Analysis of underivatized amino acids and their D/L-enantiomers by sheathless capillary electrophoresis/electrospray ionization-mass spectrometry

Article abstract of DOI:10.1021/ac0263925

Capillary electrophoresis/electrospray ionization-mass spectrometry (CE/ESI-MS) was applied to the analysis of underivatized amino acids and the separation of their D/L-enantiomers. Under full-scan mode, all standard protein amino acids were separated and detected at low-femtomole levels using a 130-cm-long, 20-μm-i.d., 150-μm-o.d. underivatized fused-silica capillary with 1 M formic acid as the background electrolyte. The CE/ESI-MS technique was also applied to the separation of L-arginine from L-canavanine (a close analogue of arginine where the terminal methylene linked to the guanidine group of arginine is replaced by an oxygen atom) in a complex mixture containing all standard protein amino acids. The utility of CE/ESI-MS in the analysis of real-world samples was demonstrated by the identification of two metabolic diseases (PKU and tyrosinemia) through blood analysis with minimal sample preparation. In addition, the on-line separation of 11 underivatized L-amino acids from their D-enantiomers was achieved by using a 30 mM solution of (+)-(18-crown-6)-2,3,11,12-tetracarboxylic acid as the background electrolyte.

Full text of DOI:10.1021/ac0263925

Anal. Chem. 2003, 75, 1508-1513
An a lys is o f Un d e riva t ize d Am in o Ac id s a n d Th e ir
D/L-En a n t io m e rs b y S h e a t h le s s Ca p illa ry
Ele c t ro p h o re s is /Ele c t ro s p ra y Io n iza t io n -Ma s s
S p e c t ro m e t ry
Ca s e y L. Sc hultz a nd Me hdi Moini*
Department of Chemistry and Biochemistry, University of Texas at Austin, Austin, Texas 78712
5
Capillary electrophoresis/ electrospray ionization-mass
spectrometry (CE/ ESI-MS) was applied to the analysis of
underivatized amino acids and the separation of their D/ L-
enantiomers. Under full-scan mode, all standard protein
amino acids were separated and detected at low-femto-
mole levels using a 1 3 0 -cm-long, 2 0 -µm-i.d., 1 5 0 -µm-
o.d. underivatized fused-silica capillary with 1 M formic
acid as the background electrolyte. The CE/ ESI-MS
of infants. Also, it has been found that glutamine concentration
in the cerebrospinal fluid of children with meningitis decreases
during the viral or bacterial infection, suggesting that this amino
acid could be very useful for monitoring the growth of patho-
3
gens. The analysis of amino acids is also very important in the
food industry where amino acids are measured to correlate flavor
trends, monitor fermentation, and assess the quality of the final
product.6
technique was also applied to the separation of
L
-arginine
Another aspect of amino acid analysis that has been particularly
challenging in the past is the separation and detection of amino
D-amino acids in
mammalian systems has surged in the past decade due to findings
from -canavanine (a close analogue of arginine where the
L
acid enantiomers.8,9 Interest in the role of
terminal methylene linked to the guanidine group of
arginine is replaced by an oxygen atom) in a complex
mixture containing all standard protein amino acids. The
utility of CE/ ESI-MS in the analysis of real-world samples
was demonstrated by the identification of two metabolic
diseases (P KU and tyrosinemia) through blood analysis
with minimal sample preparation. In addition, the on-line
that suggest neuronal and neuroendocrine roles of some D-amino
acids.1
0-12
The analysis of amino acid enantiomers is also important
for food quality analysis where the enantiomeric ratio of amino
acids can be used as a reliable parameter to assess food quality.7
An increase in the ratio of D-amino acids to L-amino acids within
separation of 1 1 underivatized
L
-amino acids from their
foods may be indicative of extensive processing, contamination,
adulteration, or aging. The enantiomeric ratio of amino acids
7
D
-enantiomers was achieved by using a 3 0 mM solution
of (+)-(1 8 -crown-6 )-2 ,3 ,1 1 ,1 2 -tetracarboxylic acid as the
background electrolyte.
within a biological sample can also be used in biological dating
applications.13
Gas chromatography,14 liquid chromatography,15 GC/ MS,16 and
3
The detection, identification, and quantitation of amino acids
is important in many areas of science including biological and
biochemical analysis,1,2 medical diagnostics,3-5 and food analy-
sis.3,6,7 For example, the determination of amino acids within
various bodily fluids such as blood and urine is critical for routine
clinical analysis. Several metabolic diseases involving amino acids,
such as phenylketonurea (PKU) and tyrosinemia (elevated ty-
rosine), are diagnosed in the neonatal period of life through the
detection of abnormal levels of amino acids in the blood or urine
capillary electrophoresis have all been used for the detection,
identification, and quantitation of amino acids. However, for
efficient separation and sensitive detection using these methods,
amino acids generally require derivatization prior to their analysis.
