Review 8 December 2017 and 1 School of Pharmacy, University of Oslo, 0316 Oslo, Norway 2 Norsk Medisinsk Syklotronsenter AS, Postboks 4950 Nydalen, 0424 Oslo, Norway 3 Realomics SFI, Department of Chemistry, University of Oslo, 0316 Oslo, Norway 4 Department of neuropsychiatry and psychosomatic medicine, Oslo University Hospital, 4950 Oslo, Norway
The decapeptide gonadotropin-releasing hormone, also referred to as luteinizing hormone-releasing hormone with the sequence (pGlu-His-Trp-Ser-Tyr-Gly-Leu-Arg-Pro-Gly-NH 2) plays an important role in regulating the reproductive system. It stimulates differential release of the gonadotropins FSH and LH from pituitary tissue. To date, treatment of hormone-dependent diseases targeting the GnRH receptor, including peptide GnRH agonist and antagonists are now available on the market. The inherited issues associate with peptide agonists and antagonists have however, led to significant interest in developing orally active, small molecule, non-peptide antagonists. In this review, we will summarize all developed small molecule GnRH antagonists along with the most recent clinical data and therapeutic applications.
Gonadotropin-releasing hormone (GnRH) or luteinizing hormone releasing hormone (LHRH) is a central regulator of the reproductive system in humans. It was first isolated from mammalian hypothalamic tissue as a linear decapeptide (pGlu-His-Trp-Ser-Tyr-Gly-Leu-Arg-Pro-Gly-NH 2) by Scally and colleagues [1]. The GnRH peptide is produced in hypothalamic neurons and released in a pulsed fashion into the portal blood-stream supplying the pituitary gland to stimulate the synthesis and secretion of gonadotropic hormones, luteinizing hormone (LH) and follicle-stimulating hormone (FSH). The gonadotropins consecutively act on the gonads to trigger the gametogenesis, and ultimately the synthesis and release of steroidal sex hormones in both male and female [2]. Gonadal steroids in turn regulate the GnRH secretion through a positive and negative feedback and consequently surge or inhibition. Given that GnRH is playing such a significant evolutionary role in various species’ reproduction, disturbance of GnRHs signaling has therefore been implied in causing a wide spectrum of diseases; i.e., reproductive hormonal dependent oncological and neurodegenerative diseases. Therefore, treatments targeting the GnRH receptor with GnRH or its agonist analogues have attracted considerable scientific interest and lead to great commercial success [3,4]. Through down-regulation, administration of GnRH agonists lead to temporary extinction of gonadotropin secretion and sex steroidogenesis and observed chemical castration. This has been therapeutically exploited in treating hormone dependent tumors, such as prostate and breast tumors [5,6,7] to inhibit their growth. Up to now, several peptide GnRH agonists including goserelin, triptorelin and leuprolide have been approved by the FDA and reached the market [8,9]. Despite GnRH agonists’ therapeutic applications, several major clinically relevant disadvantages have emerged [5,10]. Firstly, the initial surge of gonadotropin secretion (the so-called ‘flare’ effect) before the onset of desensitization can lead to an exacerbation of the symptoms in patients through elevated sex-hormone production prior to depletion of hormone secretion. Secondly, there is a risk of bone loss from long-term administration of agonists. Thirdly, parental administration of the peptide agonists is required, due to limited bioavailability of oral formulations. In contrast to agonists, GnRH antagonists result in an immediate and reversible suppression of LH and FSH secretion without evoking the initial hyper-secretion of gonadotropins [11]. Recent clinical studies have further demonstrated the ability of GnRH antagonists to suppress gonadotropins from the onset of administration [12]. However, the inherent issues of peptidic drugs targeting GnRH-R, such as low aqueous solubility and the needs for parenteral administration have limited their clinical exploitation [13,14]. Today, there are no non-peptide GnRH antagonists in clinical use in Europe.
Development of an orally bioavailable non-peptide GnRH receptor antagonist is therefore highly desirable, not only to address the drawbacks of peptidic GnRH receptor agonists but to ensure better patient compliance. In 1989, Abbott Laboratories reported the very first non-peptide GnRH antagonist as an antifungal drug—ketoconazole as a weak antagonist for rat GnRH receptor [15]. Since then, intensive research has been undertaken worldwide. To date, numerous small molecule GnRH antagonists representing various pharmacophores have been reported in the literature following the initial report and with a handful of chemical entities reached clinical developments (Table 1) [16].
