 9, Ala; 10,
Tyr-OH; 11, Tyr-NH2; 12, Bis-Lys; 13, Val; 14, Cyn; 15, Met; 16, Bis-
Tyr; 17, Ile; 18, Leu; 19, Phe; 20, Trp; U, unknown.
was 1 min. Peaks were assigned based on their unique m/z values
except for C4-NA-Ile and C4-NA-Leu, which have identical masses
and showed the same singly charged ion at m/z 293.05(+1). They
were distinguishable by both their chromatographic retention
times and tandem mass spectra. C4-NA-His was obscured in the
photodiode array trace by C4-NA arising from the hydrolysis of
C4-NA-NHS. However, it was easily recognized in the extracted
total ion chromatogram (TIC trace) when the m/z range was set
for 190-600. But there is a penalty from excluding ions under
m/z 190. Although C4-NA appears at m/z 180.0 and is eliminated
from the spectra, ammonia is also derivatized under these
conditions and gives a distinct chromatographic peak. Unfortu-
nately this product is excluded as well.
4706 Analytical Chemistry, Vol. 78, No. 13, July 1, 2006

Table 2. Comparison of Relative MS Signal Intensity
for Amino Acids Derivatized with Different Reagentsa
derivatizing reagent
amino
acid
C1-NA-
NHS
C2-NA-
NHS
C3-NA-
NHS
C4-NA-
NHS
NA-
BA-13C6-
NHSS
NHS
Glu
Lys
Trp
ndb
2.3
2.5
1.6
3.3
3.9
3.1
5.0
4.8
3.3
6.3
12.5
1.0
1.0
1.0
0.11
0.96
0.15
a Data were from two sets of experiments. See the text. Signal
intensities were normalized to NA-NHS ) 1.0. b nd, not detected.
Figure 5. Mass spectra of C4H9-NA-Phe and C4D9-NA-Phe at a
1:1 molar ratio.
surfactantlike properties. This in turn increases AA concentration
at droplet surfaces in ESI-MS and enhances ionization. The net
effect is that, in comparison to BA-13C6 derivatives, C4-NA deriva-
tives of Glu, Lys, and Trp showed 30, 6, and 83 times greater ESI-
MS signal intensity, respectively.
Comparative Quantification through Deuterium Labeling.
In vitro stable isotope coding is now widely used in proteomics
for quantitative comparisons of peptide and protein concentration
between samples. The most successful of these approaches is to
acylate R- and ꢀ-amino groups of polypeptides in a sample with a
reagent that allows global coding of all primary amine-containing
species according to sample origin. When these differentially
coded samples are then mixed and analyzed by ESI-MS, the
spectra contain doublet clusters of ions from the coded peptide
isotopomers. The difference in mass between the doublets is
simply the stable isotope mass differential between the isotopically
coded labeling agents. The difference in peak area of the clusters
is proportional to the relative concentration of the peptide in the
two samples. The relative concentration of thousands of peptides
can be determined in a single RPC-MS in this in vitro coding
method with a relative standard deviation of roughly 6%.22
A similar approach to comparative quantification was examined
in this work. A heavy form of C4-NA-NHS was synthesized by
substituting D9-n-butyl iodide for H9-n-butyl iodide in the NA-NHS
alkylation reaction, producing C4D9-NA-NHS that is 9 atomic mass
units (amu) higher than the light form of the coding agent, C4H9-
NA-NHS. Phenylalanine (Phe) was used for initial ESI-MS
comparative quantification studies. C4D9-NA-Phe and C4H9-NA-
Phe were mixed at concentration ratios varying over a 400-fold
range in concentration. Isotope ratio analyses were performed by
infusing samples into the ESI-MS through a nanospray inlet. The
characteristic double cluster of ions separated by 9 amu is seen
in the spectrum of a 1:1 concentration ratio sample (Figure 5).
The isotope ratio calibration curve was found to be linear over a
400-fold range of concentration with a linearity coefficient (r2) of
0.988 (data not shown). These data suggest that comparative
quantification of amino acids by stable isotope coding will be
possible using ESI-MS.
Figure 4. Mass spectra of NA- and C1∼4-NA derivatives of
tryptophan. The inset is the reversed-phase chromatogram of the NA-
and C1-4-NA derivatives of tryptophan with UV detection at 204 nm.
All derivatives were of the same concentration.
A more dramatic enhancement of sensitivity was observed for
the amino acids derivatized through permanently charged and
alkylated derivatizing reagents C1-4-NA-NHS. The C1-4-NA-Trp
derivative mixture along with NA-Trp was prepared and examined
by ESI-MS. Although UV absorbance of the C1-4-NA-Trp deriva-
tives was very similar, response in ESI-MS was substantially larger
with the longer alkyl chain length derivatives (Figure 4). Com-
pared to nicotinoate derivatized tryptophan (NA-Trp), C4-NA-Trp
gave 12.5 times greater MS signal intensity. Comparisons were
also made with Glu and Lys (Table 2). Although the degree of
enhancement was not as great with these amino acids, even with
Glu signal intensity was more than 3 times larger.
