Piperazine- and Piperidine-Containing Thiazolo[5,4-d]pyrimidine Derivatives as New Potent and Selective Adenosine A2A Receptor Inverse Agonists

The therapeutic use of A2A adenosine receptor (AR) antagonists for the treatment of neurodegenerative disorders, such as Parkinson and Alzheimer diseases, is a very promising approach. Moreover, the potential therapeutic role of A2A AR antagonists to avoid both immunoescaping of tumor cells and tumor development is well documented. Herein, we report on the synthesis and biological evaluation of a new set of piperazine- and piperidine- containing 7-amino-2-(furan-2-yl)thiazolo[5,4-d]pyrimidine derivatives designed as human A2A AR antagonists/inverse agonists. Binding and potency data indicated that a good number of potent and selective hA2A AR inverse agonists were found. Amongst them, the 2-(furan-2-yl)-N5-(2-(4-phenylpiperazin-1-yl)ethyl)thiazolo[5,4-d]pyrimidine-5,7-diamine 11 exhibited the highest A2A AR binding affinity (Ki = 8.62 nM) as well as inverse agonist potency (IC50 = 7.42 nM). In addition, bioinformatics prediction using the web tool SwissADME revealed that 8, 11, and 19 possessed good drug-likeness profiles.


Introduction
Adenosine is an endogenous purinergic nucleoside which interferes in many physiological states related to cardiovascular, immune, and neurological functions. Extracellular adenosine acts via four distinct G protein-coupled membrane receptors, namely A 1 , A 2A , A 2B , and A 3 adenosine receptors (ARs). The A 1 and A 3 receptors are principally coupled to G i/o proteins thus inducing an inhibitory effect on adenylyl cyclase and reducing cAMP production, while the A 2A and A 2B receptors stimulate the production of cAMP via G s proteins [1]. ARs are distributed all over in the body and elevated adenosine levels and/or upregulation of ARs have been detected in many pathological conditions [2]. The A 2A AR is located both peripherally and centrally, with the highest expression levels in the striatum, olfactory tubercle, and the immune system. The A 2A AR is a very promising target in the field of neurodegenerative pathologies, mainly Parkinson's (PD) and Alzheimer's (AD) diseases [3][4][5]. Several A 2A AR antagonists have demonstrated to improve PD motor dysfunctions in various preclinical animal models as well as in clinical studies [5]. Furthermore, neuroprotective functions were associated with Our group previously synthesized some potent human (h) A2A AR antagonists/inverse agonists belonging to different chemical classes [21][22][23][24][25][26][27][28][29][30][31]. Among these, the thiazolo [5,4-d]pyrimidine one (TP series) has been deeply investigated allowing us to delineate comprehensive structure activity relationships [21,[25][26][27]31]. This was possible because the central thiazolopyrimidine scaffold can be easily decorated by at least three different substituents at positions 2, 5, and 7, to explore diverse sites of interaction. To obtain potent and selective hA2A AR antagonists/inverse agonists, the thiazolopyrimidine core must exhibit an exocyclic amine group at position 7 and a furan-2-yl moiety Recently a large amount of research focused on the A 2A AR as a new target for cancer immunotherapy [13,14]. In fact, the A 2A AR represents an important immune checkpoint for T cells and NK cells and its activation induces suppression of immune cells response. Considering the increased A 2A AR expression in activated tumor infiltrating T cells, it is thus clear that this mechanism is important to favor tumor escape [15]. Moreover, A 2A AR is expressed also in tumor cells and its stimulation induces and increases cell proliferation, chemotaxis and migration, thus favoring tumor growth and metastasis [16]. The potential therapeutic role of A 2A AR antagonists to avoid immunoescaping of tumor cells and tumor development is evident. Indeed, four A 2A AR antagonists, including Preladenant [17], PBF-509 [18], CPI-444 [19], and AZD4635 [20] have entered clinical development as anticancer drugs alone and in combination with other agents (Figure 1).
