Analysis of intact glucuronides and sulfates of serotonin, dopamine, and their phase I metabolites in rat brain microdialysates by liquid chromatography-tandem mass spectrometry

Article abstract of DOI:10.1021/ac901320z

A method for the analysis of intact glucuronides and sulfates of common neurotransmitters serotonin (5-HT) and dopamine (DA) as well as of 5-hydroxy-3-indoleacetic acid (5-HIAA), 3,4-dihydroxyphenylacetic acid (DOPAC), and homovanillic acid (HVA) in rat b

Full text of DOI:10.1021/ac901320z

Anal. Chem. 2009, 81, 8417–8425
Analysis of Intact Glucuronides and Sulfates of
Serotonin, Dopamine, and Their Phase I
Metabolites in Rat Brain Microdialysates by Liquid
Chromatography-Tandem Mass Spectrometry
Pa¨ ivi Uutela,† Ruut Reinila¨ ,† Kirsi Harju,† Petteri Piepponen,‡ Raimo A. Ketola,§ and
Risto Kostiainen*,†
Division of Pharmaceutical Chemistry, Division of Pharmacology and Toxicology, and Centre for Drug Research
(CDR), Faculty of Pharmacy, P.O. Box 56, FI-00014 University of Helsinki, Helsinki, Finland
A method for the analysis of intact glucuronides and
sulfates of common neurotransmitters serotonin (5-HT)
and dopamine (DA) as well as of 5-hydroxy-3-indoleacetic
acid (5-HIAA), 3,4-dihydroxyphenylacetic acid (DOPAC),
and homovanillic acid (HVA) in rat brain microdialysates
by liquid chromatography-tandem mass spectrometry
(LC-MS/MS) was developed. Enzyme-assisted synthesis
using rat liver microsomes as a biocatalyst was employed
for the production of 5-HT-, 5-HIAA-, DOPAC-, and HVA-
glucuronides for reference compounds. The sulfate con-
jugates were synthesized either chemically or enzymati-
cally using a rat liver S9 fraction. The LC-MS/MS method
was validated by determining the limits of detection and
quantitation, linearity, and repeatability for the quantita-
tive analysis of 5-HT and DA and their glucuronides, as
well as of 5-HIAA, DOPAC, and HVA and their sulfate-
conjugates. In this study, 5-HT-glucuronide was for the
first time detected in rat brain. The concentration of 5-HT-
glucuronide (1.0-1.7 nM) was up to 2.5 times higher
than that of free 5-HT (0.4-2.1 nM) in rat brain microdi-
alysates, whereas the concentration of DA-glucuronide
(1.0-1.4 nM) was at the same level or lower than the free
DA (1.2-2.4 nM). The acidic metabolites of neurotrans-
mitters, 5-HIAA, HVA, and DOPAC, were found in free
and sulfated form, whereas their glucuronidation was not
observed.
The glucuronidation of DA2 and the sulfation of DA, HVA,
DOPAC, and 5-HIAA have been observed in brain.3-5 Sulfation
and glucuronidation are common phase II metabolic reactions
catalyzed by cytosolic sulfotransferases (SULT)6 and membrane-
bound uridine diphosphoglucuronosyltransferases (UGTs),7 re-
spectively. SULT1A1, 1A2, and 1A3,8 responsible for the sulfation
of phenols, and isoforms UGT1A6, 2A1, and 2B79 have been found
in human brain. In rat brain, the activity of UGT1A610 and phenol-
sulfating SULT11 has been observed. UGT1A6 and 2B7 catalyze
the glucuronidation of 5-HT, the former enzyme having a much
higher glucuronidation rate than the latter.12 These research
results suggest that phase II metabolism can affect the concentra-
tion levels of neurotransmitters in the brain. For example, the
altered concentrations of DA and 5-HT as well as their metabolites
are connected to various neurological disorders, such as schizo-
phrenia13 and Parkinson’s disease,14 and therefore reliable and
highly sensitive analytical methods are needed to study the role
of phase II metabolism in the human brain in more detail. Since
the availability of human microdialysis samples is limited, test
animals are commonly used in brain research.
(2) Uutela, P.; Karhu, L.; Piepponen, P.; Kaenmaki, M.; Ketola, R. A.; Kostiainen,
R. Anal. Chem. 2009, 81, 427–434
(3) Swahn, C. G.; Wiesel, F. A. J. Neural Transm. (1972-1989) 1976, 39,
281–290
(4) Gordon, E. K.; Markey, S. P.; Sherman, R. L.; Kopin, I. J. Life Sci. 1976,
18, 1285–1292
(5) Buu, N. T.; Duhaime, J.; Savard, C.; Truong, L.; Kuchel, O. J. Neurochem.
1981, 36, 769–772
(6) Weinshilboum, R. M.; Otterness, D. M.; Aksoy, I. A.; Wood, T. C.; Her, C.;
Raftogianis, R. B. FASEB J. 1997, 11, 3–14
.
.
.
.
Dopamine (DA) and serotonin (5-HT) are important neu-
rotransmitters, which chemically transfer the electrical impulse
from one neuron to another in brain. After transmission DA and
5-HT are metabolized through monoamine oxidase (MAO)- and
aldehyde dehydrogenase-catalyzed reactions to 3,4-dihydroxyphe-
nylacetic acid (DOPAC) and to 5-hydroxyindoleacetic acid (5-
HIAA), respectively.1 DOPAC can be further metabolized to
homovanillic acid (HVA) by catechol-O-methyltransferase (COMT).
.
(7) Mackenzie, P. I.; Bock, K. W.; Burchell, B.; Guillemette, C.; Ikushiro, S.;
Iyanagi, T.; Miners, J. O.; Owens, I. S.; Nebert, D. W. Pharmacogenet.
Genomics 2005, 15, 677–685
(8) Eisenhofer, G.; Coughtrie, M. W. H.; Goldstein, D. S. Clin. Exp. Pharmacol.
