Introduction

The overactive bladder syndrome (OAB) is a prevalent condition among adults and occurs with increasing prevalence with advancing age. Currently, no cure is available for this condition, and present symptomatic treatment is based largely on conservative treatment and the use of muscarinic receptor antagonists. However, many patients experience insufficient therapeutic benefit and/or unpleasant side effects including dry mouth and constipation from muscarinic antagonists. Hence, few patients remain on treatment for longer than a few months. In the search for potential alternative OAB treatments, β3-adrenoceptor agonists have emerged as a promising new drug class.

The potential of β3-adrenoceptor agonists in OAB treatment is based on findings that the human bladder expresses β3-adrenoceptors and that this is apparently the predominant if not exclusive β-adrenoceptor subtype mediating human detrusor relaxation; however, other subtypes may contribute to detrusor relaxation in other species such as rats (Michel and Vrydag 2006). Accordingly, β3-adrenoceptor agonists have proven effective in a wide variety of animal models of OAB (Michel et al. 2011). Against this background, several selective β3-adrenoceptor agonists have undergone clinical proof-of-concept studies including mirabegron (also known as YM 178), ritobegron (also known as KUC-7483 and as KUC-7322 for its active metabolite, which has primarily been used in the in vitro studies), and solabegron (also known as GW427353) (Fig. 1). Mirabegron has recently obtained regulatory approval in several countries. Therefore, we will shortly review the pharmacological profile of these three agonists and discuss needs for future pharmacological studies in this area.

Fig. 1
figure 1

Chemical structures of mirabegron, ritobegron (shown as active metabolite KUC-7322), and solabegron

Full size image

Biochemical and cellular studies

To characterize the molecular interaction of the β3-adrenoceptor agonists with their receptor, competition radioligand binding and cyclic AMP accumulation studies have been performed, primarily in cell lines transiently or stably transfected with the corresponding receptor. In Chinese hamster ovary (CHO) cells stably transfected with rat β3-adrenoceptors, mirabegron stimulated cyclic AMP accumulation with an EC50 of 19 nM and an efficacy of 1.0 relative to that of isoprenaline; for β1-adrenoceptors, an EC50 of 610 nM and an intrinsic efficacy of 0.6 was reported, whereas for β2-adrenoceptors efficacy was <0.1 and potency could not be quantified (Hatanaka et al. 2013b). In CHO cells transfected with cynomolgus monkey β3-adrenoceptors, mirabegron stimulated cAMP accumulation with an EC50 of 32 nM and an efficacy relative to isoprenaline of 0.8; the efficacy relative to isoprenaline at β1- and β2-adrenoceptors was only 0.2 and 0.1, respectively (Someya et al. 2010). Two studies on CHO cells transfected with human β-adrenoceptor subtypes were reported with rather similar results. Thus, mirabegron stimulated cyclic AMP accumulation via β3-adrenoceptors with EC50 values of 22 nM (Takasu et al. 2007) and 1.5 nM (Hatanaka et al. 2013b), its efficacy relative to isoprenaline was 0.8 in both studies; in contrast, efficacy at human β1- and β2-adrenoceptors was only 0.1–0.2 in both studies, which did not allow quantification of potency. In human embryonic kidney (HEK) cells stably transfected with human β3-adrenoceptors at a density of 121-fmol/mg protein, mirabegron and isoprenaline exhibited apparent affinities of 55 nM and 34 μM, respectively, in competition radioligand binding studies, and these affinities were not substantially altered in HEK cells transfected with several naturally occurring gene variants of the receptor (Vrydag et al. 2009). In cyclic AMP accumulation experiments in these HEK cells, mirabegron and isoprenaline exhibited EC50 values of 0.93 and 11.2 nM, respectively, and the efficacy of mirabegron relative to isoprenaline was 0.85, which is in good agreement with the findings in CHO cells (Hatanaka et al. 2013b; Takasu et al. 2007). The potency and efficacy of mirabegron also was not affected by genotype in these experiments. A 24-h pretreatment with 10 μM isoprenaline caused desensitization of the cyclic AMP response to freshly added isoprenaline or mirabegron, which primarily consisted of a reduced maximum response for both agonists (Vrydag et al. 2009); this desensitization also was independent of genotype.

