Abnormal Platelet Trace Amine Profiles in Migraine With and Without Aura

Abstract

Trace amines, including tyramine, octopamine and synephrine, are closely related to classic biogenic amines. In one study, where these substances were found elevated in plasma of migraineurs, it was hypothesized that trace amine metabolism is deranged in migraine. To confirm these findings, we studied, using a multichannel electrochemical high-performance liquid chromatography system, the concentrations of trace amines in platelets of migraine without aura (MoA) and migraine with aura (MA) patients in headache-free period, compared with controls. Platelet concentrations of trace amines, although elevated in both migraine types, showed a different profile in MoA and MA. Octopamine was significantly higher in MoA sufferers (0.69 ± 0.43 ng/10⁸ platelets) compared with both control subjects (0.22 ± 0.16 ng/10⁸ platelets) and MA patients (0.39 ± 0.37 ng/10⁸ platelets). Synephrine was significantly higher in MA patients (0.72 ± 0.44 ng/10⁸ platelets) with respect to both controls (0.33 ± 0.25 ng/10⁸ platelets) and MoA sufferers (0.37 ± 0.29 ng/10⁸ platelets). These results strengthen the hypothesis that tyrosine metabolism is deranged in migraine and may participate in its pathophysiology.

Introduction

In the late 1960s it was hypothesized that trace amines produced from tyrosine metabolism through decarboxylase enzyme activity (1), such as tyramine, octopamine and synephrine, all substances with a molecular structure similar of that of catecholamines (2), might play a role in the pathophysiology of migraine (3). Hanington (3) put forward this theory on the observation that ingestion of tyramine-containing food such as cheese, citrus fruits, chocolate, etc, may provoke headache attacks in migraineurs with a deficiency of monoaminoxidase (MAO), the enzyme that metabolizes tyramine and octopamine. Thereafter, in view of the delay from ingestion of provocative food to migraine attack, accumulation of octopamine, derived from tyramine metabolism, was proposed as the main mechanism of action (4). Trace amines displace other amines from neuronal secreting vesicles and the depleting agent reserpine precipitates migraine attacks (5). These results are in line with the ability of trace amines to act as false neurotransmitters. However, sound evidence in support to this hypothesis has never been found and, for decades, trace amine research fell into a ‘shadow area’. The main obstacle to carrying out experimental research in this field was the difficulty in measuring trace amines in biological fluids and tissues. Moreover, their synapses or receptors were unknown.

Now the relevance of trace amines is emerging, due to the recent identification in rat tissues, including brain, of a large new family of G protein-coupled receptors (TARs) (6), at least two of which, TAR1 and TAR2, bind to and are activated by trace amines (6). TAR1 and the other three TAR receptors have been identified in the hypothalamus, amygdala, locus coeruleus and thalamus in humans (6). In addition, by using a recently developed high-performance liquid chromatography (HPLC) method able to measure trace amines reliably (7), we have studied their concentrations in plasma of patients affected by migraine, with and without aura, and cluster headache (CH), both in active and remission periods (8). Although plasma concentrations of trace amines were increased in all the primary headache groups compared with controls, the difference was impressive (six- to ninefold higher) only in CH patients. Migraine without aura (MoA) and migraine with aura (MA) patients showed a smaller elevation of octopamine and synephrine plasma concentrations, statistically significant only for the former. Such results suggest that a metabolic trace amine abnormality may play an important role in CH, whereas its role in the pathophysiology of migraine remains uncertain.

Platelets are a useful tool to study trace amines. They take up octopamine and the other amines from the body circulation and store them in the neutral milieu of dense bodies for 14 days. Indeed, platelet trace amine concentrations are considered a good marker of the general trace amine metabolism (9). With this background in mind, we wondered whether the study of trace amine concentrations in platelets could elucidate their significance in the pathophysiology of migraine. With this aim, concentrations of tyramine, octopamine and synephrine were assessed in platelets of patients with MoA and MA during headache-free periods and compared with those of healthy control subjects.

