Histamine N-Methyltransferase in the Brain

1. Introduction

Histamine (2-[4-imidazolyl]ethylamine) was discovered by Sir Henry Hallet Dale and Sir Patrick Playfair Laidlaw in 1910 [1]. Histamine mediates a wide variety of events such as allergic reactions, gastric acid secretion, and smooth muscle contraction through interaction with four histamine receptor subtypes (H₁R–H₄R). Pharmacological research targeting histamine action in peripheral organs has led to the development of drugs such as H₁R antagonists for allergic diseases and H₂R antagonists for gastric ulcers. This amine also acts as a neurotransmitter in the brain [2]. Histaminergic neurons are located in the tuberomammillary nucleus of the hypothalamus and project their axons into various brain regions including the cerebral cortex, hypothalamus, basal ganglia, and amygdala [3]. The number of histamine-producing neurons in the human brain is estimated to be approximately 64,000. Brain histamine regulates diverse physiological functions such as sleep–wake cycles, stress response, appetite, and memory. Extensive investigations have shown the pathophysiological involvement of the histaminergic nervous system in various neuropsychiatric disorders. A loss-of-function mutation of the histidine decarboxylase (HDC) gene (EC 4.1.1.22), the essential enzyme for histamine production, is associated with Tourette’s syndrome [4]. Pathological changes in histamine neurons are involved in cognitive impairment [5,6,7]. The reduction in histamine concentration in the cerebrospinal fluid was observed in narcolepsy patients [8,9,10]. H₁R binding potential was found to be decreased in patients with depression and schizophrenia (SCZ) via positron emission tomography [11,12]. This evidence suggests that the dysfunction of the histaminergic nervous system could play a causative role in various neurological disorders and that the increase in brain histamine might have therapeutic potential.

Over the past few decades, H₃R receptors in the brain have attracted attention as therapeutic targets to modulate the histamine system. This receptor identified by Arrang et al. in 1983 is a Gi/o protein-coupled receptor and negatively regulates histamine release as a presynaptic autoreceptor at nerve end terminals [2,13]. H₃R inverse agonists, which could induce histamine release, were developed by many pharmaceutical companies and were examined for their therapeutic effects on animal models for neurological disorders and also on patients suffering from brain diseases [14,15]. Although the beneficial effect of most H₃R inverse agonists is inconclusive, pitolisant, an H₃R antagonist developed by Schwartz et al. [16], was approved for the treatment of narcolepsy by the European Medical Agency in 2016 [17], emphasising the involvement of histamine concentration in brain functions.

Neurotransmitter clearance is very important for maintaining normal neurotransmitter concentration. The neurotransmitters released into extracellular spaces are cleared by transporters and /or enzymes in the adjacent neurons or astrocytes to avoid excessive neuronal activation. Dysfunctional neurotransmitter clearance plays a causative role in various neurological disorders, including SCZ and depression. Indeed, various drugs, such as tricyclic antidepressants, serotonin norepinephrine re-uptake inhibitors, and monoamine oxidase inhibitors, block neurotransmitter clearance and exert their therapeutic actions in patients suffering from brain diseases. This evidence supports that histamine clearance machinery could be a therapeutic target for developing novel drugs that improve brain functions. However, the mechanism of brain histamine clearance had not been clarified.

Previous studies using rodents showed that astrocytes play an important role in histamine clearance [18]. We recently investigated the molecular mechanism of histamine clearance using primary human astrocytes [19]. Our in vitro experiments showed that human astrocytes transport histamine dominantly through plasma membrane monoamine transporter (PMAT) and also through organic cation transporter 3 (OCT3). Next, the histamine transported into the cytosol is metabolised by histamine N-methyltransferase (HNMT) (EC 2.1.1.8) (Figure 1). PMAT and OCT3, which are polyspecific transporters, transport various monoamines, including serotonin, dopamine, norepinephrine, and histamine [20]. PMAT, which was identified in 2004, is widely distributed in the brain [21,22]. Several reports suggested that PMAT plays a role in serotonin clearance [23,24], and that mutations in the human PMAT genes coupled with low transport activity are related to autism spectrum disorders [25], indicating the involvement of PMAT in brain monoamine concentration. However, Pmat deficiency in mice does not greatly affect brain histamine concentration under non-stressful conditions (our unpublished observation) nor induces behavioural abnormalities [26]. OCT3 is expressed in different brain regions including the cerebral cortex, hippocampus, and cerebellum [27]. The contribution of OCT3 to serotonin and dopamine concentrations has already been reported [28,29]. Zhu et al. examined the importance of OCT3 in brain histamine concentration [30]. They showed that OCT3 is not involved in brain histamine concentration in normal conditions, whereas histamine content in the brain cortex is elevated in Oct3-deficient mice after cerebral ischemia. These results might indicate the minor contribution of these transporters to brain histamine concentration, although further studies are essential to examine the importance of PMAT and OCT3 in histamine clearance. In contrast, our recent study using Hnmt knockout mice clearly showed that Hnmt plays a predominant role in brain histamine concentration and the regulation of the histaminergic nervous system [31]. In this article, we focus on HNMT function in the central nervous system (CNS).

