Journal of Neurological Sciences 156 (1998) 102–106

Plasma levels of neuroexcitatory amino acids in patients with migraine or

tension headache

Zafar Alam, Nicholas Coombes, Rosemary H. Waring*, Adrian C. Williams, Glyn B. Steventon
School of Biochemistry, The University of Birmingham, Edgbaston, Birmingham, B15 2 TT, UK

Received 5 February 1997; received in revised form 9 October 1997; accepted 26 October 1997

Abstract

Plasma amino acids were analysed in patients with migraine with (9) and without (80) aura, in patients with tension headache (14) and
in controls (62). The neuroexcitatory amino acids glutamic acid, glutamine, glycine, cysteic acid and homocysteic acid were elevated in
migraine patients while total thiols (cysteine / cystine) were reduced. Patients with tension headache had values which were similar to
those of controls. Tryptophan was elevated in migraine patients without aura only. Studies on two patients showed that the raised resting
excitatory amino acid levels became still further elevated during a migraine attack. These results show that high concentrations of
neurotransmitter amino acids occur normally in migraine patients and suggest that this profile may be a contributory factor in migraine
attacks. Tension headache, however, has different biochemical parameters.  1998 Elsevier Science B.V.

Keywords: Migraine; Neuroexcitatory; Amino acids; Tension headache

1. Introduction migraine patients as opposed to controls (Martinez et al.,

Several amino acids are known to have an excitatory
role in the central nervous system. This is particularly true
for the neurotransmitter glutamate which is excitotoxic at
high concentrations and of major importance in brain
energy metabolism. Both glutamate and aspartate are
involved in the pathophysiology of ischaemic and hypo-
glycaemic neuronal death and of epilepsy. Endogenous
release of glutamate and aspartate acts via the N-methyl-D-
aspartate (NMDA) receptor to play a significant role in the
initiation, propagation and duration of spreading depres-
sion which has been proposed as a basis for migraine
attacks (Marranes et al., 1988). In susceptible individuals,
excessive oral intake of glutamate may actually induce
migraine-like features (Reif-Lehrer, 1976). Platelet levels
of glutamic and aspartic acids and glycine have been
shown to be raised in patients with migraine with aura
(D’Andrea et al., 1991), while higher levels of aspartate
and glutamate have been found in plasma and CSF in
*Corresponding author. Tel.: 144 121 4145421; fax: 144 121
4143982; e-mail: R.H.Waring@bham.ac.uk
1993b) and in migraine patients during attacks as com-
pared with the resting state (Ferrari et al., 1990). It has
therefore been suggested that migraine patients may be
predisposed to central neuronal hyperexcitability leading to
labile responses (Martinez et al., 1993b). However, as
other amino acids besides glutamate and aspartate have
neurotransmitter or excitatory modes of action, particularly
cysteic and homocysteic acids, the present study was
conducted to see whether they too could contribute to
neuronal overstimulation. Total thiol (cysteine / cystine)
plasma levels were also measured to assess any contribu-
tion to the disease pathology by reactive oxygen species
generated as a result of raised glutamate levels.
2. Methods
2.1. Patient selection
Patients were selected from those attending the Migraine
Clinic of a single consultant. All had been diagnosed with

0022-510X / 98 / $19.00  1998 Elsevier Science B.V. All rights reserved.