Our group previously synthesized some potent human (h) A 2A AR antagonists/inverse agonists belonging to different chemical classes [21][22][23][24][25][26][27][28][29][30][31]. Among these, the thiazolo [5,4-d]pyrimidine one (TP series) has been deeply investigated allowing us to delineate comprehensive structure activity relationships [21,[25][26][27]31]. This was possible because the central thiazolopyrimidine scaffold can be easily decorated by at least three different substituents at positions 2, 5, and 7, to explore diverse sites of interaction. To obtain potent and selective hA 2A AR antagonists/inverse agonists, the Pharmaceuticals 2020, 13, 161 3 of 19 thiazolopyrimidine core must exhibit an exocyclic amine group at position 7 and a furan-2-yl moiety at position 2. In contrast, substituents endowed with variable properties, such as the steric hindrance, seems to be tolerated at position 5. In fact, good to high A 2A AR affinity was observed when an (hetero)aryl or alkyl residue was attached by diverse linkers at position 5 of the thiazolopyrimidine scaffold [21,[25][26][27]. In particular, in a recent paper by us some interesting results were obtained when the linker was a piperazine moiety directly attached to the bicyclic core or spaced by an ethylamino chain [31]. It has to be noted that piperazine derivatives are reported to elicit a broad spectrum of pharmacological activities. In fact, this heterocycle is present in many well-known drugs belonging to diverse pharmacological classes [32].
Pharmaceuticals 2020, 13, x FOR PEER REVIEW 3 of 21 at position 2. In contrast, substituents endowed with variable properties, such as the steric hindrance, seems to be tolerated at position 5. In fact, good to high A2A AR affinity was observed when an (hetero)aryl or alkyl residue was attached by diverse linkers at position 5 of the thiazolopyrimidine scaffold [21,[25][26][27]. In particular, in a recent paper by us some interesting results were obtained when the linker was a piperazine moiety directly attached to the bicyclic core or spaced by an ethylamino chain [31]. It has to be noted that piperazine derivatives are reported to elicit a broad spectrum of pharmacological activities. In fact, this heterocycle is present in many well-known drugs belonging to diverse pharmacological classes [32]. Thus, to further investigate the structure-activity relationships of the 7-amino-2-(furan-2-yl)thiazolo [5,4-d]pyrimidines as A2A AR antagonists/inverse agonists, in the present paper we describe the synthesis of the new derivatives 1-8, 10-21 ( Figure 2) bearing at position 5 a piperidine or a piperazine moiety directly attached to the bicyclic core (1, and 2-8, respectively) or spaced by an ethylamino chain (10 and 11-16, respectively). Moreover, a little set of compounds bearing at position 5 a methylamino (17) or a methylaminopiperidine chain (18)(19)(20)(21) is reported.

Pharmacological Assays
Binding affinities of compounds 1-8, 10-21 for the hA1, hA2A, and hA3 AR subtypes, expressed in Chinese Hamster Ovary (CHO) cells, were determined in radioligand competition experiments. In the binding affinity assays, the competition of ligands for specific binding of [ 3 H]DPCPX, [ 3 H]ZM241385, and [ 125 I]AB-MECA, respectively was measured to hA1, hA2A, and hA3 ARs. Activities of compounds 1-8, 10-21 at the hA2B AR subtype was determined by measuring the inhibition of NECA stimulated adenylyl cyclase activity in CHO cells expressing the hA2B receptor. Compounds 8, 11, 14-15, and 19, the best in terms of hA2A AR affinity and selectivity, were also evaluated for their functional behavior. Hence, compounds were tested to assess their ability to modulate cAMP production in hA2A CHO cells. All pharmacological data are reported in Tables 1 and 2 together with those of the reference compound ZM 241,385 [41].

