Content uploaded by Eden Tareke
Author content
All content in this area was uploaded by Eden Tareke on Aug 15, 2018
Content may be subject to copyright.
Quantification of rat brain neurotransmitters and
metabolites using liquid chromatography/electrospray
tandem mass spectrometry and comparison with
liquid chromatography/electrochemical detection
y
Eden Tareke, John F. Bowyer and Daniel R. Doerge*
National Center for Toxicological Research, U.S. Food and Drug Administration, Jefferson, AR 72079, USA
Received 13 September 2007; Revised 2 October 2007; Accepted 4 October 2007
Analytical methodology based on solid-phase extraction, polar reversed-phase liquid chromato-
graphy, and electrospray tandem mass spectrometry (LC/MS/MS) with isotope dilution was devel-
oped and validated for quantifying the neurotransmitters, dopamine and serotonin, and their major
metabolites in brain tissue. Limits of detection (0.1–20 pg/mg tissue) were sufficient for analysis of
multiple neurotransmitters in rat brain regions, including parietal cortex, hypothalamus, pituitary,
substantia nigra, and striatum. Method performance was compared with contemporaneous measure-
ments using a well-established procedure based on ion-pairing reversed-phase liquid chromato-
graphy and amperometric detection. The principal advantages of the LC/MS/MS method include a
more robust sample purification procedure, an optimized chromatographic separation, and the
qualitative and quantitative assurance that comes from coeluting isotopically labeled internal
standards; however, sensitivity did not consistently improve upon that provided by amperometric
detection. This methodology may be particularly useful for applications in which simultaneous
determinations are required for drugs and their affected neurotransmitters in specific brain regions.
Published in 2007 by John Wiley & Sons, Ltd.
Methods for determining the levels of catecholamines,
serotonin (5HT), and their metabolites in brain tissue using
high-pressure liquid chromatography (HPLC) separation
have been employed for over thirty years. The most
widespread and successful application has been the coupling
of amperometric electrochemical detection (ECD) for
analysis. Although this is an older technique, its use is still
widespread in the neurosciences for determining the levels
of aromatic monoamines with over 650 papers published
within the last five years (Pub Med).
1–7
This methodology has
been continually modified over the years and is still
undergoing improvements to increase its sensitivity to
further extend its application in brain microdialysis.
8,9
Its
attributes for analyzing the aromatic monoamines and
metabolites are that it is fast because it requires minimal
sample cleanup, is inexpensive, quantitative, and fairly
selective. It is a reliable and robust technique able to
accommodate the analysis of hundreds of samples a day
through automation of steps. Some limitations include the
interference from unretained matrix components associated
with the limited sample cleanup, the use of ion-pairing
chemistry to enhance amine retention, limited ability to
accommodate changes in mobile phase composition, and the
requirement for external standard calibrations. Another
deficiency of this method is that ECD, particularly when
using a single oxidative channel, cannot positively identify
the eluting peaks. It, therefore, relies on consistent retention
times for the HPLC peaks for analyte identity.
Mass spectrometry (MS) has often been used previously
for analysis of brain neurotransmitters by coupling with both
gas chromatography (GC) and liquid chromatography (LC).
For example, LC/electrospray tandem mass spectrometry
(ES-MS/MS) was used to quantify dopamine (DA) in brain
microdialysate without sample purification
10
and serotonin
(5HT), 5-hydroxyindole-3-acetic acid (5HIAA), and other
tryptophan metabolites have been quantified by using GC/
MS after extensive sample cleanup and derivatization.
11
LC
methods based on fluorogenic derivatization of norepi-
nephrine, serotonin, and dopamine for analysis of brain
microdialysates have also been used extensively (recently
reviewed by Yoshitake et al.
12
). However, in no known
previous studies have multiple classes of neurotransmitters
RAPID COMMUNICATIONS IN MASS SPECTROMETRY
Rapid Commun. Mass Spectrom. 2007; 21: 3898–3904
Published online in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/rcm.3295
*Correspondence to: D. R. Doerge, National Center for Toxicologi-
cal Research, U.S. Food and Drug Administration, Jefferson, AR
72079, USA.
E-mail: daniel.doerge@fda.hhs.gov
y
This article is a U.S. Government work and is in the public
domain in the U.S.A.
Contract/grant sponsor: Interagency Agreement between the
National Center for Toxicological Research/U.S. Food and Drug
Administration and the National Institute for Environmental
Health Sciences/National Toxicology Program; contract/grant
number: 224-93-0001.