Physiol. 1999, 26, S41–S53
(9) Tukey, R. H.; Strassburg, C. P. Annu. Rev. Pharmacol. Toxicol. 2000, 40,
581–616, 2 plates
(10) Suleman, F. G.; Abid, A.; Gradinaru, D.; Daval, J. L.; Magdalou, J.; Minn,
A. Arch. Biochem. Biophys. 1998, 358, 63–67
(11) Rivett, A. J.; Francis, A.; Whittemore, R.; Roth, J. A. J. Neurochem. 1984,
42, 1444–1449
(12) King, C. D.; Rios, G. R.; Assouline, J. A.; Tephly, T. R. Arch. Biochem.
Biophys. 1999, 365, 156–162
(13) Filip, M.; Frankowska, M.; Zaniewska, M.; Golda, A.; Przegalinski, E.
Pharmacol. Rep. 2005, 57, 685–700
(14) Jacintho, J. D.; Kovacic, P. Curr. Med. Chem. 2003, 10, 2693–2703
.
.
.
.
* Corresponding author. E-mail: Risto.Kostiainen@helsinki.fi. Fax: +358 9 191
59556.
.
† Division of Pharmaceutical Chemistry.
‡ Division of Pharmacology and Toxicology.
.
§ Centre for Drug Research (CDR).
(1) Cooper, J. R., Bloom, F. E. Roth, R. H. The Biochemical Basis of
Neuropharmacology, 8th ed.; Oxford University Press: New York, 2003.
.
.
10.1021/ac901320z CCC: $40.75  2009 American Chemical Society
Published on Web 09/22/2009
Analytical Chemistry, Vol. 81, No. 20, October 15, 2009 8417

Glucuronide and sulfate conjugates in rat brain or cerebrospinal
fluid (CSF) have commonly been analyzed after acid or enzymatic
hydrolysis of the conjugate by liquid chromatography electro-
chemical detection (LC-EC),15-18 LC fluorescence (FL) detec-
tion,19 or gas chromatography/mass spectrometry (GC/MS).3,4
However, the identity of the conjugate cannot be confirmed using
acid hydrolysis.15,16,18,19 The glucuronides and sulfates can be
identified with higher specificity by hydrolyzing the conjugate with
specific enzymes, ꢀ-glucuronidase, or sulfatase, respectively. The
concentrations of DOPAC, HVA, and 5-HIAA in rat brain samples
clearly increased after sulfatase hydrolysis, indicating that they
are conjugated by sulfation.3,17 5-HT-S has not been detected in
rat brain with indirect methods using sulfatase17 or acid hydroly-
sis.20 However, conjugates of 5-HT have been detected in rat CSF
after acid hydrolysis.15
glucuronide conjugates of 5-HT, 5-HIAA, DA, HVA, and DOPAC
was developed. For the first time, 5-HT-G was detected in rat brain
microdialysates. The direct LC-MS/MS method also allows the
detection of different regioisomers of sulfated and glucuronidated
compounds, which is impossible with indirect methods that use
hydrolysis.
EXPERIMENTAL SECTION
Reagents and Standards. 3,4-Dihydroxyphenylacetic acid,
3-methoxy-4-hydroxyphenylacetic acid (HVA), 5-hydroxytryptamine
(serotonin, 5-HT), and saccharic acid 1,4-lactone were purchased
from Sigma-Aldrich (St. Louis, MO). 5-Hydroxyindole-3-acetic acid
(5-HIAA), 3,4-dihydroxyphenethylamine hydrochloride (DA), and
uridine-5′-diphosphoglucuronic acid (UDPGA, trisodium salt) were
purchased from Sigma-Aldrich (Steinheim, Germany), potassium
dihydrogen phosphate from Merck (Darmstadt, Germany), and
ammonium formate and disodium hydrogen phosphate dihydrate
from Fluka (Buchs, Switzerland). Acetonitrile (ACN) was pur-
chased from Rathburn (Walkerburn, Scotland), and methanol
(MeOH) from J.T. Baker (Deventer, Holland). Ringer’s solution,
used for microdialysis and dilution of the standards, contained
147 mM NaCl, 1.2 mM CaCl2 ·2 H2O (Merck, Darmstadt,
Germany), 2.7 mM KCl (Riedel de Hae¨n, Seelze, Germany),
1.0 mM MgCl2 ·6 H2O (Merck, Darmstadt, Germany), and 0.04
mM ascorbic acid (University Pharmacy, Helsinki, Finland).
Sulfuric acid (H2SO4, BDH, 98%) was used in the chemical
synthesis of sulfate conjugates.
Chemical Synthesis of Sulfates. Cold concentrated H2SO4
(200 µL) was added to 20.5 mg of HVA, 20.4 mg of DOPAC,
25.2 mg of 5-HIAA, or 24.3 mg of 5-HT. The reaction mixture
was kept in ice for 2 h and then pipetted over 1 mL of frozen
water. The reaction mixtures were neutralized with 5 M NaOH.
Sulfates were fractionated using an Agilent HP 1100 liquid
chromatograph (Hewlett-Packard GmbH, Waldbronn, Ger-
many) equipped with a binary pump, an autosampler, a column
compartment, UV diode array detector, and fraction collector.
A Discovery HS F5 column (150 mm × 4 mm, 3 µm, Sigma-
Aldrich, Bellefonte, PA) was used in the purification. MeOH
and 5 mM ammonium formate (pH 3.4) were used as eluents
in the separation of 5-HT-S, HVA-S, and DOPAC-S and MeOH/
ACN (1:2 (v/v)) and aqueous 0.1% formic acid in the separation
of 5-HIAA-S. A linear gradient 5-35% organic solvent for 0-10
min was used for the separation of 5-HT-S and 5-HIAA-S. Similar
linear gradients were used for HVA-S and DOPAC-S, except
the first 1 min was isocratic at 5% of the organic solvent. The
flow rate was 0.9 mL/min, and a wavelength of 210 nm was
used for peak detection. The sulfate fractions were evaporated
to dryness and lyophilized. 5-HIAA-S and 5-HT-S were used as
reference standards in the LC-MS/MS analysis of brain
microdialysates.