PII S0022-510X( 98 )00023-9
 Z. Alam et al. / Journal of Neurological Sciences 156 (1998) 102 – 106
either migraine or tension headache using criteria of the
Headache Classification Committee of the International
Headache Society, they were not on concurrent medication
and had no co-existing illness. Patients with migraine had
headaches lasting less than 72 h, with frequencies between
one to five per month; blood samples were taken more than
6 days after the last headache. Controls had no history of
migraine, had no known illness and were not on medica-
tion. They were age and sex-matched and selected from
staff of the University of Birmingham and their families
(Table 1). In this study, the ratio of patients with aura to
without aura was 10%, consistent with cited prevalence
values of 10–20%.
2.2. Patients samples
Blood samples (10 ml) were taken from overnight-fasted
subjects between 0800–1000 h. These were immediately
centrifuged (2000 g for 10 min at 48C) to provide plasma,
which was stored at 2208C until used for analysis.
2.3. Plasma amino acid analysis
Amino acid standards were obtained from Sigma (Poole,
Dorset); the solvents used were HPLC grade reagents
(Fisons, Loughborough). Protein from the plasma samples
was precipitated with 2 vol. HPLC grade methanol then
centrifuged down in a microfuge. The derivatising reagent
was prepared, using O-phthalaldehyde (35 mg) in ethanol
(0.5 ml) containing 2-mercaptoethanol (0.1 ml) in 100 mM
sodium borate buffer (50 ml, pH 10.4). Equal volumes of
deproteinised plasma and derivatising reagent were mixed
for 1 min before a 20-ml flushed loop injection onto the
HPLC system. Amino acid standards gave a linear cali-
bration curve over the concentrations range used (1 nmol /
ml to 1 mmol / ml). Samples were injected onto a
Spherisorb S5 ODS2 column (150 mm34.6 mm, 5 mm
particle size) (source, Phase Separation, Deeside, Clwyd),
using a Technosphere 5 ODS2 guard column (HPLC
Technology, Macclesfield). The column temperature was
maintained at 418C (Livereel column controller,
Maidenhead, Berks.) for the analysis time. The chromato-
graphic system used was a Perkin-Elmer 250 Binary LC
pump, a dual monochromatic fluorescence detector (FD-
Table 1
Parameters of volunteers in migraine study patients
Controls Migraine without aura
n 62 80
Sex
Females 34 68
Males 28 12
Age
Females 48.6613.2 40.569.1
Males 52.1612.7 45.0611.6
103
300 Spectrovision) operating at 360 nm excitation wave-
length and 455 nm emission wavelength with a Perkin-
Elmer Nelson Integrator (Model 1020). The mobile phase
(flow rate 1 ml / min) consisted of A: 100 mM tripotassium
orthophosphate buffer (pH 7.0) with orthophosphoric acid)
containing 3% v / v) tetrahydrofuran and B: 100% acetoni-
trile. Both phases were filtered under vacuum through a
0.45-mm Millipore nylon membrane and continually de-
gassed with a helium degassing unit (ACS, Macclesfield,
Cheshire). The gradient was adapted from that of
Hirschberger et al. (1985) rising from A, 98%; B, 2% at 0
min to A, 60%; B, 40% at 25 min.
Cysteine / cystine was measured separately using the
method of Pheifer and Briggs (1995). Dithiothreitol (20
ml, 10 mM solution in 0.2 M phosphate buffer, pH 8.0 was
mixed with plasma (500 ml) for one hour. The derivatising
reagent, 10 mM DTNB (59-dithiobis (2-nitrobenzoic acid)
in 0.2 M disodium hydrogen orthophosphate buffer, pH
8.0), was added (500 ml) and the solutions were mixed for
a further 15 min. Plasma proteins were precipitated with
addition of 30 ml saturated sulphosalicylic acid (in 0.2 M
phosphate buffer, pH 8.0) then centrifuged down before a
flushed loop injection (200 ml) onto the HPLC system.
Cysteine standards taken through the same procedure gave
a linear calibration curve over a concentration range of
0.05 mmol / ml–2.0 mmol / ml.
2.4. Analysis of results
All intraindividual variations in the biochemical parame-
ters measured were within 4%. The data were tested for
significance using the Mann-Whitney test for non-normally
distributed results. There were no significant differences
(P,0.05) observed between males and females in any of
the groups.
3. Results
Table 1 shows details of the population in this study.
Plasma amino acid levels in migraine patients with and
without aura, tension headache patients and controls are
shown in Table 2. It can be seen that cysteic acid, glutamic
acid, homocysteic acid, glutamine and glycine are all
Migraine with aura Tension headache
9 14
7 13
2 1
42.069.1 29.5611.5
50.067.1 48.060
104 Z. Alam et al. / Journal of Neurological Sciences 156 (1998) 102 – 106