This is a labor-intensive and time-consuming process. Recently,
HPLC/ MS17 and sheath-flow capillary electrophoresis/ electro-
spray ionization-mass spectrometry (CE/ ESI-MS)6,18 techniques
have been introduced for the analysis of underivatized amino acids.
However, because of the high liquid flow rate (>5 µL/ min)
*
Corresponding author: E-mail: mmoini@mail.utexas.edu. Phone: (512) 471-
(8) Verleysen, K.; Sandra, P. Electrophoresis 1 9 9 8 , 19, 2798-2833.
(9) Kuhn, R.; Stoecklin, F.; Erni, F. Chromatographia 1 9 9 2 , 33, 32-36.
(10) D’Aniello, A.; Lee, J. M.; Petrucelli, L.; Maddalena Di Fiore, M. Neurosci.
Lett. 1 9 9 8 , 250, 131-134.
7
344. Fax: (512) 471-1420.
1) Teerlink, T. J. Chromatogr., B 1 9 9 4 , 659, 185-207.
(
(
2) Schegg, K. M.; Denslow, N. D.; Anderson, T. T.; Bao, Y.; Cohen, S. A.;
Mahrenholz, A. M.; Mann, K. In Techniques in Protein Chemistry VIII;
Marshak, D. R., Ed.; Academic Press: San Diego, CA, 1997; pp 207-216.
3) Prata, C.; Bonnafous, P.; Fraysse, N.; Treilhou, M.; Poinsot, V.; Couderc, F.
Electrophoresis 2 0 0 1 , 22, 4129-4138.
4) Vecchione, G.; Margaglione, G.; Grandone, E.; Colaizzo, D.; Cappucci, G.;
Fermo, I.; D’Angelo, A.; Di Minno, G. Electrophoresis 1 9 9 9 , 20, 569-574.
5) Nyhan, W. L. Abnormalities in Amino Acid Metabolism in Clinical Medicine;
Appleton Century Crofts: East Norwalk, CT, 1984.
(11) Schell, M. J.; Cooper, O. B.; Snyder, S. H. Proc. Natl. Acad. Sci. U.S.A. 1997,
94, 2013-2018.
(
(
(
(12) Schell, M. J.; Brady, R. O., Jr.; Molliver, M. E.; Snyder, S. H. J. Neurosci.
1 9 9 7 , 17, 1604-1615.
(13) Lubec, G.; Lubec, B. Amino Acids 1 9 9 3 , 4, 1-3.
(14) Amino Acid Analysis by Gas Chromatography; Zumwalt, R. W., Kuo, K. C.,
Gehrke, C. W., Eds.; CRC Press: Boca Raton, FL, 1987; Vols. I-III.
(15) Deyl, Z.; Hyanek. J.; Horikova, M. J. Chromatogr. 1 9 8 6 , 379, 177-250.
(16) Duncan, M. W.; Poljak, A. Anal. Chem. 1 9 9 8 , 70, 890-896.
(17) Kwon, J.; Moini, M. J. Am. Soc. Mass Spectrom. 2 0 0 1 , 12, 117-122.
(18) He, T.; Quinn, D.; Fu, E.; Wang, Y. K. J. Chromatogr., B 1 9 9 9 , 727, 43-52.
(
(
6) Soga, T.; Heiger, D. N. Anal. Chem. 2 0 0 0 , 72, 1236-1241.
7) Marchelli, R.; Dossena, A.; Palla, G. Trends Food Sci. Technol. 1 9 9 6 , 7,
1
13-119.
1508 Analytical Chemistry, Vol. 75, No. 6, March 15, 2003
10.1021/ac0263925 CCC: $25.00 © 2003 American Chemical Society
Published on Web 02/19/2003

associated with these techniques, their main disadvantage is a
low absolute sensitivity. Several capillary electrophoresis tech-
niques have been introduced for the on-line separation of amino
acid enantiomers.8,9,19 Nonmass spectrometric techniques, how-
ever, exhibit low selectivity and are generally limited to aromatic
amino acids (Tyr, Trp, Phe) if one wants to avoid time-consuming
derivatization procedures.
