Pharmacological targeting of the serotonergic system for the treatment of obesity

Abstract

The attenuation of food intake as induced by an increase in serotonergic (5-hydroxytryptamine, 5-HT) efficacy has been a target of antiobesity pharmacotherapies. However, the induction of tolerance and/or side-effects limited the clinical utility of the earliest serotonin-related medications. With the global prevalence of obesity rising, there has been renewed interest in the manipulation of the serotonergic system as a point of pharmacological intervention. The serotonin_2C receptor (5-HT_2CR), serotonin_1B (rodent)/serotonin_1Dβ (human) receptor (5-HT_1B/1DβR) and serotonin₆ receptor (5-HT₆R) represent the most promising serotonin receptor therapeutic targets. Canonical serotonin receptor compounds have given way to a myriad of novel receptor-selective ligands, many of which have observable anorectic effects. Here we review serotonergic compounds reducing ingestive behaviour and discuss their clinical potential for the treatment of obesity.

Early pharmacological manipulations identified an inverse relationship between the biogenic amine neurotransmitter serotonin and food intake. More specifically, a selective reduction in serotonin bioavailability was associated with hyperphagia and subsequent weight gain, whilst diminished food intake was induced by an increase in serotonin efficacy (Saller & Stricker, 1976; Fletcher & Paterson, 1989). Researchers further sought to clarify which of the 14 distinct serotonin receptors (5-HTRs) identified in vertebrates (clustered into 7 ‘families’, 5-HT₁–5-HT₇, based on sequence homology and effector pathways) are critically involved in serotonin's effects on ingestive behaviour. Using pharmacological and genetic tools, the 5-HT_1BR, 5-HT_2CR and 5-HT₆R subtypes were shown to be the principal mediators through which serotonin exerts its anorectic effects in rodents, and as such, these receptors have been investigated as pharmacotherapeutic targets for the treatment of obesity. Here we review the serotonergic compounds currently lending themselves to the treatment and investigation of obesity (Table 1).

Table 1.

Serotonergic compounds inducing hypophagia

Compound	Mechanism	Manufacturer (clinical status)	Selected reference
Inc. bioavail.
5-HTP	5HT precursor	—	Fletcher & Burton (1986)
Fenfluramine	SSRI and releaser	Servier (Withdrawn in 1997)	Halford et al. (2007)
Fluoxetine	SSRI	Eli Lilly (Withdrawn from Phase III)	Yen et al. (1987)
Paroxetine	SSRI	GlaxoSmithKline	Konkle et al. (2003)
Sertraline	SSRI	Pfizer	Nielsen et al. (1992)
Fluvoxamine	SSRI	Solvay Pharmaceuticals	Nonogaki et al. (2007)
Sibutramine	SNRI	Knoll/Abbott (Registered therapy)	Connoley et al. (1995)
5-HT1
CP-93, 129	5-HT1BR agonist	Pfizer	Macor et al. (1990)
CP-94, 253	5-HT1BR agonist	Pfizer	Koe et al. (1992)
RU-24969	5-HT1A/1BR agonist	Organon	Kennett et al. (1987)
5-HT2C
Ro 60-0175	agonist	Organon/Hoffman-La Roche	Martin et al. (1998)
Ro 60-0332	agonist	Organon/Hoffman-La Roche	Martin et al. (1998)
Org-12962	agonist	Organon (Phase II as antidepressant)	Nilsson (2006)
Org-37684	agonist	Organon	Schreiber & De Vry (2002)
VER-3323	agonist	Vernalis/Roche	Bentley et al. (2004)
VER-23779	agonist	Vernalis/Roche	Somerville et al. (2007)
VER-2692	agonist	Vernalis/Roche	Adams et al. (2006)
VER-5584	agonist	Vernalis/Roche	Bentley et al. (2004)
VER-5593	agonist	Vernalis/Roche	Bentley et al. (2004)
VER-8775	agonist	Vernalis/Roche	Dourish et al. (2004)
BVT.933	agonist	Biovitrum (Withdrawn from Phase IIb)	Svartengren et al. (2003a)
BVT-X	agonist	Biovitrum	Lam et al. (2008)
YM348	agonist	Yamanouchi Pharm.	Hayashi et al. (2004a)
APD-356	agonist	Arena Pharm. (Phase IIb completed)	Thomsen et al. (2008)
ATHX-105	agonist	Athersys	Nilsson (2006)
WAY-163909	agonist	Wyeth	Dunlop et al. (2005)
WAY-161503	agonist	Wyeth	Rosenzweig-Lipson et al. (2006)
WAY-629	agonist	Wyeth	Sabb et al. 2004
LY448100	agonist	Eli Lilly	Nilsson (2006)
IL-639	agonist	Bristol Myers Squibb	Nilsson (2006)
PNU-22394	agonist	Pharmacia (Phase I)	McCall et al. (2001)
5-HT1/2
mCPP	5-HT1B/2CR agonist	Now distributed by multiple vendors	Kennett & Curzon (1988)
TFMPP	5-HT1B/2A/2CR agonist	Now distributed by multiple vendors	Kennett et al. (1987)
MK-212	5-HT2R agonist	Now distributed by multiple vendors	Halford et al. (1997)
DOI	5-HT2R agonist	Now distributed by multiple vendors	Simansky & Vaidya (1990)
5-HT6
BVT.74316	antagonist	Biovitrum (Phase I completed)	Heal et al. (2008)
BVT.5182	antagonist	Biovitrum	Svartengren et al. (2003b)
E-6837	partial agonist	Esteve	Fisas et al. (2006)
PRX-07034	antagonist	Epix (Phase I completed)	Shacham et al. (2006)
SB-271046	antagonist	GlaxoSmithKline	Svartengren et al. (2004)
Ro 04–6790	antagonist	Organon/Hoffman-La Roche	Woolley et al. (2001)
SB-399885	antagonist	GlaxoSmithKline	Perez-Garcia & Meneses (2005)
SB-357134	antagonist	GlaxoSmithKline	Perez-Garcia & Meneses (2005)