Patients and methods

Two groups of headache outpatients (MoA and MA) from a subspecialty centre and a group of 41 healthy control subjects (blood donors) were enrolled in the study. The diagnosis of MoA and MA was made in agreement with ICHD-II criteria (10). No patients complained of any other disease. All subjects had been free from drugs known to alter platelet function or to affect MAO activity (e.g. imipramine, amitriptyline, phenelzine, tranylcipromine, selegiline and moclobemide) (11) from at least 2 weeks prior to blood sampling. In addition, to avoid biochemical interference from the diet, all subjects avoided food containing biogenic amines (e.g. chocolate, dry fruits, cheese, etc.) for at least 1 week before the haematological examination. The patients had been free from migraine prophylactic therapy for at least 1 month. Migraine attacks were treated with triptans. The patients were studied in headache-free periods at least 3 days after the last painful attack.

Peripheral venous blood (30 ml) was drawn from the antecubital vein, following overnight fasting, at 09.00 h from subjects after 5 min of resting in the supine position. The blood was drawn from the same operator and collected into 1/10 volume citric acid–citrate dextrose as anticoagulant for the estimation of trace amines. Platelet-rich plasma (PRP) was obtained by centrifugation (165 g for 15 min) of whole blood and after the platelet number was calculated by a platelet analyser, PRP was centrifuged at higher speed (2000 g for 15 min) to obtain a platelet pellet and the platelet-poor plasma (PPP). The platelet pellet was then washed with 3 ml of Tyrode's buffer solution, as previously described (12). The platelet pellet was sonicated in 1% perchloric acid solution. The mixture was centrifuged (14 100 g for 5 min) to remove protein fragments. The supernatant was removed and passed through a 0.2-µm nylon filter (Acrodisc; Waters, Milford, MA, USA). PPP was obtained by centrifugation of PRP (2000 g for 15 min). An aliquot of perchloric acid was added to PPP (total volume 4 ml) for the deproteinization. After brief centrifugation (13 900 g for 5 min), the supernatant was passed through a ultrafilter membrane.

HPLC method

The chromatographic system consisted of an autosampler (Waters) capable of variable-volume injections, equipped with a 100-µl loop and two-pump separation module (Waters 2695) connected with a C18 reverse phase (XTerra) 15 × 4.6 mm, 5-µm column, fitted with a 3.0 × 20 mm, 5-µm precolumn (XTerra). The detection device consisted of a multichannel electrochemical detector system (CoulArray; ESA, Chelmsford, MA, USA). The autosampler and pumps were monitored and controlled by Millennium 2 software. The multichannel detector was monitored and controlled by a CoulArray for Windows installed in a Pentium 3 computer (Dell). The computer system performed the data storage, analysis and report generation. The mobile phase, adjusted to pH 3 with phosphoric acid, consisted of sodium phosphate monobasic 85 mm, sodium dodecyl sulphate 0.86 mm and acetonitrile 11% v/v. The instrumental parameters of the separation method were as follows: full scale sensitivity of all channels, 100 µA; potentials of each of eight channels, starting with channel 1, −100, 0, 50, 300, 380, 780, 890, 920 mV. Three incremental doses of 0.125, 0.250 and 0.500 µg/ml of tyramine, octopamine and synephrine standard solutions (Sigma, St Louis, MO, USA) were utilized to obtain the calibration curve in order to determine the platelet and plasma concentrations of each substance. The standard solutions were prepared by dissolving 10 mg of the components in a solution of 0.1 n perchloric acid and each component was diluted with a solution of 1% perchloric acid to give a final concentration of 1 µg/ml. Aliquots of 80 µl of the standards and extracts from biological tissues were injected for identification and quantification. The retention times of tyramine, octopamine and synephrine were 18.5, 6.7 and 7.8 min (ratio accuracy 0.75, 0.88 and 0.87). The intra-assay relative standard deviations (RSD) were 1.3, 1.5 and 1.8 for octopamine, synephrine and tyramine, respectively. The interassay RSD were 2.1, 2.5 and 2.8 for octopamine, synephrine and tyramine, respectively. The detection limits of the HPLC method were 22 pg for tyramine, 22 pg for octopamine and 49 pg for synephrine.

