The histamine H₃ receptor: an attractive target for the treatment of cognitive disorders

Abstract

The histamine H₃ receptor, first described in 1983 as a histamine autoreceptor and later shown to also function as a heteroreceptor that regulates the release of other neurotransmitters, has been the focus of research by numerous laboratories as it represents an attractive drug target for a number of indications including cognition. The purpose of this review is to acquaint the reader with the current understanding of H₃ receptor localization and function as a modulator of neurotransmitter release and its effects on cognitive processes, as well as to provide an update on selected H₃ antagonists in various states of preclinical and clinical advancement. Blockade of centrally localized H₃ receptors by selective H₃ receptor antagonists has been shown to enhance the release of neurotransmitters such as histamine, ACh, dopamine and norepinephrine, among others, which play important roles in cognitive processes. The cognitive-enhancing effects of H₃ antagonists across multiple cognitive domains in a wide number of preclinical cognition models also bolster confidence in this therapeutic approach for the treatment of attention deficit hyperactivity disorder, Alzheimer's disease and schizophrenia. However, although a number of clinical studies examining the efficacy of H₃ receptor antagonists for a variety of cognitive disorders are currently underway, no clinical proof of concept for an H₃ receptor antagonist has been reported to date. The discovery of effective H₃ antagonists as therapeutic agents for the novel treatment of cognitive disorders will only be accomplished through continued research efforts that further our insights into the functions of the H₃ receptor.

Introduction

There exist four distinct histamine receptor subtypes (H₁, H₂, H₃ and H₄) that mediate the many physiologic functions of endogenous histamine. Two of these, the H₁ and H₂ receptors, have been important drug targets with highly effective and clinically beneficial therapeutic agents designed to block effects mediated by these receptors. Classical antihistamines such as chlorpheniramine, fexofenadine and desloratidine have been developed that very effectively treat allergic responses mediated by histamine activation of H₁ receptors. The histamine H₂ receptor has also proven to be a therapeutically important drug target, and selective H₂ antagonists such as ranitidine and cimetidine have been developed that treat gastric ulcers through the blockade of gastric acid secretion. The histamine H₃ receptor represents yet another histamine receptor that is a very attractive CNS drug target and has generated intense research efforts in both academic and industrial laboratories in an effort to identify potent and selective H₃ receptor antagonists. Originally described as a presynaptic autoreceptor that inhibits histamine release in the brain (Arrang et al., 1983), it was subsequently shown to also regulate the release of other important neurotransmitters via a parallel role as a heteroreceptor (Schlicker et al., 1988, 1989, 1993; Clapham and Kilpatrick, 1992; Blandina et al., 1996). To date, preclinical research with potent and selective H₃ antagonists suggests that this class of agents may offer a novel therapeutic approach for the treatment of a variety of cognitive disorders including attention deficit hyperactivity disorder (ADHD), Alzheimer's disease (AD) and schizophrenia. The aim of this paper is to review some of the important recent advances in understanding the molecular and functional aspects of the H₃ receptor with respect to the role of this receptor in cognition. In addition, the preclinical properties of some H₃ receptor antagonists that have recently advanced into human clinical studies for cognitive disorders will be highlighted.

Histamine H₃ receptor: isoforms, localization, pharmacology and signalling

Like the other members of the histamine receptor family, the histamine H₃ receptor is a G-protein-coupled receptor (GPCR, the drug and histamine receptor nomenclature used in this review conforms with the Br J Pharmacol Guide to Receptors and Channels; Alexander et al., 2007). Much recent progress has been made in the basic understanding of the structure, localization, pharmacology and signalling properties of H₃ receptor isoforms since the original cloning and characterization of the histamine H₃ receptor in 1999 (Lovenberg et al., 1999). The H₃ receptor exhibits highest homology (∼60% in the transmembrane domains) to the most recently cloned histamine H₄ receptor but much lower homology to other GPCRs including the H₁ and H₂ receptors (∼20% homology) (Hancock et al., 2003; Leurs et al., 2005).

Isoforms and localization

Whereas the full-length H₃ receptor is described as consisting of 445 amino acids, alternative splicing of the receptor gene results in at least 20 possible human H₃ receptor mRNA isoforms identified by reverse transcription (RT)-PCR. These isoforms exhibit variable amino- and carboxyl-termini lengths, truncations of the third intracellular loop and deletions of transmembrane domains. To date, eight of these recombinant human H₃ receptor isoforms (H₃(445), H₃(453), H₃(415), H₃(413), H₃(409), H₃(373), H₃(365) and H₃(329)) have been shown to be functionally competent based upon either binding or signalling assays when expressed in heterologous cell expression systems (Table 1). The pharmacology and functionality of these isoforms will be further elaborated below. All eight functional H₃ receptor isoforms share the same transmembrane domains with differences arising from modifications of the amino and carboxyl termini and truncations of the third intracellular loop. The remaining 12 human H₃ receptor isoforms are either non-functional or their biological activity is yet to be determined. Many of these receptors have alterations within their transmembrane domains due to deletions or novel stop codons and therefore would not be expected to exhibit more typical H₃ receptor pharmacology and function; however, their relevance to physiological responses has not been fully probed and so remains to be determined.

Table 1.

Summary of known functional human and rat splice variants of H₃ receptors

Isoform	Brain localizationa	Binding	Signalling	References
Human
H3(445)	Caudate, Cb, Th, Amg, Hipp, SN, FrCx, Hyp, cc, sp. cord	Yes	↑ GTPγS, ↓ cAMP, ↑ MAPK	Lovenberg et al. (1999); Coge et al. (2001); Tardivel-Lacombe et al. (2001); Wellendorph et al. (2002); Baranowski et al. (2006); Bongers et al. (2007a, 2007b)
H3(453)	ND	Yes	↓ cAMP
H3(415)	Caudate, Cb, Th, Amg, Hipp, FrCx	Yes	↑ MAPK
H3(413)	Caudate, Amg	Yes	↑ MAPK
H3(409)	Whole brain	Yes	ND
H3(373)	Th	ND	↑ R-SAT
H3(365)	Caudate, Cb, Th, Amg, Hipp, SN, FrCx, Hyp, cc, sp. cord	Yes	↑ GTPγS, ↓ cAMP, ↑ MAPK
H3(329)	Amg, SN, Cx, Hyp, Th, Cb, caudate, cc, Hipp	Yes	↑ MAPK
Rat
H3(445)	AO, Tu, CPu, Acb, Th, GrCb, Hipp, Cx, Hyp, sp. cord	Yes	↓ cAMP, ↑ MAPK	Lovenberg et al. (2000); Drutel et al. (2001); Morisset et al. (2001)
H3(413)	5 and 6b, CPu, Th, DR, VTM and VLTM neurons	Yes	↓ cAMP, ↑ MAPK
H3(397)	CPu, Tu, 5 and 6b, CA1 and CA2, DT, VMH, TMN, LC, Pk	Yes	↓ cAMP, ↑ MAPK
Monkey
H3(445)	FrCx, Hipp, Amg, caudate, Th, Hyp, Cb	Yes	↑ calcium	Yao et al. (2003); Strakhova et al. (2007)
H3(413)	Caudate	Yes	↑ calcium
H3(410)	Caudate	Yes	↑ calcium

^aAbbreviations: 5 and 6b, layers V and VIb of cortex; Acb, nucleus accumbens; Amg, amygdala; AO, anterior olfactory nucleus; CA1 and CA2, fields CA1 and CA2 of hippocampus; Cb, cerebellum; cc, corpus callosum; CPu, caudate putamen; Cx, cortex; DR, dorsal raphe; DT, dorsal thalamic nuclei; FrCx, frontal cortex; GrCb, granular cells of cerebellum; Hipp, hippocampus; Hyp, hypothalamus; LC, locus coerrulus; ND, not determined; Pk, Purkinje cell layer of cerebellum; SN, substantia nigra; sp. cord, spinal cord; Th, thalamus; TMN, tuberomammilary nucleus; Tu, olfactory tuberculum; VLTM, ventrolateral tuberomammilary nucleus; VMH, ventromedial hypothalamic nuclei; VTM, ventral tuberomammilary nucleus.