The results in Table 2 can be explained in the following way.
Clearly increasing charge in the pyridine ring through quater-
nization enhances sensitivity as seen by the difference in response
between NA-AA and C1-NA-AA. This phenomenon has also been
noted in the literature with peptides.16,17 But this does not explain
how longer alkyl chains further enhance sensitivity. It has been
noted in ESI-MS that during the ionization process species that
are both hydrophobic and cationic migrate to the droplet surface.
When present at high concentration, they even suppress the
ionization of less hydrophobic cationic species by forcing them
away from the droplet surface. In view of the fact that ionization
is thought to occur from droplet surfaces in ESI-MS, this process
has been offered as an explanation of ion suppression. Based on
this model of ESI-MS, it is likely that lengthening the alkyl chain
in C1-4-NA-AA causes the derivatized amino acid to have more
A more extensive study of comparative quantification was
undertaken by differentially coding samples containing the entire
(22) Julka, S. Regnier, F. E. J. Proteome Res. 2004, 3, 350-363.
Analytical Chemistry, Vol. 78, No. 13, July 1, 2006 4707

the concentration of amino acids between samples. The issue is
whether the chromatographic isotope effect has an impact on the
determination of isotope ratios. Put another way, if there were no
chromatographic isotope effect, which would provide the greatest
accuracy: (1) comparison of peak areas from an extracted ion
chromatogram or (2) a statistically based comparison of 10-50
direct MS-based isotope ratio measurements. The only way to test
this is to prepare a 13C coded labeling agent that shows no
chromatographic isotope effects. Preparation of such a reagent is
in progress.
CONCLUSIONS
It is concluded that derivatization of amino acids by N-acylation
with an acid that is also hydrophobic and has a quaternary amine
can facilitate RPC/MS analysis of amino acids in several ways.
One is by increasing the retention of small hydrophilic amino acids
during reversed-phase chromatography sufficiently to cause them
to elute beyond the void volume peak. This is necessary to resolve
them from large amounts of other interfering polar, nonretained
substances in complex samples. A second contribution is to aid
in electrospray ionization by increasing the charge on amino acids
through introduction of a quaternary amine group. A third way is
by combining the effect of having both a hydrophobic and
quaternary amine groups in proximity. Lengthening the alkyl chain
in the hydrophobic quaternary amine portion of derivatized amino
acids increases their surface-active properties and directs them
to the surface of electrospray droplets where ionization is more
likely to occur.
It is further concluded that during the course of in vitro
derivatization of amino acids to increase their ionization efficiency
the opportunity arises to introduce stable isotope coding of amino
acids according to sample origin. This is in effect a type of
molecular bar coding that allows a unique mass code to be placed
on samples from different sources. It also greatly facilitates
comparative quantification studies. After mixing differentially
coded samples, it is possible to determine the relative concentra-
tion of individual amino acids between these samples in a single
analysis. This approach to relative quantification should be of great
value in comparative metabolomics.
Figure 6. Chromatographic isotope effects for three amino acids
derivatized with C4H9-NA-NSH and C4D9-NA-NSH at the ratio of 1:1.
D ) deuterium-labeled derivatives; H ) nonlabeled derivatives. RD,H
refers to resolution of the heavy and light isotope-labeled derivatives.
Experimental conditions as in Figure 3.
set of amino acid standards. These sample mixtures were prepared
by splitting a mixture of amino acid standards into equal aliquots;
i.e., the amino acid concentration ratio between the two samples
was 1:1. When amino acids in these two samples were differentially
coded with the C4D9-NA-NHS and C4H9-NA-NHS labeling agents,
respectively, and analyzed by RPC-MS, it was seen in multiple
cases that the isotopomers of amino acids did not coelute (Figure
6). This meant that isotope ratio analysis could only be achieved
by comparing the chromatographic peak area of amino acid
isotopomers obtained from extracted ion chromatograms con-
structed after the analysis of the mixture was finished. Real-time
isotope ratio analysis is precluded in cases were amino acid
isotopomers are partially resolved because the isotope ratio at any
single point in time does not reflect that in the initial mixture.
This chromatographic isotope effect is common with deuterated
isotopomers.23 It has also been noted with peptide isotopomers
that the chromatographic isotope effect can be circumvented by
placing deuterium atoms in the coding agent beside a quaternary
2
amine or by using 13C coding instead of H.24
Clearly stable isotope-based quantification can be of value in
metabolomics and other fields where it is the objective to compare
Received for review January 9, 2006. Accepted March 30,
2006.
(23) Filter, C. N. J. Labeled Compd. Radiopharm. 1999, 42, 169-197.
(24) Zhang, R.; Sioma, C. S.; Thompson, R. A.; Xiong, L; Regnier, F. E. Anal.
Chem. 2002, 74, 3662-3669.
AC0600510
4708 Analytical Chemistry, Vol. 78, No. 13, July 1, 2006