Because of their low flow rates (nL/ min), sheathless CE/ ESI-
MS techniques are highly sensitive and have proved effective in
the analysis of complex biological mixtures that are available in
low quantities.20 In this study, the split-flow technique for interfac-
ing CE to MS has been applied to the analysis of underivatized
amino acids. In addition, the utility of CE/ ESI-MS in the analysis
of real-world samples has been demonstrated by the identification
of two metabolic disorders (PKU and tyrosinemia) through blood
analysis. Moreover, on-line CE/ ESI-MS separation of a multitude
of underivatized amino acid enantiomers has been demonstrated.
EXPERIMENTAL SECTION
Methods and Instrumentation. The split flow method21 was
used for interfacing CE (MDQ, Beckman-Coulter; Fullerton, CA)
to MS (LCQ, Finnigan; San Jose, CA). For the analysis of the
Fig u re 1 . CE/ESI-MS base peak electropherogram of the separa-
tion of the 20 standard protein amino acids using the 20-µm-i.d., 130-
cm-long CE capillary. No pressure was applied at the CE inlet during
this separation. Forward polarity with 30-kV separation voltage was
used. Approximately 400 fmol of each amino acid was injected.
L-amino acid standard and the separation of amino acid enanti-
omers, an underivatized 130-cm-long, 20-µm-i.d., 150-µm-o.d. fused-
silica capillary was used. For the blood analysis, an underivatized
1
15-cm-long, 20-µm-i.d., 150-µm-o.d. fused-silica capillary was used.
canavanine, giving all 21 amino acids a final concentration of 0.50
The outlet tips of all capillaries used were sharpened through
etching with 49% hydrofluoric acid while wearing gloves and
working under a fume hood. The MS was scanned in the m/ z
range of 74.5-250 under positive ionization mode. The CE was
operated in forward polarity mode (30 kV was applied to the CE
inlet electrode), and 1.1-1.5 kV was applied to the CE outlet/
ESI electrode. To speed up the analysis time during the
separation of the 21-amino acid standard and the blood analysis
experiments, and to maintain stable electrospray during the
enantiomer separation, 5, 10, and 5 psi of pressure was applied,
respectively, to the CE inlet.22
mM. The
D
/
L-amino acid enantiomer standard solution used in
the enantiomer separation contained 11
D/ L-amino acid pairs at a
concentration of 0.25 mM. The blood samples were prepared by
soaking a 3/ 16-in.-diameter dried blood spot on filter paper in 100
µL of HPLC grade water for 10 min. A 20 µL aliquot of this solution
was then removed and was diluted 10 times with a solution of
acetonitrile/ water/ formic acid (49.9/ 49.9/ 0.2; v/ v). Amino acid
recovery from dried blood spots has been extensively studied by
government and commercial laboratories23 and was, therefore, not
investigated in this study. However, since the relative intensities
of the same amino acids were compared in the blood of healthy
versus diseased infants, the issue of recovery has a minimal effect
on the comparison.
2
0
Reagents and Chemicals. All chemicals, other than HPLC-
grade water and 49% HF (Fischer Scientific; Pittsburgh, PA), were
purchased from Sigma-Aldrich (St. Louis, MO) and were used
without further purification. For the analysis of the L-amino acid
RESULTS AND DISCUSSION
standards and the blood samples, a 1 M formic acid solution was
used as the CE background electrolyte (BGE). For the separation
of amino acid enantiomers, a 30 mM (+)-(18-crown-6)-2,3,11,12-
tetracarboxylic acid (18-C-6-TCA) solution was used as the BGE/
chiral selector reagent with no formic acid added. Two standard
amino acid solutions were prepared. The first one contained all
Analysis of Underivatized Amino Acids. Figure 1 shows the
CE/ ESI-MS electropherogram of the separation of the 20 standard
protein amino acids using the underivatized 130-cm-long CE
capillary in conjunction with 1 M formic acid as the BGE. For
this experiment, ∼0.4 nL of the 20-amino acid standard solution
(
1
∼400 fmol of each amino acid) was injected. As shown in Figure
2
0 standard protein amino acids and was made by spiking a
, all amino acids were separated and detected. The use of a
commercially available 17-amino acid standard solution with the
remaining 3 standard amino acids (Asn, Gln, Trp). The final
concentration of each amino acid in the solution was 1.00 mM.