Abbreviations: SSRI, selective serotonin reuptake inhibitor; SNRI, serotonin and noradrenalin reuptake inhibitor.

Efforts to elucidate a site of action for 5-HTR-mediated ingestive behaviour highlighted the hypothalamus as a key component (Breisch et al. 1976; Fletcher & Paterson, 1989; Heisler et al. 2002). The physiological manifestations of appetite, and subsequent responses to caloric intake, involve the integration of multiple central and peripheral signals at this site. Perturbed feeding behaviour and consequential effects on body weight effectuated by surgical ablation of specific hypothalamic nuclei was the first unequivocal evidence as to the fundamentality of this brain region in modulating appetite. Much of the neurochemistry that underpins this function has begun to be elucidated. Niche populations of peptidergic neurones, principally within the arcuate (ARC), ventromedial (VMH) and paraventricular (PVH) nuclei and lateral hypothalamic area (LHA), form an intricate feeding circuit reactive to numerous appetitive signals, and initiate neuroendocrine and behavioural responses to food intake.

Acute administration of serotonergic compounds altered the expression of such peptidergic appetitive effectors within the hypothalamus, namely an increase in anorectic pro-opiomelanocortin (POMC) mRNA and a decrease in orexigenic neuropeptide Y (NPY) mRNA (Choi et al. 2006), both of which are synthesized within discrete neuronal populations of the ARC. Recently, it has been shown that manipulation of these first order hypothalamic POMC/cocaine and amphetamine regulated transcript (CART) and agouti-related protein (AgRP)/NPY neurones is a mechanism through which serotonergic compounds reduce food intake (Fig. 1). Specifically, the serotonin system concomitantly regulates the antagonistic functions of POMC/CART and AgRP/NPY neurones through neurotransmitter binding of two spatially distinct G-protein coupled receptor subtypes: depolarizing POMC/CART neurones via action at G_q-coupled 5-HT_2CRs (Heisler et al. 2002) and hyperpolarizing AgRP/NPY neurones through action at G_i-coupled 5-HT_1BRs (Heisler et al. 2006; Fig. 1). Furthermore, the anorectic effect of compounds increasing serotonergic bioavailability and 5-HT_2CR and 5-HT_1BR agonists is contingent upon the downstream activation of the melanocortin 4 receptors (MC4Rs) (Heisler et al. 2006; Lam et al. 2008). It is noteworthy that these serotonergic compounds, which are highly effective in reducing food intake, are rendered ineffective by pharmacological or genetic inactivation of this single downstream melanocortin receptor target (Heisler et al. 2006; Lam et al. 2008). These data elucidate that the melanocortin pathway is a key downstream target for serotonergic compound hypophagia. Recent research has further clarified that the key population of MC4Rs influencing appetite is expressed in the PVH and/or amygdala (Balthasar et al. 2005).

Figure 1. Proposed model of a serotonergic pathway modulating food intake.