Pharmacological Assays
Binding affinities of compounds 1-8, 10-21 for the hA 1 , hA 2A , and hA 3 AR subtypes, expressed in Chinese Hamster Ovary (CHO) cells, were determined in radioligand competition experiments. In the binding affinity assays, the competition of ligands for specific binding of [ 3 H]DPCPX, [ 3 H]ZM241385, and [ 125 I]AB-MECA, respectively was measured to hA 1 , hA 2A , and hA 3 ARs. Activities of compounds 1-8, 10-21 at the hA 2B AR subtype was determined by measuring the inhibition of NECA stimulated adenylyl cyclase activity in CHO cells expressing the hA 2B receptor. Compounds 8, 11, 14-15, and 19, the best in terms of hA 2A AR affinity and selectivity, were also evaluated for their functional behavior. Hence, compounds were tested to assess their ability to modulate cAMP production in hA 2A CHO cells. All pharmacological data are reported in Tables 1 and 2 together with those of the reference compound ZM 241,385 [41]. of compounds 1-8, 10-21 at the hA2B AR subtype was determined by measuring the inhibition of NECA stimulated adenylyl cyclase activity in CHO cells expressing the hA2B receptor. Compounds 8, 11, 14-15, and 19, the best in terms of hA2A AR affinity and selectivity, were also evaluated for their functional behavior. Hence, compounds were tested to assess their ability to modulate cAMP production in hA2A CHO cells. All pharmacological data are reported in Tables 1 and 2 together with those of the reference compound ZM 241,385 [41].

Pharmacological Assays
Binding affinities of compounds 1-8, 10-21 for the hA1, hA2A, and hA3 AR subtypes, expressed in Chinese Hamster Ovary (CHO) cells, were determined in radioligand competition experiments. In the binding affinity assays, the competition of ligands for specific binding of [ 3 H]DPCPX, [ 3 H]ZM241385, and [ 125 I]AB-MECA, respectively was measured to hA1, hA2A, and hA3 ARs. Activities of compounds 1-8, 10-21 at the hA2B AR subtype was determined by measuring the inhibition of NECA stimulated adenylyl cyclase activity in CHO cells expressing the hA2B receptor. Compounds 8, 11, 14-15, and 19, the best in terms of hA2A AR affinity and selectivity, were also evaluated for their functional behavior. Hence, compounds were tested to assess their ability to modulate cAMP production in hA2A CHO cells. All pharmacological data are reported in Tables 1 and 2 together with those of the reference compound ZM 241,385 [41].

Pharmacological Assays
Binding affinities of compounds 1-8, 10-21 for the hA1, hA2A, and hA3 AR subtypes, expressed in Chinese Hamster Ovary (CHO) cells, were determined in radioligand competition experiments. In the binding affinity assays, the competition of ligands for specific binding of [ 3 H]DPCPX, [ 3 H]ZM241385, and [ 125 I]AB-MECA, respectively was measured to hA1, hA2A, and hA3 ARs. Activities of compounds 1-8, 10-21 at the hA2B AR subtype was determined by measuring the inhibition of NECA stimulated adenylyl cyclase activity in CHO cells expressing the hA2B receptor. Compounds 8, 11, 14-15, and 19, the best in terms of hA2A AR affinity and selectivity, were also evaluated for their functional behavior. Hence, compounds were tested to assess their ability to modulate cAMP production in hA2A CHO cells. All pharmacological data are reported in Tables 1 and 2 together with those of the reference compound ZM 241,385 [41].

Pharmacological Assays
Binding affinities of compounds 1-8, 10-21 for the hA1, hA2A, and hA3 AR subtypes, expressed in Chinese Hamster Ovary (CHO) cells, were determined in radioligand competition experiments. In the binding affinity assays, the competition of ligands for specific binding of [ 3 H]DPCPX, [ 3 H]ZM241385, and [ 125 I]AB-MECA, respectively was measured to hA1, hA2A, and hA3 ARs. Activities of compounds 1-8, 10-21 at the hA2B AR subtype was determined by measuring the inhibition of NECA stimulated adenylyl cyclase activity in CHO cells expressing the hA2B receptor. Compounds 8, 11, 14-15, and 19, the best in terms of hA2A AR affinity and selectivity, were also evaluated for their functional behavior. Hence, compounds were tested to assess their ability to modulate cAMP production in hA2A CHO cells. All pharmacological data are reported in Tables 1 and 2 together with those of the reference compound ZM 241,385 [41].