Published in 2007 by John Wiley & Sons, Ltd.
and their metabolites been quantified using LC/MS/MS,
particularly in different brain regions.
The present study was undertaken to couple solid-phase
extraction (SPE) cleanup with a high efficiency reversed-
phase HPLC separation and ES-MS/MS as a way of mini-
mizing sample interferences while using the selectivity of
multiple reaction monitoring (MRM) to identify eluting
components. In this study we have evaluated the possible
benefits of using sensitive, precise, accurate, and specific LC/
ES-MS/MS methodology with isotope dilution for quanti-
fication of selected aromatic monoamine neurotransmitters
and metabolites in several regions of the rat brain. The
analytes selected included the neurotransmitters DA,
5HT, and their important MAO-derived metabolites, 3,4-
dihydroxyphenylacetic acid (DOPAC) and 5HIAA, respect-
ively, as well as the COMT-derived metabolites of DA,
5-methoxytyramine (3MT) and homovanillic acid (HVA)
(Fig. 1). The brain regions analyzed included the striatum,
parietal cortex, and hypothalamus so that tissues with the
maximal range in DA levels could be used to evaluate the
LC/ES-MS/MS technique.
EXPERIMENTAL
Reagents
Unlabeled neurotransmitters and metabolites, including DA
hydrochloride, 3MT hydrochloride, HVA, DOPAC (98%),
5HT, 5HIAA (98%), and epinephrine, were obtained from
Sigma Chemical Co. (St. Louis, MO, USA). Isotopically labeled
neurotransmitters and metabolites were acquired commer-
cially as indicated: 2-(3,4-dihydroxyphenyl-
13
C
6
)ethylamine
HCl (99 atom %
13
C; CDN Isotopes Inc., Pointe-Claire, Quebec,
Canada); 2-(4-hydroxy-3-methoxylphenyl)ethyl-10,10,20,20-D
4
(98 atom % D, CDN Isotopes); 4-hydroxy-
18
O-3-methoxy-
phenyl-
13
C
6
-3-acetic acid (99 atom %
13
C, 99 atom %
18
O; Cambridge Isotope Laboratories, Andover, MA, USA);
3,4-dihydroxyphenyl-D3-3-acetic-2,2-D2-acid (98 atom % D;
Cambridge Isotope Laboratories); 5-hydroxytryptamine-
a,a,b,b-d4 creatinine sulfate complex (99 atom %; CDN
Isotopes); 5-hydroxyindole-3-acetic-20,20-D2-acid (98 atom % D;
CDN Isotopes); and DL-epinephrine a,a,b-d
3
(97 atom % D;
Cambridge Isotope Laboratories). All other solvents were
of analytical grade and Milli-Q water was used for the study.
Standard stock solutions were prepared by accurate
weighing from each unlabeled compound (1 mg/mL in
methanol as the free base). The labeled stock solutions were
similarly prepared and the concentrations were determined
using HPLC-UV (254 nm) and electrochemical detection
(ECD) using the respective unlabeled compound as an
external standard. Diluted working solutions of the un-
labeled and labeled standards were prepared at 10 ng/mLin
aqueous 0.1% formic acid and stored frozen at 808C. All
compounds were determined to be stable for at least 1 year
using these storage conditions. For every sample set,
solutions of labeled and unlabeled standards were prepared
fresh in 0.1% formic acid at 1 ng/mL. All labeled standards
were analyzed by full scan ES-MS to determine that there
was no detectable contamination by unlabeled analytes.
Animal handling procedures
Procedures involving care and handling of rats were
reviewed and approved by the NCTR Laboratory Animal
Care and Use Committee. Male F344 rats were obtained from
the NCTR colony at weaning on postnatal day 21 and
maintained on basal diet and drinking water until approxi-
mately 84 days of age. Rats were sacrificed by decapitation,
brains were rapidly removed, placed in ice-cold buffer, brain
regions were dissected and frozen on dry ice using
procedures previously described,
13
and stored at 808C
prior to analyses. Ice-cold 0.1% M formic acid in H
2
O was
added to 50–100 mg portions of brain tissue weighed frozen
in the proportion of 10 mL/mg tissue. Internal standards
were added at 1 ng each per mg tissue. The tissues were then
homogenized on ice by using ultrasonication for 30 s and
centrifuged at 13 000 rpm 48C for 7 min. The neurotransmit-
ters were extracted from 200 mL aliquots of the supernatant
using SPE as described below.
Figure 1. Structures of selected neurotransmitters and metabolites.