Signaling key: R-SAT, Receptor Selection and Amplification Technology; calcium levels were determined by FLIPR (Fluorescence Imaging Plate Reader) in HEK cells co-expressing chimeric Gα_q/i5.

The nomenclature system is based on amino-acid number (in parenthesis).

The distribution of the histamine H₃ receptor has been characterized in the CNS largely on the basis of RNA in situ hybridization (Pillot et al., 2002; Rouleau et al., 2004) and radioligand-binding (Laitinen and Jokinen, 1998; Jansen et al., 2000) studies that have demonstrated expression throughout the brain. H₃ receptor expression is prominent in the basal ganglia, globus pallidus, hippocampus and cortex in humans (Martinez-Mir et al., 1990). Studies investigating the differential localization of human H₃ receptor isoforms using RT-PCR approaches suggest that the H₃(445) and H₃(365) isoforms predominate in many brain areas with approximately equivalent levels of expression (Bongers et al., 2007b). Brain areas in which they are more highly expressed include caudate nucleus, hippocampus, frontal cortex and hypothalamus, among others (Table 1). The levels of expression of the H₃(415), H₃(413) and H₃(329) isoforms are much lower but nevertheless can be detected in caudate nucleus and amygdala by RT-PCR (Table 1).

In rats, nine distinct recombinant H₃ receptor isoforms have been identified, with three of these (H₃(445), H₃(413) and H₃(397)) representing functional receptors (Drutel et al., 2001). The H₃(445), H₃(413) and H₃(397) isoforms as well as H₃(410) (Morisset et al., 2001) constitute four rat H₃ isoforms that differ by alterations in the third intracellular loop. Interestingly, there is no evidence for a rat H₃(365) receptor, a truncated isoform that is expressed in humans. Additionally, the H₃(397) isoform seen in rat is distinct from any seen in humans. Among the non-functional isoforms, H₃(497), H₃(465) and H₃(449) represent isoforms with transmembrane domain 7 truncations that interfere with the expression of H₃(445) but do not possess any H₃ receptor-binding activity themselves (Bakker et al., 2006). In general, the highest expression of the H₃ receptor in rodents is in the cerebral cortex, hippocampal formations, striatum and hypothalamus (Drutel et al., 2001). Studies examining the differential localization of the functional rat H₃ receptor isoforms using in situ hybridization approaches suggest that the H₃(445) and H₃(397) isoforms predominate in many brain areas (Drutel et al., 2001). Both are expressed in olfactory tubercle, but H₃(445) appears to be the major isoform in the nucleus accumbens, thalamus and caudate putamen, whereas H₃(397) is the predominate isoform in hippocampal and hypothalamic regions, locus coeruleus and cortical laminae. Conversely, H₃(413) is expressed in relatively lower abundance in striatum, thalamus and cortical regions.

Non-human primates also express multiple H₃ receptor isoforms including the functional isoforms H₃(445), H₃(413) and H₃(410) as well as an inactive H₃(335) isoform that has a truncated third intracellular loop and transmembrane 5 domain (Strakhova et al., 2007; Table 1). The monkey H₃(445) appears to be the predominant isoform, expressed in multiple brain regions such as the frontal cortex, hippocampus, caudate and hypothalamus. The monkey H₃(445), H₃(413) and H₃(410) isoforms display comparable pharmacology in both binding and functional assays (Strakhova et al., 2007).

Many human and rat brain areas that express H₃ receptor isoforms in relatively high abundance are those involved in cognition (that is, cortex and hippocampus, see below) or subcortical areas (that is, hypothalamus) that project neurons to these cognition-associated brain regions. Therefore, these receptors can function to regulate neuronal activity itself as is seen with histaminergic neurons arising from the hypothalamus or can regulate the release of neurotransmitters at the synaptic level in cognition-associated brain regions as will be elaborated further below. The observation of similar H₃ receptor expression patterns in humans and rats helps support the use of the rat as a preclinical model for testing the procognitive properties of H₃ antagonists. It should also be noted that the limited peripheral expression of the H₃ receptor is likely to reduce the potential for non-CNS side-effect liabilities that may be associated with the H₃ receptor. The potential impact of the differential expression of H₃ receptor isoforms in the brain on the activity of H₃ receptor antagonists is difficult to determine at this time given the large number of isoforms and differences in isoform types across species. Thus, there is a need to increase our understanding of the role of the multiple isoforms on neuronal activity, including the modulation of neurotransmitter release and subsequent effects on behaviour.

Isoform pharmacology and function

The pharmacology of the H₃ receptor has been extensively reviewed (Hancock et al., 2003; Cowart et al., 2004; Celanire et al., 2005; Leurs et al., 2005) and the pharmacological properties of well-characterized H₃ antagonists developed by a number of H₃ receptor research groups are highlighted below. The pharmacology of H₃ receptor ligands at the various H₃ receptor isoforms other than H₃(445) is not well described but differential pharmacological profiles have been noted for the human isoforms, most especially for agonists (Wellendorph et al., 2002; Hancock et al., 2003; Esbenshade et al., 2006b; Bongers et al., 2007b). Comparison of the potencies of H₃ receptor agonists such as histamine, R-α-methylhistamine, imetit and others at the H₃(445) and H₃(365) receptors revealed approximately from 3- to 20-fold greater potencies of these agonists at the H₃(365) receptor than H₃(445) in binding and functional assays (Wellendorph et al., 2002; Bongers et al., 2007b). Interestingly, the increase in GTPγS binding induced by the agonists is greater at the H₃(445) than at the H₃(365) isoform (Bongers et al., 2007b). Both of these findings were attributed to the higher degree of constitutive activity demonstrated by the H₃(365) isoform. The agonist pharmacological profile of the H₃(415), H₃(413) and H₃(329) isoforms closely resembles that for the H₃(445) isoform (Esbenshade et al., 2006c), whereas there appears to be little difference in the pharmacological profile of H₃ antagonists across the human H₃ receptor isoforms.

The H₃ receptor is constitutively active and capable of signalling independently of agonist both in vitro and in vivo (Morisset et al., 2000; Wieland et al., 2001). In a similar manner to the isoform-dependent coupling to signalling pathways, the level of constitutive activity of the H₃ receptor also appears to be isoform dependent. Most notably, of the human isoforms, H₃(365) is the most constitutively active, exhibiting the highest relative degree of basal activity in recombinant systems and is the isoform that provides the largest reversal of basal activity in the presence of inverse agonists (Bongers et al., 2007b; Esbenshade et al., 2007). The potential impact on the differential coupling and constitutive activity of the multiple H₃ receptor isoforms on H₃ antagonist activity is not presently known. However, it has been demonstrated that H₃ inverse agonists can reverse constitutive H₃ receptor-mediated suppression of [³H]histamine synthesis in rat brain cortical slices (Moreno-Delgado et al., 2006) and [³H]histamine release in mouse brain synaptosomes (Morisset et al., 2000). Thus, although it may be important to design H₃ antagonists that can block the agonist activity of endogenous histamine as well as act as inverse agonists to decrease H₃ receptor constitutive activity at native H₃ receptors, no clinical data as yet have demonstrated whether H₃ receptor inverse agonists are superior to antagonists in blocking an agonist response.

