Allosteric Modulation of Adenosine A2A Receptors as a New Therapeutic Avenue

The therapeutic potential of targeting adenosine A2A receptors (A2ARs) is immense due to their broad expression in the body and central nervous system. The role of A2ARs in cardiovascular function, inflammation, sleep/wake behaviors, cognition, and other primary nervous system functions has been extensively studied. Numerous A2AR agonist and antagonist molecules are reported, many of which are currently in clinical trials or have already been approved for treatment. Allosteric modulators can selectively elicit a physiologic response only where and when the orthosteric ligand is released, which reduces the risk of an adverse effect resulting from A2AR activation. Thus, these allosteric modulators have a potential therapeutic advantage over classical agonist and antagonist molecules. This review focuses on the recent developments regarding allosteric A2AR modulation, which is a promising area for future pharmaceutical research because the list of existing allosteric A2AR modulators and their physiologic effects is still short.

The therapeutic potential of targeting A 2A Rs has prompted the development of numerous antagonist and agonist molecules to selectively control A 2A R function. The myriad A 2A R agonists and antagonists are considered potential therapeutic agents for inflammation, sickle cell disease, ischemia-reperfusion injury, and central nervous system (CNS) diseases [15,16]. The A 2A R agonist regadenoson was approved by the US Food and Drug Administration to boost blood flow during cardiac stress tests [16]. Many other agonists and antagonists are undergoing clinical trials.
Medicinal chemists have made many efforts to develop small molecules as allosteric modulators in recent years. Unlike agonist and antagonist molecules, allosteric modulators evoke a selective physiologic response only where and when the orthosteric ligand is