The potency, efficacy, and selectivity of KUC-7322, the active metabolite of ritobegron, have been tested for cyclic AMP accumulation in CHO cells transiently transfected with each of the three human β-adrenoceptor subtypes (Maruyama et al. 2012a). These studies revealed EC50 values of 22,000, 2,300, and 73 nM for KUC-7322 at β1-, β2-, and β3-adrenoceptors, respectively, as compared to 0.91, 0.67, and 12 nM, respectively, for isoprenaline. While KUC-7322 was a full agonist relative to isoprenaline at each of the three subtypes, its selectivity β3- vs. β1- or β2-adrenoceptors was 301- and 32-fold, respectively.

Two series of experiments have been reported from stably transfected CHO cells measuring cyclic AMP accumulation via human β-adrenoceptor subtypes as stimulated by solabegron. In the first series, the EC50 of solabegron at human β1-, β2-, and β3-adrenoceptors was 1,259, 3,981, and 3.98 nM, respectively, as compared to 1.0, 0.16, and 3.16 nM, respectively, for isoprenaline, i.e., solabegron was 1,000- and 316-fold selective for β3- vs. β1- or β2-adrenoceptors (Uehling et al. 2006); in these experiments, the efficacy of solabegron relative to isoprenaline was 0.79 at β3-adrenoceptors but 0.02 and 0.05 at β1- or β2-adrenoceptors, respectively. In a second series of experiments, solabegron stimulated cyclic AMP accumulation via human β3-adrenoceptors with an EC50 of 6.9 nM as compared to 3.3 nM for isoprenaline, and the efficacy of solabegron relative to isoprenaline was 0.89 (Hicks et al. 2007). In these experiments in concentrations up to 10 μM, solabegron caused cyclic AMP accumulation of only 4.2 and 8.8 % of isoprenaline values at β1- and β2-adrenoceptors, respectively. Of note, this assay, if anything, had been biased against the detection of β3-adrenoceptor agonism because β1-, β2-, and β3-adrenoceptor expression densities were 1,070, 605, and 164 fmol/mg of protein, respectively.

Absolute potencies of the three β3-adrenoceptor agonists are somewhat difficult to compare across studies due to differences in the cell lines, expression levels, and methods being used. However, it appears that mirabegron was slightly more potent than isoprenaline at human β3-adrenoceptors, whereas KUC-7322 and solabegron were slightly less potent. Relative to isoprenaline, mirabegron, and solabegron were strong partial agonists with efficacies of 0.8–0.9, whereas KUC-7322 appeared to be a full agonist. Whether these minor potency and efficacy differences are of clinical relevance remains unclear. Probably more importantly, all three agonists exhibited considerable selectivity over β1- and β2-adrenoceptors. However, mirabegron and solabegron had only poor efficacy at β1- and β2-adrenoceptors, whereas KUC-7322 had low potency but appeared to be a full agonist. These data imply that in vivo KUC-7322 may be more likely to activate β1- or β2-adrenoceptors than the other two agonists in tissues where a large receptor reserve exists, i.e., where occupation of only a small receptor fraction may already elicit a relevant response; however, this also remains to be tested experimentally as the receptor reserve in human atrium for β-adrenergic stimulation of inotropy is only small (Brown et al. 1992).

Isolated tissue studies

Isolated tissue experiments with β3-adrenoceptor agonists in clinical development have largely been performed with isolated animal (rat, dog, and monkey) and human detrusor preparations (Table 1), but, in some cases, other tissues have also been examined. Similar to isoprenaline, mirabegron was somewhat more potent in human than in rat isolated bladder but, in both tissues, had an efficacy which was at least as large as that of isoprenaline (Takasu et al. 2007). In another study in rat bladder, a comparable potency of mirabegron was detected, but, in that study, the efficacy relative to isoprenaline was slightly smaller (Vrydag and Michel 2009). Another study has compared the potency and efficacy of mirabegron and isoprenaline in isolated detrusor strips obtained from patients without bladder dysfunction, bladder outlet obstruction, and bladder outlet obstruction with documented detrusor overactivity (Svalo et al. 2013). While mirabegron was consistently less potent than isoprenaline in all three groups, its efficacy was comparable to that of isoprenaline; while the potency of isoprenaline was significantly lower in patients with than without bladder outlet obstruction, smaller numerical differences between groups with mirabegron did not reach statistical significance. In the comparison of these data sets, it should be considered that some experiments were performed against carbachol-induced tone, whereas others were performed against KCl-induced tone; a previous work has demonstrated that β-adrenoceptor agonists generally are less potent and less effective against bladder contraction induced by muscarinic agonists than against that induced by KCl (Michel and Sand 2009; Witte et al. 2011).