Signalling pathways coupled to H₃(445) receptors have been identified using recombinantly expressed receptors where they have been shown to modulate multiple signal transduction pathways (Bongers et al., 2007a). Activation of H₃ receptors can mediate Gα_i/o-protein-coupled inhibition of adenylate cyclase (Lovenberg et al., 1999) and the Na⁺/H⁺ exchanger (Silver et al., 2001) as well as stimulation of GTPγS binding (Morisset et al., 2000; Wulff et al., 2002), phospholipase A₂ (Morisset et al., 2000), mitogen-activated protein kinase (MAPK) (Drutel et al., 2001), GSK-3β and Akt (Bongers et al., 2007a). It should be noted that isoform-dependent H₃ receptor differential activation of signalling pathways (MAPK and adenylate cyclase) has also been shown for both human and rat H₃ receptors (Drutel et al., 2001; Esbenshade et al., 2006c, 2007; Bongers et al., 2007b). Interestingly, several of these signalling pathways have been associated with potential roles in various CNS processes including long-term plasticity (MAPK), neuronal cell death (PLA₂) and neuronal migration/neuroprotection (Akt/GSK-3β) (Bongers et al., 2007a). It should also be noted that direct coupling of these signalling events to H₃ receptors has been demonstrated not only in recombinant systems but also in brain tissues expressing native H₃ receptors. Much remains to be determined concerning the role, whether directly through H₃ receptor activation or indirectly through the modulation of the release of multiple neurotransmitters (see below), of this important CNS receptor on these signalling pathways and their associated central functions.

It has been demonstrated that native H₃ receptors couple to Gα_i/o proteins, activating GTPγS binding in brain tissues (Clark and Hill, 1996; Humbert et al., 2007) and inhibiting adenylate cyclase in striatal slices (Sanchez-Lemus and Arias-Montano, 2004). Additionally, the native H₃ receptor modulates the synthesis and release of histamine (Arrang et al., 1983; Gomez-Ramirez et al., 1998, 2002) and the release of a variety of other neurotransmitters, including ACh, norepinephrine and others (Schlicker et al., 1988, 1989, 1993; Clapham and Kilpatrick, 1992; Blandina et al., 1996). The precise signalling events that contribute to this modulation in neurotransmitter release by H₃ receptors are not well defined; however, it has been demonstrated that histamine suppresses N- and P-type Ca²⁺ channels in dissociated rat tuberomammillary nucleus histaminergic neurons through an H₃ receptor-coupled pertussis toxin-sensitive G-protein-mediated mechanism (Takeshita et al., 1998) Additionally, recent work examining norepinephrine release from cardiac synaptosomes suggests the involvement of protein kinase A and voltage-operated calcium channels (Seyedi et al., 2005). Despite the limited understanding of the neuronal intracellular signalling associated with native H₃ receptors, the role of H₃ receptors in the modulation of neurotransmitter release and the ability of H₃ antagonists to enhance the release of multiple neurotransmitters is well established and is highlighted below.

H₃ receptor modulation of neurotransmitter release

It has been hypothesized that H₃ receptors are specifically located on axon terminals by neurons of multiple neurochemical phenotypes. Although originally described as a presynaptic autoreceptor controlling histamine release (Arrang et al., 1983), the H₃ receptor is also thought to function as a postsynaptic heteroreceptor involving axoaxonic synapses that regulate the release of other neurotransmitters. Whereas their neuronal soma resides exclusively in the posterior hypothalamus, specifically the tuberomammillary nucleus, histaminergic fibres project throughout most regions of the brain, including cortex, striatum, thalamus, hippocampus, hypothalamus, locus coeruleus and spinal cord. By forming synapses with other axon terminals expressing H₃ receptors, release of histamine from these projections can modulate the release of neurotransmitters contained within the postsynaptic terminal. Consistent with autoreceptor inhibition, the release and interaction of histamine with Gα_i-protein-coupled H₃ heteroreceptors on axoaxonic postsynaptic terminals leads to inhibition of neurotransmitter release. Conversely, the inverse agonism associated with H₃ receptor antagonists has been shown to increase release of neurotransmitters that include ACh, dopamine, norepinephrine and serotonin, as supported by growing numbers of in vitro and/or in vivo neurotransmitter release studies (summarized in Table 2).

Table 2.

Summary of reported in vitro and in vivo H₃ receptor ligand-mediated neurotransmitter release

Compounds	ACh	Dopamine	Norepinephrine	Serotonin	Histamine
Histamine		↓ stimulated in vitro (Schlicker et al., 1993)		↓ stimulated in vitro (Schlicker et al., 1988)	↓ stimulated in vitro (Arrang et al., 1983)
RAMH	↓ stimulated in vitro (Clapham and Kilpatrick, 1992)	↓ stimulated in vitro (Schlicker et al., 1993)	↓ stimulated in vitro (Schlicker et al., 1989)		↓ stimulated in vitro (Arrang et al., 1983)
	↓ stimulated in vivo (PFC) (Blandina et al., 1996)
Imetit	↓ stimulated in vivo (PFC) (Blandina et al., 1996)
Immepip	↓ stimulated in vivo (PFC) (Blandina et al., 1996)
Ciproxifan					↑ basal in vivo (PFC) (Horner et al., 2007)
Thioperamide	↑ stimulated in vitro (Clapham and Kilpatrick, 1992)	∅ RAMH ↓ in vitro (Schlicker et al., 1993)	∅ RAMH ↓ in vitro (Schlicker et al., 1989)
		↑ stimulated in vivo (NA) (Munzar et al., 2004)	∅ RAMH ↓ in vivo (Di Carlo et al., 2000)
Clobenprobit		↑ stimulated in vivo (NA) (Munzar et al., 2004)
Burimamide				∅ histamine ↓ in vitro (Schlicker et al., 1988)
Impromidine				∅ histamine ↓ in vitro (Schlicker et al., 1988)	↑ basal in vivo (ant hyp) (Mochizuki et al., 1991)
					↑ basal in vivo (amyg) (Cenni et al., 2004)
A-304121					∅ histamine ↓ in vitro (Esbenshade et al., 2003)
A-317920					∅ histamine ↓ in vitro (Esbenshade et al., 2003)
ABT-239	↑ basal in vivo (PFC) (Fox et al., 2005)	↑ basal in vivo (PFC) (Fox et al., 2005)			∅ histamine ↓ in vitro (Esbenshade et al., 2005)
BF2.649	↑ basal in vivo (PFC) (Ligneau et al., 2007b,)	↑ basal in vivo (PFC) (Ligneau et al., 2007b)
GSK189254	↑ basal in vivo (PFC) (Medhurst et al., 2007a)	↑ basal in vivo (PFC) (Medhurst et al., 2007a)	↑ basal in vivo (cing ctx) (Medhurst et al., 2007a)

Abbreviations: amyg, amygdala; ant hyp, anterior hypothalamus; cing ctx, cingulate cortex; NA, nucleus accumbens; RAMH, R-α-methylhistamine; PFC, prefrontal cortex.

Keys: ↑, increased NT release; ↓, decreased NT release; ∅, blocked pharmacological effect.