The second amino acid standard contained 21 amino acids and
was made by combining an aliquot of the 20-amino acid standard
nonderivatized column is certainly a big advantage since capillary
derivatization techniques are usually long, labor-intensive, expen-
sive, and short-lived. In addition, the low molecular weight of
formic acid ensures minimal background chemical noise and,
therefore, lower detection limits.
Table 1 summarizes the extracted ion detection limits and
linearity correlations of the 20 standard protein amino acids except
for cysteine (Cys), which was observed as its oxidized form (Cys-
with an equal volume aliquot of a 1.00 mM solution of
L-
(
19) Verleysen, K.; Bosch, T. Van den; Sandra, P. Electrophoresis 1999, 20, 2650-
655.
2
(
(
(
20) Moini, M. Anal. Bioanal. Chem. 2 0 0 2 , 373, 466-480.
21) Moini, M. Anal. Chem. 2 0 0 1 , 73, 3497-3501.
22) Cao, P.; Moini, M. Electrophoresis 1 9 9 8 , 19, 2200-2206.
(23) Adam, B. W.; Alexander, J. R.; Smith, S. J.; Chace, D. H.; Loeber, J. G.;
Elvers, L. H.; Hannon, W. H. Clin. Chem. 2 0 0 0 , 46, 126-128.
Analytical Chemistry, Vol. 75, No. 6, March 15, 2003 1509

Ta b le 1 . P ro p e rt ie s a n d Fig u re s o f Me rit o f t h e Tw e n t y S t a n d a rd Am in o Ac id s
pKa values
NH3+
detection
limit (fmol)
correlation
coefficient
amino acid
MW
COOH
R group
pI
nonpolar, aliphatic
R groups
Gly, G
Ala, A
Val, V
Leu, L
Ile, I
Met, M
75
89
117
131
131
149
2.34
2.34
2.32
2.36
2.36
2.28
9.60
9.69
9.62
9.60
9.68
9.21
5.97
6.01
5.97
5.98
6.02
5.74
14.0
13.5
3.0
2.7
2.7
0.9994
0.9997
0.9996
0.9971
0.9988
0.9935
2.1
aromatic
R groups
Phe, F
Tyr, Y
Trp, W
polar, uncharged
R groups
Ser, S
Pro, P
Thr, T
165
181
204
1.83
2.20
2.38
9.13
9.11
9.39
5.48
5.66
5.89
1.7
1.0
1.0
0.9979
0.9974
0.9998
10.07
8.18
105
115
119
121
132
146
2.21
1.99
2.11
1.96
2.02
2.17
9.15
10.96
9.62
10.28
8.80
5.68
6.48
5.87
5.07
5.41
5.65
6.9
3.9
5.3
-
7.9
3.2
0.9993
0.9998
0.9997
-
0.9981
0.9999
Cys, C
Asn, N
Gln, Q
9.13
positively charged
R groups
Lys, K
His, H
Arg, R
146
155
174
2.18
1.82
2.17
8.95
9.17
9.04
10.53
6.00
12.48
9.74
7.59
10.76
2.9
4.1
2.4
0.9993
0.9995
0.9994
negatively charged
R groups
Asp, D
133
147
1.88
2.19
9.60
9.67
3.65
4.25
2.77
3.22
5.5
3.2
0.9962
0.9981
Glu, E
Cys). As is shown, the detection limits are in the low-femtomole
range. The detection limits reported in Table 1 were achieved
under full-scan mode (m/ z 74.5-250.0) and are 3-4 orders of
magnitude lower than one sheath-flow CE/ MS study reported.18
The linear quantitation range for all amino acids was ∼2 orders
of magnitude. The high sensitivity achieved was due to the low
flow rate of the narrow CE capillary used. Since ESI is a
concentration sensitive ionization technique, the lower flow rates
of narrow capillaries provide higher sensitivities.
From the results of Figure 1, two general features of the
experimental results should be discussed. One is the order of
migration of the amino acids and the other is the relative intensity
of each amino acid.
the BGE used (1 M formic acid, pH ∼1.8), all amino acids had a
net positive charge and moved toward the CE outlet under forward
polarity mode. Their degree of positive charge, however, depended
on their pK values. As shown in Figure 1, amino acids that were
highly charged (the amino acids with basic side chains) migrated
first, moderately charged amino acids (aliphatic amino acids with
unionizable side chains) migrated second, and both larger amino
acids (aromatic amino acids, Cys-Cys) and amino acids with a
very low degree of positive charge (amino acids with acidic side
chains and proline) migrated last.