A 2A R and Its Physiologic Roles
The four types of adenosine receptors, A 1 R, A 2A R, A 2B R, or A 3 R, react with extracellular adenosine [13]. The activation of A 2B Rs reportedly requires a high adenosine concentration. Unlike A 2B Rs, adenosine levels under basal physiologic conditions are adequate to activate A 1 Rs, A 2A Rs, and A 3 Rs with relatively equal potency. The pharmacologic strength of an endogenous ligand or agonist at its receptor, however, relies on the number of receptors on the cells. Higher concentrations of adenosine are needed to show an effect in the presence of only a few receptors. Local expression of the A 1 Rs and A 2A Rs in the brain is suggested to be relatively higher than that of the other two adenosine receptors [6,29].
A 2A Rs were first identified by Libert and colleagues when they cloned several orphan GPCRs from the dog thyroid [30]. Afterward, A 2A Rs were cloned from other species, including guinea pigs, mice, rats, and humans [31][32][33][34]. As with the other GPCRs, A 2A Rs induce classical secondary messenger pathways. The A 2A R signaling pathway may vary depending on the cell and tissue type in which the receptors occur. For example, Gs is the major G-protein associated with A 2A Rs in the peripheral system. On the other hand, A 2A Rs in the striatum, where they are highly expressed, mediate their effects mainly through Golf activation in the rat. Active Gs and Golf proteins stimulate adenylyl cyclase ( Figure 1) which increases cellular cAMP levels and activates protein kinase A (PKA) which then phosphorylates and promotes cAMP-responsive element-binding protein 1 (CREB1) [16,35]. Activation of A 2A Rs also activates extracellular signal-regulated kinases (ERK) and several other kinases of the mitogen-activated protein kinase (MAPK) family which trigger specific cellular responses [36]. A 2A Rs form heterodimer structures with other GPCRs (e.g., metabotropic glutamate type 5 receptor (mGluR5)/A 2A R, cannabinoid receptor type 1 (CB 1 )/A 2A R, dopamine D 2 receptor (D 2 R)/A 2A R, dopamine D 3 receptor/A 2A R), and even CB 1 /A 2A R/D 2 R heterotrimers [37][38][39][40][41].
induce classical secondary messenger pathways. The A2AR signaling pathway may vary depending on the cell and tissue type in which the receptors occur. For example, Gs is the major G-protein associated with A2ARs in the peripheral system. On the other hand, A2ARs in the striatum, where they are highly expressed, mediate their effects mainly through Golf activation in the rat. Active Gs and Golf proteins stimulate adenylyl cyclase ( Figure  1) which increases cellular cAMP levels and activates protein kinase A (PKA) which then phosphorylates and promotes cAMP-responsive element-binding protein 1 (CREB1) [16,35]. Activation of A2ARs also activates extracellular signal-regulated kinases (ERK) and several other kinases of the mitogen-activated protein kinase (MAPK) family which trigger specific cellular responses [36]. A2ARs form heterodimer structures with other GPCRs (e.g., metabotropic glutamate type 5 receptor (mGluR5)/A2AR, cannabinoid receptor type 1 (CB1)/A2AR, dopamine D2 receptor (D2R)/A2AR, dopamine D3 receptor/A2AR), and even CB1/A2AR/D2R heterotrimers [37][38][39][40][41]. Neuronal A2AR signaling cascades. A2AR is a Gs(olf)-protein-coupled receptor involved in various physiologic processes. (1) The allosteric modulation sites may be pharmacologically relevant for avoiding adverse effects on the cardiovascular and other peripheral systems. (2) Binding of adenosine and an allosteric modulator to A2ARs enhances the activation of cyclic adenosine monophosphate (cAMP) and protein kinase A (PKA), resulting in the phosphorylation of calcium ion channels and increased influx of Ca +2 into the cytoplasm. (3) The PKA pathway also promotes neural progenitor cell (NPC) survival, proliferation, and differentiation; and activation of the mitogen-activated protein (MAP)-kinase pathway. (4) PKA-mediated phosphorylation of the cAMP-responsive element binding protein 1 (CREB-1) regulates the expression of genes such as c-fos, enkephalin (ENK), neurotensin, and zinc finger protein 268 (zif268). (5) The secretion of brain-derived neurotrophic factor (BDNF) and activation of tropomyosin receptor kinase B (TrkB) receptors in response to A2AR activation in