Published in 2007 by John Wiley & Sons, Ltd. Rapid Commun. Mass Spectrom. 2007; 21: 3898–3904
DOI: 10.1002/rcm
LC/MS/MS analysis of brain neurotransmitters 3899
Solid-phase extraction
Strata X polymeric reversed-phase cartridges (33 mm, 60 mg,
3 mL; Phenomenex Co., Torrence, CA, USA) were used for
extraction of neurotransmitters from 200 mL aliquots. The
cartridge was washed with 1 1 mL of 0.1% formic acid in
acetonitrile and 1 1 mL methanol and equilibrated using
11mLH
2
O. To brain homogenates were added 1 ng each
of the internal standards at the level of 1 ng of each internal
standard per mg of tissue. Samples were loaded onto the SPE
cartridges and eluted with 4 300 mL of 0.1% formic in 50%
acetonitrile in methanol. The elutes were dried using a
centrifugal vacuum concentrator without applied heat and
reconstituted in 200 mL of 0.1% formic acid in H
2
O for
analysis.
Liquid chromatography
The LC separations were performed on an Alliance 2795
system (Waters Corp., Milford, MA, USA) using either an
Atlantis column dC18 (3 mm particle size, 2.1 150 mm;
Waters) or an Aquasil C18 column (5 mm particle size,
2.1 150 mm; Thermo Hypersil-Keystone, Bellefonte, PA,
USA) at a flow rate of 0.2 mL/min. The mobile phase
consisted of aqueous 0.1% formic acid and acetonitrile.
Samples were eluted with a linear gradient from 0–4%
acetonitrile in 2 min and then stepped to 50% acetonitrile at
8 min for an additional 2 min before equilibrating the column
with the initial conditions for an additional 5 min. Total run
time was 15 min. The injection volume of 100 mL (5–10 mg
equivalent) was the maximum possible without producing
unacceptable peak broadening.
Mass spectrometry
A Quattro Micro triple quadrupole mass spectrometer
(Waters Corp., Milford, MA, USA) equipped with an ES
interface was used with a source block temperature of 1008C
and desolvation temperature of 4008C. Nitrogen gas was
used as desolvation gas (750 L/h). Argon was used as
collision gas, at a collision cell pressure of 4.5 10
3
mbar.
Resolution was set to give peak widths at half-height of
0.9 Th for product and precursor ions. The base peak in each
compound spectrum was either the protonated molecule
[MþH]
þ
or the deprotonated molecule, [MH]
, which were
subsequently used as precursor ions for the respective
positive and negative ion multiple-reaction monitoring
(MRM) transitions. The MS acquisition was split into two
timed scan functions in order to maximize sensitivity of
detection: 0–9.6 min and 8.9–17 min. Similar sensitivity
requirements constrained the acquisitions to a single product
ion transition for each analyte. Positive ion transitions were
acquired in MRM mode using the collision energies and cone
voltages during the two time windows and negative ion
transitions were also acquired in the second time window as
shown in Table 1. Fraction 1 for positive ions used a retention
window of 1–9.6 min and a dwell time of 0.07 s. Fraction 2
used a retention window for positive and negative ions of
8.9–15 min and a dwell time of 0.2 s.
Calibration
Responses for varying amounts of unlabeled neurotrans-
mitter standards (0.01–1 ng) on-column for DA, 5HT and
0.01–10 ng on-column for 5HIAA, 3MT, HVA, and DOPAC)
were plotted against a fixed amount of the corresponding
labeled internal standard (0.5 ng for DA, 3MT, and 5HT and
5 ng for 5HIAA, DOPAC, and HVA). The plots were linear
(r
2
>0.99 for all analytes) and the response factors were 1.5,
1.13, 1.06, 1.09, 0.75, and 0.92 for DA, 5HT, 5HIAA, 3MT,
HVA, and DOPAC, respectively. The use of double labeled
5HIAA internal standard required that a correction of 0.85%
be made for the contribution of Mþ2 ions from unlabeled
5HIAA into the labeled channel (m/z 194 >148). For quality
control purposes, blanks and a mixed standard containing all
labeled and unlabeled analytes were run after every four
samples.
LC/ECD procedures
Thawed samples of brain tissue were homogenized on ice in
0.2 M perchloric acid using ultrasonication for 30 s, centri-
fuged for 10 min at 2000 g, passed through a 0.2 micron filter,
and 20 mL aliquots were injected directly onto the HPLC
system similar to that described by Stephens et al.