Table 1 Potency and efficacy of selective β3-adrenoceptor agonists for relaxation of isolated bladder preparations
Full size table

The potency and efficacy of ritobegron, assessed by experiments with its active metabolite KUC-7322, for relaxation of rat bladder has been determined in three studies against passive tension (Maruyama et al. 2012a; Yamazaki et al. 2002) and (Cernecka et al. unpublished observations), and against carbachol- and KCl-induced tone in one study (Cernecka et al. unpublished observations). One of these studies demonstrated that the relaxing effects in rat bladder were not affected by co-application of the β1-adrenoceptor antagonist CGP 20,712A and the β2-adrenoceptor antagonist ICI 118,551 but inhibited when the β3-adrenoceptor antagonist SR 58,894A was added on top of ICI 118,551, yielding an apparent pA2 value of 6.43 for the latter, in line with the relaxation response in rat bladder being solely mediated by β3-adrenoceptors (Maruyama et al. 2012a). Across all of these studies, the potency ratio of isoprenaline/KUC-7322 was 8.1–29.5 (Table 1), whereas it was 80,000 for acceleration of rat atria beating rate and 2,200 for reduction of rat myometrial relaxation, representing prototypical β1- and β2-adrenoceptor responses, respectively (Maruyama et al. 2012a). Accordingly, KUC-7322 was much more potent in the rat bladder than in rat atrium and uterus, whereas the opposite was the case for isoprenaline. KUC-7322 also depressed spontaneous contractions of rat colon with an EC50 of 4.3 nM, and this was inhibited by SR 58,894A, indicating mediation via a β3-adrenoceptor (Yamazaki et al. 2002). In the isolated cynomolgus monkey bladder, both KUC-7322 and isoprenaline were less potent than in the rat, but KUC-7322 remained a full agonist (Maruyama et al. 2012b). Similar to the rat studies, KUC-7322 was considerably more potent for bladder relaxation than atrial beating rate or tracheal relaxation in the monkey, whereas, if anything, the opposite was true for isoprenaline (Maruyama et al. 2012b). In isolated human bladder, KUC-7322 was slightly less potent than in the monkey bladder and much less potent than in the rat bladder in two independent studies performed against either passive tension of KCl-induced tone but remained a full agonist relative to isoprenaline (Igawa et al. 2012). The response against KCl-induced tone was not affected by the β1-adrenoceptor antagonist CGP 20,712A or the β2-adrenoceptor antagonist ICI 118,551,but surprisingly also not by the general β-adrenoceptor antagonist SR 59,230A (Niclauß et al. 2006).

In contrast to the other agonists, bladder relaxation by solabegron has not been reported in rats as the primary animal species but has been in dogs (Hicks et al. 2007). In this species, solabegron caused concentration-dependent relaxation against KCl-induced tone, but a smaller but substantial relaxation was also seen in time control experiments with vehicle addition. While solabegron caused significantly greater relaxation than vehicle, this made quantitative analysis of the solabegron concentration–response curve difficult; however, it appears that the EC50 of solabegron in isolated dog bladder was <10 nM. In single-concentration experiments, relaxation by solabegron was not affected by the β1-selective atenolol or the β2-selective ICI 118,551 but attenuated by the non-selective antagonist bupranolol. The non-selective antagonist SR 59,230A (Niclauß et al. 2006) attenuated the response to solabegron but also caused some relaxation in the absence of solabegron (Hicks et al. 2007). Relaxation responses to SR 59,230A in the absence of exogenous agonist had also been reported in rat and human bladder (Frazier et al. 2011). In one study with isolated human bladder strips, solabegron produced concentration-dependent relaxation with an efficacy comparable with that of isoprenaline, but the resulting curves did not allow calculation of an EC50 value (Biers et al. 2006); in that study, SR 59,230A produced a right shift of the solabegron concentration–response curve. Within that study, spontaneous contractile activity was also noted in some strips and that was also reduced by solabegron. Moreover, solabegron also inhibited field stimulation-induced contraction, which was not mimicked by isoprenaline. In another study with human isolated detrusor strips, solabegron produced relaxation with an EC50 of 1.9 nM with an efficacy similar to that of isoprenaline (Tyagi et al. 2009). Within that study, both solabegron and isoprenaline also inhibited human detrusor tone induced by field stimulation.

