Qualitative and quantitative analysis of small amine molecules by MALDI-TOF mass spectrometry through charge derivatization

Article abstract of DOI:10.1021/ac035537k

A pair of isotopically coded light and heavy reagents, tris-(2,4,6- trimethoxyphenyl)phosphonium acetic acid-N-hydroxysuccinimide esters (1 and 2) were synthesized and used to derivatize low molecular weight (<500 Da) molecules containing primary or secon

Full text of DOI:10.1021/ac035537k

Anal. Chem. 2004, 76, 4888-4893
Qualitative and Quantitative Analysis of Small
Amine Molecules by MALDI-TOF Mass
Spectrometry through Charge Derivatization
Peter J. Lee, Weibin Chen, and John C. Gebler*
Life Science Chemistry, Waters Corporation, 34 Maple Street, Milford, Massachusetts 01757
of high-proton-affinity functional groups. Second, the presence of
a variety of abundant matrix-related ions in the low-mass range
clutters the spectrum below 500 Da. Also, matrix ions could
suppress analyte signals and may cause isobaric overlay with
analyte peaks in the MALDI spectrum. To widen the applicability
and to fully take advantage of the MALDI-TOF technique, several
alternative approaches have been proposed to improve the
measurability of low molecular weight analytes by MALDI-TOF
MS.4-8 Gou et al.4 utilized a surfactant in an attempt to suppress
the matrix signals and therefore differentiate them from the
analyte signals. Siuzdak et al.6 reported qualitative analysis results
on small molecules using the desorption/ ionization on silicon
(DIOS) approach. Overall, qualitative and quantitative MALDI-
TOF analysis of small molecules remains a challenge.
Our strategy to analyze small molecules via MALDI-TOF MS
was to use a simple and efficient derivatization reaction to add a
large positively charged tag to the analytes. Many potential
advantages can be realized by derivatizating low-molecular weight
molecules with a large charged tag for MALDI analysis. These
include increased signal intensities due to improved ionization
efficiency, a mass shift toward higher mass range to differentiate
analyte ions from matrix ions and an unambiguous identification
of a class of compounds in a mixture. Further, by incorporation
of isotopes into the derivatizing reagents, reliable quantitative
analysis of small molecules could be achieved.
A pair of isotopically coded light and heavy reagents, tris-
(2 ,4 ,6 -trimethoxyphenyl)phosphonium acetic acid N-hy-
droxysuccinimide esters (1 and 2 ) were synthesized and
used to derivatize low molecular weight (<5 0 0 Da)
molecules containing primary or secondary amine func-
tional groups for MALDI-TOF MS analysis. The light and
heavy reagents added a 5 7 3 and 6 0 0 Da positively
charged tag to each analyte, respectively. In the presence
of 1 0 times excess of tag reagents, the coupling reactions
reached near completion within 1 0 min. The derivatiza-
tion greatly facilitated MALDI analysis of small molecules
and significantly improved the sensitivity of analysis,
allowing a limit of detection in the low femtomole range.
Additionally, the reaction mixtures were directly analyzed
by MALDI without sample cleanup. The quantification of
small molecules by MALDI-TOF MS was successfully
achieved by analysis of isotopically coded light and heavy
derivatives. MALDI-TOF quantitative analysis of a mixture
of antibiotics yielded calibration curves in the concentra-
tion range from 0 .3 to 3 0 pmol/ µL with r 2 values greater
than 0 .9 9 9 5 .