2. Histamine-Metabolising Enzymes

There are two different enzymes for the inactivation of histamine: diamine oxidase (DAO) (EC 1.4.3.22) and HNMT. DAO, also known as histaminase, is the product of the AOC1 gene. DAO acts as a homodimeric protein to oxidatively deaminate various amines including histamine, putrescine, and spermidine [35,36]. DAO is highly expressed in the digestive tract. The K_m (Michaelis constant) value of human intestinal DAO to histamine was calculated as 19 µM [37]. DAO plays a role in detoxification of dietary histamine to reduce histamine uptake through enterocytes. Thus, impaired DAO activity results in the increase of histamine absorption and the elevation of blood histamine concentration. Although DAO is also highly expressed in the kidneys and placenta, DAO expression in the CNS is low or absent [38], indicating that DAO metabolises histamine in the peripheral organs but not the CNS.

HNMT is an enzyme catalysing the transfer of a methyl group from S-adenosyl-l-methionine (SAM) to histamine, yielding N^τ-methylhistamine and S-adenosyl-l-homocysteine (Figure 2). The human HNMT gene was cloned by Girard et al. in 1994 and encodes a 33 kDa protein consisting of 292 amino acids [39] (Table 1). Although HNMT is widely observed in vertebrates including humans, rodents, birds, lizards, and amphibians, the expression of HNMT has not been confirmed in invertebrates and plants. In mammals, HNMT is widely expressed in various organs including liver, kidney, and brain [40]; and methylated histamine metabolites are abundantly excreted in urine [41], suggesting the important role of HNMT in histamine metabolism. Brain HNMT activity in the CNS was first detected in the soluble supernatant fraction from guinea pig brain in 1959 [42]. Schayer and Reilly confirmed the existence of methylated histamine inactivated by HNMT in guinea pig brain [43]. Human HNMT activity has also been detected in the frontal, temporal, parietal, occipital, and cerebellar cortices [44]. The Human Protein Atlas project showed a high expression of HNMT in the cerebellum and medium expression in the cerebral cortex, hippocampus, and caudate [45]. The project also found that both neurons and glial cells express HNMT. An in situ hybridisation study in Allen Mouse Brain Atlas showed the highest mRNA expression of HNMT in the cortical subplate [46]. Northern blot analysis using mouse and rat brains revealed ubiquitous expression of Hnmt except in rat cerebellum and mouse striatum [47]. Immunohistochemical analysis using bovine brain revealed that several neurons including the oculomotor nucleus, red nucleus, facial nucleus, and dorsal vagal nucleus strongly express Hnmt [33]. However, the detailed distribution of brain HNMT in mice, rats, and humans is still unknown. Further immunohistochemical analysis using specific antibodies is essential to understanding detailed HNMT distribution in the CNS.