Histamine

Functioning as an excitatory neurotransmitter involving postsynaptic stimulation of H₁ and H₂ receptors throughout the CNS, histamine plays a key role in attention and vigilance (Passani et al., 2000, 2004; Blandina and Passani, 2006). Activation of secondary pathways involved in attention may also be linked to histaminergic neurotransmission, in particular the noradrenergic reticular formation evolving from the locus coeruleus that receives histamine terminal projections from the tuberomammillary nucleus. In this regard, pharmacological-evoked histamine release may afford efficacy in attentional disorders such as ADHD. Whereas release of several different neurotransmitters can be mediated through the H₃ receptor, described below, histamine was first hypothesized to control its own release through the interaction with presynaptic H₃ autoreceptors. In initial studies, incubation with histamine or H₃ receptor agonists inhibited potassium-evoked release of [³H]histamine from rat cortical slices, whereas H₃ receptor antagonists had a facilitatory effect on the stimulated release (Arrang et al., 1983). Several subsequent studies have similarly demonstrated H₃ receptor regulation of histamine release in vitro, including the selective H₃ receptor antagonists A-304121, A-317920 and ABT-239 that competitively reversed histamine-mediated inhibition of [³H]histamine release from rat brain cortical slices (Esbenshade et al., 2003, 2005).

The first report of histamine release in the whole animal was demonstrated in the hypothalamus of thioperamide-treated rats (Itoh et al., 1991; Mochizuki et al., 1991). H₃ receptor antagonism produced by systemic administration of GT-2016 was subsequently reported to increase histamine in the parietal cortex of awake, freely moving rats (Tedford et al., 1995). Ciproxifan, along with the ADHD agents methylphenidate and atomoxetine, was shown to increase extracellular histamine levels in rat prefrontal cortex (Horner et al., 2007). These later findings raised the intriguing possibility that efficacy associated with ADHD agents such as methylphenidate and atomoxetine may in part involve increased histaminergic tone, supporting the therapeutic potential of H₃ antagonists in the treatment of attention disorders. In studies using H₁ receptor knockout mice, ciproxifan increased wakefulness only in wild type, yet in both genotypes ciproxifan increased histamine release in the frontal cortex, supporting H₃ receptor antagonist-evoked histamine release and subsequent H₁ receptor-mediated vigilance (Huang et al., 2006). Additionally, evoked histamine release has been demonstrated in the basolateral amygdala following local thioperamide administration (Cenni et al., 2004). Taken together, these studies suggest that increased release of histamine by H₃ receptor antagonists may act as indirect H₁ and H₂ receptor agonists enhancing histaminergic neurotransmission within the brain with the potential to augment attention in cognitive disorders such as ADHD and AD.

Acetylcholine

Cholinergic transmission represents an essential neurophysiological component in cognitive functioning. One recognized therapeutic approach to improve cognitive deficits associated with neurodegenerative disorders such as AD is the development of agents capable of increasing extracellular concentrations of ACh in brain regions associated with cognition (for example, hippocampus and prefrontal cortex). The clinical success of this approach is exemplified by acetylcholinesterase inhibitors such as donepezil (Aricept), widely used in the treatment of AD. Early in vitro evidence for H₃ receptor-mediated regulation of ACh neurotransmission was demonstrated in experiments examining potassium-stimulated tritium release from slices of entorhinal cortex preloaded with [³H]choline (Clapham and Kilpatrick, 1992). Whereas the H₃ receptor agonist R-methylhistamine inhibited release, the H₃ receptor antagonist thioperamide augmented potassium-stimulated [³H]ACh release. Blandina et al. (1996) later provided the first in vivo evidence for a role of histamine H₃ receptors in regulating ACh release in rat cortex, which receives cholinergic input originating primarily from the nucleus basalis. In a series of in vivo microdialysis experiments, it was demonstrated that histamine and the H₃ receptor agonists R-α-methylhistamine, imetit and immepip locally administered through the microdialysis probe inhibited potassium-evoked ACh release in the frontoparietal cortex (Blandina et al., 1996). The inhibition was prevented by the H₃ antagonist clobenpropit, but not by an H₁ antagonist (tripolidine) or H₂ antagonist (cimetidine). In addition, R-α-methylhistamine and imetit inhibited potassium-evoked ACh cortical release when administered systemically (i.p.) at doses shown to disrupt short-term memory performance, suggesting a potentially important role for the H₃ receptor as a target for neurodegenerative disorders associated with impaired cognitive function. H₃ receptors also regulate ACh release in other brain regions including the hippocampus where systemic administration of R-α-methylhistamine decreased electrically evoked ACh release, whereas thioperamide enhanced ACh release in the hippocampus (Mochizuki et al., 1994). Similarly, when administered locally into the medial septum diagonal band, R-α-methylhistamine decreased, whereas thioperamide augmented hippocampal ACh release (Bacciottini et al., 2002). Studies have also shown that in the basolateral amygdala, local administration of H₃ receptor agonists enhance ACh release from this brain region at doses corresponding with enhanced memory retention in a contextual fear-conditioning paradigm (Cangioli et al., 2002), whereas H₃ receptor antagonists reduce ACh release (Passani et al., 2001) with a dose-associated impairment in memory retention. Since these initial studies, there have been reports of novel histamine H₃ receptor antagonists increasing ACh release as demonstrated by in vivo microdialysis associated with procognitive efficacy in behavioural animal models. The selective histamine H₃ receptor antagonist ABT-239 increased ACh release in the frontal cortex and to a lesser extent in the hippocampus at doses (0.1–3 mg kg⁻¹) similar to those producing efficacy in rat cognition models (Fox et al., 2005), as described below. Similarly, the novel histamine H₃ receptor antagonists BF2.649 (Ligneau et al., 2007b) and GSK189254 (Medhurst et al., 2007a) increased ACh release in the frontal cortex and/or dorsal hippocampus.

Dopamine

Aberrant dopaminergic neurotransmission has long been recognized as a major aetiological component of schizophrenia psychopathology. The primary pharmacological approach to schizophrenia has employed the use of dopamine receptor antagonists for treating the hyperdopaminergic transmission associated with positive symptoms (hallucinations, delusions and thinking disturbances), In contrast, the negative symptoms (apathy, blunted affect and inattention) and cognitive deficits also observed in schizophrenia do not respond well to dopamine receptor antagonists, which in fact are considered to manifest through hypodopaminergic transmission, specifically in the prefrontal cortex. Pharmacological stimulation of dopamine release in the prefrontal cortex is being considered a viable approach in treating negative symptoms and cognitive impairment in schizophrenia, symptoms that are currently not well treated and thus currently representing a significant unmet medical need. There have been several reports indicating that histamine H₃ receptors can regulate dopamine release. The H₃ agonists histamine and R-α-methylhistamine were shown to inhibit preloaded [³H]dopamine release from mouse striatal slices and this effect was blocked by the H₃ antagonist thioperamide, but not by H₁ or H₂ receptor antagonists (Schlicker et al., 1993). In the whole animal, H₃ receptor antagonism produced by systemic administration of either thioperamide or clobenprobit potentiated methamphetamine-induced dopamine release in the nucleus accumbens shell, but had no effect on extracellular dopamine when given alone (Munzar et al., 2004). Administration of the H₃ antagonist ABT-239 by itself increased extracellular dopamine concentrations of dopamine in rat prefrontal cortex, but not in the striatum (Fox et al., 2005). Enhanced dopamine release in rat prefrontal cortex has also been demonstrated with both BF2.649 (Ligneau et al., 2007b) and GSK189254 (Medhurst et al., 2007a). Taken together, these studies support the therapeutic potential of H₃ receptor antagonists for treating negative symptoms and cognitive deficits associated with schizophrenia as defined by hypodopaminergic function in prefrontal cortex.