Matrix-assisted laser desorption/ ionization time-of-flight (MALDI-
TOF) mass spectrometry has become a widely used and powerful
tool for analysis of biomolecules and synthetic polymers.1 Protein
identifications via the peptide mass fingerprinting (PMF) approach
employ MALDI-TOF to analyze protein digests.2 MALDI has also
been utilized for genotyping single nucleotide polymorphisms
(SNP) generated from PCR reactions.3 The wide application of
MALDI-TOF MS has been attributed to some of the major
advantages of the technique, such as the simplicity of usage, high
sensitivity, reasonable tolerance against impurities, and the
potential of high sample throughput. Despite the vast success of
MALDI-TOF MS in the analysis of large molecules, difficulties
may arise for the MALDI analysis of low molecular mass
compounds (so-called small molecules, MW < 500 Da). There
are several obstacles for successful application of MALDI-TOF
MS to the analysis of small molecules. First, the analytes of
interest may have very poor ionization efficiency due to the lack
In this work, we report our investigation on derivatizing small
molecules with N-hydroxysuccinimide ester 1 and its heavy-
isotope labeled analogue 2 for MALDI-TOF MS. Reagent 1 was
first reported by Huang et. al. as a reagent for N-terminal peptide
modification of tryptic digests.9-13 For this study, a variety of
(4) Guo, Z.; Zhang, Q.; Zou, H.; Guo, B.; Ni, J. Anal. Chem. 2 0 0 2 , 74, 1637-
1641.
(5) Cohen, L. H.; Gusev, A. I. Anal. Bioanal. Chem. 2 0 0 2 , 373, 571-586.
(6) Wei, J.; Buriak, J. M.; Siuzdak, G. Nature 1 9 9 9 , 399, 243-246.
(7) Ayorinde, F. O.; Garvin, K.; Saeed, K. Rapid Commun. Mass Spectrom.
2 0 0 0 , 14, 608-615.
(8) Tholey, A.; Wittmann, C.; Kang, M. J.; Bungert, D.; Hollemeyer, K.; Heinzle,
E. J. Mass Spectrom. 2 0 0 2 , 37, 963-973.
(9) Huang, Z.; Shen, T.; Wu, J.; Gage, D. A.; Watson, J. T. Anal. Biochem. 1999,
268, 305-317.
(10) Sadagopan, N.; Watson, J. T. J. Am. Soc. Mass Spectrom. 2 0 0 0 , 11, 107-
119.
(11) Sadagopan, N.; Watson, J. T. J. Am. Soc. Mass Spectrom. 2 0 0 1 , 12, 399-
409.
(12) Adamczyk, M.; Gebler, J. C.; Wu, J. Rapid Commun. Mass Spectrom. 1 9 9 9 ,
* To whom correspondence should be addressed. E-mail: John_Gebler@
waters.com. Fax: 508-482-3100. Tel: 508-482-2786.
(1) Karas, M.; Hillenkamp, F. Anal. Chem. 1 9 8 8 , 60, 2999-2301.
(2) Aebersold, R.; Mann, M. Nature 2 0 0 3 , 422, 198-207.
(3) Sauer, S.; Gut, I. G. J. Chromtogr. B 2 0 0 2 , 782, 73-87.
13, 1413-1422.
(13) Adamczyk, M.; Gebler, J. C.; Wu, J.; Yu, Z. J. Immunol. Methods 2 0 0 2 ,
260, 235-249.
4888 Analytical Chemistry, Vol. 76, No. 16, August 15, 2004
10.1021/ac035537k CCC: $27.50 © 2004 American Chemical Society
Published on Web 07/16/2004

Scheme 1
Scheme 2
Scheme 3
Scheme 4
primary and secondary amines including alkylamines, drugs,
antibiotics, and stimulants were derivatized by 1 and 2 , adding
573 and 600 Da of a positively charged tag to each analyte (3 and
4 , Scheme 1). The derivatization procedure was tested and
optimized with respect to simplicity, efficiency, and measurability
of the derivatives by MALDI-TOF MS. In addition, the quantitative
analysis of small molecules by MALDI-TOF MS was performed.
