Introduction
Gamma-aminobutyric acid (GABA) is a non-proteogenic amino acid present in bacteria, plants and animals.Reference Watanabe, Maemura and Kanbara1 In mammalian cells, GABA is synthesised from glutamic acid (glutamate) by glutamic acid decarboxylase and from polyamines with the aid of diamine oxidase.Reference Watanabe, Maemura and Kanbara1, Reference Tillakaratne, Medina-Kauwe and Gibson2 Gamma-aminobutyric acid is a major inhibitory neurotransmitter in the adult mammalian brain, acting at either ionotropic (GABAA and GABAC) or metabotropic (GABAB) receptors.Reference Watanabe, Maemura and Kanbara1, Reference Bormann3, Reference Watanabe, Maemura, Oki, Shiraishi, Shibayama and Katsu4 Apart from its role as an inhibitory neurotransmitter, GABA is also thought to regulate different stages of brain neurogenesis, including proliferation, migration, differentiation and synaptic integration of newborn neurons.Reference Watanabe, Maemura, Oki, Shiraishi, Shibayama and Katsu4, Reference Ge, Pradhan, Ming and Song5
Glutamic acid decarboxylase, diamine oxidase, GABA and GABA receptors can also be present in non-neuronal tissues, including heart, liver, gut, pancreas, salivary glands, ovary and testis.Reference Watanabe, Maemura and Kanbara1, Reference Watanabe, Maemura, Oki, Shiraishi, Shibayama and Katsu4, Reference Ortega6 Although the exact role of GABA in peripheral tissues remains to be established, it is accepted that GABA and GABA receptors are involved in cell proliferation and differentiation.Reference Watanabe, Maemura and Kanbara1, Reference Watanabe, Maemura, Oki, Shiraishi, Shibayama and Katsu4 Recently, it has been found that GABA negatively regulates proliferation of pluripotent stem cells in peripheral embryonic and adult tissues.Reference Andäng, Hjerling-Leffler, Moliner, Lundgren, Castelo-Branco and Nanou7, Reference Young and Bordey8
In line with the involvement of GABA in neuronal and non-neuronal cell proliferation,Reference Watanabe, Maemura, Oki, Shiraishi, Shibayama and Katsu4, Reference Ge, Pradhan, Ming and Song5, Reference Andäng, Hjerling-Leffler, Moliner, Lundgren, Castelo-Branco and Nanou7, Reference Young and Bordey8 it has been reported that GABA levels are significantly higher in malignant tissue of the thyroid,Reference Andersson, Henningsson and Jarhult9 colonReference Kleinrok, Matuszek, Jesipowicz, Matuszek, Opolski and Radzikowski10, Reference Maemura, Yamauchi, Hayasaki, Kanbara, Tamayama and Hirata11 and mammary glands,Reference Mazurkiewicz, Opolski, Wietrzyk, Radzikowski and Kleinrok12 compared with normal tissue. Higher GABA levels have also been found in benign colonic adenomas, compared with healthy control tissue.Reference Maemura, Yamauchi, Hayasaki, Kanbara, Tamayama and Hirata11 Tatsuta and colleaguesReference Tatsuta, Iishi, Baba, Nakaizumi, Ichii and Taniguchi13, Reference Tatsuta, Iishi, Baba and Taniguchi14 were the first group to suggest a functional link between GABA and oncogenesis, by showing that GABA and the clinically utilised GABAB agonist baclofen suppressed gastric carcinogenesis in the rat. More recently, it has been suggested that GABA acts as a near-universal signal which can inhibit the mitotic activity and migration of tumour cells.Reference Ortega6, Reference Young and Bordey8 When considering the possible clinical applications of GABA concentration measurement, it is worthy of note that elevated urinary GABA concentrations have been found in ovarian cancer patients.Reference Nicholson-Guthrie, Guthrie, Sutton and Baenziger15
The parotid gland is the most common site of salivary tumours.Reference Ansari16–Reference Madani and Beale18 Eighty per cent of salivary neoplasms occur in the parotid, of which up to 80 per cent are benign. Pleomorphic adenoma is the most common benign parotid neoplasm, comprising approximately 70 per cent of all parotid tumours, and Warthin's tumour is the second most common.Reference Ansari16–Reference Madani and Beale18
To the best of our knowledge, GABA levels have never been assessed in human salivary glands or salivary tumours. Given the above, the aim of the present study was to evaluate GABA concentrations in samples of the most common parotid tumours and adjacent non-tumourous salivary tissue taken during parotid surgery. As GABA can be synthesised from glutamate in the rat parotid glandReference Sawaki, Ouchi, Sato and Kawaguchi19 and in other tissues,Reference Tillakaratne, Medina-Kauwe and Gibson2 we decided to analyse glutamate concentrations in the tumourous and non-tumourous samples. Aspartic acid (aspartate) levels were examined to control for possible non-specific alterations in amino acid concentrations in the tumourous tissue.