Norepinephrine

Noradrenergic neurotransmission within the CNS plays an important role in attentional processing and affective behaviours, which is highly regulated through norepinephrine release in cortical and hippocampal regions from axon terminals of neurons located in the locus coeruleus. Several psychiatric therapeutics lead to enhanced noradrenergic transmission through various pharmacological means, including inhibition of synaptic reuptake or increased release of norepinephrine. Histamine H₃ receptors expressed on noradrenergic terminals innervating cortical and hippocampal regions may represent a potential target in modulating norepinephrine release. Support for this potential originates from studies demonstrating that R-α-methylhistamine inhibition of [³H]norepinephrine release from rat cortical slices was prevented by the H₃ receptor antagonist thioperamide (Schlicker et al., 1989). Initial rat in vivo microdialysis studies involving both systemic and local administration of thioperamide did not stimulate basal norepinephrine release in the hippocampus, but did prevent the reduction of norepinephrine that was produced by R-α-methylhistamine (Di Carlo et al., 2000). Although these results suggested that norepinephrine release mediated through histamine H₃ heteroreceptors located on noradrenergic terminals may only play a minor role in regulating hippocampal norepinephrine release, it was subsequently demonstrated that oral administration the novel H₃ receptor antagonist GSK189254 increased basal norepinephrine levels in the cingulate cortex of freely moving rats at doses improving cognitive performance (Medhurst et al., 2007a).

Serotonin

Similar to norepinephrine, pharmacological augmentation of extracellular brain serotonin represents a viable approach in the treatment of affective disorders, in particular unipolar depression, as evidenced by the clinical efficacy of selective serotonin reuptake inhibitors. Located in the midbrain nuclei, serotonergic neurons project axons throughout cortical and hippocampal forebrain regions where histamine H₃ receptors are located. Reports of H₃ receptor-mediated serotonin release have been primarily limited to in vitro studies. Inhibition of electrically evoked [³H]serotonin from rat cortical slices by histamine was antagonized by the mixed H₂/H₃ receptor agonist/antagonists burimamide and impromidine, the later evoking release alone (Schlicker et al., 1988). Additionally, in studies utilizing rat midbrain slices, it has been shown that H₃ receptors regulate serotonin release in the substantia nigra pars reticulata where electrically evoked serotonin release was inhibited up to 60% by H₃ receptor agonists such as R-α-methylhistamine and immepip (Threlfell et al., 2004). Interestingly, this effect was reversed by the H₃ receptor antagonist thioperamide but not by antagonists of GABA or glutamate receptors, strongly suggesting a role for histamine and H₃ receptors in the function of the substantia nigra pars reticulata and a potential target for basal ganglia therapies (Threlfell et al., 2004). Whether in vitro demonstration of H₃ receptor-mediated serotonin release translates to significant in vivo effects remains to be determined. Whereas the selective H₃ receptor antagonist GSK189254 was shown to evoke ACh, dopamine and norepinephrine release in the rat cingulate cortex, there was no effect on serotonin (Medhurst et al., 2007a). On the other hand, there is substantial interest in the field for agents that combine H₃ receptor antagonism with serotonin uptake inhibition to increase neuronal serotonin levels (see below).

In summarizing the role of H₃ receptor-mediated neurotransmitter release, experimental results over the last 10 years support the hypothesis that the behavioural effects of H₃ receptor antagonists likely involves the release of various neurotransmitters in brain regions associated with cognitive function. Thus, functioning as ‘indirect' agonists at multiple receptor classes within the CNS provides the potential for H₃ antagonists to treat psychiatric pathologies resulting from reduced neurotransmission. However, much remains unknown as to how the interaction and crosstalk between different neurotransmitters affected by H₃ receptor antagonism contribute to the potential efficacy afforded by these novel CNS agents. Sophisticated microdialysis studies capable of assessing several neurotransmitters at multiple sites simultaneously are warranted and may provide improved understanding of such mechanisms.

H₃ receptors and cognition

Histamine is a biogenic amine that exhibits high affinity for the H₃ receptor and has a demonstrated role in CNS activities including learning and memory. For example, the histaminergic system has been implicated in arousal and attention (influenced in ADHD), AD and schizophrenia. Increases in histamine levels in the brain of patients with AD have been reported in key brain areas such as the frontal cortex and hippocampus (Cacabelos et al., 1989), whereas others have reported a significant reduction in the content of histamine in the hippocampus and other areas (Mazurkiewicz-Kwilecki and Nsonwah, 1989; Panula et al., 1998). These differences between studies could reflect differences in the amount of neuronal damage or disease state (Fernandez-Novoa and Cacabelos, 2001), although no clear relationship between histamine levels and AD, for example, have been demonstrated. Interestingly, recent work has further supported a role for H₃ receptors in AD, demonstrating that H₃ receptor expression remains prevalent in the medial temporal cortex of patients diagnosed with AD, even in advanced stages of the disease (Medhurst et al., 2007a). A substantial body of evidence supports that increasing histaminergic tone can facilitate cognition (De Almeida and Izquierdo, 1986; Kamei et al., 1993; Miyazaki et al., 1995), supporting the utility of drugs that increase histaminergic activity in key brain regions, although there is also some evidence that indicates a decrease in histaminergic tone can increase cognition (Huston et al., 1997). These conflicting reports make it difficult to conclusively demonstrate that the beneficial effects of H₃ receptor antagonists in diverse cognition models are mediated solely by histamine or whether other neurotransmitter systems previously mentioned (for example, ACh, dopamine, etc.) play equally or more important roles.

The neuronal histamine system and specifically H₃ receptors have been suggested as modulators of the sleep–wake cycle and cognitive processes (Passani et al., 2004; Esbenshade et al., 2006a). In general, literature data indicate that the administration of H₃ receptor agonists can impair cognition (see Table 3 for a summary), although see also Rubio et al. (2002). These data, taken together with evidence from histamine H₃ receptor knockout animals that demonstrate enhanced spatial learning and memory in the Barnes maze (Rizk et al., 2004), support a role for H₃ receptors in cognition. As described above, histamine H₃ receptors are an attractive drug target as these receptors modulate neurotransmitter release and the localization and neurochemistry of H₃ receptors make this system uniquely poised to play a role in aspects of learning and memory. Given preclinical evidence suggesting that blockade of histamine H₃ receptors can decrease impulsivity, improve attention, and enhance learning and memory, research has focused on the ability of H₃ receptor antagonists to potentially treat cognitive disorders. These cognitive disorders can be further subdivided into different cognitive domains and much of the preclinical research in these areas has focused on cognition assays that measure the different learning and memory domains thought to be most affected in diseases such as AD. Other domains, such as attention and impulsivity, are likely to be of importance in AD and especially in other patient populations, especially those with ADHD (see Table 3). Broad efficacy has been demonstrated across these multiple domains with H₃ receptor antagonists, even if not all compounds have been tested in each assay or not every H₃ receptor antagonist tested was demonstrated active in all tasks.

Table 3.