Swahn et al.3 studied the glucuronidation of DOPAC, HVA,
and 5-HIAA in brain with an indirect method by hydrolyzing the
conjugate with ꢀ-glucuronidase. However, the increase in the
concentration of aglycone after hydrolysis was very small, and
the presence of glucuronides was ambiguously confirmed. In a
recent study, the glucuronidation of DA in rat and mouse brain
was observed when intact DA-glucuronide (DA-G), without hy-
drolysis, was detected in brain microdialysates using LC-MS/
MS.2 Also neurosteroid glucuronides have been directly detected
in mouse brain by using an LC-MS/MS method.21
The metabolism of neurotransmitters in the brain can be
studied by microdialysis, where physiological perfusion fluid is
pumped through a dialysis membrane, which is surgically
implanted into a region of interest in an animal’s brain.22-24 The
extracellular fluid of brain contains synaptically released neu-
rotransmitters and their metabolites, as well as compounds from
nonsynaptic sources.24 These low-molecular-weight compounds
in the extracellular fluid are extracted to the perfusion fluid by
passive diffusion, and the perfusion fluid is subsequently analyzed.
So far, indirect analytical methods employing either acid or
enzymatic hydrolysis of the 5-HT-, 5-HIAA-, DOPAC-, and HVA-
conjugates have been used, since glucuronide and sulfate stan-
dards are commercially unavailable. Indirect analysis methods,
however, are prone to errors due to a hydrolysis step that also
complicates the analysis. Even though the identity of the conjugate
can be investigated with greater certainty after enzymatic hy-
drolysis as opposed to acid hydrolysis, the presence of the
conjugate remains difficult to confirm with indirect methods,
especially if the concentration of the conjugate is low compared
to that of free aglycone. In this study, a specific LC-electrospray
(ESI)/MS/MS method for the quantification of intact sulfate and
(15) Hammond, D. L.; Yaksh, T. L.; Tyce, G. M. J. Neurochem. 1981, 37, 1068–
1071
(16) Sarna, G. S.; Hutson, P. H.; Curzon, G. Eur. J. Pharmacol. 1984, 100, 343–
350
(17) Warnhoff, M. J. Chromatogr. 1984, 307, 271–281
(18) Curzon, G.; Hutson, P. H.; Kantamaneni, B. D.; Sahakian, B. J.; Sarna, G. S.
J. Neurochem. 1985, 45, 508–513
(19) Dedek, J.; Baumes, R.; Tien-Duc, N.; Gomeni, R.; Korf, J. J. Neurochem.
1979, 33, 687–695
(20) Korf, J.; Sebens, J. B. J. Neurochem. 1970, 17, 447–448
(21) Kallonen, S. E.; Tammima¨ki, A.; Piepponen, P.; Raattamaa, H.; Ketola, R. A.;
Kostiainen, R. Anal. Chim. Acta 2009, 651, 69–74
(22) Westerink, B. H. C. J. Chromatogr., B: Biomed. Sci. Appl. 2000, 747, 21–
32
(23) Plock, N.; Kloft, C. Eur. J. Pharm. Sci 2005, 25, 1–24
(24) Bourne, J. A. Clin. Exp. Pharmacol. Physiol. 2003, 30, 16–24
.
.
.
.
.
Enzymatic Synthesis of Sulfates. The incubation mixture
contained 2 mM 5-HT, 5-HIAA, DOPAC, or HVA, 100 µM
3-phosphoadenosine-5-phosphosulfate (PAPS, 99% H. Glatt, Ger-
man Institute for Human Nutrition, Potsdam, Germany), 10 mM
phosphate buffer (pH 7.4), and 30% v/v rat liver S9 fraction (male
Sprague-Dawley rats induced by Aroclor 1254, a mixture of
.
.
.
.
.
8418 Analytical Chemistry, Vol. 81, No. 20, October 15, 2009

Figure 1. Structures of 5-HT, 5-HIAA, DA, HVA, and DOPAC. The possible glucuronidation sites are phenol, amine, and carboxylic acid
groups, whereas the phenolic hydroxyl group is most favorably sulfated in the reaction catalyzed by SULT.
polychlorinated biphenyls) as described previously.25 The treat-
ment of the rats was approved by the local Ethics Committee for
Animal Studies. The reaction mixture was incubated (37 °C) for
2 h and centrifuged for 15 min (16 100g). To the supernatant, 5%
aqueous ZnSO4 (Sigma-Aldrich, Steinheim, Germany) was
added (1:1, v/v), and then MeOH was added to the final
concentration of 50%. The mixture was centrifuged for 15 min
(16 100g), and the supernatant was filtered (0.45 µm). The
filtrate was diluted with 0.1% aqueous formic acid until the
concentration of MeOH was 10% (v/v) and then purified by
solid phase extraction using mixed-mode strong cation ex-
change and reversed-phase cartridges (Oasis MCX, Waters,
MA). The cartridges were conditioned with MeOH and 0.1%
aqueous formic acid. After the sample load, the sulfates were
eluted from the cartridge with 0.1% aqueous formic acid and a
mixture of MeOH/0.1% aqueous formic acid (20:80). The
solvent was evaporated by lyophilization. The DOPAC-S and
HVA-S were fractionated by LC as described for chemically
synthesized 5-HIAA-S.
The concentrations of enzymatically synthesized DOPAC-S and
HVA-S solutions were determined indirectly by hydrolyzing the
conjugate with 0.5 or 0.75 M HCl at 90 °C for 20 min and analyzing
the concentration of free aglycone by LC-MS/MS (method
described later). The recovery for 800 nM free HVA and DOPAC,
heated in acid similarly to sulfates, was 87% and 90%, respectively.
The solutions containing HVA-S and DOPAC-S were used for the
preparation of calibration standards in the LC-MS/MS analysis.
The quantification of both DOPAC-S regioisomers was done using
the calibration curve of the first eluting DOPAC-S (retention time
(rt) ) 6.6 min, SRM pair 247 f 123).
phosphate buffer (pH 7.4), and 2% ACN in a total volume of 30
mL. After 3 h of incubation (37 °C), an additional 6 mg of 5-HT,
5-HIAA, DOPAC, or HVA and 50 mg of UDPGA was added.