An increase in serotonin bioavailability (due to food intake or pharmacological compounds such as sibutramine and fenfluramine) or direct agonism of 5HT_2CRs and 5HT_1BRs modulates firing of pro-opiomelanocortin (POMC)/cocaine and amphetamine regulated transcript (CART) and agouti related protein (AgRP)/neuropeptide Y (NPY) neurones within the arcuate nucleus of the of the hypothalamus (ARC). Anorectic POMC neurones expressing 5HT_2CR depolarize on receptor activation and release α-melanocyte-stimulating hormone (α-MSH), which in turn activates second-order melanocortin 4 receptor (MC4R) expressing neurones, principally within the paraventricular nucleus of the hypothalamus (PVH; Balthasar et al. 2005). Concomitant activation of 5HT_1BRs expressed on orexigenic AgRP/NPY neurones within the ARC causes membrane hyperpolarization and subsequent inhibition of neuropeptide release. Inhibitory 5HT_1BR activation also attenuates inhibitory postsynaptic currents onto POMC/CART neurones further potentiating anorexigenesis. Subsequent downstream neuroendocrine signalling promotes satiety and the cessation of food intake.

In addition to serotonergic modulation of neurones containing the melanocortin agonist and antagonist, serotonin may also directly influence the activity of MC4R-containing cells. For example, infusion of serotonin into the PVH of rats reduces food intake (Fletcher & Paterson, 1989). Corticotropin-releasing hormone (CRH) neurones expressed within the PVH are directly innervated by serotonergic projections and both c-fos induction and CRH expression are stimulated by compounds increasing serotonergic efficacy (Liposits et al. 1986; Bovetto et al. 1996; Choi et al. 2006). A subpopulation of CRH neurones have been reported to express MC4Rs (Lu et al. 2003), and it is possible that serotonin may directly influence the activity of these CRH MC4R-expressing cells. Additional research is required to determine if this is the case. Such further research investigating the down- and upstream pathways through which serotonin influences appetite may yield additional pharmacological targets for the treatment of obesity.

Compounds increasing 5-HT bioavailability

Fenfluramine, an amphetamine derivative lacking psychostimulant properties, was synthesized in the 1970s, and followed 20 years later by the more efficacious enantiomer, dexfenfluramine. Both compounds were successfully prescribed (often in combination with phentermine) as anorectic treatments for obesity until their withdrawal from clinical use in 1997, due to corollary incidences of cardiopulmonary complications. Mechanistically, these drugs are analogous to amphetamine, causing reversal and blockade of the serotonin transporter and a consequential increase in serotonin efflux and synaptic persistence (Crespi et al. 1997). Furthermore, the major metabolite of fenfluramine, norfenfluramine, is a 5-HT_2CR agonist (Gibson et al. 1993). Genetic and pharmacological studies demonstrated that action at the 5-HT_2CRs and 5-HT_1BRs is required for fenfluramine to influence ingestive behaviour (Lucas et al. 1998; Vickers et al. 1999, 2001; Simansky & Nicklous, 2002).

The effect of fenfluramine on body weight has been convincingly established by a number of experimental and clinical analyses (for a more extensive review see Halford et al. 2007). In obese rodents and humans, fenfluramine administration attenuated body weight gain in a manner consistent with a more rapid onset of satiety and, in the case of humans, suppressed premeal hunger (Fisler et al. 1993; Halford et al. 2007). In chronic fenfluramine trials, obese patients attained maximal weight reduction by 24 weeks, but demonstrated a considerable re-accumulation of body weight upon cessation of treatment after 3.5 years (Fernstrom & Choi, 2008). A similar effect was seen during intermittent chronic fenfluramine administration in rodents (5 days of drug followed by 5 days no drug, repeated 10 times), inasmuch as hypophagia and weight loss were only observed for the first 2 days of each treatment period (Choi et al. 2002). During no-drug periods animals regained body weight, due to increased food intake, before losing it again during the successive treatment period (Choi et al. 2002).

Administration of exogenous serotonin and its precursors also elicits a potent anorectic effect in humans and rodents. Injections of serotonin directly into the PVH of rats resulted in a significant decrease in food intake due to a reduction in meal size and duration (Fletcher & Paterson, 1989). However, serotonin itself fails to penetrate the blood–brain barrier and as such is of little therapeutic salience. By contrast, peripheral administration of 5-hyroxytryptophan (5-HTP), the carboxylated precursor of serotonin, successfully suppressed food intake (Fletcher & Burton, 1986). Within a clinical context, 5-HTP has demonstrated observable hypophagic effects in obese subjects and an associated reduction in body weight; moreover, the attenuation of caloric intake was correlated with an enhancement of within-meal satiety (Ceci et al. 1989). Despite these results, 5-HTP is not a registered antiobesity therapy. However, its amino acid precursor, tryptophan, has demonstrated anorectic effects in a clinical setting and is available as an adjunctive antidepressant medication within the UK (Halford et al. 2007).