MALDI-TOF MS Analysis. Spectra were acquired with a
Micromass M@LDI LR time-of-flight mass spectrometer (Waters
Corparation, Milford, MA) using a 337-nm nitrogen laser and an
accelerating voltage of 15 kV. The instrument has delayed
extraction capabilities, and the delay time setting was 500 ns. All
the data were acquired using positive reflectron mode, and each
spectrum represented a sum of 100 laser shots. The ion extraction
pulse voltage was fixed at 2000 V. Data collection and processing
were controlled by the manufacture’s software.
High-purity CHCA was used as the MALDI matrix throughout
this study. It was dissolved in 50% acetonitrile and 50% ethanol at
a concentration of 10 mg/ mL. Samples for mass spectrometry
were prepared by mixing the acidified reaction mixture with the
matrix solution in a 1:1 ratio (v/ v) and spotting 1 µL of this onto
a stainless steel target plate. The applied samples were allowed
to dry at ambient temperature before transferring into the mass
spectrometer.
P reparation of Stock Solutions. Stock solutions of 1 and 2
(12 nmol/ µL) were prepared by placing 13.1 mg of 1 and 13.7
mg of 2 into 1400 and 1415 µL of anhydrous acetonitrile,
respectively, in a 1.5-mL screw cap polypropylene microcentrifuge
tube. N-hydroxysuccinimide esters 1 and 2 were reacted with
water to form carboxylic acids 8 and 9 when they were exposed
to water (Scheme 4). At ambient temperature, the half-life time
of 1 in unbuffered CH3CN/ H2O (50/ 50 v/ v) solution is 18 h and
the half-life in pH 9.2 buffer solution is 40 s. Care must be taken
to avoid exposing the stock solution to moisture or buffer solutions
until required. The stock solutions can be stored in a desiccator
at room temperature for 3 months without discernible decomposi-
tion as determined by ESI-MS or MALDI-MS.
EXPERIMENTAL SECTION
Reagents. Anhydrous acetonitrile (CH3CN), cyclooctylamine,
dimethyl-d6 sulfate, dioctylamine, enoxacin, lomefloxacin, nor-
floxacin, naphthalenemethylamine, phosphorus trichloride, tri-
ethylamine, triethanolamine, trihydroxybenzene, triethylaminon-
ium bicarbonate buffer (1.0 M), tris(2,4,6-trimethoxyphenyl)-
phosphine (5 ), and zinc chloride undecylamine were purchased
from Sigma-Aldrich (St. Louis, MO). Phentermine was purchased
from Alltech (Deerfield, IL) and ciprofloxacin was purchased from
Serologicals (Canada). Trifluoroacetic acid (TFA) was purchased
from Fisher Scientific (Pittsburgh, PA) and ultrapure water (18.2
MΩ) was used. R-Cyano-4-hydroxycinnamic acid was obtained
from Waters (MassPREP MALDI Matrix CHCA). Compounds 1
and 2 were prepared by reacting 5 or tris(2,4,6-trimethoxy-d9)-
phenyl)phosphine (6 ) with bromoacetic acid N-hydroxysuccin-
imide ester, respectively, in toluene for 30 min using published
procedures (Scheme 2).7,9 Compound 6 for preparing 2 was
synthesized using published methods (Scheme 3).14,15 Basically,
trihydroxybenzene, dimethyl-d6 sulfate, and potassium carbonate
were refluxed in acetone for 3 h to yield 7 .14 The isolated 7 was
mixed with ZnCl2 and refluxed in phosphorous trichloride solution
for 8 h to yield 6 .15
Three buffer solutions (16 mM triethylamine, triethanolamine,
and triethylaminonium bicarbonate), each containing 20% (v/ v)
(14) Clark-Lewis, J. W. Aust. J. Chem. 1 9 5 7 , 10, 505-506.
(15) Protopopov, I. S.; Kraft, M. Y. Zh. Obshch. Khim. 1 9 6 3 , 33, 3050-3052.