In conclusion, mirabegron appears to be somewhat more potent in the human as compared to rodent bladder, whereas the opposite may be the case of the active metabolite of ritobegron. In the human bladder, solabegron (potency in rodents not reported) appears to be more potent than mirabegron and ritobegron. All three agonists appear to have an efficacy comparable to that of isoprenaline (Table 1).

In vivo animal studies

Several in vivo studies have been reported for each of the three agonists, mainly with regard to bladder function. In the urethane-anesthetized rat, i.v. administration of mirabegron produced a dose-dependent reduction in the frequency of rhythmic bladder contractions but did not significantly decrease contraction amplitude in doses up to 3 mg/kg; in contrast, the muscarinic antagonist oxybutynin increased contraction frequency and reduced contraction amplitude (Takasu et al. 2007). In pentobarbital-anesthetized rats, i.v. administration of mirabegron dose-dependently reduced intravesicular pressure and spontaneous bladder contractions; this was accompanied by a dose-dependent heart rate increase of up to 11 % and a mean blood pressure decrease of up to 29 % (Hatanaka et al. 2013a). In rats with cerebral infarction-induced bladder dysfunction, mirabegron dose-dependently increased voided volume and partly normalized it (Hatanaka et al. 2013b). In rats with bladder outlet obstruction, i.v. administration of mirabegron reduced the frequency of non-voiding bladder contractions but did not affect voiding contractions, micturition pressure, threshold pressure, voided volume, residual volume, or bladder capacity (Hatanaka et al. 2013a). Within that study, oxybutynin or tolterodine did not affect the frequency of non-voiding contractions, but oxybutynin significantly increased residual volume, and tolterodine reduced voided volume. In another study in rats with bladder outlet obstruction, non-voiding activity consisted of small and large transients and i.v. administration of both mirabegron and tolterodine reduced the cumulative activity of the large non-voiding contractions but had little effect on the small transients (Gillespie et al. 2012). In that study, mirabegron primarily reduced the frequency of such contractions, whereas tolterodine mainly reduced their amplitude. The authors proposed a model in which both agents act on a pacemaker-like mechanism with cholinergic excitatory and adrenergic inhibitory inputs. In water-loaded, conscious cynomolgus monkeys, oral mirabegron at 1–3 mg/kg reduced micturition frequency and increased voided volume per micturition but did not affect total voided volume (Someya et al. 2010). To improve the mechanistic understanding on mirabegron effects in the bladder, a study was performed in urethane-anesthetized rats in which single-unit mechanosensitive bladder afferent Aδ- and C-fiber activities (identified based upon conduction velocities) were identified at the L6 dorsal root by electrical stimulation of the pelvic nerve and bladder distension (Aizawa et al. 2012). Mirabegron dose-dependently inhibited mechanosensitive bladder afferent activities of both Aδ- and C-fibers during bladder filling, which was more remarkable for Aδ-fibers. In addition, under isovolumetric conditions of the bladder, mirabegron inhibited both the bladder microcontractions and Aδ-fiber activity at doses that did not decrease the bladder pressure. At higher doses, which also decreased the bladder pressure, mirabegron inhibited C-fiber activity. These effects were not observed with oxybutynin.

Intraduodenal administration of ritobegron or i.v. administration of its active metabolite KUC-7322 dose-dependently reduced intravesicular pressure in urethrane-anesthetized male rats (Maruyama et al. 2012a; Yamazaki et al. 2002); within these studies, ritobegron did not affect heart rate and reduced blood pressure by about 10 mm Hg. In conscious female rats in which bladder irritation was induced by intravesical prostaglandin E2 instillation, intragastric administration of ritobegron prolonged the micturition interval and increased the micturition volume (Yamazaki et al. 2002). In enflurane-anesthetized cynomolgus monkeys intraduodenal administration of ritobegron dose-dependently reduced intravesicular pressure without significantly affecting heart rate or mean blood pressure, whereas i.v. administration of isoprenaline after ritobegron increased heart rate and reduced blood pressure (Maruyama et al. 2012b).