hippocampal neurons may be relevant for cognitive functions such as (2) Binding of adenosine and an allosteric modulator to A 2A Rs enhances the activation of cyclic adenosine monophosphate (cAMP) and protein kinase A (PKA), resulting in the phosphorylation of calcium ion channels and increased influx of Ca +2 into the cytoplasm. (3) The PKA pathway also promotes neural progenitor cell (NPC) survival, proliferation, and differentiation; and activation of the mitogenactivated protein (MAP)-kinase pathway. (4) PKA-mediated phosphorylation of the cAMP-responsive element binding protein 1 (CREB-1) regulates the expression of genes such as c-fos, enkephalin (ENK), neurotensin, and zinc finger protein 268 (zif268). (5) The secretion of brain-derived neurotrophic factor (BDNF) and activation of tropomyosin receptor kinase B (TrkB) receptors in response to A 2A R activation in hippocampal neurons may be relevant for cognitive functions such as learning and memory. (6) A 2A R activation may be a counter mechanism to control the activation and expression of dopamine D 2 receptors (D 2 Rs). Long-term imbalance of D 2 R signaling leads to impairments in cognitive and motor functions and the development of Parkinson's and Huntington's diseases. The development of electron microscopy, selective radioligands, and antibodies has greatly contributed to A 2A R distribution mapping. Furthermore, advancements in electron microscopy helped to determine the interactions of agonists and antagonists with their receptors and receptor density in particular regions. A 2A Rs are concentrated on GABAergic medium-sized spiny neurons of the striatum, core and shell regions of the nucleus accumbens (NAc), olfactory tubercule, and dopamine-rich areas of the brain [16].
A 2A Rs play a significant role in regulating the indirect pathways of the basal ganglia in the brain (Figure 2) [16]. The basal ganglia have an evolutionarily conserved essential role in learned habits, goal-directed movements, and locomotion [42]. The basal ganglia carry out their functions through direct and indirect circuits, originating in conspicuous populations of striatal medium spiny neurons that project to different output structures. Direct pathway neurons express excitatory dopamine D 1 receptors (D 1 Rs) and inhibitory A 1 Rs, whereas indirect pathway neurons express inhibitory D 2 Rs and excitatory A 2A Rs [43]. Studies in mice revealed that both direct-and indirect pathway medium spiny neurons are active during mouse locomotion but quiescent during inactive phases [44], and chemogenetic activation of direct and indirect pathways neurons increases and decreases locomotor activity, respectively [45]. Moreover, recent findings indicate that optogenetic activation of indirect pathway neurons in the NAc, a part of the brain that is associated with motivation and pleasure, induces slow-wave sleep, whereas inhibition suppresses slow-wave sleep [46]. Other observations show that when an action does not result in a reward, increased activity of indirect pathways occurs, suggesting a role of the indirect pathways in controlling goal-directed behavior [47].   Apart from medium-sized spiny neurons in the basal ganglia, A 2A Rs are expressed in various other tissues, including smooth muscle cells, thymus, blood platelets, endothelial and lymphoid cells, leukocytes, spleen, blood vessels, lung, heart, and neurons in sympathetic and parasympathetic nervous systems [14,48] (Figure 2). A 2A Rs have a wide range of physiologic functions in the body, such as protecting tissues from inflammatory damage, mediating vasodilation, and supporting the formation of new blood vessels.
Several A 2A R agonists and antagonists are currently in clinical trials. The selective A 2A R agonist regadenoson was approved by the US Food and Drug Administration to increase blood flow in cardiac nuclear stress tests. On the other hand, the effects of A 2A R antagonists for the treatment of Parkinson's disease (PD) are promising. Other trials have been conducted with several agonists and antagonists aimed at treating infectious disease, ischemia-reperfusion injury, cancer, inflammation, sickle cell disease, diabetic nephropathy, and other CNS disorders. The increasing number of reports and patents demonstrates the growing interest in targeting the A 2A R [16].