The incubation was continued for an additional 4 h with 5-HT
and for 19 h with the other compounds. The reaction mixture
was acidified with formic acid (98-100%, Riedel de Hae¨n,
Seelze, Germany) to pH 3 and centrifuged for 14 min (26 700g,
6 °C).
The supernatant containing glucuronides was filtrated (0.45
µm). Filtrates containing 5-HT-G or 5-HIAA-Gs were purified by
solid phase extraction using mixed-mode strong cation exchange
and reversed-phase cartridges (Oasis MCX, 20 cc, Waters, MA).
For HVA-Gs and DOPAC-Gs, reversed-phase cartridges (Oasis
HLB, 20 cc, Waters, Milford, MAA) were used. The cartridges
were conditioned with 15 mL of MeOH and 15 mL of 0.1%
aqueous formic acid. After the addition of 10 mL of filtrate
containing glucuronides, the cartridge was washed with 15 mL
of 2% (MCX) or 0.1% (HLB) aqueous formic acid. For 5-HT-G,
the wash was continued with 10 mL of MeOH, and the
glucuronides were eluted from the cartridge with 10 mL of
MeOH containing 5% ammonia. 5-HIAA-G was eluted with
MeOH (2 × 10 mL) and MeOH containing 2% ammonia (1 ×
10 mL). DOPAC-G and HVA-G were eluted with 50% or 25%
MeOH in water, respectively. The solvent was evaporated with
a rotary evaporator, and glucuronides were fractionated from
the residue as described for sulfates; 0.1% formic acid and ACN
(5-HIAA-G and DOPAC-G) or ACN/MeOH (2:1) (HVA-G and
5-HT-G) were used as eluents. A linear gradient 5-35% organic
solvent for 0-10 min was used to separate the glucuronides
and aglycones. The collected glucuronide fractions were
evaporated to dryness and lyophilized.
Liquid Chromatography-Mass Spectrometry. The Discov-
ery HS F5 column (150 mm × 4 mm, 3 µm, Sigma-Aldrich,
Bellefonte, PA) was used for the separation of DA, DOPAC, HVA,
5-HT, and 5-HIAA as well as their sulfates and glucuronides
(Figure 1) in brain microdialysates. Aqueous 0.1% formic acid and
ACN/MeOH (1:1, quantitative analysis) or (2:1, qualitative analy-
sis) were used as eluents. Different linear gradients were used
for basic (DA, 5-HT, DA-G, 5-HT-G) and acidic compounds (phase
I metabolites and their sulfates). Gradient 1 (acidic compounds):
5% organic solvent for 0-1 min, 5-20% organic solvent for 1-11
min, 20-85% organic solvent for 11-13 min, 85% organic solvent
for 13-15 min, 85-5% organic solvent for 15-15.1 min, and 5%
Enzymatic Synthesis of Glucuronides. The enzymatic
synthesis of 5-HT-G, 5-HIAA-G, DOPAC-G, and HVA-G were
performed using microsomes prepared from rat liver (male
Sprague-Dawley rats induced by Aroclor 1254; a mixture of
polychlorinated biphenyls) as described previously.25 The syn-
thesis of DA-G is described elsewhere.2 Protein concentrations
of the microsomes were determined with the BCA Protein Assay
Kit (Pierce Chemical, Rockford, IL). In addition to rat liver
microsomes (protein concentration 1 mg/mL), the incubation
mixture contained 2 mM 5-HT, 5-HIAA, DOPAC, or HVA, 5 mM
saccharic acid 1,4-lactone, 5 mM UDPGA, 5 mM MgCl2, 50 mM
(25) Luukkanen, L.; Elovaara, E.; Lautala, P.; Taskinen, J.; Vainio, H. Pharmacol.
Toxicol. 1997, 80, 152–158
.
Analytical Chemistry, Vol. 81, No. 20, October 15, 2009 8419

organic solvent for 15.1-19 min. Gradient 2 (basic compounds):
similar to gradient 1 but 20-65% organic solvent for 11-13 min
and 65% organic solvent for 13-15 min. The flow was directed to
waste for the first 3.8 min to prevent the inorganic ions of Ringer’s
solution to enter the mass spectrometer. The flow was split with
a ratio of 1:10, 10% entering the mass spectrometer. The brain
microdialysates and standards diluted to Ringer’s solution were
injected as such without sample pretreatment, and the injection
volume was 100 µL.
An Agilent 6410 triple-quadrupole mass spectrometer (Agilent
Technologies, Santa Clara, CA) was used in the quantitative
analysis of brain microdialysates. Nitrogen (Parker Balston N2-
22 nitrogen generator, Parker Hannifin Corporation, Haverhill)
was used as a nebulizer (40 psi), curtain (12 L/min, 350 °C), and
collision gas. The ESI needle and fragmentor voltages as well as
the collision energy were optimized for each compound. Agilent
Mass Hunter software version B.01.03 was used for data acquisi-
tion and processing. An API3000 triple-quadrupole mass spec-
trometer (Applied Biosystems/MDS Sciex, Concord, Canada) with
a turbo ion spray source was used in the qualitative analysis of
rat brain microdialysates. Purified air (Atlas Copco CD 2, Wilrijk,
Belgium) was used as a nebulizing gas, and nitrogen generated
with a Whatman 75-72 generator (Haverhill, MA) was used as a
turbo, curtain, and collision gas. The turbo gas flow rate was 6
L/min and was heated to 280 °C. The ion spray and orifice
(declustering) voltages, the collision energy, and the collision cell
exit potentials for different selected reaction monitoring (SRM)
pairs were individually optimized for each compound. The data
were collected and processed by Analyst 1.4.2 software.
and approved by the Animal Experiment Board in conformance
with current legislation. The rats were housed in groups of four
to five per cage and had free access to chow and water. They
were maintained under a 12:12 h light/dark cycle with lights on
from 06:00 to 18:00 at an ambient temperature of 20-22 °C.