Analytical Chemistry, Vol. 76, No. 16, August 15, 2004 4889

of CH3CN, were used in this study. Buffer A (pH 11.7) was
prepared by mixing 0.810 g of triethylamine with 400 mL of water
and 100 mL of acetonitrile. Buffer B (pH 9.9) was prepared by
mixing 1.194 g of triethanolamine with 400 mL of water and 100
mL of acetonitrile. Buffer C (pH 9.2) was prepared by by mixing
8 mL of triethylaminonium bicarbonate (1.0 M) solution with 392
mL of water and 100 mL of acetonitrile.
Derivatization of Amines. The method described below is a
generic procedure used for derivatization of amines. Analytes were
dissolved in the buffer solutions (A, B, or C). To 400 µL of analyte
solution (4 pmol/ µL) was added 8 µL of 1 or 2 (12 nmol/ µL)
stock solution. The solution was mixed with a vortex mixer for
10 s and incubated at room temperature for 10 min. At the end of
the reaction, the reaction mixture was acidified with TFA to pH
2.0.
Comparison of Coupling Yield at Different pH. The amines
1 0 -1 5 (Figure 1) were placed in buffer A, B, and C to make 4
pmol/ µL of analyte solution D, E, and F, respectively. Analyte
solutions D and E were derivatized with 1 , and analyte solution
F was derivatized with 2 , using the derivatization method
described to yield solutions D-1 , E-1 , and F-2 , respectively. Equal
volumes of solutions D-1 and F-2 were mixed to make sample I
and equal volumes of solutions E-1 and F-2 were mixed to make
sample II. Samples I and II were acidified with TFA and analyzed
by MALDI-TOF MS. The peak intensity of derivatives in D-1 and
F-2 of sample I and derivatives in E-1 and F-2 of sample II were
recorded in Tables 1 and 2, respectively. The peak intensity of
derivatives in D-1 , E-1 , and F-2 in the spectra of samples I and II
were further normalized using the following equations ((1)-(3)):
Figure 1. Structure of analytes.
Table 1. Comparison of pH 11.7 and 9.2 for Coupling
Reaction
peak intensity
derivative of 1
(pH 11.7)
derivative of 2
analytes
(pH 9.2)
primary
amine
1 0
(n + m)
8374
3278
712
1390
1013
J )
K )
L )
(1)
(2)
(3)
1 1
2
1 2
3924
1 3
10326
13990
w(n + m)
2n
secondary
amine
1 4
251
1909
8342
1 5
5020
z(n + m)
2m
Table 2. Comparison of pH 9.9 and 9.2 for Coupling
Reaction
where J is the normalized peak intensity of derivatives in F-2 , K
is the normalized peak intensity of derivatives in D-1 , L is the
normalized peak intensity of derivatives in E-1 , n is the peak
intensity of derivatives in F-2 in the sample I spectrum, m is the
peak intensity of derivatives in F-2 in the sample II spectrum, w
is the peak intensity of derivatives in D-1 in the sample I spectrum,
and z is the peak intensity of derivatives in E-1 in the sample II
spectrum.
Stoichiometry. Antibiotics 1 6 , 1 7 , and 1 8 were dissolved in
buffer C to prepare 9 pmol/ µL of antibiotic stock solution. To six
microcentrifuge tubes containing 1 mL of the antibiotic solution
was added 10, 20, 30, 40, 50, and 100 µL of 1 (12 nmol/ µL),
respectively. Each solution was mixed with a vortex mixer for 10
s and incubated at ambient temperature for 10 min to yield
derivatives of 1 . To those microcentrifuge tubes, 50 µL of 2 (12
nmol/ µL) was added and incubated for 10 min to convert any
unreacted antibiotic to derivatives of 2 . Then, each solution was
acidified with TFA and added matrix and subjected to MALDI-
TOF analysis. The peak intensities of the derivatives of 1 and
peak intensity
derivative of 1
derivative of 2
analytes
(pH 9.9)
(pH 9.2)
primary
amine
1 0
1 1
1 2
1943
2962
9752
647
1516
987
1 3
16308
14475
secondary
amine
1 4
3520
6313
3349
6442
1 5
their corresponding derivatives of 2 on each spectrum were
examined and used to calculate the percent yield of the coupling
reaction of 1 of those six solutions as shown in eq 4.