The Concept of Allosteric Modulation
The most common method to stimulate receptors in pharmacology and biochemistry is to target orthosteric sites with their endogenous ligand, agonists, or antagonists. On the other hand, studies show that receptor activity can be altered by small molecules that bind to an allosteric site different from the site where the endogenous ligand, agonists, or antagonists would bind [49]. The small molecules that bind to the allosteric sites of the receptors are termed allosteric modulators. Unlike endogenous ligands, agonists, or antagonists, an allosteric modulator cannot itself activate or inactivate receptors but alters the receptor's response to substrates that bind to orthosteric sites in two ways: (1) increase or decrease affinity, i.e., the ability of orthosteric substances to bind receptors, and (2) increase/decrease efficacy, i.e., the ability of orthosteric substances to activate receptors [50]. Allosteric modulators reportedly change the receptor conformation, which alters the effect of the endogenous ligand, agonist, and antagonist binding [51]. The concept of receptor modulation is not straightforward with respect to practical implementation. Allosteric modulators do not necessarily equally alter the affinity and efficacy of endogenous ligands, agonists, or antagonists of the receptors. An allosteric modulator may alter the efficacy or affinity of the endogenous ligand, but not that of the agonist or antagonist of the receptors or vice versa [52].
The term 'allostery' was first used in enzymology studies in the early 1960s [53][54][55]. Subsequently, allosteric modulation has been identified for all receptor superfamilies, including GPCRs, nuclear hormone receptors [56,57], receptor tyrosine kinases [58,59], and ligand/voltage-gated ion channels [60][61][62][63][64]. The term "allosteric" began to be used increasingly in the literature, and a broad spectrum of allosteric modulators was described. Consequently, the classification of allosteric modulators was necessary to avoid possible confusion [65][66][67]. Three properties are considered in the classification of allosteric modulators: (1) affinity modulation of the orthosteric ligand, (2) modulation of the signaling effect of the orthosteric ligand, and (3) direct effects of the allosteric modulator in the absence of the orthosteric ligand. Moreover, allosteric modulators are classified in terms of their effects on orthosteric ligands as positive allosteric modulators (PAM), negative allosteric modulators (NAM), or silent allosteric modulators, also known as neutral allosteric ligands [68]. PAMs enhance the agonist/antagonist affinity and efficacy, whereas NAMs decrease orthosteric ligand affinity and efficacy. Unlike PAMs and NAMs, silent allosteric modulators do not affect the agonist or antagonist activity of orthosteric ligands, but bind to the allosteric site of the receptors and prevent PAMs or NAMs from binding to the same site, thereby inhibiting the activity of positive/negative allosteric modulators [52]. It is important to note that activities of allosteric modulators are therefore limited by where and when the orthosteric ligand is released. Thus, in contrast to agonists or antagonists, allosteric modulators promise greater safety and fewer side effects in therapeutic applications.