The second feature of Figure 1 that needs to be explained is
the relative intensities of the protonated amino acids. Since
approximately equal molar amounts of each amino acid were
injected (∼400 fmol) and analyzed under constant BGE and ESI
voltage, their relative intensities reflect their ionization efficiency
under ESI. Their relative intensities can be divided into three
general categories: high intensity, medium intensity, and low
intensity. While a detailed explanation of the relative sensitivities
of the amino acids is under investigation, the general categories
observed are consistent with the proposed mechanism of ESI.
Under ESI, droplets with a surface excess charge are created.
The ESI equilibrium partitioning model suggests that ions within
a droplet partition between the surface of the droplet (where there
is much excess charge) and the interior of the droplet and that
the relative abundances of ions in the mass spectrum are
In CE, the electrophoretic velocity (νep) of a charged species
is proportional to the electric field strength (E ) V/ L, where V is
the separation voltage and L is the column length). The propor-
tionality constant is called electrophoretic mobility (µep, cm2/
V‚s) and is determined by the following equation:
µep ) ν / E ) q/ 6πηr
(1)
ep
where q is the analyte’s net charge, η is the viscosity of the BGE,
and r is the analyte’s ionic radius. From eq 1, it is apparent that
small, highly charged species generally have high electrophoretic
mobilities while large, minimally charged species generally have
low electrophoretic mobilities. Table 1 summarizes some of the
important properties of the 20 standard protein amino acids
including their molecular weights, pK values, and pI values.24 As
shown in Table 1, all amino acids have a pI > 2. Therefore, with
proportional to the relative concentrations of the different charged
species at the droplet surface.25,26 This partitioning is influenced
(
24) Lehninger, A. L.; Nelson, D. L.; Cox, M. M. Principles of Biochemistry; Worth
Publishers: New York, 1993.
(25) Enke, C. G. Anal. Chem. 1 9 9 7 , 69, 4885-4893.
1510 Analytical Chemistry, Vol. 75, No. 6, March 15, 2003

Fig u re 2 . CE/ESI-MS base peak electropherogram (top panel) of
the separation of the 21-amino acid mixture using the 20-µm-i.d., 130-
cm-long CE capillary. To shorten the analysis time, the separation
voltage of 30 kV was augmented with 5 psi of inlet pressure. As is
shown, L-canavanine (Can) was baseline separated from arginine (R).
Middle and bottom panels show mass spectra of arginine and
canavanine, respectively.
Fig u re 3 . (A) Amino acid analysis of the blood afflicted with PKU
and its comparison to that of a healthy individual (inset). Analysis
preformed using a 20-µm-i.d., 115-cm-long CE capillary. (B) Amino
acid analysis of the blood of an infant afflicted with tyrosinemia and
its comparison to that of a healthy individual (inset). Analysis
preformed using a 20-µm-i.d., 115-cm-long CE capillary. To shorten
the analysis time, in both experiments, the separation voltage of 30
kV was augmented with 10 psi of inlet pressure.
by a number of properties of the ions and the solvent including
solvation energy, ion-pairing energy, charge density, and hydro-
phobicity.26 This is consistent with our data in which a relatively
high ESI response was observed for amino acids that have highly
hydrophobic areas such as aromatic rings, and a relatively low
ESI response was observed for aliphatic amino acids with small
side chains (or no side chain in the case of Gly).
There are more than 900 amino acids, in addition to the 20
standard protein amino acids, that exist in nature (mostly in plants)
and are not typically incorporated in proteins. The ingestion of
some of these non-protein amino acids can result in acute and
chronic poisoning or can result in the incorporation of non-protein
amino acids into proteins, thereby interfering with protein
the separation to reduce analysis time. As shown in Figure 2, the
use of long and narrow capillaries provides high separation
efficiency; however, a disadvantage of using long capillaries is long
analysis times.