Table 2
Levels of plasma amino acids in controls and headache patients a
Amino acid Control Migraine without aura Migraine with aura Tension
(nmol / ml) headache
Cysteic acid 49626 94656 b 87638 b 62621
Glutamic acid 277687 4856129 b 454698 b 306676
Homocysteic acid 2664 67619 52615 b 2266
Asparagine 48616 49620 46613 53615
Serine 155641 148659 128663 136648
Glutamine 6946211 8476214 b 8686194 b 7246159
Glycine 4516114 7456239 b 7136168 b 497685
Threonine 4476116 4236140 3966127 4636122
Alanine 6846171 6086184 6576101 6336151
Taurine 132659 133636 124645 137625
Tyrosine 122637 238692 155680 145647
Valine 207644 228663 174645 196662
Tryptophan 184642 260682 372680 b 248646
Lysine 354695 333659 251675 320680
a
Results6SD.
b
P,0.05.

elevated in both types of migraine (P,0.05), although the
other amino acids were within control values. Tryptophan
was elevated in patients without aura; the levels in patients
with aura were not significantly different from controls.
Migraine patients had lower total thiol (cysteine1cystine;
80623 nmol / ml without aura, 82621 nmol / ml with aura)
levels than normal patients (179624 nmol / ml) (P,0.05);
tension headache patients’ thiol levels were 140619 nmol /
ml. Amino acid values from control and tension headache
patients were not generally significantly different from
each other, although different from migraine patients. Total
thiol levels in tension headache tended to fall between
those of controls and migraine patients.
Plasma amino acid levels were measured in four patients
before and during a migraine attack and the results are
shown in Table 3. Levels of plasma thiols, cysteic acid,
glutamic acid, homocysteic acid, glutamine, glycine and
tryptophan were all found to be elevated during an attack,
although the other amino acid values were essentially
unchanged.
4. Discussion
Although non-neurotransmitter amino acid concentra-
tions were normal in plasma from migraine patients, the
levels of those amino acids which are neurotransmitters or
neurotransmitter precursors were elevated. Other workers
have shown raised resting levels of platelet and plasma
glutamate and glycine which potentiates the action of
glutamate by acting as an agonist at the NMDA sub-type
receptor (D’Andrea et al., 1991; Martinez et al., 1993b). In
this study, both plasma glutamate and glutamine con-
centrations were higher in migraine patients; other workers