Another class of drugs that augment serotonin bioavailability are the serotonin reuptake inhibitors. Via the blockade of monoamine transporters, these drugs promote the synaptic persistence of serotonin and thus sustained activation of postsynaptic receptors. Such compounds are more generally associated with the treatment of depression or anxiety; however, in light of their indiscriminate effect on serotonin concentrations, many also exhibit anorectic properties in rodents. For example, the selective serotonin reuptake inhibitor (SSRI) fluoxetine (Prozac®) significantly reduced food intake and body weight gain in the rat (Yen et al. 1987; Heisler et al. 1997; Heisler et al. 1999). Other SSRIs demonstrating therapeutic potential for obesity treatment include fluvoxamine (Nonogaki et al. 2007), paroxetine and sertraline (Halford et al. 2007).

Sibutramine (Reductil®) is a serotonin and noradrenalin reuptake inhibitor (SNRI) and a registered antiobesity treatment. Rodent studies demonstrated that sibutramine enhanced satiety (Halford et al. 1995) and induced significant hypophagia and weight loss (Connoley et al. 1995). Interestingly, studies using obese models (ob/ob mice and Zucker rats) revealed that chronic administration of sibutramine can elicit significant weight loss without a prolonged effect on food intake, suggesting that this compound also influences energy expenditure (Connoley et al. 1995; Day & Bailey, 1998; Golozoubova et al. 2006). Consistent with rodent work, clinical studies of sibutramine administration demonstrated that through the enhancement of satiety, chronic dosing leads to attenuated food intake and subsequent weight loss in the order of 10% (for review see Halford et al. 2007). Furthermore, these effects could be potentiated by a conjunctive low calorie diet, cogently demonstrating the need for life-style counselling in addition to pharmacological intervention (Apfelbaum et al. 1999). In long-term human studies, maximal weight loss was observed at 6 months followed by a stabilized body weight significantly lower than baseline levels or placebo controls (McNeely & Goa, 1998). During one such trial, food intake was still significantly reduced after 10 months of treatment, indicating the continued efficacy of the drug (Barkeling et al. 2003). At present, sibutramine represents the standard in serotonergic antiobesity therapeutics and is the yard stick by which the efficacy of newly developed compounds will be judged.

Compounds targeting serotonin receptors

5-HT_1AR and 5-HT_1BR compounds

5-HT_1AR activation results in membrane hyperpolarization via G_i mediated inhibition of cAMP production (for review see Lam & Heisler, 2007). 5-HT_1ARs, principally located on serotonergic efferents and cell bodies, are characterized autoreceptors which suppress action potential firing. Consistent with this function, administration of the 5-HT_1AR agonist 8-OH-DPAT elicited an increase in rodent feeding (Hutson et al. 1988). A perhaps more critical role for the 5-HT_1ARs in the serotonergic regulation of anxiety and depression has been supported by murine genetic studies (Abenhaim et al. 1996; Heisler et al. 1998; Parks et al. 1998; Ramboz et al. 1998; Gross et al. 2002), and drug discovery efforts related to this receptor have thus far not been focused on obesity.

Of the rodent 5-HT₁Rs, it is the 5-HT_1BR subtype that is the most intimately associated with appetitive control. The human homologue of the rodent 5-HT_1BR is the 5-HT_1DβR. 5-HT_1BR knockout mice exhibit an approximate 9–17% elevation in body weight compared to wild-type littermates and an increase in feeding that is consistent with increased body weight (Bouwknecht et al. 2001; see Table 2). 5-HT_1BR knockout mice also displayed attenuated responses to the anorectic effect of fenfluramine and the classic 5-HT_1A/1BR agonist RU24969 (Lucas et al. 1998). Complementing these genetic studies, the selective 5-HT_1BR pyridine agonists CP-93,129 and CP-94,253 have both demonstrated significant acute anorectic effects in rodents (Macor et al. 1990; Koe et al. 1992), effects blocked by pretreatment with 5-HT_1BR antagonists (Lee et al. 2002; Heisler et al. 2006). Prolonged administration of CP-94,253 decreased food intake in rats and reduced body weight gain, although drug desensitization was apparent within a week (Koe et al. 1992). CP-93,129, despite being significantly more selective than CP-94,253, demonstrated poor blood–brain barrier penetration and thus a lack of anorectic properties when peripherally administered (Macor et al. 1990; Lee & Simansky, 1997). The anorectic effect of these 5-HT_1BR agonists appears to involve hyperpolarization of ARC AgRP/NPY neurones and a disinhibition of POMC/CART neurones, and ultimately, activation of the downstream MC4Rs (Heisler et al. 2006).