Allosteric A 2A R Modulation
Adenosine receptors are among the first known allosterically regulated GPCRs. Early studies demonstrated that amiloride and its analogs are allosteric A 2A R inhibitors [17,18,69]. Subsequent studies revealed that the amiloride analog 5-(N,N-hexamethylene)-amiloride (HMA) is a potent allosteric A 2A R inhibitor. The other amiloride analogs, benzamil, 5-(Nmethyl-N-isobutyl)amiloride (MIBA), 5-(N-methyl-N-guanidinocarbonyl-methyl)amiloride (MCGMA), and phenamil, were found to be more effective allosteric inhibitors than amiloride at rat A 2A Rs [17,70]. Moreover, amiloride and its analogues do not affect the dissociation rate of the agonist [71]. By contrast, sodium ions, for example, deteriorate the dissociation rate of the antagonist [ 3 H]ZM241385 from A 2A Rs in a dose-dependent manner [17]. It is important to note that other adenosine receptor agonists and antagonists are differentially affected by amilorides [70]. A new approach specifically targeting the sodium ion pocket, known as fragment-screening based on affinity mass spectrometry, led to the discovery of fragment Fg754 as a new A 2A R NAM carrying a novel azetidine moiety and exhibiting inhibitory potency comparable to HMA. Subsequent simulations of the molecular dynamics, structureactivity relationship studies of the ligand, and nuclear magnetic resonance analyses in solution revealed the unique binding mode and antagonistic properties of Fg754, which is distinctly different from HMA [72]. In addition, cholesterol is reported to be a weak PAM of A 2A Rs [73].
Identification of binding sites for allosteric modulators on A 2A Rs based on the crystal structure of the receptor is critical for the development of new allosteric modulators. Two tightly linked residues, histidine residue number 278 (His 278 ) in transmembrane domain 1 and glutamic acid 13 in transmembrane domain 7 of the human A 2A R, are reported to be the most crucial components for agonist recognition and play a partial role in the allosteric regulation by sodium ions [70,71,[74][75][76][77]. Studies of the crystal structure of the antagonist-bound adenosine A 2A R revealed that a highly conserved aspartate (Asp) residue in the second transmembrane domain is involved in sodium modulation of GPCRs [78]. Comparative studies of crystal structures in which a sodium ion bound in the allosteric site of human protease-activated receptor 1 [79], the β1-adrenergic receptor [80,81], the human δ-opioid receptor [82], and the human adenosine A 2A R [78] show that sodium ions interact with the common residues Asp 2.50 (superscript numbers refer to the Ballesteros and Weinstein residue numbering system [83]) serine 3.39 , tryptophan 6.48 , asparagine (Asn) 7.45 , and Asn 7.49 , either directly or through water-mediated hydrogen bonding [83]. Pre-crystal structure studies revealed that the positively charged sodium ion forms a permanent salt bridge with the negatively charged amino acid Asp 2.50 , suggesting that this residue represents the most conserved sodium ion binding site among GPCRs [84].
Subsequent studies on the crystal structure of the A 2A R at 1.8 Å resolution provided sufficient resolution to confirm that Asp 2.50 interacts directly with sodium ions via the salt bridge [78]. The crystal structures of agonist complexes for two variants in the first sodium coordination shell of the human A 2A adenosine receptor have also been reported [85]. A fluorine-19 nuclear magnetic resonance spectroscopy study suggested that A 2A Rs have four distinct activation states; a partial agonist that favors the population of an active state (S 3 ), an active state induced by full agonists (S 3 ), and two inactive states (S 1-2 ); this study also demonstrated that sodium ions enhance the inactive states of A 2A Rs [86]. In contrast, partial agonists and HMA induce active states, indicating that HMA competes with sodium ions for interaction with A 2A Rs [84]. Moreover, all-atom simulations of molecular dynamics have shown that Fg754 can steadily enter the transmembrane domain core and form contacts with transmembrane helices 2, 3, 6, and 7, and extracellular loop 2. Particularly, the azetidine moiety of Fg754 may occupy the sodium ion-binding site by forming a salt bridge [72]. Another molecular dynamics simulation study described the allosteric effects of a mini-Gs protein on A 2A Rs [87].
In conclusion, the effects of amiloride and its derivatives on A 2A Rs are well studied. While the findings indicate that amiloride competes with sodium ions at the allosteric site of the A 2A R, with Asp being the crucial amino acid, the allosteric binding site(s) of other small molecules selective for A 2A Rs remain unknown.