Microdialysis. The animals were implanted with a guide
cannula (BAS MD-2250, Bioanalytical Systems Inc., IN) using a
stereotaxic device (Stoelting, Wood Dale, IL) under isoflurane
anesthesia (4.5% during induction for 5 min and then 3.5% during
surgery). The guide cannula was aimed above the rat dorsal
striatum (A/P ) +1.0, L/M ) 2.7, D/V ) -6.0) according to the
atlas by Paxinos and Watson.26 The cannula was fastened to the
skull with dental cement (Aqualox, Voco, Germany) and three
stainless steel screws. Buprenorphine (0.05 mg/kg subcutane-
ously) was given for postoperative pain. The animals were placed
into individual test cages and allowed to recover for 5-7 days
before the experiment.
A microdialysis probe was inserted into the striatum (BAS MD-
2204, 4 mm membrane, Bioanalytical Systems Inc., IN) through
the guide cannula on the morning of the experimental day. The
collection of microdialysis samples (2.5 µL/min) began 2.5-3 h
after insertion of the probe. The microdialysis samples were stored
in a freezer (-70 °C) before analysis with LC-MS/MS. The
microdialysates (100 µL) were injected as such without sample
pretreatment. After completion of the experiments, the positions
of the probes were verified histologically by fixing coronal brain
sections on gelatin/chrome-coated slides and confirming the
respective placements of the implanted probes in the striatum.
Mass spectrometric detection in an ESI positive ion mode was
carried out using SRM with the following reactions: 5-HT (m/z
177 f 115, 117, 132, 160), 5-HT-G (m/z 353 f 160, 177, 336),
5-HT-S (m/z 257 f 115, 160, 240), DA (m/z 154 f 91, 137), and
DA-G (m/z 330 f 91, 137, 154). A negative ion mode was used in
the analysis of HVA (m/z 181 f 122, 137), HVA-G (m/z 357 f
113, 193), HVA-S (m/z 261 f 80, 181, 217), DOPAC (m/z 167 f
93, 95, 108, 123), DOPAC-G (m/z 343 f 113, 123), DOPAC-S (m/z
247 f 80, 123, 167, 203), 5-HIAA (m/z 190 f 116, 144, 146),
5-HIAA-G (m/z 366 f 113, 146), and 5-HIAA-S (m/z 270 f 80,
131, 146, 226). Data in the positive and negative ion modes were
acquired in separate runs. The SRM pairs used for the quantitative
analysis are shown in Table 3.
NMR Spectroscopy. The nuclear magnetic resonance (NMR)
spectra of 5-HT and synthesized and purified 5-HT-G and 5-HT-S
were measured in D2O with a Varian Mercury Plus 300
spectrometer (Varian, Palo Alto, CA) at 24 °C; 5-HIAA and
synthesized and purified 5-HIAA-S were dissolved in CD3OD.
Chemical shifts (δ) are denoted in parts per million (ppm)
relative to the internal standard acetone added to the samples
(1H NMR 2.2 ppm and 13C NMR 30.2 ppm). The protons and
carbons were assigned with heteronuclear single quantum
correlation (HSQC) and heteronuclear multiple bond correla-
tion (HMBC) experiments.
RESULTS AND DISCUSSION
Sulfation and glucuronidation in the brain have usually been
studied using indirect analytical methods employing acid or
enzymatic hydrolysis. One reason for this is the lack of commercial
sulfate and glucuronide standards. In this study, sulfates and
glucuronides of HVA, DOPAC, 5-HT, and 5-HIAA (Figure 1) were
chemically and enzymatically synthesized to provide reference
compounds for a direct LC-MS/MS method. Unlike with indirect
methods employing hydrolysis of the conjugates, the direct
analysis method used in this study permits the detection of
different regioisomers of intact glucuronides and sulfates.
Synthesis and Characterization of Glucuronides. The
glucuronides of 5-HT, 5-HIAA, HVA, and DOPAC were synthe-
sized enzymatically using rat liver microsomes as a biocatalyst,
and the products were characterized by LC-MS. The synthesis
and characterization of DA-G are described elsewhere.2 LC-MS
ion chromatograms of the deprotonated molecules [M - H]-
corresponding to monoglucuronidated compounds in the nega-
tive ion mode showed one peak for 5-HT-G, two peaks for
5-HIAA-Gs, and four peaks for DOPAC-Gs and HVA-Gs (Figure
2
). The respective MS/MS spectra of [M - H]- showed product
ions at m/z 175 [M - H - aglycone]-, m/z 157 [M - H -
aglycone - H2O]-, and m/z 113 [M - H - aglycone - H2O -
CO2]- derived from the glucuronic acid moiety confirming the
presence of the glucuronic acid moiety in the molecule (Table
Animals. Male Wistar rats were used at 8-16 weeks of age.
All procedures with the animals were performed according to the
European Community Guidelines for the use of experimental
animals (European Communities Council Directive 86/609/EEC)
and reviewed by the State Provincial Office of Southern Finland
1). A product ion at m/z 193 (glucuronic acid) was observed in
the MS/MS spectrum of three HVA-Gs (rt 6.65, 7.20, and 7.35
(26) Paxinos, G.; Watson, C. The Rat Brain in Stereotaxic Coordinates, 2nd ed.;
Academic Press: San Diego, CA, 1986.