Table 1. Thieno[2,3-b]pyridine-4-one derivatives and their binding affinities at human GnRH-R.
Research activity directed towards discovery of novel non-peptidic GnRH antagonists remain high and recent reviews on this topic being published nearly one decade ago [17,18,19,20,21]. We now wish to review the developments in the field from the very first non-peptide antagonist developments, indications and the most recent available clinical data.
Based on the ‘direct screening’ of in-house chemical libraries, researchers at Takeda Chemical Industries in Japan reported the first orally-active nonpeptidic GnRH receptor antagonist for human GnRH receptor in 1998 [22]. The disclosed thiophene-based bicyclic scaffold with crucial substituents on position 2, 7, 5 and 3 was designed to mimic the (Tyr 5-Gly 6-Leu 7-Arg 8) β-turn of GnRH, a dominant structure responsible for receptor binding. Structure-activity studies (SARs) led to identification of T-98475 (compound 1, Table 1), a thieno [2,3-b]pyridine-4-one analogue, which demonstrated high potency to inhibit GnRH-activation of cloned human receptor (IC 50 = 0.2 nM). Nevertheless, species dependent selectivity is in fact is a common pattern for several classes of non-peptide GnRH-R antagonists. Compound 1 was investigated in terms of in vivo antagonism and was found to suppress plasma level of LH in Cynomolgus monkeys in a time-dependent manner after oral administration. Two different strategies were subsequently employed looking for potent GnRH antagonists with improved in vivo efficacy, and these strategies were (1) further optimization on each substituent of compound 1 and (2) replace the existing theinopyridin-4-one scaffold with other heterocyclic isosteres. The former strategy eventually led to the identification of compound 2 [23], which displayed subnanomolar in vitro activity and an improved in vivo efficacy in castrated monkeys. A more effective and sustained LH suppression (greater than 24 h) at low doses (10 and 30 mg/kg) was also observed after oral admission of compound 2. The latter strategy looking for alternative bicyclic scaffold had resulted in a new series of molecules based on thieno[2,3-d]pyrimidine-2,4-diones, exemplified by TAK-013 and TAK-385 (Table 2; compounds 3 and 4, respectively) [24,25]. Compound 3 showed high potency at both human and monkey GnRH-R receptor, with IC 50 values of 0.1 and 0.6 nM [24]. Moreover, it also demonstrated comparable in vivo efficacy to 2 in castrated monkey. The improved oral bioavailability of compound 3 was attributed to the introduction of the terminal methoxyureido moiety at the R3-position. Through molecular modeling studies, formation of an intramolecular hydrogen bond between the aniline NH and the methoxy oxygen atom was suggested, and it was believed to shield the hydrogen bonding moieties from the solvent and therefore resulted in improvement of lipophilicity/membrane permeability [24]. Further chemical modification of compound 3 (at R 1 and R 2) led to the discovery of compound 4, a highly potent and orally active GnRH antagonist [25]. In castrated monkeys, compound 4 exhibited a suppressive effect on plasma LH levels at a dose as low as 1 mg/kg for more than 24 h. Based on the biochemical and pharmacological results, both compounds 3 and 4 have been selected as candidates for clinical trials for treating sex-hormone-dependent diseases. Parallel to the literature publications, Takeda has continuously filed claims on various thiophene-based bicyclic compounds and their application as human GnRH receptor antagonists for treating hormone-dependent diseases since mid-90s. Potent analogues, such as compounds 3 and 4 were filed for protections in their preparations, indications and formulations [26,27,28,29,30].
Table 2. Thieno[2,3-d]pyrimidine-2,4-one derivatives and their binding affinity at human GnRH.