100x
(x + y)
yield(% ) )
(4)
4890 Analytical Chemistry, Vol. 76, No. 16, August 15, 2004

Table 3. Comparison of pH 11.7, 9.9, and 9.2 for
Coupling Reaction
normalized peak intensity
derivative of 1
(pH 11.7)
derivative of 1
derivative of 2
analytes
(pH 9.9)
(pH 9.2)
primary
amine
1 0
7992
3425
2040
2839
679
1453
1 1
1 2
3872
9883
1000
1 3
10504
16035
14232
secondary
amine
1 4
345
4448
2763
7245
2629
7392
Figure 2. MALDI-TOF spectrum of 2 pmol of derivatized 10 (C+
)
1 5
700.33), 12 (C+ ) 722.32), 13 (C+ ) 730.30), 11 (C+ ) 744.39), 14
(C+ ) 814.43), and 15 (C+ ) 893.32).
where x is the peak intensity of the derivative of 1 and y is the
peak intensity of its corresponding derivative of 2 .
RESULTS AND DISCUSSIONS
A mixture of six amines, 1 0 -1 5 (4 pmol/ µL), was derivatized
by 1 in buffer A (pH 11.7). The selected compounds represent
some primary and secondary amines with a selection of aliphatic
and aromatic moieties coupled to various combinations of chemical
functional groups as shown in Figure 1. The N-hydroxysuccin-
imide ester of 1 was reacted with an amine functional group to
form a stable amide linkage, adding 573 Da to each analyte. Figure
2 shows the MALDI-TOF spectrum of the resulting derivatives.
The peak intensities of the derivatized 1 0 , 1 1 , and 1 3 ions were
among the strongest, followed by the derivatized 1 2 and 1 5 ions.
The peak intensity of the derivatized 14 ion was the lowest among
the six derivatives, which is possibly a result of the lower
conversion yield of the coupling reaction due to the steric
hindrance of the two long alkyl chains surrounding the amine
group, or from poor cocrystallization with the matrix (CHCA).
The competitive reactions of amidation (Scheme 1) and
hydrolysis of the N-hydroxysuccinimide esters 1 and 2 (Scheme
4) occur spontaneously during derivatization in aqueous solution.
The basic buffer solutions were used to facilitate the condensation
reaction, and a large excess of 1 was frequently added to ensure
high yields of derivatization of the amine analytes.9 Several buffer
systems have previously been reported for the charge derivati-
zation reaction,9-11 including phosphate, Tris-HCl, BICINE, and
HEPES. The use of these buffers requires sample purification prior
to mass analysis since they are not volatile. Also, buffers such as
Tris-HCl can compete with analytes to react with 1 and 2 during
derivatization. Because MALDI-TOF was chosen as the analytical
technique for the analysis of derivatized compounds in this study,
it was desirable to choose votatile buffers that can provide buffer
capacity in the desired pH range and also minimize interferences
during MALDI analysis.
Figure 3. Plots of the percent conversion yield versus the molar
ratio of 1 to antibiotics 16 ([), 17 (9), and 18 (2).
method, which analyzes the relative amount of isotopically coded
light 1 and heavy 2 derivatives, was developed to compare and
optimize the coupling reaction conditions.