Table 2.

Feeding and body weight phenotypes of 5-HTR knockout mice

Receptor	Feeding & body weight phenotypes of mutant animals	Additional comments	References
5-HT1A	Increased intake of sucrose solution in females (potentially sex-hormone related). No alterations in homecage feeding or body weight reported.	Increased anxiety related behaviours and reduced depression-related behaviours	Heisler et al. (1998); Parks et al. (1998); Ramboz et al. (1998); Bechtholt et al. (2008)
5-HT1B	Mildly increased body weight and relative increase in food intake	Reduced sensitivity to d-fen and mCPP induced hypophagia. Compensatory reduction in 5HT2CR function	Lucas et al. (1998); Bouwknecht et al. (2001); Clifton et al. (2003); Lee et al. (2004)
5-HT2A	No alterations in homecage feeding, novelty suppressed feeding or body weight reported.	Increased anxiety related behaviours	Weisstaub et al. (2006)
5-HT2B	Not reported	Mutants died perinatally due to incomplete heart development	Nebigil et al. (2000)
5-HT2C	Marked hyperphagia throughout life and increased body weight gain from around 12 weeks	Increased locomotor activity and subsequent age-dependent reductions in energy cost of physical activity. Dysregulation of HPA axis and reduced anxiety-related behaviour	Tecott et al. (1995); Nonogaki et al. (2003); Heisler et al. (2007a,b)
5-HT3	No observed differences in body weight or food intake	Dysregulation of the HPA axis in response to stress	Bhatnagar et al. (2004)
5-HT4	Modestly reduced weight gain in homecage environment, despite normal food intake. Food intake increased after restraint induced stress	This model of stress induced anorexia has been associated with an increase in CART expression	Compan et al. (2004); Jean et al. (2007)
5-HT5	Normal body weight. No data on food intake reported	Increased exploratory behaviour that was independent of an effect on anxiety	Grailhe et al. (1999)
5-HT6	Normal body weight. No data on food intake reported	Altered responses to ethanol	Bonasera et al. (2006)
5-HT7	Normal body weight. No data on food intake reported	Failed to exhibit the expected hypothermic response to serotonin administration	Hedlund et al. (2003)

Abbreviations: d-fen, dexfenfluramine; HPA, Hypothalamic-Pituitary-Adrenal; CART; cocaine and amphetamine regulated transcript.

5-HT_2CR compounds

In light of compelling pharmacological and genetic evidence, manipulation of 5-HT_2CR signalling has been a focus of serotonin-related obesity drug discovery efforts. The established involvement of this particular receptor subtype in the broader pharmacokinetic activities of dexfenfluramine, sibutramine and 1-3(chlorophenyl)piperazine (mCPP) elegantly highlight its therapeutic pertinence. Furthermore, genetic abrogation of 5-HT_2CR expression in mice engendered marked hyperphagia and middle age onset obesity (Tecott et al. 1995; see Table 2).

Two older compounds, Ro 60-0175 and Ro 60-0332, both described as full 5-HT_2CR agonists, demonstrated significant anorectic properties in a paradigm of palatable food consumption in normal rats (Martin et al. 1998). Continuous subcutaneous infusion of the former revealed that hypophagia only ensued for the first 10 days of a 2 week regime. Moreover, body weight accumulation was markedly reduced compared to vehicle from 2 days postimplantation and persisted throughout the duration of the study (Vickers et al. 2000). The authors suggested that the prolonged weight-reducing effect in the absence of sustained hypophagia was due to drug-induced hyperthermia, leading to increased energy expenditure. Such a thermogenic effect was also observed by Hayashi et al. (2005) during chronic oral administration of Ro 60-0175, although interestingly in this case, attenuated weight gain and hypophagia persisted concomitantly. Consistent with these effects being transmitted via the 5-HT_2CR, SB242084 antagonist pretreatment negated the hypophagic and hyperthermic properties of Ro 60-0175 (Hayashi et al. 2005). However, Ro 60-0175 also appears to have binding affinity for the 5-HT_2BRs, which may limit its therapeutic potential given that action at peripheral 5-HT_2BRs is thought to underlie the cardiopulmonary complications associated with some serotonergic drugs (Martin et al. 1998; Fitzgerald et al. 2000).