Parotid tumours can manifest as abnormalities of the gland itself and/or changes in saliva chemistry.Reference Lee17 Hence, we decided to assess GABA, glutamate and aspartate levels in unstimulated parotid saliva taken from the same patients before surgery.
The present study was an extension of our previous report showing that the most common benign parotid tumours do not profoundly alter unstimulated saliva secretion from the affected gland.Reference Jezewska, Scinska, Kukwa and Bienkowski20
Materials and methods
Patients
The study was carried out in accordance with the Declaration of Helsinki of the World Medical Association. Each participant signed an informed consent form after study procedures had been fully explained. The protocol for the study was reviewed and approved by the ethics committee of the Warsaw Medical University.
We enrolled in the study patients with primary parotid tumour who were admitted consecutively to the otolaryngology department of the Czerniakowski Hospital for routine surgical procedures. Potential participants were excluded if they had: xerostomia; sialadenitis; sialadenosis; other significant pathology in the head and neck region (except unilateral parotid tumour); or a history of severe liver, renal or endocrine disorder. Patients were also eliminated if they were taking medication with anticholinergic effects or GABA-ergic medication (e.g. benzodiazepines, anticonvulsants and baclofen), as these may affect GABA levels.Reference Nicholson-Guthrie, Guthrie, Sutton and Baenziger15, Reference Jezewska, Scinska, Kukwa and Bienkowski20, Reference Sreebny21
Twenty-five patients were screened for eligibility. Five patients were excluded after initial screening; thus, results from 20 patients are reported. All study participants were Caucasian and aged 18 to 79 years (see Table I for sociodemographic and clinical characteristics). A painless, palpable mass was a major presenting symptom in all cases. There were no signs of facial nerve palsy or local inflammation. The diagnosis of pleomorphic adenoma or Warthin's tumour was confirmed by histopathological examination of the tumour tissue.
Table I Baseline characteristics of study participants
*Total n = 14. †Total n = 6. SEM = standard error of the mean; BMI = body mass index (kg/m2); mth = months
Saliva collection
The swab method was used to collect unstimulated parotid saliva from both the involved and non-involved sides.Reference Ahlner and Lind22 Saliva was collected for 15 minutes with the aid of Salivette cotton rolls (Sarstedt, Rommelsdorf, Germany), as described by Jezewska et al. Reference Jezewska, Scinska, Kukwa and Bienkowski20 The saliva-soaked rolls were placed in single-use syringes equipped with filter units (0.45 µm; Millex®-HV, Millipore Corporation, Bedfont, Massachusetts, USA) and squeezed until 0.2–0.3 ml of filtered saliva was collected into an Eppendorf tube. Saliva samples were stored at −70°C until further analysis.Reference Scinska-Bienkowska, Wrobel, Turzynska, Bidzinski, Jezewska and Sienkiewicz-Jarosz23
Tissue sampling and preparation
Samples of tumour and adjacent salivary tissue were collected from each patient during tumour resection. All specimens were immediately washed in sterile, ice-cold physiological saline (0.9 per cent NaCl), frozen on dry ice and stored at −70oC. Before further analysis, each sample was weighed, placed in an ice-cooled Eppendorf tube and homogenised for 60 seconds in 0.2 M perchloric acid. Homogenates were centrifuged at 10 000 rpm at 4°C for 10 minutes and supernatants filtered through 0.45-μm Millex-HV filters.
Biochemical analysis
The relevant biochemical procedures have been described in detail by Maciejak et al. Reference Maciejak, Czlonkowska, Bidzinski, Walkowiak, Szyndler and Lehner24 and Scinska-Bienkowska et al. Reference Scinska-Bienkowska, Wrobel, Turzynska, Bidzinski, Jezewska and Sienkiewicz-Jarosz23 Concentrations of GABA, glutamate and aspartate in saliva and tissue samples were determined with the aid of a gradient reversed-phase high performance liquid chromatography system (Shimadzu, Tokyo, Japan), with pre-column derivatisation with o-phtalaldehyde and electrochemical detection.
Statistical analysis
Differences in GABA, glutamate and aspartate levels were analysed with the aid of Student's t-test, unpaired. A p value of less than 0.05 was considered significant.
Results and analysis
Figure 1 shows GABA, glutamate and aspartate concentrations in unstimulated parotid saliva taken from patients with benign parotid tumours. No statistically significant differences were found for GABA (p = 0.07), glutamate (p = 0.4) or aspartate concentrations (p = 0.7), comparing the involved and non-involved sides.
Fig. 1 Concentration of (a) γ-aminobutyric acid (GABA) and (b) glutamate and aspartate, in unstimulated saliva collected from the non-involved (healthy) and involved (tumourous) sides. Bars represent mean values and outliers represent standard error of the mean, for 20 patients with benign parotid tumour.