Summary of H₃ receptor-mediated cognitive effects across cognitive domains

Compounds	Recognition memory	Spatial memory	Memory consolidation	Working memory	Executive function	Attention/impulsivity	Representative references
Histamine
Histamine	↑ (SR)	↑ (scop. def. RM)	↑ (PA)				De Almeida and Izquierdo (1986); Prast et al. (1996); Chen (2000); Chen and Kamei (2000)
Histidine	↑ (SR)	↑ (scop. def. EPM)	↑ (AA, chronic)				Kamei et al. (1993); Miyazaki et al. (1995)
H3 receptor agonists
Immepip	↓ (SR)						Prast et al. (1996)
RAMH	↓ (OR)	↑ (facilitated SM)	↓ (PA)				Rubio et al. (2002)
Imetit	↓ (OR)		↓ (PA)				Blandina et al. (1996)
H3 receptor antagonists
Thioperamide	↑ (SR)	↑ (scop. def. WM; no effect BM)	↑ (PA senescence, SHR pup models)	↑(scop. def. y-maze)	↑ (scop. def. RM)	↑ (SHR 5-trial IA)	Meguro et al. (1995); Miyazaki et al. (1995); Prast et al. (1996); Chen (2000); Orsetti et al. (2002); Komater et al. (2003, 2005)
Ciproxifan	↑ (SR)	↑ (BM, scop. def. WM)				↑ (SHR 5-trial IA; 5-CSRTT)	Fox et al. (2002, 2005); Day et al. (2007)
GT-2331						↑ (SHR 5-trial IA)	Fox et al. (2002)
ABT-239	↑ (SR, adult/aged)	↑ (scop. def. WM)	ND	↑ (y-maze)		↑ (SHR 5-trial IA)	Fox et al. (2005)
BF2.649	↑ (OR; naïve and scop. def.)						Ligneau et al. (2007b)
JNJ-10181457						↑ (SHR 7-trial IA)	Bonaventure et al. (2007)
GSK189254	↑ (OR)	↑ (aged WM)	↑ (PA)		↑ (executive set shifting)		Medhurst et al. (2007a)

Abbreviations: 5-CSRTT, five-choice serial reaction time test; AA, active avoidance; BM, Barne's maze; EPM, elevated plus maze; IA, inhibitory avoidance; OR, object recognition; PA, passive avoidance; RAMH, R-α-methyl-histamine; RM, radial maze; scop. def., scopolamine deficit; SM, spatial memory; SR, social recognition; WM, water maze.

Keys: ↑, improvement; ↓, impairment.

Attention/impulsivity

Whereas attentional deficits span multiple disease states including AD, schizophrenia and ADHD, of particular relevance to ADHD are deficits in impulsivity. Tests that measure aspects of attention and impulsivity that have been used to evaluate efficacy of H₃ receptor antagonists are the five-trial (or seven-trial) inhibitory avoidance paradigm in spontaneously hypertensive rat (SHR) pups, as well as the five-choice stimulus reaction time test (5-CSRTT). As detailed in Table 3 as well as in multiple publications (Fox et al., 2002; Hancock and Fox, 2004; Esbenshade et al., 2006a), a number of H₃ receptor antagonists, including thioperamide, ciproxifan, ABT-239 and GT-2331, are efficacious in the five-trial inhibitory avoidance in SHR pups (with a recent report also describing the efficacy of JNJ-10181457 in a seven-trial inhibitory avoidance version of the model (Bonaventure et al., 2007)). SHRs exhibit many behaviours commonly observed in patients with ADHD, and as such are often used as a model of ADHD (Davids et al., 2003; Russell, 2007). SHR pups are normotensive at the age of testing in the five-trial and seven-trial inhibitory avoidance assay and thus cognitive deficits are independent of hypertension. Deficits in the SHR may be linked to a reduction in nicotinic-ACh receptors observed in a number of brain regions including cortex, hippocampus, thalamus and striatum (Gattu et al., 1997; Terry et al., 2000). It is also possible that behavioural abnormalities in SHRs may be due to an impaired release of dopamine from nerve terminals in the prefrontal cortex (Davids et al., 2003). As H₃ receptors have been shown to regulate the release of both ACh and dopamine, blockade of H₃ receptors with antagonists would be expected to improve function in either or both cases. The terms attention and impulsivity are quite general, referring to multiple processes. As an example, attention covers selective and divided attention, vigilance and distractability. Impulsivity is typically defined as the inability to withhold a response.

Another way of modelling aspects of attention and impulsivity is using the 5-CSRTT in which individual measures of attention, impulsivity, motivation and motor function can be quantified (Robbins, 2002). The 5-CSRTT is analogous to a test used to assess humans (the Continuous Performance Test), thus data from 5-CSRTT may serve as a useful translational assay for efficacy in these behavioural domains. The 5-CSRTT uses visual cues to predict a food reward; a reward that is presented only when the animal correctly responds to the appropriate stimuli. By measuring the percent correct or incorrect choices or the number of missed responses, attention can be measured. Impulsivity can be measured by assessing the number of responses in between trials (when it is inappropriate to respond). Motor function can be assessed by measuring variables such as latency to respond. Previous studies have demonstrated some conflicting results with H₃ receptor antagonists in this task, with one report indicating efficacy (Ligneau et al., 1998), whereas another study found thioperamide did not reverse a scopolamine-induced deficit (Kirkby and Higgins, 1998). A recent publication found that ciproxifan is efficacious on measures of impulsivity, with some efficacy on measures of attention (Day et al., 2007).

Recognition memory

One of the early domains studied with H₃ receptor antagonists was social memory (a form of short-term recognition memory). Social recognition is frequently impaired in patients with AD, and as such efficacy in this domain may be relevant for domains impaired in patients with AD. Social recognition relies on the retention of memory in rats, in which an adult animal uses olfactory cues to recall a social interaction with a conspecific juvenile. Evidence supporting a role for the histaminergic system in social memory came from an early study demonstrating that i.c.v. administration of histamine facilitated social memory; an effect that was also observed with thioperamide (Prast et al., 1996). Further, recognition recall can be blocked by the inhibition of histamine synthesis (Prast et al., 1996). Another form of short-term recognition memory, object recognition, has been used to evaluate short-term recognition memory. In this task, rodents are assessed for their ability to remember a familiar vs an unfamiliar test object (that is, this type of recognition is less likely to involve the possibility of social performance). As detailed in Table 3, several H₃ receptor antagonists have demonstrated efficacy in the recognition memory domain.

Recognition memory involves multiple brain regions and neuromodulatory systems. Evidence has implicated the hippocampus (Kogan et al., 2000) and entorhinal cortex (Bannerman et al., 2002) in recognition memory, but research suggests that the parahippocampal region (including the entorhinal, perirhinal, parahippocampal/postrhinal cortices) is important in this cognitive domain (Mishkin, 1978). ACh (Winslow and Camacho, 1995) and histamine (Prast et al., 1996) are important neurotransmitters involved in social memory. Taken together, it is likely that the observed improvement of recognition memory with H₃ receptor antagonists in the current studies reflects enhanced ACh and histamine release in these key brain regions and it is possible that this will translate into improved social memory and cognition in patients.

Spatial memory

Spatial memory is another aspect of cognition that is impaired in patients with AD. There are a number of assays that have been used to measure spatial memory, including the Barnes maze or variants of the water maze or radial arm maze. All of these tasks used for evaluating spatial memory have the shared feature of requiring the animal to use local information in navigating through their environment. In addition to the results with H₃ receptor antagonists listed in Table 3, in the Barnes maze, mice lacking H₃ receptors learn spatial cues more readily than do their wild-type counterparts (Rizk et al., 2004). Taken together, these data indicate that blockade of H₃ receptors enhances spatial memory.

There are many different types of spatial memory tasks that have been used to evaluate H₃ receptor antagonists, but in general these assays rely on the hippocampus and the septal–hippocampal pathway. It is likely that the improvements in spatial memory observed with H₃ receptor antagonists, therefore, are due to effects on ACh.