Optimization of pH for Coupling Yield. Three volatile
buffers, A (pH 11.7, 16 mM triethylamine), B (pH 9.9, 16 mM
triethanolamine), and C (pH 9.2, 16 mM triethylaminonium
bicarbonate), were selected to provide a pH range for investigating
the efficiency of coupling 1 and 2 with amines 1 0 -1 5 . The first
experiment was to compare pH 11.7 with pH 9.2. As the results
show in Table 1, the primary amines 1 0 , 1 1 , and 1 2 reacted
more efficiently at pH 11.7 with 1 . The naphthalenemethylamine
(1 3 ) and secondary amines (1 4 and 1 5 ) reacted better at pH
9.2 with 2 . The second experiment was to compare pH 9.9 with
pH 9.2. The results show that amines 1 0 -1 4 reacted more
efficiently at pH 9.9 with 1 , but antibiotic 1 5 still reacted slightly
better at pH 9.2 with 2 . For further comparison, the results of
Tables 1 and 2 were normalized. As shown in Table 3, the best
derivatization pH for 1 0 and 1 1 was 11.7 (buffer A), for 1 2 , 1 3 ,
and 1 4 it was 9.9 (buffer B), and for 1 5 it was 9.2 (buffer C).
Stoichiometry. To optimize the stoichiometry of derivatization,
a 9 pmol/ µL solution of antibiotics 1 6 , 1 7 , and 1 8 in buffer C
(pH 9.2) was reacted with varying amounts of 1 from 4 to 44 molar
excess. After 10 min, an excess amount of 2 (22x) was added to
each reaction to convert any unreacted antibiotic to derivatives
of 2 . The relative amount of derivatives with 1 and 2 in each
Since the rate of reaction is likely to be analyte-specific,
optimizing the pH of buffers and the stoichiometry will increase
the yield of derivatization and decrease the amount of hydrolyzed
N-hydroxysuccinimide ester in the reaction solution. HPLC
analysis was used to analyze the yield of N-hydroxysuccinimide
ester 1 derivatized analytes. However, it was a time-consuming
process for obtaining suitable internal standards and optimizing
separation conditions.9,12 A fast and simple MALDI-TOF MS
Analytical Chemistry, Vol. 76, No. 16, August 15, 2004 4891

to derivatize 1 6 and 1 7 to near completion. However, more than
20x molar excess of 1 was required for 1 8 to achieve the same
conversion yield. The difference in reaction yields among the
antibiotics may be due to the steric hindrance caused by the
methyl group next to the secondary amine of 1 8 .
Quantitative MALDI-TOF Analysis. The isotopically coded
light and heavy reagents (1 and 2 ) were used to explore the
potential for quantitative analysis of small analytes by MALDI-
TOF MS. The evaluation was carried out in buffer C (pH 9.2)
using antibiotics 1 6 , 1 7 , and 1 8 as model compounds. Figure 4
shows the specturm of a mixture of two equal volumes of antibiotic
solutions (3 pmol/ µL) that were treated with 40x molar excess of
1 and 2 . The spectrum shows that the peak intensities of the
derivatives of 1 and their corresponding derivatives of 2 are
essentially the same. This indicates that antibiotics derivatized
by both 1 and 2 have the same cocrystallization behavior with
the matrix. To examine the dynamic range of the quantitative
analysis by MALDI-TOF MS through derivatization, a series of
test solutions containing 0.3, 0.6, 0.9, 3, 9, 24, and 30 pmol/ µL (2
orders of magnitude) of antibiotics 1 6 , 1 7 , and 1 8 in buffer C
were treated with 40x molar excess of 1 . An antibiotic solution
containing 3 pmol/ µL of 1 6 , 1 7 , and 1 8 in buffer C was reacted
with 40x molar excess of 2 to form a 2 derivatized antibiotic
standard solution (3 pmol/ µL). Equal volumes of 2 derivatized
antibiotic standard solution and 1 derivatized test solution were
mixed, and the resulting solution was analyzed by MALDI-TOF
MS. Equation 5 was used to obtain the observed antibiotic
concentration (O) in the test solutions.
Figure 4. MALDI-TOF spectrum of 750 fmol of 1 and 2 derivatized
16 (C+ ) 892.33 and 919.50), 17 (C+ ) 904.32 and 931.49), and 18
(C+ ) 924.34 and 951.49).