The Yamanouchi Pharmaceutical Company's 5-HT_2CR agonist YM348 also elicited a robust dose-dependent decrease in food consumption in Zucker rats, with almost complete aphagia at some concentrations (Hayashi et al. 2004a). However, continuous infusion failed to sustain a hypophagic effect, with treated animals returning to control levels by day 10, although body weight gain remained significantly lower than controls (Hayashi et al. 2004a). This observation is potentially explained by augmented energy and calorie expenditure, as these physiological indices were not desensitized by prolonged YM348 treatment (Hayashi et al. 2004a,b). However, despite promising antiobesity properties, the affinity of this drug for the 5-HT_2BR is only 2-fold lower than that for 5-HT_2CR, again alluding to the potential for adverse clinical side-effects.

Vernalis, in collaboration with Hoffmann-La Roche, have synthesized a number of 5-HT_2CR compounds with demonstrable anorectic activity. VER-3323, VER-5593 and VER-5348 are all indoline-based analogues similar in structure to Ro 60-175, although with greater 5-HT_2CR selectivity. Both subcutaneous and oral administration of these compounds elicited hypophagia in food deprived rats (Bentley et al. 2004). A subsequent study in Siberian hamsters revealed that VER-3323, whilst decreasing food intake, had no observable effect on energy expenditure in the form of thermogenesis or locomotion (Schuhler et al. 2005). A more recent Vernalis compound, VER-2692, exhibited a greater selectivity profile than the indoline derivatives and also resulted in hypophagia when administered acutely to food deprived rats, a response blocked by 5-HT_2CR antagonist pretreatment (Adams et al. 2006). A perhaps more promising Vernalis compound is VER-8775, a piperazine derivative with 15- and 476-fold selectivity for the 5-HT_2CRs over 5-HT_2ARs and 5-HT_2BRs, respectively (Nilsson, 2006). VER-8775 significantly reduced food intake in fasted rats and mice and induced weight loss in diet-induced obese (DIO) mice during a 28-day oral regime (Dourish et al. 2004).

The Wyeth compound WAY-161503 has exhibited potent anorectic and weight reducing effects in both mouse and rat models of obesity (Rosenzweig-Lipson et al. 2006). During chronic trials in Zucker rats, drug tolerance was not observed, with both food intake and weight accumulation remaining significantly decreased throughout the 15 day study. While these results were promising, functional selectivity assays demonstrated that WAY-161503 offered no apparent selectivity over 5-HT_2BRs, although SB215505 (5-HT_2BR antagonist) pretreatment did not ablate the anorectic properties of this drug (Rosenzweig-Lipson et al. 2006). A more recent Wyeth compound, WAY-163909, is characterized as a full 5-HT_2CR agonist and partial 5-HT_2BR agonist, despite being 46-fold more selective for the former (Dunlop et al. 2005). In acute feeding paradigms, WAY-163909 reduced food intake in lean and obese rodent models. Significantly, no tolerance to this compound was observed after a 10-day regime and furthermore animals exhibited a 46% reduction in body weight as compared to controls (Dunlop et al. 2005).

Preclinical data on BVT.933 (generated by Biovitrum) indicated that chronic dosing of this compound was successful in attenuating both food intake and weight accumulation in DIO rats, and furthermore, that these results were correlated with a reduction in adiposity (Svartengren et al. 2003a). The compound was entered into clinical trials, but was withdrawn from Phase IIb. A subsequent compound developed by Biovitrum, BVT.X, has recently been characterized. Lam et al. (2008) demonstrated that acute administration of BVT.X was effective in reducing food intake in two models of murine obesity. However, chronic dosing via osmotic minipumps in leptin deficient obese ob/ob mice revealed that this anorectic effect was relatively short lived, with treated animals returning towards control levels just 2 days through the 7 day trial. Interestingly, despite desensitization to hypophagia, body weight gain was significantly attenuated at the end of the trial. Lam et al. (2008) investigated the central pathways associated with BVT.X's effect on energy balance and observed an up-regulation of POMC expression in the ARC and further determined that downstream activation of MC4Rs was required for this compound to exert its anorectic effect (see Fig. 1). The stimulation of this melanocortin anorexigenic pathway is consistent with a previously asserted model of dexfenfluramine and CP-94 253 anorexia (Heisler et al. 2002, 2006).