8420 Analytical Chemistry, Vol. 81, No. 20, October 15, 2009

Figure 2. LC-ESI-MS extracted ion chromatograms of the glucuronides (API 3000) and sulfates of 5-HT, 5-HIAA, DOPAC, and HVA (Agilent
6410) analyzed in the negative ion mode. The ions extracted (m/z) corresponding to the [M - H]- ion appears in the figure. Compounds were
synthesized enzymatically, except 5-HT-S and 5-HIAA-S were synthesized chemically. Linear gradients were used: glucuronides, 5-35% ACN/
MeOH (2:1) for 0-10 min; sulfates, 5% ACN/MeOH (1:1) for 0-1 min, and 5-20% ACN/MeOH (1:1) for 1-11 min.
min), two DOPAC-Gs (rt 5.18 and 5.76 min), and one 5-HIAA-G
(rt 7.31 min). It has been reported earlier that in the MS/MS
spectra of aliphatic hydroxyl glucuronides, a product ion m/z 193
is observed, whereas it is not observed in the MS/MS spectra of
phenol-linked glucuronides.27-29 It is therefore reasonable to
assume that those HVA-Gs, DOPAC-Gs, and 5-HIAA-G showing
a product ion at m/z 193 in the MS/MS spectra were glucu-
ronidated to the carboxylic group, thus producing acyl glucu-
ronides. The DOPAC-Gs with retention times of 6.23 and 6.47 min,
HVA-G with a retention time of 6.79 min, and 5-HIAA-G with a
retention time of 6.56 min were most probably glucuronidated at
phenolic groups since m/z 193 was not observed in the MS/MS
spectra. The acyl glucuronides are prone to acyl migration30 (i.e.,
the aglycone is moving from one hydroxyl group of glucuronic
acid to another through molecular internal reactions forming a
new regioisomer). This explains why the number of glucuronides
formed for HVA and DOPAC (four peaks) exceeded the expected
number of products (two and three, respectively) based on the
possible glucuronidation sites (Figure 1). It is known that the acyl
migration is reduced at low temperatures and under acidic
conditions.30 However, basic conditions (MeOH containing 2%
ammonia) were needed in the purification step of 5-HIAA-G
standard when it was eluted from the mixed-mode SPE cartridge,
otherwise acidic conditions were used throughout the study. In
the analysis of brain microdialysates, the acyl migration was
controlled by freezing the brain microdialysates immediately after
sampling and thawing the samples just before LC-MS/MS
analysis. The brain microdialysates were injected as such without
any sample pretreatment which also minimized the possible acyl
migration. The glucuronidation site of 5-HT-G was determined by
NMR (Table 2). The HMBC experiment showed 3J coupling of
an acetal CH proton (H1′) to the quaternary aromatic carbon
(27) Wen, Z.; Tallman, M. N.; Ali, S. Y.; Smith, P. C. Drug Metab. Dispos. 2007,
35, 371–380
(28) Niemeijer, N. R.; Gerding, T. K.; De Zeeuw, R. A. Drug Metab. Dispos. 1991,
19, 20–23
(29) Fenselau, C.; Johnson, L. P. Drug Metab. Dispos. 1980, 8, 274–283
(30) Johnson, C. H.; Wilson, I. D.; Harding, J. R.; Stachulski, A. V.; Iddon, L.;
Nicholson, J. K.; Lindon, J. C. Anal. Chem. 2007, 79, 8720–8727
.
.
.
.
Analytical Chemistry, Vol. 81, No. 20, October 15, 2009 8421

Table 1. m/z Values of Main Ions in the LC-MS/MS Spectra of 5-HT-, 5-HIAA-, HVA-, and DOPAC-Sulfates and
Glucuronides Analyzed in the Negative Ion Mode
precursorion
compound
rta (min)
[M - H]-
m/z in the product ion spectrum of [M - H]- (abundance %)
Glucuronides
5-HT-G
6.82
6.56
7.31
5.18
5.76
6.23
6.47
6.65
6.79
7.20
7.35
351
366
366
343
343
343
343
357
357
357
357
113 (33), 157 (10), 175 (100)
5-HIAA-G
113 (41), 146 (54), 157 (26), 175 (94), 190 (100)
113 (95), 131 (19), 133 (19), 146 (29), 157 (24), 175 (52), 190 (86), 193 (100)
113 (100), 123 (100), 157 (20), 167 (11), 175 (27), 193 (25)
113 (77), 123 (100), 131 (18), 157 (18), 167 (50), 175 (45), 193 (77)
113 (100), 123 (92), 157 (24), 175 (22)
DOPAC-G
113 (100), 123 (98), 157 (20), 167 (16), 175 (27)
HVA-G
113 (100), 133 (10), 157 (10), 175 (22), 193 (10)
113 (100), 137 (23), 157 (13), 175 (25), 181 (25)
113 (100), 131 (16), 133 (15), 157 (13), 175 (19), 193 (12)
113 (55), 131 (23), 137 (21), 157 (18), 175 (23), 181 (22), 193 (100)
Sulfates
5-HT-S
9.40
8.37
9.64
6.64
7.04
7.97
255
270
270
247
247
261
175 (100), 131 (42), 80(26)
5-HIAA-S
226 (100), 147 (14), 146 (14), 80 (62)
226 (58), 146 (100), 131 (26), 80 (19)
203 (18), 167 (11), 123 (78)
DOPAC-S
203 (38), 167 (10), 123 (75)
181 (51), 137 (100), 122 (44), 80 (8)
HVA-S
a rt ) retention time.
Synthesis and Characterization of Sulfates. 5-HT-S and
5-HIAA-S were synthesized chemically and DOPAC-S and HVA-S
enzymatically using the rat liver S9 fraction. The chemical
Table 2. 1H and 13C NMR Chemical Shifts of 5-HT and
5-HT-G
1H
5-HT
5-HT-G
13C
5-HT
5-HT-G
synthesis method was also applied for DOPAC and HVA, but
according to the LC-MS/MS measurements, the method failed
to produce those sulfate isomers of DOPAC and HVA found in
brain microdialysates (see results below). LC-MS ion chromato-
grams of the synthesis products showed one peak for 5-HT-S and
HVA-S and two peaks for HIAA-Ss and DOPAC-Ss (Figure 2). All
the negative ion mass spectra showed abundant [M - H]-, thus
corresponding to monosulfated compounds. The product ion
spectra of the [M - H]- of 5-HIAA-Ss, DOPAC-Ss, and HVA-S
showed abundant product ions at m/z 226, m/z 203, and m/z
217, respectively, formed by the direct loss of the carbon
dioxide (44 Da) from the [M - H]- ion, thus indicating that
no sulfate is attached to the carboxylic acid group of the
substrate (Table 1). It is known that a hydroxyl group is a
common acceptor group of sulfate in the reaction catalyzed by
SULT,32 and that the most obvious sulfatation site is therefore a
phenolic hydroxyl group. The sulfation sites of 5-HT-S and
5-HIAA-S (rt ) 9.64 min) were verified by NMR. The peak in the
extracted ion chromatogram of 5-HIAA-S with the retention time
of 8.37 min (Figure 2) is a side product formed in the chemical
synthesis of sulfates. It was not observed in the brain microdi-
alysates and therefore its structure was not studied. The chemical
shifts of the H4 and H6 protons of 5-HT-S and 5-HIAA-S (rt )
9.64 min) clearly moved downfield compared to those of the free
substrates, thus indicating that sulfation occurred at the phenolic
group. The yields of HVA-S and DOPAC-Ss produced by the
enzymatic synthesis method were too low to study the sulfation
site by NMR. The yields for 5-HT-S and 5-HIAA-S (rt ) 9.6 min)
were 2.6 (9%) and 2.4 mg (7%), respectively.