3. HNMT and Human Brain Diseases

The effects of genetic polymorphism of human HNMT on enzymatic activity was first reported by Preuss et al. in 1998 [40]. They showed that a C-to-T transition at nucleotide 314 (C314T) in exon 4 replaces threonine with isoleucine at codon 105 (Thr105Ile), leading to decreased enzymatic activity. Although amino acid residue 105 is distal from the substrate binding sites, the C314T polymorphism has a great impact on active site structure and dynamics, resulting in the increased K_m value of human HNMT for histamine and SAM by 1.3- and 1.8-fold, respectively [52,53]. Several groups examined the association of C314T polymorphism with Parkinson’s disease (PD) [54,55,56,57], and the meta-analysis of these four studies suggested a protective role of this polymorphism against the development of PD [58] (Table 2). Other genetic association studies suggested the involvement of the C314T substitution in SCZ [57], attention deficit hyperactivity disorder (ADHD) [59] and migraine [60] but not in Alzheimer’s disease (AD) [61], amyotrophic lateral sclerosis [62], multiple sclerosis [63], or restless legs syndrome [64]. The effect of the C314T polymorphism on alcoholism was inconclusive [65,66]. An A-to-G polymorphism at nucleotide 939 in the 3′ untranslated region of the human HNMT gene increases HNMT mRNA stability and increases HNMT protein, and enhances enzymatic activity [67]. An A939G polymorphism is related to several brain disorders such as myasthenia gravis and ADHD [59,68]. Recently, Heidari et al. reported that two novel mutations in the human HNMT gene (G179A and T632C) impairs its enzymatic activity, leading to intellectual disability [69,70].

Several postmortem studies examined the alteration of HNMT expression in neurological disorders. HNMT expression was increased at the inferior frontal gyrus in Huntington’s disease [71], at the substantia nigra and putamen in PD [72], and at the frontal cortex in Pick’s disease [73]. HNMT mRNA expression was also elevated in the prefrontal cortex of female AD patients [74]. On the other hand, HNMT expression was reduced at the anterior cingulated cortex in depression [75] and the frontal cortex in Down syndrome [73]. Although these results indicate the possible involvement of HNMT functions in neuropsychiatric disorders, it is still unknown whether alterations in HNMT activity exert the causative role in disease progression, play a compensatory function for impaired brain functions, or is a secondary outcome accompanied by primary pathological change. Therefore, these associations should be validated in larger studies and possibly by prospective clinical studies in the near future.

4. Pharmacological Analysis Using HNMT Inhibitors

Duch and colleagues discovered that metoprine, a derivative of 2,4-diaminopyrimidine, inhibited HNMT activity with a Ki value of 100 nM [76,77]. Because metoprine can cross the blood–brain barrier (BBB) due to its hydrophobicity (logP = 2.82) and increase brain histamine concentration [78], many researchers have used this inhibitor to investigate the role of HNMT in the CNS. Pharmacological studies using metoprine indicated that the activation of the histaminergic system in the CNS affects a wide variety of brain functions such as antinociception [79], suppression of energy intake [80], hyperglycaemic action [81], improvement of cognitive function [82], antiepileptic effect [83,84,85], and attenuation of methamphetamine-induced behavioural abnormalities [86]. However, metoprine can also inhibit mammalian dihydrofolate reductase (EC 1.5.1.3) and decrease cellular folate metabolism, resulting in the attenuation of cell growth [87,88]; therefore, it cannot be ruled out that the low specificity of metoprine affected the results. Another HNMT inhibitor, SKF91488, developed by Beaven et al. can specifically inhibit the enzymatic activity devoid of histamine receptor agonist activity [89]. Unfortunately, due to the poor BBB permeability of SKF91488, the brain research using SKF91488 has been limited [90,91,92]. Therefore, novel HNMT inhibitors with high specificity and sufficient BBB permeability are expected to accelerate to pharmacological research on brain HNMT. Properties of commercially available HNMT inhibitors are summarised in Table 3 and Figure 3.