Structure-Activity Relationships
Binding and potency data of the newly synthesized compounds 1-8, 10-21, and of the previously reported derivative 9 [31] are summarized in Table 1.

Structure-Activity Relationships
Binding and potency data of the newly synthesized compounds 1-8, 10-21, and of the previously reported derivative 9 [31] are summarized in Table 1.
Compounds 1-9 bear a piperidine (1) or a piperazine (2-9) substituted ring directly linked to the bicyclic thiazolopyrimidine core. Comparison of the hA 2A AR binding activity of the piperidine substituted 1 (K i = 594 nM) and of its corresponding piperazine analogue 3 (K i = 58 nM), both bearing an appended benzyl group, indicates that the piperazine linker is preferred. Analyzing the effect of different substituents on the piperazine ring, the data indicate that while an appended phenyl (2) or benzyl residue (3) was equally tolerated, a longer phenylethyl group (4) or a para-substituent (OCH 2 CH 2 OCH 3 , COOEt) on the phenyl ring of 2 (compounds 5 and 9, respectively), produced a drop in the binding activity. Introduction of a furan-2-yl methanone residue on the piperazine ring gave compound 6 which shows good hA 2A AR affinity even if lower than that of 2 and 3. In contrast, the presence of an ethylamine chain yielded derivatives 7-8 endowed with higher hA 2A AR affinity than that of 2 and 3. Moreover, the (pyrrolidin-1-yl)ethyl derivative 8 is also highly selective toward this receptor subtype.
With respect to derivatives 1-6 and 9, the piperidine or piperazine residue at position 5 of compounds 10-16 was shifted from the thiazolopyrimidine core by an ethylamino linker thus increasing chain flexibility. In general this structural change leads to an improved binding affinity with only two exceptions. In fact, while derivative 12 is slightly less active than its homologue 3, the ethylbenzoate derivative 16 is equiactive to 9. Among the herein reported compounds, the phenylpiperazine derivative 11 possesses the highest hA 2A AR affinity displaying a K i value of 8.6 nM. Compared to the latter, the (furan-2-yl)methanonepiperazine derivative 15 shows a similar binding activity (K i = 10.8 nM) but is more selective toward the hA 2A AR. Compound 14, characterized by the same side chain of Preladenant, possesses high hA 2A AR affinity (K i = 18.3 nM) similar to that of 11 and 15, and is also highly selective.
Selected compounds 8, 11, 14-15, and 19, the best in terms of hA 2A AR affinity and selectivity, were also evaluated in functional assays to assess their ability to modulate cAMP production in hA 2A CHO cells (Table 2, Figure S1). All the tested compounds behaved as inverse agonists since they were able to inhibit basal cAMP accumulation. In particular, according to their nanomolar hA 2A AR affinities, compounds 8, 11, 14-15, and 19 show IC 50 values spanning from 15.2 to 7.42 nM and also in this assay derivative 11 is the most active.