Figure 2 shows tissue concentrations of GABA, glutamate and aspartate. Levels of GABA and glutamate were significantly higher in the tumourous tissue samples compared with the control tissue (p < 0.01). There was no significant difference in aspartate concentration between the tumourous and control tissue samples (p = 0.6).
Fig. 2 Concentration of (a) γ-aminobutyric acid and (b) glutamate and aspartate in tumourous and non-tumourous human parotid tissue. Bars represent mean values and outliers represent standard error of the mean, for 20 patients with benign parotid tumour. *p < 0.01.
Discussion
To the best of our knowledge, this is the first study to show that the most common benign parotid tumours (i.e. pleomorphic adenoma and Warthin's tumour) contain more GABA than healthy parotid tissue. Increased GABA concentrations in the tumour tissue samples were accompanied by a significant rise in glutamate levels but not in aspartate levels. The latter observation suggests that our findings were not secondary to any non-specific increase in amino acid content of the tumour tissue samples.
There was no statistically significant difference in GABA levels in unstimulated saliva taken from the affected versus unaffected salivary glands. This finding was not completely unexpected, as Warthin's tumours and pleomorphic adenomas are slow-growing tumours with a fibrous capsule of varying thickness,Reference Ansari16–Reference Madani and Beale18 and may thus exert little effect on saliva secretion and chemistry. It remains to be established whether salivary GABA levels can rise in patients with malignant parotid tumours.
• Gamma-aminobutyric acid (GABA) is thought to regulate various stages of cell proliferation and differentiation in the brain and periphery
• Gamma-aminobutyric acid may also be a near-universal signal capable of altering tumour cells and modifying mitotic activity
• This study assessed GABA, glutamate and aspartate concentrations in benign parotid tumour tissue and unstimulated parotid saliva
• Concentrations of GABA and glutamate, but not aspartate, were significantly higher in the tumour vs non-tumour tissue samples; no significant difference was found between saliva from the involved vs non-involved side
• Gamma-aminobutyric acid may be involved in the growth of benign parotid tumours
Our results provide further support for previous reportsReference Watanabe, Maemura, Oki, Shiraishi, Shibayama and Katsu4, Reference Ortega6, Reference Kleinrok, Matuszek, Jesipowicz, Matuszek, Opolski and Radzikowski10–Reference Mazurkiewicz, Opolski, Wietrzyk, Radzikowski and Kleinrok12 (see Introduction) that tumours of various histological origin contain more GABA than healthy tissue. In addition, the present study provides more evidence that GABA levels may rise not only in malignant but also in benign tumours.Reference Maemura, Yamauchi, Hayasaki, Kanbara, Tamayama and Hirata11, Reference Okada, Taniguchi and Schimada25 Interestingly, Maemura et al. Reference Maemura, Yamauchi, Hayasaki, Kanbara, Tamayama and Hirata11 have found a positive correlation between GABA concentration and the degree of atypia in benign colonic tumours. Further studies with larger groups of patients are needed to determine whether GABA concentrations predict the histological characteristics and/or clinical course of parotid tumours.
Sawaki et al. Reference Sawaki, Ouchi, Sato and Kawaguchi19 have confirmed the presence of GABA and its metabolic enzymes in the rat salivary gland. The present results, taken together with this finding,Reference Sawaki, Ouchi, Sato and Kawaguchi19 suggest that the most common parotid tumours contain the molecular machinery required for GABA-ergic transmission. More studies are needed to address this hypothesis and to delineate a possible role for GABA and its receptors in salivary oncogenesis. For example, it may be desirable to assess whether clinically utilised GABA-ergic medications (e.g. baclofen, valproic acid and benzodiazepines) alter the growth of salivary tumours.
Our finding that glutamate concentrations were higher in tumourous than non-tumourous tissue samples was not completely unexpected. Gamma-aminobutyric acid can be synthesised from glutamate by glutamic acid decarboxylase in neuronal and non-neuronal tissues,Reference Watanabe, Maemura and Kanbara1, Reference Watanabe, Maemura, Oki, Shiraishi, Shibayama and Katsu4 including the rat parotid gland.Reference Sawaki, Ouchi, Sato and Kawaguchi19 Thus, increased GABA levels in benign parotid tumours could be secondary to increased substrate (i.e. glutamate) availability. Although glutamate is mainly considered to be a major excitatory neurotransmitter,Reference Hinoi, Takarada, Ueshima, Tsuchihashi and Yoneda26 a large body of evidence has accumulated over the last decade indicating a role for glutamate as an extracellular signal mediator in peripheral autocrine and paracrine systems. Peripheral cells express all the molecular machinery required for glutamatergic transmission, including glutamate receptor and transporter proteins.Reference Hinoi, Takarada, Ueshima, Tsuchihashi and Yoneda26 The present report could be a starting point for further studies on the role of glutamate in signalling cascades regulating salivary oncogenesis.
Acknowledgements
The study was supported by the Warsaw Medical University (EJ, AS and WK) and the Institute of Psychiatry and Neurology, Warsaw, Poland (PB).