8
For the
split-sample investigation of LC/MS/MS vs. LC/ECD, the
tissue was homogenized in ice-cold 0.1% formic acid in
water. The mobile phase consisted of 40 mM sodium acetate,
25 mM citric acid, 1.5 mM octanesulfonic acid, 1.5 mM EDTA,
and 6% methanol at a pH of 3.8. A Supelcosil LC18 analytical
column (Supelco, Bellefonte, PA, USA) was used for
separation and a BAS-LC4B amperometric detector with a
BAS-LC-17 oxidative flow cell was used for detection with
the working electrode potential set at 0.65 V. Analytes were
quantified by comparing detector responses in sample with
those produced by external standards prepared in mobile
phase.
RESULTS
Method validation and performance
The LC/MS/MS method was validated by adding 10 mLof
1 ng/mL solution with labeled and unlabeled standards to
quadruplicate portions of 10 mg of parietal cortex sample
giving a concentration of 1 ng/mg tissue. Cortex tissue was
chosen because of its relatively low endogenous concen-
Table 1. Mass spectral parameters used for analysis of neu-
rotransmitters
Analyte
MRM transition
(m/z)
Cone
voltage (V)
Collision
energy
(eV)
Retention
time (min)
DA 153.9 >136.9 13.0 10.0 3.65
DA (
13
C
6
) 159.9 >142.9 13.0 10.0 3.68
3MT 167.9 >151.0 12.0 10.0 5.70
3MT (D4) 171.9 >154.9 12.0 10.0 5.60
5HT 176.9 >159.9 10.0 10.0 6.32
5HT (D4) 180.9 >163.9 10.0 10.0 6.26
5HIAA 191.9 >145.9 15.0 15.0 14.86
5HIAA (D2) 193.9 >147.9 17.0 17.0 14.83
DOPAC 166.9 >123.0 10.0 10.0 12.86
DOPAC (D5) 171.9 >128.0 10.0 10.0 12.73
HVA 180.9 >137.9 17.0 7.0 15.41
HVA (
13
C
6
,
18
O) 188.9 >145.9 17.0 7.0 15.41
Published in 2007 by John Wiley & Sons, Ltd. Rapid Commun. Mass Spectrom. 2007; 21: 3898–3904
DOI: 10.1002/rcm
3900 E. Tareke, J. F. Bowyer and D. R. Doerge
trations of 5HT and its very low concentrations of DA. For all
determinations of spiked samples, the endogenous levels of
all neurotransmitter analytes determined in blank tissues
were subtracted. The homogenized cortex samples were then
purified using SPE and analyzed by LC/MS/MS on two
separate occasions. Interday and intraday variability in
analyte recoveries from spiked cortex were determined by
comparing SPE-processed spiked samples with those in
which the standards were fortified after SPE (Table 2).
Interday and intraday variability in suppression was
determined by comparing responses for neat standards
with those from fortified cortex homogenates (i.e., after SPE)
and was <10% for all analytes on day 1 and <12% on day 2.
Interday and intraday variability in method accuracy,
determined using quadruplicate parietal cortex samples
spiked with 1.0 ng/mg DA, 5HIAA, 3MT, HVA, 5HT, and
DOPAC, is shown in Table 2. Interday and intraday
variability in method precision values from the quadrupli-
cate analyses (relative standard deviations) were determined
to be: <5% for DA, 5HIAA, 3MT, HVA and 5HT, and 6% for
DOPAC on both days (Table 2).
A bridging study was conducted to directly compare the
performance of the LC/MS/MS method with LC/ECD.
8,9
Striatum samples were collected from saline-treated and
amphetamine-treated male Sprague-Dawley rats as pre-
viously described,
13
homogenized as described in the
Experimental section, and split into equal portions. One
portion was processed for LC/MS/MS analysis as described
in the Experimental section and the other portion was
immediately frozen at 808C for further analysis using LC/
ECD (Table 3). In addition, replicate striatum samples from
untreated male Fischer 344 rats were analyzed for these
neurotransmitters and metabolites using both LC/MS/MS
and LC/ECD (Table 4). The retention times for the LC/ECD
procedure were 4.49 min (DOPAC); 7.15 min (DA); 9.25 min
(5HIAA); 12.75 min (HVA); 20.96 min (5HT); and 21.90 min
(3-MT); the respective retention times for the LC/MS/MS
method using the Atlantis column were: 12.73 min (DOPAC);
3.68 min (DA); 14.80 min (5HIAA); 15.43 min (HVA);
6.32 min (5HT); and 5.70 min (3-MT) (Fig. 2), and the
retention times for the LC/MS/MS method using the
Aquasil column were: 10.10 min (DOPAC); 3.60 min (DA);
10.91 min (5HIAA); 11.12 min (HVA); 8.17 min (5HT);and
7.98 min (3-MT) (data not shown). The quantitative results
from these comparative analyses are shown in Tables 3 and 4.