Allosteric A 2A R Modulators and Their Potential Clinical Application
Allosteric A 2A R modulation could be a new target for drug discovery [88]. Allosteric modulators can selectively elicit a physiologic response where and when the orthosteric ligand is released, thereby reducing the risk of an adverse effect of A 2A R activation. Moreover, the possibility of saturating allosteric effects offers greater potential for fine-tuning the physiologic response in a positive or negative direction. As allosteric modulators have no pharmacologic effect beyond the saturation dose, these molecules are associated with a lower risk for adverse effects than orthosteric ligands, giving them a potential therapeutic advantage over classical agonists and antagonists [18,89].
Some compounds act as allosteric A 2A R modulators, such as sodium ions, amiloride, and potassium-sparing diuretics, that also modulate other GPCRs than A 2A Rs [90]. For example, PD120918 is reported to enhance the activity of A 2A R agonists in the rat striatum [91]. In contrast, thiadiazoles such as SCH-202676 alter the binding characteristics A 2A R agonists and antagonists [92]. Some studies, however, suggest that thiadiazoles act as binding or oxidizing agents for SH groups rather than as allosteric modulators [92]. To date, only a relatively small number of selective allosteric A 2A R modulators have been reported (Table 1) [93].  Induced slow wave sleep without affecting cardiovascular function or body temperature in wild-type male mice [94,95].

AEA061
Allosteric enhancer/modulator Enhanced adenosine and inosine signaling and increased effect of the A2AR agonist CGS 21680.
Inhibited the binding of antagonists and agonists at the A2AR orthosteric site [93].

Unknown
Induced slow wave sleep without affecting cardiovascular function or body temperature in wild-type male mice [94,95].

AEA061
Allosteric enhancer/modulator Enhanced adenosine and inosine signaling and increased effect of the A 2A R agonist CGS 21680.

Cholesterol Allosteric modulator
Decreased the dissociation rate of the agonist NECA at A2ARs-embedded nanodiscs [73].

Unknown
Unknown Cholesterol Allosteric modulator Decreased the dissociation rate of the agonist NECA at A 2A Rs-embedded nanodiscs [73].

Cholesterol Allosteric modulator
Decreased the dissociation rate of the agonist NECA at A2ARs-embedded nanodiscs [73].
Unknown Unknown

Allosteric A 2A R Modulation Related to Inflammation
Adenosine is present in high concentrations in inflamed areas due to cell activation and breakdown [98][99][100]. The intracellular concentration of cAMP has a regulatory role in immune and inflammatory cells [101] and specifically, A 2A Rs are responsible for the anti-inflammatory effects of adenosine [102,103]. The anti-inflammatory effects of A 2A R agonists are well known. Their therapeutic benefit, however, is not a given due to the potential adverse effects of A 2A R agonists following systemic administration [7].
AEA061, which has an undisclosed structure, promotes the anti-inflammatory effects of adenosine by allosterically enhancing the activity of endogenous adenosine at A 2A Rs [96]. AEA061, which has no activity at a rat or human A 2A Rs in the absence of adenosine, inhibits the production of cytokines such as interleukin-1α, macrophage inflammatory protein-1α, 1β, and 2, keratinocyte chemokine, RANTES (regulated upon activation, normal T cell expressed and presumably secreted), and tumor necrosis factor-α in monocytes and splenocytes in a mouse model of lipopolysaccharide-induced inflammation. Therefore, positive allosteric modulators of A 2A Rs may represent a potential therapeutic approach to inflammation.
Inosine and inosine analog 6-S-[(4-nitrophenyl)methyl]-6-thioinosine (NBMPR) selectively and dose-dependently activate human A 2A Rs. NBMPR and inosine inhibit the production of pro-inflammatory cytokines and chemokines in splenic monocytes of wild-type mice, but not A 2A R knockout mice. The positive allosteric A 2A R modulator AEA061 enhances inosine-mediated A 2A R activation, inosine-mediated inhibition of pro-inflammatory cytokines, and chemokine production by splenic monocytes [97].