Approval was obtained from local ethics committees. Signed informed consent was obtained from each patient.

The statistical analysis was performed by one-way anova test followed by Tamhane test for multiple comparisons using the SPSS 11.5 statistical package (SPSS Inc., Chicago, IL, USA).

Results

Trace amine platelet concentrations were measured in 32 patients with MoA, 19 patients with MA and 41 healthy control subjects. The demographic and clinical characteristics of the population studied are shown in Table 1. No significant differences between genders were detected in trace amine concentrations. Octopamine was significantly more elevated in MoA patients than in controls (P < 0.0001) and MA patients (P = 0.04) (Table 1). Synephrine was significantly higher in MA sufferers than in both control subjects (P = 0.005) and MoA patients (P = 0.01) (Table 2).

Table 1

Characteristics of the population studied

	MoA (n = 32)	MA (n = 19)	Control subjects (n = 41)
Gender, F, n (%)	18 (56.3)	14 (73.7)	20 (48.8)
Age, years
Mean ± SD	36.4 ± 8.92	37.4 ± 9.14	39.4 ± 8.66
Range	21–53	18–60	22–56
Current smokers, n (%)	7 (21.9)	3 (15.8)	10 (24.4)
Current oral contraceptive intake, n (%)	5/15∗ (33.3)	2/13∗ (23.1)	5/15∗ (33.3)
Number of migraine attacks per month, mean ± SD	3.1 ± 1.1	0.9 ± 1.4
Type of migraine with aura
Typical aura with migraine headache, n (%)		15 (78.9)
Typical aura with non-migraine headache, n (%)		4 (21.1)

∗

Women of fertile age.

MoA, Migraine without aura; MA, migraine with aura.

Table 2

Platelet levels of trace amines in migraine without aura (MoA), migraine with aura (MA) and control subjects

	MoA, n = 32	MA, n = 19	Control subjects, n = 41	P
Tyramine	0.33 ± 1.31	0.55 ± 0.89	0.04 ± 0.07	MoA vs. ctrl: P = NSMA vs. ctrl: P = NSMoA vs. MA: P = NS
Octopamine	0.69 ± 0.43	0.39 ± 0.37	0.22 ± 0.16	MoA vs. ctrl: P < 0.0001MA vs. ctrl: P = NSMoA vs. MA: P = 0.04
Synephrine	0.37 ± 0.29	0.72 ± 0.44	0.33 ± 0.25	MoA vs. ctrl: P = NSMA vs. ctrl: P = 0.005MoA vs. MA: P = 0.01

Values are expressed as ng/10⁸ platelets ± SD.

Discussion

This study clearly shows that the concentrations of some trace amines are elevated in platelets of migraine patients. MoA and MA, however, show a different metabolic amine profile. In MoA only the concentrations of octopamine were significantly higher than those found both in controls and MA subjects, while in MA only the concentrations of synephrine were significantly higher compared with those of controls and MoA patients. The concentrations of tyramine, which is a very short-lived substance (it is readily transformed into octopamine (7)), were less consistent than the concentrations of other trace amines and, in many cases, were below the sensitivity threshold of our methodology. In fact, tyramine was not detectable in 23 out of 41 control subjects (56.1%), in 13 out of 32 MoA patients (40.6%) and in seven out of 19 MA patients (36.8%). Octopamine and synephrine, in contrast, were detected in almost all the subjects. This may explain the large SD of tyramine and, probably, the lack of significance between migraine patients and controls, although the tyramine concentrations were much higher in patients than in controls.