In Silico ADME Prediction
Compounds 8, 11, 14-15, and 19 were also evaluated in silico to test their "drug-likeness" profiles on the basis of the absorption, distribution, metabolism, and excretion (ADME) properties. Calculations were performed by the SwissADME web service (http://www.swissadme.ch developed by the Molecular Modeling Group of the Swiss Institute of Bioinformatics) that gives free access to a pool of fast yet robust predictive models for small molecules pharmacokinetic properties [42]. The data evaluated for the selected compounds are summarized in the Supplementary Materials (Table S1).
Investigated molecules possessed several favorable ADME properties. All compounds obeyed the Lipinsky's rule of five indicating drug-likeness. Moreover, they possessed good probability to have at least 10% oral bioavailability in rat or measurable Caco-2 permeability. SwissADME returns warnings if the molecule under evaluation contains fragments that could yield a false positive biological output (PAINS Pan Assay Interference Structures) . Compounds 8, 11, 15, and 19 had no PAINS alerts, while 14 showed one alert. The topological surface area (TPSA) measures the drug ability to permeate cells .  Compounds 8, 11, and 19 showed similar TPSA values less than 140 Å 2 suggesting that they could permeate cell membranes. The Consesus log P o/w (octanol/water partition coefficient) values indicated rather a reasonable absorption (1.71 < Consensus log P o/w < 3.34), while the log S values defined moderate solubility in the body.
The bioavailability radars (Figure 3) are the drug-likeness graphs of analyzed compounds presented in the form of a hexagon with each of the vertices representing a parameter (lipophilicity, size, polarity, solubility, flexibility, and saturation) that define a bioavailable drug. The pink region is the suitable physicochemical space for oral bioavailability. The radar plot of the molecule, represented by the red distorted hexagon, has to fall entirely in the pink area to be considered drug-like. From the graphs in Figure 3, it was found that while compounds 8, 11, and 19 were orally bioavailable, compounds 14 and 15 were not, because of being too polar and 14 also too flexible.  Finally, the BOILED-egg (Brain Or IntestinaL EstimateD) method ( Figure 4) allows predicting simultaneously two keys in vivo ADME parameters, i.e., the passive gastrointestinal absorption (HIA) and brain access (BBB) [43]. While all studied compounds had no BBB permeability (none in the yellow region), compounds 8, 11, and 19 exert high HIA (in the white region) and compounds 14 and 15 were not permeable (in the grey region). Moreover, they all were predicted as actively effluxed by Pgp (blue dots = PGP+).  Finally, the BOILED-egg (Brain Or IntestinaL EstimateD) method ( Figure 4) allows predicting simultaneously two keys in vivo ADME parameters, i.e., the passive gastrointestinal absorption (HIA) and brain access (BBB) [43]. While all studied compounds had no BBB permeability (none in the yellow region), compounds 8, 11, and 19 exert high HIA (in the white region) and compounds 14 and 15 were not permeable (in the grey region). Moreover, they all were predicted as actively effluxed by Pgp (blue dots = PGP+).
Finally, the BOILED-egg (Brain Or IntestinaL EstimateD) method ( Figure 4) allows predicting simultaneously two keys in vivo ADME parameters, i.e., the passive gastrointestinal absorption (HIA) and brain access (BBB) [43]. While all studied compounds had no BBB permeability (none in the yellow region), compounds 8, 11, and 19 exert high HIA (in the white region) and compounds 14 and 15 were not permeable (in the grey region). Moreover, they all were predicted as actively effluxed by Pgp (blue dots = PGP+).

General Methods
The microwave-assisted syntheses were performed using an Initiator EXP Microwave Biotage instrument (frequency of irradiation: 2.45 GHz). Analytical silica gel plates (Merck F254, Kenilworth, NJ, USA), preparative silica gel plates (Merck F254, 2 mm), and silica gel 60 (Merck, 70-230 mesh) were used for analytical and preparative TLC, and for column chromatography, respectively. All melting points were determined on a Gallenkamp melting point apparatus and are uncorrected. Elemental analyses were performed with a FlashE1112 Thermofinnigan elemental analyzer for C, H, N and the results were within ±0.4% of the theoretical values. All final compounds revealed a purity not less than 95%. Compounds were named following IUPAC rules as applied by ChemDrawUltra 9.0. The IR spectra were recorded with a Perkin-Elmer Spectrum RX I spectrometer in Nujol mulls and are expressed in cm −1 . NMR spectra were recorded on a Bruker Avance 400 spectrometer (400 MHz for 1 H-NMR and 100 MHz for 13 C-NMR). The chemical shifts are reported in δ (ppm) and are relative to the central peak of the solvent which was CDCl 3 or DMSOd 6 . The following abbreviations are used: s: Singlet, d: Doublet, t: Triplet, m: Multiplet, br: Broad, and ar: Aromatic protons.