H1
7.25
3.08
3.28
7.07
6.86
7.39
7.24
C1
125.3
108.4
22.6
125.5
109.1
22.5
H2R
H2ꢀ
H4
3.05
C2
3.24
C2R
C2ꢀ
C3
7.31
39.6
39.6
H6
7.03
127.1
102.4
148.7
111.8
112.8
131.6
126.6
105.5
150.5
113.6
112.7
133.0
101.9
72.9
H7
7.41
C4
H1′
5.04
C5
H5′
3.83
C6
H2′,3′, 4′
3.60 - 3.61
C7
C8
C1′
C2′
C3′
C4′
C5′
C6′
75.4
71.7
76.2
175.3
ArCO (C5), indicating that the glucuronide group is attached
to phenolic oxygen. An earlier study has shown that the
glucuronidation of 5-HT with mouse liver microsomes occurred
at the phenol group.31 The yield of 5-HT-G synthesized with
rat liver microsomes was good, being 20.6 mg (59%). The yields
for HVA-G, DOPAC-G, and 5-HIAA-G were clearly worse than
for 5-HT-G, being 0.5-1.5 mg (1.4-4.4%) for phenol-glucu-
ronides and 0.04-0.1 mg (0.1-0.4%) for acyl glucuronides, and
therefore their structure was not determined by NMR.
Analysis of Rat Brain Microdialysates by LC-MS/MS. The
glucuronides and sulfates of 5-HT, 5-HIAA, DA, HVA, and
(31) Krishnaswamy, S.; Duan, S. X.; Von Moltke, L. L.; Greenblatt, D. J.;
Sudmeier, J. L.; Bachovchin, W. W.; Court, M. H. Xenobiotica 2003, 33,
(32) Blanchard, R. L.; Freimuth, R. R.; Buck, J.; Weinshilboum, R. M.; Coughtrie,
169–180
.
M. W. H. Pharmacogenetics 2004, 14, 199–211.
8422 Analytical Chemistry, Vol. 81, No. 20, October 15, 2009

Table 3. Results of the Validation Tests of the LC-MS/MS Method for the Analysis of Monoamines and Their
Metabolites in Rat Brain Microdialysatesa
linearity
range
(nM)
repeatability
100 nM,
n ) 9
SRM pairs,
LOD (nM), LOQ (nM),
S/N ) 3-5 S/N > 10
% RSD 10 nM,
compound
m/z
r2
accuracy %
n ) 6
5-HT
177 f 160 (132, 117, 115)
353 f 160 (336, 177)
190 f 146 (144, 116)
270 f 226 (146, 131, 80)
154 f 137 (91)
0.1
0.1
1
0.5
0.5
5
0.5-10
0.5-10
0.997
0.999
92-123
99-100
91-116
92-107
95-129
94-125
97-102
96-111
95-105
97-109
3.7
2.3
2.4
3.1
2.9
3.6
3.5
2.4b
3.9
2.5b
nm
5-HT-G
5-HIAA
2.9
10-260 0.998
15-160 0.996
nm
5-HIAA-S
DA
DA-G
DOPAC
4
15
<LOQ
2.6
0.3
0.1
0.8
0.5
30
5
60
2
0.8-20
0.5-10
180-1000 0.999
50-500 0.999
70-1000 0.999
50-500 0.999
0.998
0.998
330 f 137 (154, 91)
167 f 123 (108, 95, 93)
1.9
10
<LOQ
nm
DOPAC-S (rt ) 6.6 min) 247 f 123 (203, 167, 80)
HVA
HVA-S
1
181 f 137 (122)
261 f 181 (217, 80)
20
<LOQ
nm
0.5
a The qualifier ions are shown in parentheses. nm ) not measured. b n ) 5.
Figure 3. Positive ion SRM chromatograms of (A) 5 nM 5-HT, 5-HT-G, DA, and DA-G standards diluted in Ringer’s solution; (B) a rat brain
microdialysate; and (C) Ringer’s solution (blank). In addition, qualitative SRM pairs for each compound were followed (Table 3). The analyses
were performed with an Agilent 6410.
DOPAC in rat brain microdialysates were analyzed using a
direct LC-MS/MS method. The adequate retention and good
separation of different glucuronide and sulfate regioisomers
were obtained with a pentafluorophenylpropyl column. ESI in
the positive ion mode provided high ionization efficiency for
5-HT, DA, 5-HT-G, 5-HT-S, and DA-G, whereas the negative ion
mode provided better ionization efficiency for DOPAC, HVA,
5-HIAA, and their glucuronides and sulfates than did the
positive ion mode. The analytes in brain microdialysates were
detected by comparing the retention times and the intensity-
ratios of at least two SRM pairs of each compound (Table 3) to
those of the standards diluted in Ringer’s solution. The variation
of the intensity ratio of the monitored product ions was less
than 15%, and the variation of the relative retention times was
less than 1.3% in all positive identifications. The SRM chro-
matograms of the blank Ringer’s solution, run before every
microdialysis sample, showed no memory effects.