While many measurements were significantly different
(p<0.05 using the two-tailed t-test), the values determined
for DA, 3MT, HVA, and 5HT were generally comparable
between the two methods for both sample comparisons
made; however, the values determined using LC/MS/MS
were consistently much higher for 5HIAA and DOPAC.
The neurotransmitter/metabolite levels were also deter-
mined in several other untreated male F344 rat brain regions
using LC/MS/MS (Table 5). These tissues were selected to
demonstrate method performance over a physiological range
of dopaminergic innervation in different brain regions. These
tissues include the cortex with very low DA levels, the
striatum with high levels, and the hypothalamus, pituitary,
and substantia nigra, with intermediate levels.
The sensitivity of the LC/MS/MS method was estimated
using parietal cortex because this tissue had the lowest levels
of most neurotransmitters of all brain regions examined. The
limited amounts of cortex tissue available limited the
validation study to a single spiking study using a level
(1 ng/mg) in the middle of expected brain region concen-
trations expected for these neurotransmitters and metab-
olites. The signal-to-noise (S/N) ratios for the endogenous
neurotransmitters and metabolites were determined and the
limits of detection (LODs) were estimated as those levels that
would produce an S/N ratio of 3 (Table 6). In general, the
LODs were lowest for the positive ion analytes (0.1–3 pg/mg
tissue for DA, 3MT, 5HT, 5HIAA) but significantly greater for
the later eluting negative ions, DOPAC and HVA (10–20 pg/mg
tissue). For an injection of 10 mg equivalents, the detectable
levels of analytes would be 1–30 pg on-column. The
sensitivity of the LC/MS/MS method did not consistently
improve upon that previously reported from the LC/ECD
method where LODs ranged from 0.7–4 pg/mg tissue.
14
Table 2. Method validation parameters for analysis of cat-
echolamines in control cortex. Control cortex was spiked with
neurotransmitters and metabolites at 1 ng/mg tissue and
analyzed on two separate days to determine accuracy and
precision. The neurotransmitter values shown were corrected
for the endogenous concentrations. Intraday and interday
precision values are expressed as RSD (relative standard
deviation)
Analyte
Recovery
(Day 1/Day 2, %)
Accuracy
(Day 1/Day 2, %)
Precision
(RSD, %)
DA 107/98 125/115 <5
5HIAA 113/123 93/87 <5
3MT 102/106 94/91 <5
HVA 97/98 78/79 <5
5HT 109/106 97/103 <5
DOPAC 79/81 102/129 6
Table 3. Neurotransmitter and metabolite concentrations in saline- and amphetamine-treated male Sprague-Dawley rat striatum
determined using either LC/MS/MS or LC/ECD. The values, expressed in as mean SD in units of ng/mg tissue, were determined
as indicated using n ¼4 brains.
Significantly different from respective ECD measurement by two-tailed t-test (p<0.05).
Significantly different from respective amphetamine-treated striatum measurement by two-tailed t-test ( p<0.05)
Striatum DA 5HIAA 3MT HVA 5HT DOPAC
Saline-treated Sprague-Dawley (LC/MS/MS)
18 2.9 1.1 0.13
0.19 0.08
0.64 0.10 0.55 0.06
2.3 0.19
Amphetamine-treated Sprague-Dawley (LC/MS/MS) 8.4 1.6 1.0 0.11
0.15 0.04 0.44 0.07
0.54 0.06
1.3 0.13
Saline-treated Sprague-Dawley (LC/ECD)
17 2.7 0.11 0.02 0.23 0.15 0.61 0.10 0.30 0.02
0.89 0.14
Amphetamine-treated Sprague-Dawley (LC/ECD) 9.5 1.5 0.14 0.03 0.18 0.09 0.60 0.02 0.34 0.04 0.56 0.05
Published in 2007 by John Wiley & Sons, Ltd. Rapid Commun. Mass Spectrom. 2007; 21: 3898–3904
DOI: 10.1002/rcm
LC/MS/MS analysis of brain neurotransmitters 3901
DISCUSSION
The described LC/MS/MS methodology incorporates sig-
nificant advantages over the traditional LC/ECD method,
which employs minimal sample preparation (e.g., protein
precipitation, centrifugation, and filtration), minimal sample
loading capability (1–2 mg equivalents on-column maxi-
mum), an ion-pairing LC separation that selectively retains
the cationic analytes, and amperometric detection whose
oxidative potential is optimized for the catechol moiety. By
contrast, the LC/MS/MS method uses an efficient solid-
phase extraction to purify and concentrate sample extracts
without observable analyte losses from artifactual oxidation,
is insensitive to sample injection volume up to 10 mg
equivalent on-column, uses a polar reversed-phase LC
column to retain and separate cationic and acidic metab-
olites, permits specific mass spectrometric detection using
MRM transitions, and incorporates coeluting isotopically
labeled internal standards. The use of labeled internal
standards is critical to the enhanced method performance
because the isotope ratio measurements provide a measure
of quality control for each analyte in every sample by
compensating for changes in analyte retention time,
recovery, degradation, and changes in detector responses
caused by coeluting contaminants (i.e., suppression). No
examples of a similarly comprehensive quantitative analysis
of neurotransmitters and metabolites in brain regions using
either LC/MS or GC/MS techniques were present in the
literature tracked by PubMed from 1987 to the present.