Memory consolidation

One important behavioural assay that has been commonly used to assess cognitive function, especially components of longer term memory consolidation, is the inhibitory avoidance assay. Thioperamide improves memory in this assay in a model of senescence (Meguro et al., 1995) as well as in SHRs (Hancock and Fox, 2004). Further, thioperamide can prevent the deficit induced by scopolamine in this assay (Giovannini et al., 1999), and mice lacking the H₃ receptor are known to be insensitive to the cognitive-impairing effects of scopolamine in this test (Toyota et al., 2002). Therefore, the observation of efficacy of H₃ receptor antagonists in this task implies that they may be acting via modulation of cholinergic function in this cognitive domain. One point of interest is that in this particular assay H₃ receptor antagonists generally have efficacy when the administration is given before training, rather than post-training (Giovannini et al., 1999), suggesting a role for H₃ receptors specifically in memory acquisition, rather than memory recall. Another rodent model where H₃ receptors appear to play a role in memory consolidation is contextual fear conditioning. Local post-training administration of H₃ receptor agonists and antagonists into the basolateral amygdala augment and reduce ACh release, respectively, with corresponding enhancement and impairment of memory retention in this model (Passani et al., 2001; Cangioli et al., 2002), suggesting a role for a H₃ receptor-regulated cholinergic component in this cognitive domain.

Working memory

Working memory is a cognitive domain that is impaired in schizophrenia as well as AD and ADHD. Working memory generally refers to a memory system that is used to temporarily store information required to successfully complete complex cognitive tasks. H₃ receptor antagonists have demonstrated efficacy in this domain (Table 3), and examples of frequently used assays include the y-maze, variations of the water maze, radial arm maze and tests of delayed match-to-sample memory. Increased error rates (radial arm maze) have been associated with declining histamine content in key brain regions, and these deficits can be reversed by either i.c.v. histamine or thioperamide (Chen et al., 1999). As another example, thioperamide can reverse scopolamine-induced deficits in the y-maze task (Orsetti et al., 2002).

Executive function

A recent report by Medhurst et al. (2007a) is the first support for efficacy of H₃ receptor antagonists in the cognitive domain of executive function, a domain impaired in patients with schizophrenia. In attentional set-shifting assays, or models of cognitive flexibility, rats are required to learn that a previously rewarded rule no longer is the correct strategy and they must now use a new rule to obtain a reward. The H₃ receptor antagonist GSK189254 was found efficacious; improving both reversal learning and attentional set-shifting (Medhurst et al., 2007a). The demonstration of efficacy with H₃ receptor antagonists in a model of cognitive flexibility and executive function could have implications for the treatment of cognitive deficits of schizophrenia.

H₃ receptor antagonists have demonstrated efficacy across a number of cognitive domains relevant to disorders such as AD, cognitive deficits associated with schizophrenia and ADHD. Many groups in the H₃ field are currently evaluating the safety and efficacy of H₃ receptor antagonists in the clinic, and the results of these studies will provide more specific evidence of the role of H₃ receptors, and potentially the broader role of histamine in cognitive functioning. A selection of prominent H₃ antagonists (Table 4), which have been used preclinically as tool compounds to elucidate the effects of H₃ receptor blockade on cognition have been identified in patents of companies with extensive activity in the H₃ antagonist arena, or those H₃ antagonists identified as currently being in Phase I/Phase II trials are highlighted in the following section.

Table 4.

Chemical structures of histamine H₃ receptor agonists and antagonists described in the scientific and patent literatures

Properties of H₃ receptor antagonists

ABT-239

ABT-239 (Table 4) is a 2-ethylaminobenzofuran that has high potency in rat (K_i=1.3 nM) and at all other H₃ receptor species (Cowart et al., 2005). The compound is selective for H₃ receptors vs over 80 other CNS receptor sites tested, exhibits good pharmacokinetic properties in several animal species and is broadly effective in diverse behavioural models (Esbenshade et al., 2005; Fox et al., 2005). ABT-239 improves short-term recognition memory in adult rats (0.01 mg kg⁻¹ i.p.) and aged rats (0.3 mg kg⁻¹), and it ameliorates attention deficits in the five-trial inhibitory avoidance model of ADHD in SHR rat pups (0.1 mg kg⁻¹). At higher doses, ABT-239 partially reversed scopolamine-induced deficits in spatial memory in rats in the two-platform water maze (3 mg kg⁻¹) and enhanced N40 gating in DBA2 mice (1 and 3 mg kg⁻¹).

Whereas the specific neurotransmitters mediating the behavioural effects of ABT-239 cannot yet be stated with certainty, in vivo microdialysis studies show that ABT-239 increases cortical and hippocampal ACh at doses (0.1–3.0 mg kg⁻¹) and time courses (30–120 min) that parallel the behavioural efficacy in cognitive models. Significantly, both the in vivo ACh release and behavioural efficacy are retained upon chronic (5 day) dosing in rats. ABT-239 increases the release of histamine in vitro from rat brain synaptosomes, indicating that the release of either or both ACh and histamine could modulate the procognitive effects of ABT-239 in vivo. On the basis of drug exposures (14–507 nM), rat binding potency and protein binding, ABT-239 is effective in the ADHD model at calculated H₃ receptor occupancies from 37 to 96%, values consistent with other close analogues (Cowart et al., 2007). ABT-239 exhibited an unfavourable compound-related cardiovascular profile associated with QT prolongation in monkeys that precluded further clinical development (Hancock, 2006). However, due to its potency and selectivity for the H₃ receptor, it is an excellent tool compound to investigate the role of H₃ receptors in the CNS.

BF2.649

BF2.649 (Table 4) is a piperidinylpropoxyalkylphenyl H₃ antagonist that exhibits potent binding to the rat (K_i=2.7 nM) and mouse (K_i=14 nM) histamine H₃ receptors (Ligneau et al., 2007a, 2007b). It also binds to the human cortical H₃ receptor with IC₅₀=5.3 nM. Systemic injections of BF2.649 increased ACh and dopamine release in the rat cortex at 10 mg kg⁻¹ i.p., and in the object recognition model BF2.649 (15 mg kg⁻¹, i.p.) reversed the cognitive deficits induced by scopolamine, consistent with the ability to release ACh in the rat brain. Behavioural studies also demonstrated that BF2.649 (5 mg kg⁻¹, i.p.) attained a 63% attenuation of methamphetamine-induced hyperlocomotion, but only at lower but not higher doses of methamphetamine (Ligneau et al., 2007a). BF2.649 induced a modest attenuation of MK-801 hyperactivity, had limited efficacy against apomorphine-induced prepulse inhibition deficits and it did not affect apomorphine-induced climbing. Despite the claim that BF2.649 exhibits antipsychotic potential (Ligneau et al., 2007a), based on these preclinical data, the overall effects of BF2.649 in these models were minimal.

Pharmacokinetic parameters measured by a radioreceptor assay indicated that BF2.649 exhibits 84% oral bioavailability in mice. However, these data were not replicated in our lab (with 30, 5 and 2% bioavailability in mice, rat and dog). One recent report indicated a 2-h half-life and good brain penetration in mice using a superior analytical assay but no detailed pharmacokinetic parameters were provided (Ligneau et al., 2007a). This limited oral bioavailability questions the data related to the ability of BF2.649 to increased histamine brain levels after oral administration in mice and the EEG studies conducted in cats. Further studies are needed to demonstrate the role of BF2.649 or its metabolites in preclinical models, with detailed analysis of the plasma levels of parent as well as the main metabolites in these species. BF2.649 is presently under clinical investigation in several Phase II trials for the treatment of schizophrenia, ADHD, dementia and Parkinson's disease. (www.stanleyresearch.org/programs/trialgrants.htm). From the development point of view, our laboratory findings suggest that CYP2D6 inhibition, potent hERG binding and the potential for phospholipidosis would likely be important hurdles for this novel compound.