These results strengthen our previous findings obtained in plasma of MoA patients (8), where octopamine concentrations were significantly higher than those found in controls. At variance with previous findings in plasma (8), the platelet concentrations of synephrine were significantly higher in MA patients than in controls and MoA patients. Since platelets store octopamine and the other amines over time, the high concentrations of octopamine in MoA and synephrine in MA suggest that an increased release of these substances occurs. Since the major source of trace amines is the sympathetic system (13), where they are synthesized and stored in the organelle compartment of neuronal endings around vessels and from where they are released in blood upon stimulation, it is conceivable that an anomaly of trace amine synthesis in the hypothalamus and autonomic system may occur. The reason why the amine metabolic pathway differs between the two migraine types remains to be determined. One could hypothesize that the activity of enzymes of trace amine pathways, such as tyrosine decarboxylase, dopamine β-hydroxylase and phenylethanolamine N-methyltransferase have a different range of activity in MoA and MA. High platelet concentrations of octopamine and synephrine could be distinctive markers of MoA and MA, respectively, but this hypothesis has to be substantiated by further studies.

Our findings support the hypothesis, proposed by Welch in the late 1980s (14), that a possible new link between trace amines and migraine exists. He suggested that migraine is a bio-behavioural disorder, in which a possible shift in tyrosine metabolism toward a decarboxylation may alter the physiological balance between the synthesis of octopamine and that of noradrenalin in the intrinsic noradrenergic nervous system and its putative orbito-frontal connections. The present data, together with our previous studies, may give a more complete picture of this presumed metabolic derangement in migraine. In fact, we have previously demonstrated that dopamine concentrations are increased in platelets of MoA patients (15), whereas those of norepinephrine are low (12). The reason why platelet dopamine is increased is unknown. A possible explanation is the reduced activity of dopamine-beta-hydroxylase (d.b.h.), the enzyme that transforms dopamine into norepinephrine, in MoA (16). The low concentration of norepinephrine reported in migrainous women at menses (17), together with an altered allelic distribution of the d.b.h. gene found in migraine patients (18), supports this hypothesis.

The reduced activity of d.b.h. may divert dopamine toward the trace amine metabolic pathway, since some enzymes may transform dopamine into tyramine or directly into octopamine (19). The very high concentrations of platelet octopamine found in MoA may hypothetically be due also to these metabolic enzymatic activities. All these findings taken together suggest that a derangement of tyrosine metabolism characterizes migraine.

Trace amines are nowadays considered neuromodulators (20) and neurotransmitters (6). Neuromodulator is a chemical released from a neuron which causes no change in the excitability of postsynaptic cells in the absence of neurotransmitter. The released neuromodulator acts to modify the action (increase or decrease) of a coexisting neurotransmitter. The current hypothesis is that trace amines are endogenous neuromodulator compounds that can act as a fine-tuneing mechanism to maintain basal neuronal tone within physiological limits. An increase of these neuromodulators, such as that recently reported in primary headaches (8), together with an abnormal concentration of neurotransmitters, may therefore cause a dysfunction in the synaptic activity of dopaminergic and noradrenergic systems at different hierarchical levels in the central nervous system. In addition, octopamine and the other trace amines are neurotransmitters, since they bind and sensitize some of a newly discovered family of G-coupled receptors, such as TAR1 and TAR2 (6), mainly located in amygdala, hypothalamus and some brainstem nuclei (locus coeruleus, raphe magno). Moreover, TAR1 and TAR2 are present in leucocytes (7). As neurotransmitters, trace amines may concur to the regulation of several functions, related to the autonomic system, the pain threshold and immunity. The excessive amount of trace amines is consistent with the autonomic system symptomatology (21), the impairment of pain threshold (22) and the interleukin dysfunction (23) found in migraine.

In conclusion, we have demonstrated that some trace amines are elevated in platelets of MoA and MA patients. In particular, octopamine is very high in MoA and synephrine very elevated in MA. This abnormal concentration of trace amines, together with other possible metabolic abnormalities, may contribute to the pathophysiology of migraine.