Attempts to look for new scaffolds of non-peptidic antagonists based on compound 1 (T-98475) successfully led to the discovery of pyrrolo[1,2-a]pyrimidines (Table 3) by Neurocrine Bioscience in the USA [31,32,33]. Initial SAR studies revealed that all non-basic compounds of this series were inactive. Combination of a 2-fluorobenzyl group at position 4 and a hydrophobic aromatic ring with an additional hydrogen bond acceptor at R 2 position proved to improve the binding activity significantly. Subsequent optimization of aromatic ring at R 3 demonstrated that introduction of a lipophilic isobutoxy group on the para position further increased potency and also led to the discovery of compound 5 [31]. Continuous modifications of this series resulted in compound 6, which differs from 5 in removing a cyano substitute at the 3-position and introduction of an amide functionality at R 3 [32]. Despite its high affinity, preliminary studies with compound 6 and its closed analogues revealed a liability under acidic conditions. This degradation resulted from an acid catalyzed retro-Mannich reaction on the basic side-chain at 1-position. It was reasoned that the presence of an electron withdrawing group at 3-position, such as the cyano group at compound 5, would reduce electron density in the bicyclic system and thus helped in maintaining stability toward acid. With this rationale, Tucci et al. went ahead and introduced a fluoride at position 3 [33]. As expected, compound 7(K i = 9 nM) was not only stable toward acid, but also maintained high affinity toward human GnRH receptor. Moreover, the electron deficient aromatic ring at position 4 was believed to be of great importance, as it interacted with electron-rich tyrosine residues in the receptor, either via π-stacking or edge-to-face interactions and led to higher binding affinity. In addition, pyrrolopyrimidones, such as compounds exemplified in Table 3 were also patented by Neurocrine Biosciences as GnRH receptor antagonists in 2002 [34].
Table 3. Pyrrolo[1,2-a]pyrimidin-7-one derivatives and their binding affinity at human GnRH-R.
In 2002, parallel to the identification of pyrrolo[1,2-a]pyrimidines, Takeda and Neurocrine Biosciences both introduced bicyclic imidazolopyrimidinone scaffolds as new class of non-peptidic antagonists for human GnRH receptor (Table 4). As seen from compounds 8–12, the imidazolo[1,2-a]pyrimidin-5-one analogues retain the heterocyclic 5,6-ring system but substitute the pyrrolo ring with an imidazole moiety. Compound 8, which was identified by Takeda, was not only a potent antagonist in vitro but it also exhibited comparable potency to T-98475 (compound 1). Based on these positive results, a heterocyclic 5,6-fused ring system bearing a phenyl group on the five-membered ring was postulated to be a key structural feature for small molecular GnRH antagonist scaffolds [35]. In addition, patent application featuring compound 8 as potential therapeutic agent for hormone-dependent disease was disclosed by Takeda prior to the literature publication [36].
Table 4. Imidazolo[1,2-a]pyrimidin-5-one and their binding affinity at human GnRH-R.
Simultaneously, Neurocrine Biosciences also disclosed a series of imidazolo[1,2-a]pyrimidines as potent GnRH receptor antagonists in a patent publication, where the preparations and pharmaceutical composition were also described [37]. The high potency of imidazopyrimidines, exemplified by compounds 9–12 in Table 4, was attributed the existence of both the basic tertiary amine and the adjacent pyridine ring at the 3-position of the bicyclic system [38,39]. The tertiary amine, confirmed by modeling, initiated an interaction between ligand and an acidic residue within putative helical domains while the pyridine ring itself provided aromatic π-π interaction with a phenyl residue on the receptor [40]. Subsequent modifications focused on the bulky ester substituents at the 6-position, as the ester groups are prone to hydrolysis in vivo. Gross et al. reasoned that by substituting the ester group functioning as a lipophilic group and a hydrogen bond acceptor, with an arene bearing one or more hydrogen bound acceptors could circumvent the acid-liability issue. Different ring systems, such as methylenedioxyphenyl and 3-methoxyphenyl were then introduced. Combination of the new ring systems at R 1 together with the 3-methoxyphenyl substituent at R 3 resulted in potent compounds 10 and 11 with K i value at 11 and 4.6 nM, respectively [39]. Later SAR studies revealed that once the ester group at R 1 was substituted with a 3-methoxyphenyl, the importance of a para-substituted aromatic ring at R 3 in binding was somehow diminished. Evidenced by compound 12, a highly potent molecule (K i = 5.2 nM), which had a tert-butyl moiety incorporated at R 3 position [40].
The SAR results from the bicyclic imidazolopyrimidinones led to the discovery of a potent GnRH receptor antagonist with reduced molecular weight (c