Blood Analysis and Metabolic Disorder Identification. The
utility of CE/ ESI-MS in the analysis of real-world samples was
demonstrated by the detection of abnormal levels of amino acids
in the blood of two newborns afflicted with metabolic disorders
(one with PKU and one with tyrosinemia). In this study, the
relative intensities of glutamine and phenylalanine (PKU) and
glutamine and tyrosine (tyrosinemia) were compared in the blood
of healthy versus diseased infants. Figure 3 shows significant
elevation of phenylalanine and tyrosine, respectively, in the blood
of the diseased infants. In Figure 3, the migration times were
normalized to the migration time of glutamine in the normal
infant’s blood sample. The results demonstrate that this technique
can be applied to the qualitative analysis of amino acids in real-
world samples at low concentrations and low volumes. Virtually
no sample preparation (other than dilution) was required for these
experiments.The use of a narrow capillary in conjunction with
dilute blood minimizes memory effects and prevents column
overload. It should be noted that the data presented are only
qualitative and simply compare the relative amounts of the amino
acids present in the blood. These analyses would have to be
structure and function. One such amino acid is L-canavanine (see
inset of Figure 2 for chemical structure), whose accidental
incorporation into proteins can disrupt a broad range of funda-
mental life processes.27 The separation and detection of these non-
protein amino acids within a mixture of standard protein amino
acids is therefore an important task. To demonstrate the selectivity
of CE/ ESI-MS in the detection of a non-protein amino acid within
a complex amino acid mixture, a 21-amino acid mixture containing
the 20 standard protein amino acids plus the non-protein amino
acid
was completely separated from its close analogue
though 5 psi of pressure had been applied at the CE inlet during
L-canavanine was analyzed (Figure 2). As shown,
L-canavanine
L-arginine, even
(
(
26) Sjoberg, Per J. R.; Bokman, C. F.; Bylund, D.; Markides, K. E. Anal. Chem.
0 0 1 , 73, 23-28.
27) Rubenstein, E. Medicine 2 0 0 0 , 79, 80-89.
2
Analytical Chemistry, Vol. 75, No. 6, March 15, 2003 1511

ring system (C* represents the chiral center of the amino acid).
The carboxyl group substituents of 18-C-6-TCA allow for enantio-
recognition. They themselves lie perpendicular to the plane
defined by the polyethylene ring system and differentially impede
9
complexation through stereospecific steric hindrance. Thus,
D
-
and
L-amino acid enantiomers have different affinities for com-
plexation with 18-C-6-TCA.
Figure 4 demonstrates the separation of 11 amino acid
enantiomers using a 30 mM 18-C-6-TCA solution as the back-
ground electrolyte/ chiral selector reagent. Approximately 2 nL
of the amino acid enantiomer solution (∼500 fmol of each amino
acid) was injected. As is shown (Figure 4), almost all
D/ L-
enantiomers are baseline separated. The 30 mM solution of 18-
C-6-TCA was found to be compatible with ESI and suitable as the
BGE without the need for additional formic acid. However,
compared to the absolute intensities of the amino acids analyzed
under 1 M formic acid (Figure 1), the absolute intensities of the
amino acids analyzed under 30 mM 18-C-6-TCA (Figure 4) were
on the average ∼4 times lower (after normalization with respect
to the amount injected in each experiment). To our knowledge,
this is the first demonstration of enantiomeric separation of
underivatized amino acids using CE/ ESI-MS. However, the on-
line enantiomer separation of three pharmaceutical compounds
with structures similar to that of the standard amino acids (all
containing amine groups located near a chiral center) has been
achieved using sheath-flow CE/ ESI-MS in conjunction with various
cyclodextrin derivatives.32 Comparison between the results ob-
tained in these studies (Figure 4 of this article and Figures 5-7
of ref 32) demonstrates both sensitivity (after calculating the
approximate amount of sample injected in each study) and
resolution advantages of using 18-C-6-TCA in conjunction with
sheathless CE/ ESI-MS. Moreover, the technique presented here
was able to separate 11 pairs of amino acid enantiomers in one
run under identical BGE/ chiral selector reagent conditions (30
mM 18-C-6-TCA), while the use of different chiral selector reagents
Fig u re 4 . Selected ion electropherogram of the separation of 11
amino acid enantiomers using a 30 mM 18-C-6-TCA solution as the
BGE/chiral selector reagent and a 20-µm-i.d., 130-cm-long CE
capillary. This analysis was carried out with 5 psi of inlet pressure.
This pressure was necessary to maintain stable ESI due to a lowered
electroosmotic flow under 30 mM 18-C-6 TCA conditions. Ap-
proximately 500 fmol of each amino acid was injected. Inset (top left)
shows the chemical structure of 18-C-6-TCA.
quantitatively validated, and the analysis time would have to be
significantly reduced in order for this technique to be applicable
to clinical or newborn screening laboratories.