One of the most recently reported 5-HT_2CR compounds is Arena's APD-356 (Lorcaserin®). This benzapine molecule exhibited robust functional selectivity over 5-HT_2ARs and 5-HT_2BRs (18- and 104-fold, respectively; Thomsen et al. 2008). Acute oral dosing of APD-356 significantly reduced cumulative food intake in non-deprived rats for up to 22 h, although post hoc analysis attributed this observation to potent hypophagia during the first 2 hours and the absence of subsequent rebound hyperphagia (Thomsen et al. 2008). This effect was readily ablated by 5-HT_2CR antagonist pretreatment, suggesting the functional selectivity of the compound for the 5-HT_2CR. In chronic assessments, APD-356 successfully attenuated both food intake and body weight gain in DIO rats (Thomsen et al. 2008). Although anorectic desensitization was observed at 13 days, body weight remained significantly lower than controls throughout the course of the study, indicating further effects on additional modalities of weight loss. During the 4 week withdrawal period, treated animals exhibited rebound hyperphagia and a subsequent return to control body weight (Thomsen et al. 2008). Significantly, high dosing of both Sprague–Dawley rats and cynomolgus monkeys over 3 months failed to induce any cardiopulmonary side-effects, consistent with the compound's low selectivity for peripheral 5-HT_2BR (Smith et al. 2006). APD-356 is currently in Phase 3 clinical trials having successfully demonstrated weight reducing effects during a 12 week Phase IIb study. Drug treated obese patients exhibited a statistically significant 1.3 kg reduction in body weight. No valvopathies or pulmonary complications were observed in these patients or other clinical cohorts (Smith et al. 2006).

Additional 5-HT_2CR compounds with observed antiobesity properties include PNU-22394, a relatively non-selective Pharmacia/Pfizer compound with anorectic effects in rodents and humans (McCall et al. 2001); ATHX-105 and ATH-188651, from Athersys; and IL639 from Bristol-Myers Squibb which demonstrated a 250-fold selectivity over 5-HT_2A/2BRs and is orally efficacious in rat feeding paradigms (Nilsson, 2006).

5-HT_1/2R compounds

Compounds that demonstrate a degree of 5-HT₁R and 5-HT₂R cross-talk have often proved highly effective in modulating feeding behaviour, most likely due to their concerted action at anorexigenic and orexigenic populations of hypothalamic neurones. For example, mCPP exhibits a preferential affinity for 5-HT_1BR and 5-HT_2CR subtypes and induced robust hypophagia during both acute and chronic regimes (Kennett & Curzon, 1988; Vickers et al. 2003). Significant attenuation of body weight gain was also observed during a 28 day study, with no apparent drug tolerance (Vickers et al. 2003). Furthermore, the use of pair-fed animals in this study strongly implicated the hypophagic effect of the compound as underlying the observed reduction in weight accumulation. In human subjects, mCPP was capable of reducing caloric intake at a test meal by 30% when administered to healthy women, an effect associated with a decrease in premeal hunger ratings (Halford et al. 2007). A similar decline in premeal hunger and body weight was observed in obese men receiving mCPP over a 2 week period, although assessment of food consumption was not reported (Halford et al. 2007). 1-[3-(Trifluoromethyl) phenyl]piperazine (TFMPP), a structurally related analogue of mCPP, also elicited observable hypophagic and weight reducing effects, although drug desensitization was observed during chronic administration (Kennett et al. 1987; Rouru et al. 1993). Hypophagia induced by mCPP and TFMPP is consistent with enhancement of satiety, although at higher doses both drugs have shown to induce hypoactivity (Kennett & Curzon, 1988).

5-HT₆R compounds

5-HT₆R distribution within the CNS includes hypothalamic regions of immediate salience to a role in appetitive control, including the ARC, PVH and VMH (Heal et al. 2008). Concordant with this expression profile, manipulation of 5-HT₆R signalling has been demonstrated to have potent effects on both food consumption and body weight (for review see Heal et al. 2008). It is of note that it is the antagonism of this receptor that is generally associated with its antiobesity function (Heal et al. 2008). One of the earliest 5-HT₆R antagonists reported to induce hypophagia was Ro 04–6790 (Organon/Hoffman-La Roche), which when administered to rats on three consecutive days, resulted in a suppression of body weight gain that persisted for at least 10 days postwithdrawal (Woolley et al. 2001). Interestingly, the same study also demonstrated that food consumption and weight gain were decreased following antisense knockdown of 5-HT₆R, a phenotype absent in traditional 5-HT₆R knockout mice (Woolley et al. 2001; Bonasera et al. 2006).