The SRM chromatograms showed for the first time that 5-HT-G
exists in rat brain (Figure 3). In previous studies, an acid-
hydrolyzable conjugate of 5-HT was detected in superfusates of
rat spinal cord15 but no 5-HT-G has been detected in brain
samples. 5-HT has proved to be a good substrate for UGT 1A6,
which is expressed in human12 and rat brain,10 thus supporting
the presence of 5-HT-G in rat brain microdialysates found in this
Analytical Chemistry, Vol. 81, No. 20, October 15, 2009 8423

Figure 4. Negative ion SRM chromatograms of (A) 100 nM 5-HIAA, 5-HIAA-S, and HVA-S, 300 nM DOPAC and HVA, and 360 nM DOPAC-S
standards diluted in Ringer’s solution; (B) a rat brain microdialysate; and (C) Ringer’s solution (blank). In addition, qualitative SRM pairs for
each compound were followed (Table 3). The analyses were performed with an Agilent 6410.
study. The glucuronidation of DA in rat brain has been shown
before,2 and DA-G was also detected in rat brain microdialysates
Table 4. Analytical Results of Rat Brain
Microdialysates by LC-MS/MS
in this study (Figure 3). No glucuronidation of 5-HIAA, DOPAC,
concentration (nM)
or HVA in rat brain was observed, whereas 5-HIAA-S, HVA-S, and
two DOPAC-S-regioisomers, most probably DOPAC-3- and DOPAC-
4-O-S, were detected in the microdialysates (Figure 4). The
sulfation of DOPAC, HVA, and 5-HIAA in rat brain has also been
observed in previous studies with indirect methods where a clear
increase in the concentration of free substrate after enzymatic
sulfatase hydrolysis was observed.3,17 5-HT-S was not detected
(LOD ) 0.1 nM) in rat brain microdialysates in this study.
The quantitative LC-MS/MS method was validated in terms
of LOD (signal-to-noise ratio (S/N) of at least 3) and the limit
of quantification (LOQ, S/N of at least 10), linearity, and
repeatability by using standards diluted in Ringer’s solution
used in the brain microdialysis (Table 3). The LODs for 5-HT,
DA, 5-HT-G, and DA-G analyzed in the positive ion mode were
0.1-0.3 nM (10-30 fmol injected on the column, Table 3). The
LODs for HVA-S, DOPAC-S, 5-HIAA-S, and their free substrates
analyzed in the negative ion mode varied between 0.5 and 20
nM (50-2000 fmol injected on the column). The calibration
curves for each compound were measured, and the linearity
was excellent, r2 being >0.996 for each compound. The within-
day repeatability of the method was studied at two concen-
tration levels: 10 and 100 nM (Table 3). The relative standard
deviations (% RSD) were typically below 4%, indicating good
repeatability of the analysis.
The concentrations of free and glucuronidated 5-HT and DA
and of free and sulfated DOPAC, HVA, and 5-HIAA in rat brain
microdialysates are shown in Table 4. Note that the recoveries of
the analytes across the microdialysis membrane were not deter-
mined in this work. The concentration of 5-HT-G (1.0-1.7 nM)
was over 2 times higher than that of 5-HT (0.4-2.1 nM) in three
of four rat brain microdialysates analyzed. The concentration of
DA-G was at the same level or lower than that of the free DA as
reported previously.2 It was observed that around 10, 15, and 30%
of the 5-HIAA, HVA, and DOPAC, respectively, was in the sulfated
form in the brain microdialysates. In previous studies, the
compound
animal 1 animal 2 animal 3 animal 4
5-HT
0.4a
1.0
110
7a
1.2
1.1
890
40
0.5
1.3
140
11a
2.0
1.4
910
40
0.7
1.7
120
12a
2.4
1.0
700
40
2.1
1.5
150
13a
2.1
1.0
980
40
5-HT-G
5-HIAA
5-HIAA-S
DA
DA-G
DOPAC
DOPAC-S (rt ) 6.6 min)
DOPAC-S (rt ) 7.0 min)
HVA
280
510
90
290
510
90
280
360
80
400
540
100
HVA-S
a <LOQ, only an estimation.
conjugated phase I metabolites were extracted from brain samples,
and the amounts of conjugated 5-HIAA, HVA, and DOPAC after
an acid hydrolysis were 10%, 20-36%, and 12-32%, respectively,
of the total amount (the concentration of conjugated and free) in
rat brain striatum.16,19,33 These concentration ratios are quite
similar to those obtained in this study in microdialysates collected
from rat brain striatum also.
CONCLUSIONS
5-HT-, 5-HIAA-, DA-, DOPAC-, and HVA-sulfates and glucu-
ronides were synthesized and analyzed by an LC-MS/MS
method. Even low concentrations of conjugates can be measured,
and the regioisomers can also be separated with the direct method
used in this study, which is impossible with indirect methods that
use hydrolysis. 5-HT-G was detected in rat brain microdialysates
for the first time. It seems that neurotransmitters are glucu-
ronidated and that their phase I metabolites are sulfated. However,
further experiments must be carried out to study the impact of
these phase II metabolites on brain function. The concentration
(33) Elchisak, M. A.; Maas, J. W.; Roth, R. H. Eur. J. Pharmacol. 1977, 41,
369–378
.
8424 Analytical Chemistry, Vol. 81, No. 20, October 15, 2009

of neurotransmitters and their phase I metabolites are commonly
used as a measure of neuronal activity. However, the concentration
of glucuronides is at the same level (DA) as or higher (5-HT)
than that of free neurotransmitters, and up to 30% of phase I
metabolites are sulfated. Therefore the quantitative analysis of
these phase II metabolites should also be included in brain studies.
The recoveries of neurotransmitters, and their phase I and II
metabolites across the microdialysis membrane should also be
studied in future experiments to get an estimate of the real
extracellular concentrations of analytes in the brain.
ACKNOWLEDGMENT
We are grateful to Minna Baarman for her technical assistance
and to Anna Alonen, Katriina Ita¨aho, and Taina Sten for their
helpful discussions. This study was financially supported by the
Academy of Finland (Projects 115543 and 1130992).
Received for review June 18, 2009. Accepted August 31,
2009.
AC901320Z
Analytical Chemistry, Vol. 81, No. 20, October 15, 2009 8425