On-Line Separation of Underivatized Amino Acid Enan-
tiomers. Another important feature of amino acid analysis is the
(
DM-â-CD and HP-â-CD) at varying concentrations (5, 15, and
0 mM) was required to efficiently separate the three pairs of
2
enantiomers.32
To determine which enantiomer (D or L) has a higher affinity
for 18-C-6-TCA, a solution containing three pairs of amino acid
enantiomers with all -enantiomers at a concentration of 1.0 mM
and all -enantiomers at a concentration of 0.5 mM was analyzed
data not shown). It was found that the peak height of the
separation and detection of
D- and L-enantiomers. One advantage
D
of CE is that many -amino acid enantiomers can be separated
D/ L
L
on-line by using a BGE that contains a chiral selector. It has
previously been found (and was experimentally verified under ESI-
MS; data not shown) that a 30 mM solution of 18-C-6-TCA provides
(
enantiomers that eluted first were approximately half that of the
peaks that eluted second. This demonstrated that in each case
the optimum resolution for the separation of amino acid enanti-
9
the
based on their differential migration time,
acids have a stronger affinity for complexation with 18-C-6-TCA
than their - counterparts.
L-enantiomer eluted first. It was therefore concluded that,
omers under CE. The separation of
D
- and
L-amino acid enanti-
D
-enantiomers of amino
omers was, therefore, studied using a 30 mM 18-C-6-TCA solution
as the chiral selector/ BGE.
L
1
8-C-6-TCA is a macrocyclic polyether ring system consisting
The on-line separation and detection of nonaromatic amino acid
enantiomers without the requirement of precolumn derivatization
of several oxygens joined by ethylene bridges. The structure of
this molecule is shown in the inset of Figure 4. The polyethylene
ring forms a cavity with the oxygens roughly forming a plane on
9
(28) Kyba, E. P.; Timko, J. M.; Kaplan, L. J.; de Jong, F.; Gokel, G. W.; Cram, D.
the inner side of the cavity. The cavity formed by the ring of
J. J. Am. Chem. Soc. 1 9 7 8 , 100, 4555.
18-C-6-TCA can form complexes with cations of suitable size. The
(
29) Sousa, L. R.; Sogah, G. D. Y.; Hoffman, D. H.; Cram, D. J. J. Am. Chem. Soc.
1 9 7 8 , 100, 4569.
ammonium cation of an amino acid can form a complex with the
+
(30) Behr, J.-P.; Girodeau, J.-M.; Heyward, R. C.; Lehn, J.-M.; Sauvage, J. P. Helv.
Chim. Acta 1 9 8 0 , 63, 2096.
polyethylene ring of 18-C-6-TCA through three NH‚‚‚O hydrogen
28-31
bonds in a tripod arrangement.
The NsC* bond lies perpen-
(
31) Behr, J.-P.; Lehn, J.-M.; Vierling, P. Helv. Chim. Acta 1 9 8 2 , 65, 1853
dicular to the plane defined by the oxygens of the polyethylene
(32) Lu, W.; Cole, R. B. J. Chromatogr., B 1 9 9 8 , 714, 69-75.
1512 Analytical Chemistry, Vol. 75, No. 6, March 15, 2003

demonstrates an important advantage of CE/ ESI-MS over previ-
ously reported methods of on-line amino acid enantiomer separa-
tion.9,19
sample pretreatment (dilution). By using a 30 mM solution of 18-
C-6-TCA, 11 amino acid enantiomers were separated and detected
with ease.
CONCLUSIONS
ACKNOWLEDGMENT
CE/ ESI-MS is a powerful technique for the analysis of
Dr. Donald Chace of Neo Gen Screening (Pittsburgh, PA) is
gratefully acknowledged for supplying the blood samples used in
this study.
underivatized amino acids. By using a nonderivatized CE capil-
lary and utilizing the split-flow technique for interfacing CE to
MS, all 20 standard amino acids plus one nonstandard amino
acid (L-canavanine) were separated and detected at low-femto-
Received for review December 3, 2002. Accepted January
mole levels under full-scan mode. Because a nonderivatized
column is used, the technique is very easy to implement. Blood
samples were analyzed for high levels of amino acids with minimal
24, 2003.
AC0263925
Analytical Chemistry, Vol. 75, No. 6, March 15, 2003 1513