Allosteric A 2A R Modulation Related to Sleep and Neurologic Disorders
A 2A Rs are also expressed in the CNS, with the highest levels in the ventral and dorsal striatum [104]. A 2A Rs are present in the pre/postsynaptic compartment of neurons and microglia, oligodendrocytes, astrocytes, and capillary endothelial cells [12,[105][106][107][108][109][110]. A growing number of reports illustrate that A 2A Rs play a critical role in emotional and cognitive processes, motivation, and voluntary movements [111]. Moreover, A 2A R-expressing neurons in the NAc regulate sleep [8,47,112]. Therefore, A 2A R stimulation should be considered a potential treatment approach for insomnia. Insomnia is a sleep disorder that affects millions of people worldwide and frequently co-occurs with a wide range of psychiatric disorders [113][114][115] Although A 2A R agonists have strong sleep-inducing effects [116][117][118][119], they also have adverse cardiovascular effects and thus cannot be used clinically to treat sleep disorders. Moreover, the development of adenosine analogs to treat CNS disorders, including insomnia, is hampered by the poor transport of these drugs across the blood-brain barrier. In mice, a small blood-brain barrier-permeable monocarboxylate (3,4-difluoro-2-((2-fluoro-4-iodophenyl)amino) benzoic acid, denoted as A 2A R PAM-1, was recently found to induce sleep by enhancing A 2A R signaling in the brain (Figure 3) but, surprisingly, did not exhibit the typical unwanted cardiovascular and body temperature effects of A 2A R agonists [94,95]. More specifically, A 2A R PAM-1 dose-dependently enhanced A 2A R signaling in A 2A R-expressing Chinese hamster ovary (CHO) cells but not in CHO cells lacking A 2A R expression or in the absence of adenosine (Figure 3). The A 2A R PAM-1 did not alter the activity of the A 2A R agonist CGS 21680 [120]. Intracerebroventricular infusion and intraperitoneal injection of A 2A R PAM-1 induced prolonged slow-wave sleep, but not rapid-eye-movement sleep, in wild-type mice, but not A 2A R knockout mice. Further testing revealed that A 2A R PAM-1, unlike A 2A R agonists, had no effects on blood pressure, cardiac function, or body temperature, suggesting that adenosine or A 2A R expression levels in the cardiovascular system are insufficient to elicit an A 2A R PAM-1 response under normal physiologic conditions. Therefore, molecules that allosterically enhance A 2A R signaling may be developed to help people with insomnia fall asleep more easily. Moreover, A 1 Rs play a crucial role in the resolution of sleep need by modulating slow-wave activity, a slow, oscillatory neocortical activity that intensifies in correlation with wake duration and declines during sleep [121]. Slow-wave activity is widely used as a marker of mammalian sleep homeostasis and is necessary for sleep function. Therefore, dual allosteric A 1 R/A 2A R modulators may be useful for improving not only the maintenance of sleep but also its function. with wake duration and declines during sleep [121]. Slow-wave activity is widely used as a marker of mammalian sleep homeostasis and is necessary for sleep function. Therefore, dual allosteric A1R/A2AR modulators may be useful for improving not only the maintenance of sleep but also its function. A2ARs have roles in neurodegenerative, neurodevelopmental, and psychiatric diseases. The potential therapeutic use of A2AR agonists and antagonists for specific conditions such as Niemann Pick disease, schizophrenia, autism-spectrum disorders, depression, anxiety, Alzheimer's disease, attention-deficit hyperactivity disorder, PD, and fragile X syndrome is comprehensively discussed in the literature [122]. Allosteric A2AR modulators may provide alternative therapeutic options for neurologic disorders to circumvent the complexity of central and peripheral adenosine signaling. For example, dopaminereplacement therapy in PD is potentiated by blocking A2ARs due to the adenosine-dopamine antagonism in the striatum [123]. Decade-long preclinical studies of A2AR antagonists in PD models led to clinical trials of the A2AR antagonist istradefylline, which confirmed its clinically significant motor benefit in advanced PD patients and resulted in the approval of istradefylline for the treatment of PD patients in Japan and the US. The complexity of adenosine signaling contributed at least partially to the debilitating side effects The potential therapeutic use of A 2A R agonists and antagonists for specific conditions such as Niemann Pick disease, schizophrenia, autism-spectrum disorders, depression, anxiety, Alzheimer's disease, attention-deficit hyperactivity disorder, PD, and fragile X syndrome is comprehensively discussed in the literature [122]. Allosteric A 2A R modulators may provide alternative therapeutic options for neurologic disorders to circumvent the complexity of central and peripheral adenosine signaling. For example, dopamine-replacement therapy in PD is potentiated by blocking A 2A Rs due to the adenosine-dopamine antagonism in the striatum [123]. Decade-long preclinical studies of A 2A R antagonists in PD models led to clinical trials of the A 2A R antagonist istradefylline, which confirmed its clinically significant motor benefit in advanced PD patients and resulted in the approval of istradefylline for the treatment of PD patients in Japan and the US. The complexity of adenosine signaling contributed at least partially to the debilitating side effects and suspension of the clinical phase III trial of the A 2A R antagonist tozadenant for PD, which resulted in the death of five patients due to inflammatory complications. Thus, there is also a critical need to develop safer and more effective means of suppressing A 2A R signaling; for example, by negative allosteric modulation. Whereas the most potent PD medication is levodopa (L-3,4dihydroxyphenylalanine), clinicians try to limit levodopa doses to the extent possible to avoid various adverse effects occurring with chronic use, such as dyskinesia and dopamine dysregulation. A 2A R PAM, when administered concomitantly with levodopa, may mitigate some of these side effects, but strong evidence is currently lacking.
In addition, positive allosteric modulators of A 2A Rs may also alleviate various symptoms in neuropsychologic disorders. For example, psychotic symptoms such as delusions are caused by impaired discrimination of environmental stimuli. Recent evidence shows that D 2 Rs mediate discrimination learning in the NAc, but A 2A Rs expressed together with D 2 Rs in the NAc are required for discrimination learning. While normal mice can discriminate between reward-predictive and non-reward-predictive tones several days after generalized reward conditioning (when any tone is reward-predictive), mice in which A 2A Rs are blocked in the NAc do not show this ability [124]. In addition, hypofunction of NMDA-type glutamate receptors is thought to be involved in schizophrenia, as NMDA receptor antagonists such as phencyclidine and dizocilpine (MK-801) cause psychotic and cognitive disorders in humans and animals [125]. Deleting A 2A Rs in NAc astrocytes leads to motor and memory impairments relevant to schizophrenia, namely exacerbation of the MK-801-induced psychomotor response and impaired working memory [126]. Thus, the enhancement of A 2A R signaling may be helpful to treat sleep disorders as well as schizophrenia and other psychotic disorders by overcoming dopaminergic hyperactivity or glutamatergic hypoactivity.

Concluding Remarks
Here, we discussed recent developments regarding allosteric A 2A R modulation. Although numerous allosteric modulators of A 2A Rs have been identified, the physiologic functions of only a few of them have been established. The sleep-promoting effects and inflammatory process-modulating roles of allosteric A 2A R modulators open the doors for the potential therapeutic use of these molecules for treating diseases. Allosteric modulators exert their effects only where and when the orthosteric ligand is released, conferring a potential therapeutic advantage over classical antagonists and agonist molecules. Thus, allosteric A 2A R modulation could provide patients with an effective and safe treatment for various diseases.
Finally, A 2A Rs form heterodimer structures with other receptors such as D 2 Rs and mGluR5 in the CNS. Receptor heterodimers may be an applicable target for developing A 2A R PAMs with high specificity for the heterodimer and thus limited adverse effects [38,[127][128][129][130][131].

Conflicts of Interest:
The authors declare no competing interest associated with the manuscript.