5. Phenotyping of Hnmt-deficient Mice

Recently, we phenotyped Hnmt-knockout (KO) mice for a better understanding of the role played by HNMT in brain function [31]. First, we showed that histamine content from the cortices, diencephalons, brainstems, and cerebella of KO mice was at least 5-fold higher than that of wild-type (WT) mice. Histamine concentration was consistently higher in whole brain homogenates from neonatal, adolescent, and adult KO mice compared with those of WT mice. Higher extraneuronal histamine concentration in KO mice was also confirmed by in vivo microdialysis assay. However, other monoamines and their metabolites were not changed in KO mouse brains. Histamine concentration in several organs such as the skin and stomach was not elevated. These results demonstrated that Hnmt affected extracellular and intracellular histamine concentration of mouse brains throughout development and adult life. Next, we performed comprehensive behavioural testing to investigate the impact of elevated histamine on mouse behaviours. Although KO mice did not show anxiety-like behaviours, depression-like behaviours, impaired memory, or impaired motor function, highly aggressive behaviours and decreased locomotor activity in home cages were observed in KO mice. High aggression in KO mice was attenuated by pre-treatment of H₂R antagonist zolantidine but not by H₁R antagonist pyrilamine, indicating that elevated histamine by Hnmt deficiency activated H₂R, thereby driving aggressive behaviours in mice. The decreased locomotor activity in home cages suggested that the sleep–wake cycle was disrupted in KO mice, resulting in extended immobility during periods that should have been active. Sleep analysis with electroencephalography and electromyography showed prolonged wakefulness during the inactive period and compensatory sleep during the active period. The average wake bout duration was increased in KO mice during zeitgeber time (ZT) 0–6 compared to WT mice, although the total number of wake and sleep bouts was not changed by Hnmt deficiency. Pyrilamine injection at ZT0 normalised both the prolonged wakefulness during ZT0–6 and the decreased wakefulness during ZT12–18 in KO mice, indicating that excessive H₁R activation disrupted the sleep–wake cycle in KO mice. These results demonstrated that HNMT played an essential role in regulating brain histamine concentration and accordingly controls aggression and sleep–wake cycles. In this study, we deleted whole body Hnmt to understand the histaminergic nervous system using conventional knockout technology. We could not rule out the possibility that alteration of peripheral organ function might have affected brain functions. Thus, focal disruption of Hnmt in adult mice using an adeno-associated virus vector should be utilised to determine brain regions responsible for the aggression and sleep–wake abnormalities. It is also of interest to examine the behavioural changes of KO mice under stressful conditions such as sleep deprivation and social defeat stress so that better understanding of the role of HNMT in psychiatric diseases can be gained. Moreover, the role of Hnmt in rodent models of neurodegenerative disorders including PD and AD should be examined because several clinical studies showed alteration of the histaminergic nervous system in these disorders. Further studies are essential to confirm and extend our findings in rodent models and, eventually, in humans.

6. Future Perspectives

As mentioned above, dysfunction of the histaminergic nervous system is associated with various neuropsychiatric disorders including narcolepsy, AD, Tourette’s syndrome, eating disorders, and depression. Animal experiments using H₃R inverse agonists or HNMT inhibitors suggest the therapeutic effect of histamine elevation on brain function [96]. Dietary intake of histidine, as a precursor of histamine, also improved memory functions via the histamine system in rodents (unpublished observation) and ameliorated feelings of fatigue in humans [97]. The marked effect of HNMT on brain histamine concentration indicates the strong therapeutic potential of HNMT inhibitors against brain diseases. Hnmt disruption enhanced brain histamine levels by >6-fold [31], whereas the impact of H₃R antagonists on brain histamine was quite weak (<2-fold) [98]. Injection of a H₃R inverse agonist ciproxifan transiently increased histamine release in mouse brains; however, the reduced histamine release was observed 24 h after the injection, possibly due to the development of receptor hypersensitivity or endogenous histamine deletion [99]. Theoretically, HNMT inhibitors do not induce histamine depletion. These results indicate that HNMT inhibitors might have some advantages over H₃R inverse agonists. Nevertheless, HNMT inhibitors might increase peripheral tissue histamine levels and exacerbate histamine-related diseases such as allergic rhinitis, urticaria, and gastric ulcers, as HNMT is also expressed in peripheral tissues. Several reports have shown that C314T polymorphisms of the HNMT gene is not associated with allergic asthma and rhinitis [100,101,102], although a case-control study in Poland indicated the association of this polymorphism with asthma [103]. We confirmed that Hnmt disruption did not affect histamine concentration of the skin and stomach [31]. Thus, adverse effects induced by HNMT inhibition in the peripheral tissues might be negligible, although we could not rule out the possibility that HNMT inhibitors might aggravate allergic rhinitis, skin diseases, and peptic ulcers. Taken together, it is intriguing to find novel HNMT inhibitors with high specificity and sufficient BBB permeability. We hope that prominent HNMT inhibitors will help many patients who suffer from brain diseases.

Author Contributions

Wrote the manuscript: T.Y., T.N. and K.Y.

Funding

This work was supported by a Grant-in-Aid for Scientific Research (A) (26253016) from the Japan Society for the Promotion of Science (JSPS) and by a Grant-in-Aid for Scientific Research (C) (18K06886), also from JSPS. We also acknowledge the support of the Nishinomiya Basic Research Fund, Japan and Takeda Science Foundation, Japan.

Conflicts of Interest

The authors declare no conflict of interest.