Analysis of additional important brain neurotransmitters,
the more polar epinephrine and nor-epinephrine, was
initially investigated. Although the detection sensitivity
was similar to the other cationic analytes (Table 6), the overall
performance was deemed unacceptable because of the
chromatographic separation. The minimal retention of
epinephrine and nor-epinephrine under these polar reversed-
phase LC conditions led to large and variable suppression of
MS signals, resulting in unacceptable accuracy and precision
even though an isotopically labeled standard was available
for epinephrine. Future LC/MS/MS method enhancements
would ideally include these two important neurotransmit-
ters, probably through use of alternate chromatographic
specificity and synthesis of suitably labeled nor-epinephrine.
It should be noted that the ion-pairing chromatography used
for the LC/ECD method is also problematic in providing
adequate retention of epinephrine and nor-epinephrine
away from matrix interferences necessary for routine
quantification along with the later eluting analytes. In
addition, the use of typical gradient elution methods
necessary to accommodate early- and late-eluting analytes
is precluded with LC/ECD because of continually changing
background responses.
The LC/MS/MS methodology was used to measure the
levels of neurotransmitters and metabolites in five brain
regions including cortex, hypothalamus, substantia nigra,
pituitary, and striatum from untreated Fischer 344 rats
(Table 5). The levels of DA, HVA, and 5HT reported for
striatum and hypothalamus in Table 5 are similar to those
previously reported from LC/ECD measurements in Fischer
344 rat brain regions.
15,16
The LC/MS/MS method provided
good precision of measurements over a wide physiological
range of DA concentrations in these selected brain regions
(i.e., from 0.01 ng/mg in cortex to 10.4 ng/mg in the
striatum). The levels of DOPAC and 5HIAA from the LC/
MS/MS method are consistently higher as discussed above.
While the labile nature of 5HIAA makes its accurate and
precise determination difficult for any detector, this problem
should be more acute for the LC/ECD method because it
does not incorporate an internal standard that can compen-
sate for artifactual changes in concentration (i.e., oxidation
during sample workup). Similarly, the ion-pairing chromato-
graphic separation used in the LC/ECD method causes
DOPAC to be the first peak to elute in the region where
coeluting interferences from unretained tissue components
would be expected to be more significant than for analytes
with greater retention; however, the polar reversed-phase
separation used in the LC/MS/MS method significantly
retains DOPAC and its internal standard into a later region of
the chromatogram that is less susceptible to interferences. It
is therefore likely that the systematic bias to higher brain
concentrations measured for 5HIAA and DOPAC in tissues
using the LC/MS/MS method is the result of more selective
and reliable method performance and not from an artifact.
The performance of the LC/MS/MS method vis a vis the
LC/ECD method was tested using sample sets comprised of
treated and control brains using two different rat strains (i.e.,
Sprague-Dawley, Table 3; Fischer 344, Table 4). For another
experiment comparing amphetamine- and saline-treated
male Sprague-Dawley rat striatum (Table 3), the LC/MS/MS
results recapitulated the DA-depleting action of amphet-
amine in the major dopaminergic brain region,
13,14
as
reflected by significant decreases in DA, HVA, and DOPAC.
By comparison, only DA and DOPAC were significantly
different using the LC/ECD data (Table 3). Similarly, the lack
of an amphetamine effect on serotonergic neurons in the
striatum
13
was reflected by no significant changes in 5HT or
5HIAA levels from either LC/MS/MS or LC/ECD data
(p>0.05 using the two-tailed t-test, see Table 3).
