Past, present and future of A(2A) adenosine receptor antagonists in the therapy of Parkinson's disease

Abstract

Several selective antagonists for adenosine A(2A) receptors (A(2A)R) are currently under evaluation in clinical trials (phases I to III) to treat Parkinson's disease, and they will probably soon reach the market. The usefulness of these antagonists has been deduced from studies demonstrating functional interactions between dopamine D₂ and adenosine A(2A) receptors in the basal ganglia. At present it is believed that A(2A)R antagonists can be used in combination with the dopamine precursor L-DOPA to minimize the motor symptoms of Parkinson's patients. However, a considerable body of data indicates that in addition to ameliorating motor symptoms, adenosine A(2A)R antagonists may also prevent neurodegeneration. Despite these promising indications, one further issue must be considered in order to develop fully optimized antiparkinsonian drug therapy, namely the existence of (hetero)dimers/oligomers of G protein-coupled receptors, a topic that is currently the focus of intense debate within the scientific community. Dopamine D₂ receptors (D₂Rs) expressed in the striatum are known to form heteromers with A(2A) adenosine receptors. Thus, the development of heteromer-specific A(2A) receptor antagonists represents a promising strategy for the identification of more selective and safer drugs.

Figure 1. A_2A adenosine receptor antagonists: xanthine derivatives

A major metabolite of caffeine in humans, paraxanthine (3) is an A_2AR antagonist as potent as caffeine that contributes to the in vivo activity of caffeine (Arnaud, 2011). A synthetic analogue of caffeine, 3,7-dimethyl-1-propargylxanthine (DMPX, 4) exhibits a somewhat higher affinity for the A_2A versus the A₁ receptor, and has thus been used in vitro and in vivo as the first “A_2A-selective” AR antagonist (Seale et al., 1988). However, DMPX is only very moderately selective with respect to A₁ and it is not selective for the A_2B receptor (see Table 1). Truly A_2AR-selective xanthine derivatives include the 8-styrylxanthine derivatives istradefylline (KW6002, 5: Kase, 2003), 8-(m-chlorostyryl) caffeine (CSC, 6: Jacobson et al., 1993), and MSX-2 (7: Sauer et al., 2000; Hockemeyer et al., 2004). In contrast to the 8-unsubstituted xanthine derivatives (1-4), which exhibit acceptable to good water-solubility, the 8-styrylxanthines (5-7) are relatively insoluble in water. Therefore, water soluble prodrugs of MSX-2 have been developed, such as the phosphate prodrug MSX-3 (7a: Hockemeyer et al., 2004) and the L-valine ester prodrug MSX-4 (7b: Vollmann et al., 2008), which represent very valuable pharmacological tools, particularly for in vivo studies (e.g., Randall et al., 2011; Collins et al., 2010; Mott et al., 2009; Bilkei-Gorzo et al., 2008; Schindler et al., 2005; Blum et al., 2003; Hauber et al., 1998). These drugs are very water-soluble but readily cleaved by enzymatic hydrolysis. Both compounds can be applied by injection, but are also bioavailable after peroral administration (unpublished results). CSC (5) acts as a dual compound, inhibiting monoamine oxidase B (MAO-B) and blocking the A_2AR to a similar extent (K_i A_2A: 54 nM, K_i MAO-B: 80.6 nM). By contrast, istradefylline (5) and MSX-2 (7) do not inhibit MAO-B, or only at concentrations that are >100-fold higher than those required for A_2A receptor blockade (see Table 1). Istradefylline is the only A_2AR -selective xanthine derivative being evaluated in clinical trials (see Figure 2).

Figure 2. A_2AR antagonists: non-xanthine derivatives

Several classes of non-xanthine A_2AR antagonists have been developed (see Figure 2 and Table 1). The antagonists frequently used in pharmacological experiments include the non-selective CGS-15943 (8), and the A_2AR -selective compounds ZM-241385 (9) and SCH-58261 (10). All of these amino-substituted heterobi- or –tricyclic compounds are structurally related to adenosine, although they lack the ribose sugar moiety (see Figure 2). The X-ray structure of the A_2AR protein complexed with antagonist 9 has been obtained, revealing the binding site for antagonists (Jaakola et al., 2008). Very recently an agonist-bound X-ray structure of the A_2AR was published showing that the agonist, an adenosine derivative, occupies virtually the same binding site as antagonist 9 (Xu et al., 2011). A further optimized compound derived from SCH-58261 (10) has been generated, SCH-442416 (11), which exhibits improved selectivity and is used in a ¹¹C-labelled form as a tracer for positron emission tomography (PET: see below). Another analogue of 10 is preladenant (12), which is currently being evaluated in clinical trials for the treatment of PD (see below: Salamone, 2010). Two additional aminopurine or aminoazapurine derivatives that have been tested in clinical trials are vipadenant (BII014, V2006, 13 : Gillespie et al., 2009) and ST-1535 (14: Stasi et al., 2006). While preladenant is highly A_2AR -selective, the latter compounds (13 and 14) are less selective (see Table 1). The first selective A_2AR antagonist to be tested clinically that is not structurally related to xanthine or adenine was the benzothiazole derivative SYN-115 (15: Black et al., 2010). Compound 15 also shows high selectivity for the A_2AR.