Figure 5. Plots of the observed and calculated concentrations of
antibiotics 16 (9, r2 ) 0.9997), 17 ([, r2 ) 0.9996), and 18 (2, r2 )
0.9997) in the test solutions.
P
O ) × (3 pmol/ µL)
(5)
Q
reaction was evaluated by MALDI-TOF MS. Equation 4 was used
to estimate the percent yield of the coupling reaction when a
certain amount of 1 was added. The calculation assumes that the
reaction will reach completion when a large molar excess of
reagents are used. Figure 3 is a plot of the percent conversion
yield of coupling reaction against the molar ratios of 1 to
antibiotics. Approximately, a 10x molar excess of 1 was sufficient
where P is the peak intensity of the derivatives of 1 and Q is the
peak intensity of the corresponding derivatives of 2 in each
spectrum. Figure 5 shows the plots of the calculated and observed
values for the concentration of three antibiotics in the test
solutions. The linear correlation coefficients (r2) for all three
analytes are greater than 0.9995, which demonstates the utility of
Figure 6. MALDI-TOF spectrum of 3 fmol (A) and 30 fmol (B) of 1 and 2 derivatized 16, 17, and 18.
4892 Analytical Chemistry, Vol. 76, No. 16, August 15, 2004

spotted directly on a stainless steel target plate. The limit of
detection for this class of compounds is in the low femtomole
range. Contrastly, direct MALDI analysis of these antibiotics at
picomole loadings without derivatization yielded very weak signals
that were buried in the abundant matrix-related peaks (data not
shown).
Interestingly, 1 - and 2 -derivatized naphthalenemethylamine
(1 3 ) could be observed without using matrix by direct laser
desorption analysis (Figure 7). A high signal-to-noise ratio was
achieved at 200-fmol loading. It is presumed that the aromatic
nature of the analyte was responsible for this observation.
CONCLUSIONS
Isotopically coded light and heavy N-hydroxysuccinimide esters
Figure 7. Laser desorption spectrum of 200 fmol 1 and 2 derivatized
13 (C+ ) 730.37 and 757.55) without using a matrix.
(1 and 2) were successfully employed as an aid for analyzing small
molecules by MALDI-TOF MS. Esters 1 and 2 rapidly reacted
with primary and secondary amines in mild conditions to add a
large positively charged tag (573 and 600 Da, respectively) to the
analytes. This increased their limit of detection due to improved
ionization efficiency and a mass shift away from the matrix
clusters. The peak intenisty ratio of derivatives of 1 and 2 obtained
from the same MALDI target spot were directly related to the
ratio of analyte concentrations in compared samples. It was proven
that there is potential for using isotopically coded light and heavy
reagents such as 1 and 2 for conducting quantitative MALDI-
TOF MS analysis of small molecules and antibiotic mixtures.
quantitative MALDI-TOF MS analysis coupled with derivatization
using isotopically coded 1 and 2 .
Although MALDI has a moderate tolerance for contaminants
in the sample, it is also well-known that the ionization efficiency
can be severely diminished by the presence of salts or buffers in
the sample. For any derivatization reaction, there is always a
probability of introducing contaminants that may interfere with
downstream MALDI analysis. Among all possible contaminantion
resources in this study, the primary ones are the derivatizing
reagents and the buffers. N-Hydroxysuccinimide esters 1 and 2
and their derivatives were observed to be highly compatible with
MS analysis, and all the reaction mixtures were analyzed without
prior cleanup. Figure 6 shows the MALDI spectra for 3 and 30
fmol of reaction mixture of 1 and 2 derivatized 1 6 , 1 7 , and 1 8
Received for review December 24, 2003. Accepted June 2,
2004.
AC035537K
Analytical Chemistry, Vol. 76, No. 16, August 15, 2004 4893