Table 3
Plasma amino acids in patients before and during a migraine attack without aura
Amino acid Patient 1 Patient 2 Patient 3 Patient 4
(nmol / ml)
Normal Migraine Normal Migraine Normal Migraine Normal Migraine
attack attack attack attack
Cysteic acid 81 103 99 126 78 111 102 127
Glutamic acid 520 691 389 431 428 467 417 493
Homocysteic acid 71 86 53 79 49 72 84 98
Asparagine 59 62 67 66 62 64 68 66
Serine 187 176 151 158 184 187 165 162
Glutamine 909 1061 874 1012 895 939 860 987
Glycine 835 983 665 881 693 808 852 1004
Threonine 509 487 503 571 481 493 476 460
Alanine 665 649 746 734 734 728 752 760
Taurine 127 131 147 130 133 130 122 124
Tyrosine 101 98 181 186 137 133 166 156
Valine 198 214 221 238 241 244 184 192
Tryptophan 292 405 238 335 267 378 306 417
Lysine 331 342 326 346 320 328 318 310
Cysteine / cystine 87 135 48 113 62 110 93 156
 Z. Alam et al. / Journal of Neurological Sciences 156 (1998) 102 – 106
have reported normal plasma glutamine and raised gluta-
mate (Cananzi et al., 1995) though this may reflect
differences in methodology as glutamine readily breaks
down to give glutamate. Increased tryptophan, a precursor
of serotonin, was found in migraine patients, but only
those without aura. Although abnormalities of serotonin
metabolism have been implicated in migraine (Rajur et al.,
1989), tryptophan levels were normal in patients with aura
in this study, which suggests that this may not be a primary
mechanism. Homocysteic acid and cysteic acid, both
elevated in this study, are excitotoxins (Do et al., 1988;
Kim et al., 1987; Porter and Roberts, 1993) hyper-
homocysteinemia being a recognised risk factor for cere-
brovascular disease and probably toxic to the vascular
endothelium (van den Berg et al., 1995). Cysteine itself is
neurotoxic and is known to act at the NMDA sub-type of
glutamate receptor, high levels causing prolonged calcium
release and cell damage (Flint, 1992). Thiols have a
protective antioxidant effect (Bridges et al., 1991) but in
this study, total thiol levels (cysteine and cystine) were
reduced in migraine patients, increasing during an attack
and potentiating the raised levels of glutamate. The basal
lower levels of cysteine could be a response to continuous
oxidant stress generated by high concentrations of ex-
citotoxic amino acids. If plasma values of neurotransmitter
amino acids correlate with those in brain tissue, it would
result in migraine patients having permanently increased
electrical signal activity in the CNS with a persistent
neuronal hyperexcitability. The increases in neurotrans-
mitter amino acids during an attack would then magnify
the clinical dysfunction and probably lengthen its duration.
It may be argued that peripheral amino acid levels would
not affect brain function. However, recent work has shown
that stress situations greatly increase the permeability of
the blood–brain barrier and that plasma levels of drugs and
amino acids can affect the CNS at least under some
circumstances (Friedman et al., 1996). Conversely, raised
levels of excitatory amino acids in brain tissue which have
been shown to occur in dysfunctional states such as
hypoxia and epilepsy (McDonald and Johnston, 1990)
could reach the systemic circulation.
These results show that migraine patients generally have
major imbalances of neurotransmitter amino acids and that
even when they are between attacks, and apparently
normal, they are in fact metastable, with different bio-
chemistry from controls. As migraine has a strong
heritability, these faults in processing of amino acids and
their derivatives would appear to have a genetic basis.
Defective receptor feedback mechanisms, probably at the
NMDA subtype, may be a primary cause of migraine
headaches and drugs which block the NMDA receptor
could therefore be clinically useful. As migraine patients
have been found to have a variety of abnormalities,
including sulphotransferase activity (Littlewood et al.,
1982), serotonin metabolism (Ferrari et al., 1989) and
taurine levels (Martinez et al., 1993a), it seems probable
105
that these individuals have contributions from a number of