JNJ compounds

Several novel series of H₃ antagonists have been reported by Johnson & Johnson (Table 4). JNJ-5207852 is a potent dibasic amine antagonist that binds potently to rat H₃ receptors (K_i=1.2 nM), and has good brain penetration. In ex vivo binding studies in mice, the compound had an ED₅₀ of 0.13 mg kg⁻¹, subcutaneously (Barbier et al., 2004). It promotes wakefulness in rodents at 10 mg kg⁻¹ s.c. but not at 1 mg kg⁻¹, and significantly, this effect was absent in H₃ receptor KO mice. This compound appears to have not advanced to the clinic, possibly due to a long brain residency and/or induction of phospholipidosis. JNJ-10181457 is also a dibasic amine antagonist that exhibits high-affinity binding for the rat H₃ receptor (K_i=7.1 nM), promoting wakefulness in rodents and reducing cataplectic attacks in narcoleptic dogs (Bonaventure et al., 2007). JNJ-10181457 improved cognitive performance in SHR pups at 10 mg kg⁻¹ s.c., consistent with data obtained with JNJ-5207852 that reversed pentylenetetrazol-induced memory deficits in several learning and memory tests (Jia et al., 2006).

The data obtained with these potent H₃ receptor antagonists demonstrate that they can promote wakefulness and improve cognition in preclinical animal models. Some of these aforementioned agents did not advance to the clinic (due to different reasons), but JNJ-17216498 is reportedly in Phase II studies in patients with narcolepsy (www.clinicaltrials.gov). A recent publication described the pharmacology of a new class of compounds exemplified by JNJ-28583867, a combined H₃ antagonist and serotonin reuptake inhibitor that increases serotonin, norepinephrine and dopamine release in rat brain (Barbier et al., 2007). This compound showed antidepressant-like activity in mice and promoted wakefulness in rats. In view of these combined behavioural effects in animals, the authors proposed that JNJ-28583867 might be useful for the treatment of several symptoms in depressed patients.

GSK189254

The benzo[d]azepine H₃ receptor antagonist GSK189254 (Table 4) binds with high affinity to the rat and human histamine H₃ receptor (K_i=3 and 0.2 nM, respectively) and increases the release of ACh, norepinephrine and dopamine in rat cortex after oral administration of 1–3 mg kg⁻¹ (Medhurst et al., 2007a). It reversed scopolamine-induced amnesia in the inhibitory avoidance assay at the same dose range and it was also efficacious in other cognitive models (i.e., water maze and object recognition test). Interestingly, despite the high affinity of this compound for the rat H₃ receptor and ex vivo binding studies showing that the ED₅₀ for cortical H₃ receptor occupancy is 0.17 mg kg⁻¹ (oral), efficacy in animal models of cognition is reportedly achieved only at 10-fold higher doses. The published preclinical data are consistent with the ability of H₃ antagonists to improve cognition. However, available clinical information indicates that GSK189254 is presently under clinical evaluation in patients suffering narcolepsy and in an electrical hyperalgesia model in healthy volunteers as a translational model of neuropathic pain (www.clinicaltrials.gov).

Preclinical data on pain models have not been disclosed for GSK189254 but a recent paper described the effects of GSK207040 and GSK334429 in animal models of cognition and pain (Medhurst et al., 2007b). These compounds are potent antagonists at the rat H₃ receptor (K_i=1 and 0.8, respectively) that reversed scopolamine-induced amnesia in the inhibitory avoidance test and significantly reversed capsaicin-induced reduction in the paw withdrawal threshold, indicating that these H₃ antagonists can reduce tactile allodynia. Duloxetine (Cymbalta) has recently been approved for the treatment of neuropathic pain and it has been suggested that its efficacy may be related to its ability to increase serotonin and norepinephrine levels in the brain. As H₃ antagonists can increase neurotransmitter release in the brain, H₃ antagonists may increase these or other relevant neurotransmitters, and be useful for the treatment of neuropathic pain in humans. Despite this initial finding in the capsaicin model, evidence for efficacy in other models of neuropathic pain such as the Chung and Bennett models is needed to support to this notion. Other advanced H₃ antagonists from GSK include GSK239512 in a brain imaging study (www.clinicaltrials.gov), GSK357868 and GSK678103.

MK-0249

Merck is conducting three Phase II clinical trials to determine the efficacy of the H₃ antagonist MK-0249 in AD, ADHD and cognitive deficits of schizophrenia (www.clinicaltrials.gov). The chemical structure of this compound has not been disclosed but several series have been disclosed in patents applied for by Banyu/Merck (Nagase et al., 2005, 2006). A representative compound of the quinazolinone series filed by Banyu is shown in Table 4.

Lilly

Lilly has filed a number of patent applications. One recent application names a compound (Table 4) with a K_i of 11.7 nM for antagonism of R-α-methylhistamine-stimulated [³⁵S]GTPγS binding to the human H₃ receptor (Beavers et al., 2007). The presence and preference for the (S)-2-pyrrolidin-1-ylmethyl-pyrrolidine moiety is reminiscent of another H₃ antagonist chemical series from Novo-Nordisk, including NNC 0038-0000-1202 (Peschke et al., 2006).

Pfizer

Pfizer has more than two dozen published patent applications on diverse genera with claimed therapeutic targets that include CNS diseases. At least one application focused on only a single compound (Table 4) that also names treatment of inflammation and respiratory diseases and combinations with anti-inflammatory agents (Lunn, 2007). The compound had a K_i of 2.6 nM in blocking imetit-dependent inhibition of forskolin-stimulated cAMP synthesis in HEK-293 cells transfected with the human H₃ receptor.

Concluding remarks

There has been considerable progress made in our understanding of the complex biology and properties of the H₃ receptor that has correspondingly led to an increased interest in developing H₃ antagonists to treat cognitive disorders. Although there is indeed great complexity associated with the H₃ receptor including the heterogeneity of isoforms as well as their corresponding differential localization, pharmacological and signalling properties that can complicate drug discovery efforts, considerable efforts have been expended by academic and industrial laboratories to identify potent and selective H₃ antagonists for the treatment of cognitive disorders. Much of the interest in the therapeutic potential of H₃ antagonists arises from the ability of H₃ antagonists to enhance the release of key neurotransmitters such as histamine, ACh, norepinephrine and dopamine that play critical roles in cognitive processing. Additionally, the cognitive-enhancing effects of H₃ antagonists across multiple cognitive domains in a wide variety of preclinical cognition models also bolster confidence in this therapeutic approach for the treatment of ADHD, AD and schizophrenia. Despite these many advances, to date no clinical proof of concept for an H₃ receptor antagonist has been reported. However, a number of clinical studies examining the efficacy of H₃ receptor antagonists for a variety of cognitive disorders are currently underway, so the first reports of the efficacy of these compounds may be reported soon. In the mean time, research efforts are sure to continue to gain further insights into the functions of the H₃ receptor in the quest to discover selective therapeutic H₃ antagonists for the novel treatment of cognitive disorders.

Abbreviations

5-CSRTT

five-choice stimulus reaction time task

AD

Alzheimer's disease

ADHD

attention deficit hyperactivity disorder

RAMH

R-α-methylhistamine

RT

reverse transcription

SHR

spontaneously hypertensive rat

Conflict of interest

We are all employees of Abbott Laboratories.