Biovitrum's BVT.5182 5-HT₆R antagonist has also demonstrated significant anorectic properties during both chronic and acute studies in rat and mouse models of obesity. In DIO mice, repeated subcutaneous doses elicited a 9% and 11% decrease in food consumption and body weight, respectively (Svartengren et al. 2003b). Subsequent studies in obese ob/ob mice and non-obese rats demonstrated that the observed hypophagia was attributable to enhanced satiety (Svartengren et al. 2004). This later study also reported that acute doses of the GlaxoSmithKline 5-HT₆R antagonist SB-271046 dose-dependently suppressed food consumption. However, Phase I trials of SB-271046 were apparently halted due to poor blood–brain barrier permeability. Little has been reported about Biovitrum's second 5-HT₆R compound, BVT.74316, though it did complete a Phase Ia study as an antiobesity therapy (press release of Biovitrum, 19 December 2007). PRX-07034 from Epix Pharmaceuticals has also recently completed Phase I trials (press release of Epix-Pharmaceuticals, 29 October 2007). In preclinical studies, chronic administration successfully reduced food intake and body weight in DIO rats; moreover, it would appear that no drug tolerance was observed over the course of the 5 week study (Shacham et al. 2006).

E-6837 (Esteve) represents a curious pharmacological paradox in that it is classified as a 5-HT₆R partial agonist, but like 5-HT₆R full antagonists, it reduces body weight (for discussion on this issue, see Heal et al. 2008). Specifically, chronic oral administration of E-6837 significantly attenuated body weight by day 3 and continued to do so until reaching a plateau at day 20. At the culmination of the trial, E-6837 treated animals were approximately 6% lighter than a sibutramine control group. Furthermore, whilst the sibutramine group regained control weight by day 44, E-6837 treated animals were still significantly lighter at day 71 (Fisas et al. 2006). Interestingly, E-6837-induced hypophagia lasted for only the first 3 weeks after which significant rebound hyperphagia, commencing at week 1 of the withdrawal analysis, was observed (Fisas et al. 2006). Although no thermogenic effects were reported, further elucidation of E-6837 regulated energy expenditure and metabolism is warranted.

Conclusion

Pharmacological and genetic studies have consistently demonstrated the fundamentality of the serotonergic system in the regulation of appetite and ingestive behaviour. Moreover, they have highlighted the involvement of specific serotonin receptor subtypes in mediating these effects. Historically, pharmacological compounds that augment endogenous serotonergic signalling have proved effective in suppressing food consumption and attenuating body weight gain, physiological effects of immediate clinical salience to the ever increasing incidence of obesity. Furthermore, with many antiobesity compounds falling short of optimal therapeutic efficacy (i.e. due to minimal weight loss, adverse side-effects or long-term resistance), there is a substantial unmet clinical need for novel pharmacological therapies. Currently, the 5-HT_2CR and 5-HT₆R have garnered the most attention in serotonin drug discovery efforts. Both 5-HT_2CR agonists and 5-HT₆R compounds have entered into clinical trials. Critically, the success of any 5-HT_2CR compound is contingent upon its functional selectivity, in particular over the peripherally expressed 5-HT_2BRs. In this regard, 5-HT₆R compounds hold an advantage, inasmuch as they demonstrate little affinity for other serotonin receptor subtypes. Although the neuro-molecular/pharmacological aspects of 5-HT₆R function remain to be fully elucidated, a number of specific compounds have demonstrated significant antiobesity potential in preclinical studies. Given the apparent independence of 5-HT₆R mediated appetite from that of 5-HT_2CR, the possibility of combinational therapies that exploit both pathways bears considering.

Energy homeostasis is a critical balance between the intake and utilization of energy, such that a decrease in the former or an increase in the latter engenders a negative energy balance and an associated loss of weight in the form of fat. Although the weight-reducing effects of most serotonergic compounds are principally mediated via their anorectic properties, evidence from a number of pharmacological studies has indicated that augmented energy expenditure may also play a role (Connoley et al. 1995; Vickers et al. 2000; Hayashi et al. 2004b; Golozoubova et al. 2006; Lam et al. 2008). The melanocortin pathway, which is downstream of the homeostatic serotonergic system, has been demonstrated to promote energy expenditure. The necessity for intact MC4R signalling in the anorectic actions of a number of serotonergic compounds (Heisler et al. 2002, 2006; Lam et al. 2008) may also be of pertinence to the promotion of serotonin mediated energy expenditure. Murine genetic studies have reported the involvement of the MC4R in modulation of a number of expending processes, including locomotion, diet-induced thermogenesis and lipolysis (reviewed in Ellacott & Cone, 2006). Furthermore, the role of these mechanisms in regulating energy homeostasis can be spatially dissociated from the anorectic properties of MC4R (Balthasar et al. 2005). Whether the MC4Rs are a downstream component of serotonergic-mediated energy expenditure remains to be fully clarified.

In conclusion, it is apparent that the serotonergic system offers therapeutic potential in the unremitting battle against obesity. Continued pharmaceutical investment in this cause, together with more spatially refined genetic models, will afford researchers more specific tools with which to unravel the neurology of this increasingly pervasive condition.