General Procedure for the Synthesis of 1-8, 10-21
The proper amine 23-42 (3 mmol) was added to a solution of the 5-chloro-2-(furan-2-yl)thiazolo [5,4-d]pyrimidin-7-amine derivative 22 [21] (1 mmol) in n-BuOH (2 mL). The reaction mixture was microwave irradiated at 200 • C for 20 min, then cooled at room temperature and basified with an aqueous KOH solution (50%). Addition of water afforded a solid which was collected by filtration and washed with Et 2 O. The crude material was purified by crystallization or by chromatography.
2-(4-Benzylpiperidin-1-yl)ethan-1-amine (31 In a 50 mL flask, equipped with a magnetic stirrer and reflux condenser, the proper phthalimide derivatives 47-50 (7 mmol), hydrazine hydrate (10 mmol), and methanol (50 mL) were added. The resulting mixture was refluxed for 2 h, cooled down to room temprature, and concentrated under reduced pressure. The remaining residue was treated with diethyl ether and the solid was filtered and used without further purification in the following step.

General Procedure for the Synthesis of 39-42
To a solution of 58-61 (5.6 mmol) in dichloromethane (60 mL), oxalic acid (6.3 mmol) was added. The solution was diluted with water (30 mL) and refluxed under vigorous stirring for 3 h. After cooling, the aqueous layer was isolated, washed twice with dichloromethane (30 mL), added with a NaOH aqueous solution (1 M, pH 9-10), and extracted with chloroform (50 mL × 3). The organic layer was dried over Na 2 SO 4 , evaporated under reduced pressure, and the obtained oily residue was used as such in the next step.

Cyclic AMP Assays
CHO cells transfected with hAR subtypes were washed with phosphate-buffered saline, detached with trypsin, and centrifuged for 10 min at 200× g. Cells were seeded in a 96-well white half-area microplate (Perkin Elmer, Boston, USA) in a stimulation buffer composed of Hank Balanced Salt solution, 5 mM HEPES, 0.5 mM Ro 20-1724, 0.1% BSA, 1 IU/mL adenosine deaminase. cAMP levels were then quantified by using the AlphaScreencAMP detection kit (Perkin Elmer, Waltham, MA, USA) following the manufacturer's instructions [47]. At the end of the experiments, plates were read with the Perkin Elmer EnSight Multimode Plate Reader.

Data Analysis
The protein concentration was determined according to a Bio-Rad method with bovine albumin as a standard reference. Inhibitory binding constant (Ki) values were calculated from those of IC 50 according to the Cheng and Prusoff equation Ki = IC 50 /(1 + [C*]/K D *), where [C*] is the concentration of the radioligand and K D * is its dissociation constant [46]. K i and IC 50 values were calculated by the non-linear regression analysis using the equation for a sigmoid concentration-response curve (Graph-PAD Prism, San Diego, CA, USA).

Conclusions
In conclusion, the herein reported structural investigation has led to a good number of new 7-amino-2-(furan-2-yl)-thiazolo [5,4-d]pyrimidines, featuring piperidine or piperazine substituents at position 5, endowed with potent and selective hA 2A AR inverse agonist activities. Among them, compound 11 bearing a phenylpiperazine-ethylamino chain at position 5, showed the highest hA 2A AR binding affinity and potency. Furthermore, the SwissADME prediction indicated that compounds 8, 11, and 19 exhibited good drug-likeness properties.

Conflicts of Interest:
The authors declare no conflict of interest.