The originally reported in vivo data with solabegron relate to its potential for the treatment of type 2 diabetes and/or obesity. Thus, a 14-day oral treatment of db/db mice with solabegron dose-dependently lowered plasma glucose, HbA1c, and insulin levels and elicited a thermogenic response from the brown fat-rich interscapular region (Uehling et al. 2006). Given the differential role of β3-adrenoceptors in glucose and lipid metabolism between rodents and humans (Arch 2008), the relevance of such findings for OAB patients is difficult to predict. In a placebo-controlled, double-blind study in 36 healthy volunteers, oral administration of solabegron doses of 50 or 200 mg twice daily for 7 days did not significantly alter gastrointestinal or colonic transit or bowel function (Grudell et al. 2008). In propofol-anesthetized female dogs, bladder irritation was induced by intravesical acetic acid instillation; in this model, i.v. administration of solabegron increased micturition volume threshold but did not significantly affect voided volume, voiding efficiency, bladder contraction amplitude, and duration (Hicks et al. 2007). Within that study, the solabegron dose increasing micturition volume threshold decreased mean arterial pressure by 35 mm Hg and increased heart rate by 97 bpm. Pretreatment of the dogs with the antagonist bupranolol significantly attenuated the solabegron effects on micturition volume threshold and heart rate but not those on mean arterial pressure.

In conclusion, all three agonists have proven efficacious in one or more in vivo bladder function models. The differences in models being used make comparisons across compounds difficult. It is noteworthy that in doses effective on the bladder, ritobegron had little effect on cardiovascular function in monkeys, whereas as mirabegron and solabegron lowered blood pressure and increased heart rate in rats and dogs, respectively. However, it should be noted that the role of β3-adrenoceptors in cardiovascular regulation may be highly species-dependent (Rozec and Gauthier 2006), which implies that comparisons between species and models are difficult, and cardiovascular effects in experimental animals do not necessarily predict their presence in humans.

Perspective

While the above studies already provide considerable insight into the pharmacological properties of selective β3-adrenoceptor agonists currently in clinical development, a number of questions remain to be answered for a more comprehensive characterization of these compounds and the entire drug class. For example, the effects of β3-adrenoceptor agonists at the cellular level have mostly been studied using cAMP formation/accumulation as a functional read-out. However, it has now become clear that β-adrenoceptors in general and β3-adrenoceptors in particular can couple to multiple signaling pathways and that some ligands preferentially activate some of these pathways as compared to others, a phenomenon called “biased agonism” or “ligand-directed signaling” (Evans et al. 2010). Whether and how mirabegron, ritobegron ,and solabegron are affected by this phenomenon remains to be studied.

While most of the in vitro bladder function studies with β3-adrenoceptor agonists have looked at smooth muscle function, some of the in vivo findings already indicate that afferent nerves may also contribute to their overall profile. While β3-adrenoceptors are also expressed in human urothelium (Ochodnicky et al. 2012), it remains controversial whether such receptors contribute to β3-adrenoceptor agonist effects on detrusor function (Kullmann et al. 2011; Masunaga et al. 2010; Propping et al. 2009). Other potential effects of β3-adrenoceptor agonists on bladder function should also be considered. For example, there is an interaction between muscarinic and β-adrenergic systems in the bladder (Witte et al. 2011), which may be important pathophysiologically when acetylcholine is released non-neuronally during the storage phase but also therapeutically if a combination treatment of muscarinic antagonists and β-adrenergic agonists is being considered. In this regard, we are still trying to understand why some muscarinic antagonists exhibit a certain degree of bladder selectivity (Yamazaki et al. 2011), and it is largely unknown whether a similar phenomenon exists for β3-adrenoceptor agonists. Moreover, it is well established that the function of the systemic renin–angiotensin system is in part regulated by β-adrenoceptors, in humans β1-adrenoceptors, but a local renin–angiotensin system also exists in the urogenital tract (Comiter 2012). β-Adrenoceptors are also known to affect the function of gap junctions in the heart (Salameh and Dhein 2011), and gap junctions also play a role in the regulation of bladder function (Roosen et al. 2009). Moreover, adrenergic agonists can affect expression of uroplakin in the bladder, and this may play a role in cystitis (Kyung et al. 2012). A potential role of β3-adrenoceptors in all of these effects has not yet been explored. Such work will be added by a better characterization of the localization of expression of β3-adrenoceptors in the bladder and other tissues, but tools to characterize such expression at the protein level are just emerging (Cernecka et al. 2012). Finally, some of the above studies have already explored β3-adrenoceptor agonist effects outside the bladder as a potential source of side effects, but they may also be leading to potential future indication, e.g., with retinal disease (Mori et al. 2011). Thus, plenty of research remains to be done to obtain a more complete picture of β3-adrenoceptor agonist pharmacology.