Table 4. Neurotransmitter and metabolite concentrations in untreated male F344 rat striatum determined using either LC/MS/MS
or LC/ECD. The values, expressed as mean SD in units of ng/mg tissue, were determined as described using n ¼6 or 10 brains
(LC/ECD and LC/MS/MS, respectively).
Significantly different from respective ECD measurement by two-tailed t-test ( p<0.05)
Striatum DA 5HIAA 3MT HVA 5HT DOPAC
LC/MS/MS 10.42.1 1.0 0.23
0.19 0.09
0.68 0.27
0.73 0.16 3.8 1.8
LC/ECD 11.8 0.63 0.23 0.02 0.13 0.04 0.54 0.06 0.65 0.08 0.77 0.04
Published in 2007 by John Wiley & Sons, Ltd. Rapid Commun. Mass Spectrom. 2007; 21: 3898–3904
DOI: 10.1002/rcm
3902 E. Tareke, J. F. Bowyer and D. R. Doerge
FORTIFIED CORTEX
5.00 10.00 15.00
Time
0
100
%
0
100
%
0
100
%
0
100
%
0
100
%
0
100
%
0
100
%
0
100
%
0
100
%
0
100
%
0
100
%
0
100
%
188.9 > 145
1.00e5
180.9 > 137
8.54e4
171.9 > 128
8.24e4
166.9 > 123
1.06e5
193.9 > 147.9
8.51e5
191.9 > 145.9
9.67e5
180.9 > 163.9
5.05e6
6.26
497361
176.9 > 159.9
7.57e6
6.32
696923
171.9 > 154.9
8.43e6
5.60
904401
167.9 > 151
7.59e6
5.70
911236
159.9 > 142.9
9.42e6
153.9 > 136.9
9.45e6
CONTROL CORTEX
5.00 10.00 15.00
Time
0
100
%
0
100
%
0
100
%
0
100
%
0
100
%
0
100
%
0
100
%
0
100
%
0
100
%
0
100
%
0
100
%
0
100
%
188.9 > 145
9.59e4
180.9 > 137
6.05e3
171.9 > 128
6.84e4
166.9 > 123
1.58e3
12.73
68
193.9 > 147.9
7.52e5
191.9 > 145.9
5.28e5
180.9 > 163.9
3.99e6
6.26
437157
176.9 > 159.9
1.98e6
6.32
197241
171.9 > 154.9
9.12e6
5.63
1014775
167.9 > 151
3.00e4
159.9 > 142.9
6.19e6
153.9 > 136.9
1.60e5
3.65
450328
517079
3.65
14.86
14.83
12.86
12.73
15.41
15.41
76865
75316
12080
9694
5961
6696
15.43
6082
12.60
8096
52958
14.77
10049
14.80
15.43
43
5.70
668
381812
8322
3.68
3.68
DA
DA-IS
3MT-IS
5HT
5HT-IS
3MT
5HIAA
5HIAA-IS
DOPAC
DOPAC-IS
HVA
HVA-IS
Figure 2. Analysis of dopamine, serotonin, and metabolites in rat parietal cor tex. Cortex tissue was analyzed using LC/MS/MS
as described in the Experimental section. Right: MRM chromatograms of endogenous neurotransmitters and metabolites from
cortex tissue to which was added only labeled internal standards (1 ng/mg tissue each). Left: MRM chromatograms from
fortified cortex tissue to which was added unlabeled and labeled internal standards (1 ng/mg tissue each).
Published in 2007 by John Wiley & Sons, Ltd. Rapid Commun. Mass Spectrom. 2007; 21: 3898–3904
DOI: 10.1002/rcm
LC/MS/MS analysis of brain neurotransmitters 3903
CONCLUSIONS
This paper describes a sensitive and specific LC/ES-MS/MS
method for the simultaneous quantification of neurotrans-
mitters and metabolites from brain tissue. The principal
advantages for using the LC/MS/MS method derive from a
more robust sample purification procedure, an optimized
chromatographic separation, and the quantitative and con-
firmatory assurance that comes from coeluting isotopically
labeled internal standards. This technique, like the classical
technique using LC with ECD, provides adequate sensitivity
for the quantification of DA, serotonin, and their metabolites
in dissected rat brain regions. The precision, accuracy, and
consistency of LC/MS/MS results obtained using the
demonstrated method suggest that the significant differences
observed in levels of brain region neurotransmitters
observed vis a vis contemporaneous LC/ECD measurements
should be seriously considered. Finally, the ready incorp-
oration of additional analytes into an MRM acquisition
confers additional utility to the LC/MS/MS method vis a vis
LC/ECD by facilitating concurrent quantification of drugs
levels and the affected neurotransmitters in specific brain
regions (e.g., DA and cocaine in rat nucleus accumbens
microdialysate
10
).