dysfunctional pathways leading to a high endogenous CNS
excitability which is readily triggered by a variety of
exogenous factors, including stress.
Patients with tension headaches can be clearly differen-
tiated from those with migraine if biochemical markers are
used, even though the clinical presentations may be
similar. Their plasma amino acid levels do not differ
significantly from controls while thiols seem to be slightly
lower than control value, which may suggest release of
reactive oxygen species. Investigation of biochemical
parameters can therefore not only prove helpful in sug-
gesting lines of treatment but can also aid classification
and diagnosis particularly between patients with migraine
and with headache.
References
Bridges, R.J., Koh, J.Y., Hatalski, C.G., Cotman, C.W., 1991. Increased
excitotoxic vulnerability of cortical cultures with reduced levels of
glutathione. Eur. J. Pharmacol. 192, 199–220.
Cananzi, A.R., D’Andrea, G., Perini, F., Zamberlan, F., Welch, K.M.,
1995. Platelet and plasma levels of glutamate and glutamine in
migraine with and without aura. Cephalalgia 15 (2), 132–135.
D’Andrea, G., Canduzi, A.R., Joseph, R., Morra, M., Zamberlan, F.,
Ferro Milone, F., Grunfeld, S., Welk, K.M., 1991. Platelet glycine,
glutamate and aspartate in primary headache. Cephalalgia 11 (4),
197–200.
Do, K., Herling, P., Streit, P., Guenod, M., 1988. Release of neuroactive
substances; homocysteic acid as an endogenous agonist of the NMDA
receptor. J. Neurol. Trans. 72, 185–190.
Ferrari, M., Odink, J., Capparelli, C., 1989. Serotonin metabolism in
migraine. Neurology 39, 1239–1242.
Ferrari, M.O., Odink, J., Malessy, M.J.A., Bruyn, G.W., 1990. Neuro-
excitatory plasma amino acids are elevated in migraine. Neurology 40,
1582–1586.
Flint, B.M., 1992. Mechanisms of excitotoxicity in neurologic disease.
FASEB J. 6, 3338–3344.
Friedman, A., Kaufer, D., Shemer, I., Hendler, I., Soreq, H., 1996.
Pyridostigmine brain penetration under stress enhances neuronal
excitability and induced early immediate transcriptional response.
Nature Med. 2, 1382–1385.
Hirschberger, L.L., Delavasa, J., Stipamuk, M.H., 1985. Determination of
cysteine sulfinate, hypotaurine and taurine in physiological samples by
reversed-phase high performance liquid chromatography. J. Chroma-
togr. 343, 303–313.
Kim, J.P., Koh, J., Choi, D.W., 1987. L-Homocysteate is a potent
neurotoxin in cultured cortical neurons. Brain Res. 437, 103–110.
Littlewood, J.T., Glover, V., Sandler, M., Petty, R., Reatfield, R., Rose,
F.C., 1982. Platelet phenolsulphotransferase deficiency in dietary
migraine. Lancet 1, 953–985.
Marranes, R., Willems, R., De Prius, E., Wanquier, A., 1988. Evidence for
a role of the N-methyl-D-aspartate (NMDA) receptor in cortical
spreading depression in the rat. Brain Res. 457, 226–240.
Martinez, F., Castillo, J., Leira, R., Prieto, J.M., Lema, M., Noya, M.,
1993. Taurine levels in plasma and cerebrospinal fluid in migraine
patients. Headache 33 (6), 324–327.
Martinez, F., Castillo, J., Rodriguez, J.R., Leira, R., Noya, M., 1993.
Neuroexcitatory amino acid levels in plasma and cerebrospinal fluid
during migraine attacks. Cephalalgia 13, 89–93.
McDonald, J.W., Johnston, M.W., 1990. Physiological and
106 Z. Alam et al. / Journal of Neurological Sciences 156 (1998) 102 – 106

pathophysiological roles of excitatory amino acids during central
nervous system development. Brain Res. Rev. 14, 41–70.
Pheifer, J.H., Briggs, D.E., 1995. The estimation of thiols and disulfides
in barley. J. Inst. Brew. 101 (1), 5–10.
Porter, R.H., Roberts, P.J., 1993. Glutamate metabotropic receptor
activation in neonatal rat cerebral cortex by sulphur-containing excitat-
ory amino acids. Neurosci. Lett. 154 (1-2), 78–80.
Rajur, J., Welch, K., D’Andrea, G., 1989. Serotinergic hypo-function in
migraine. Cephalalgia 9, 293–294.
Reif-Lehrer, L., 1976. Chinese restaurant syndrome. Fed. Proc. 35,
2205–2211.
van den Berg, M., van der Knaap, M.S., Boers, G.H., Stehouwer, C.D.,
Rauwerda, J.A., Valk, J., 1995. Hyperhomocysteinaemia: with refer-
ence to its neuroradiological aspects. Neuroradiology 37 (5), 403–411.