Acknowledgements
This research was supported in part by Interagency Agree-
ment #224-93-0001 between the National Center for Toxico-
logical Research/U.S. Food and Drug Administration and
the National Institute for Environmental Health Sciences/
National Toxicology Program. ET acknowledges support of a
fellowship from the Oak Ridge Institute for Science and
Education, administered through an interagency agreement
between the U.S. Department of Energy and the U.S. Food
and Drug Administration. The views presented in this article
do not necessarily reflect those of the U.S. Food and Drug
Administration.
REFERENCES
1. Van der Gugten J, Palkovits M, Wijnen HL, Versteeg DH.
Brain Res. 1976; 107: 171.
2. Palkovits M, Saavedra JM, Jacoboqitz DM, Kizer JS,
Zaborszky L, Brownstein MJ. Brain Res. 1977; 130: 121.
3. Saraswat LD, Holdiness MR, Justice JB, Salamone JD, Neill
DB. J. Chromatogr. 1981; 222: 353.
4. Sasa S, Blank CL. Anal. Chem. 1977; 49: 354.
5. Sasa S, Blank CL, Wenke DC, Sczupak CA. Clin. Chem. 1978;
24: 1509.
6. Westerink BH, Korf J. Eur. J. Pharmacol. 1976; 38: 281.
7. Westerink BH, Korf J. Brain Res. 1976; 113: 429.
8. Stephans SE, Whittingham TS, Douglas AJ, Lust WD, Yama-
moto BK. J. Neurochem. 1998; 71: 613.
9. Stephans SE, Yamamoto BK. Synapse 1994; 17: 203.
10. Hows MEP, Lacroix L, Heidbreder C, Organ AJ, Shah AJ.
J. Neurosci. Methods 2004; 138: 123.
11. Markey SP, Boni RL, Yergey JA, Heyes MP. Adv. Exp. Med.
Biol. 1991; 294: 41.
12. Yoshitake T, Kehr J, Todoroki K, Nohta H, Yamaguchi M.
Biomed. Chromatogr. 2006; 20: 267.
13. Bowyer JF, Harris AJ, Delongchamp RR, Jakab RL, Miller
DB, Little AR, O’Callaghan JP. Neurotoxicology 2004; 25:
555.
14. Tor-Agbidye J, Yamamoto B, Bowyer JF. Toxicol. Sci. 2001; 60:
103.
15. Ali SF, Hong JS, Wilson WE, Uphouse LL, Bondy SC. Arch.
Toxicol. 1983; 52: 35.
16. Ali SF. Toxicol. Lett. 1983; 17: 101.
Table 5. Neurotransmitter and metabolite concentrations in untreated male F344 rat brain regions determined using LC/MS/MS.
The values, expressed as mean SD in units of ng/mg tissue, were determined as described for the different rat brain regions
using n ¼10 brains
Brain region DA 5HIAA 3MT HVA 5HT DOPAC
Parietal cortex 0.01 0.003 0.31 0.22 <LOD 0.02 0.01 0.41 0.08 0.01 0.02
Hypothalamus 0.34 0.07 1.3 0.32 <LOD 0.04 0.04 0.82 0.16 0.35 0.10
Substantia nigra 0.41 0.13 1.44 0.17 0.01 0.001 0.08 0.03 0.51 0.12 0.38 0.13
Pituitary 0.29 0.06 0.33 0.07 0.03 0.03 0.04 0.001 0.17 0.08 0.51 0.24
Striatum 10.4 2.1 1.0 0.23 0.19 0.09 0.68 0.27 0.73 0.16 3.8 1.8
Table 6. Limits of detection (LODs) for neurotransmitter and
metabolite estimated in untreated male F344 rat parietal
cortex using LC/MS/MS and LC/ECD.
Approximate LODs
for neurotransmitters using LC/ECD cited from Tor-Agbidye
et al.
14
Analyte
LOD (pg/mg tissue)
LC/MS/MS
LOD (pg/mg tissue)
LC/ECD
DA 0.1 0.7
5HIAA 3 0.7
3MT 3 4
HVA 20 1.5
5HT 2 1.5
DOPAC 10 0.7
Published in 2007 by John Wiley & Sons, Ltd. Rapid Commun. Mass Spectrom. 2007; 21: 3898–3904
DOI: 10.1002/rcm
3904 E. Tareke, J. F. Bowyer and D. R. Doerge
