C3 amino-substituted chalcone derivative with selective adenosine rA1 receptor affinity in the micromolar range.

ABSTRACT: To identify novel adenosine receptor (AR) ligands based on the chalcone scaffold, herein the synthesis, characterization and in vitro and in silico evaluation of 33 chalcones (15-36 and 37-41) and structurally related compounds (42-47) are reported. These compounds were characterized by radioligand binding and GTP shift assays to determine the degree and type of binding affinity, respectively, against rat (r) A1 and A2A ARs. The chalcone derivatives 24, 29, 37 and 38 possessed selective A1 affinity below 10 µM, and thus, are the most active compounds of the present series; compound 38 was the most potent selective A1 AR antagonist (K i (r) = 1.6 µM). The structure-affinity relationships (SAR) revealed that the NH2-group at position C3 of ring A of the chalcone scaffold played a key role in affinity, and also, the Br-atom at position C3' on benzylidene ring B. Upon in vitro and in silico evaluation, the novel C3 amino-substituted chalcone derivative 38-that contains an α,ß-unsaturated carbonyl system and easily allows structural modification-may possibly be a synthon in future drug discovery. GRAPHIC ABSTRACT: C3 amino-substituted chalcone derivative (38) with C3' Br substitution on benzylidene ring B possesses selective adenosine rA1 receptor affinity in micromolar range. © Institute of Chemistry, Slovak Academy of Sciences 2020.

Introduction

In the human body, adenosine takes part in both physiologic and pathophysiologic processes (Boison 2018). Most, if not all, of these actions are mediated by four cell surface receptors, denoted A1, A2A, A2B and A3 (Fredholm et al. 1994, 2001, 2011). Ensuing almost 100 years of research on adenosine and its receptors as well as the ligands that bind to these adenosine receptors (ARs), the potential of the adenosine system as a drug discovery target is evident (de Lera Ruiz et al. 2013).

The A1 AR has been cloned from different mammalian species, including humans (Müller 2001) and show structural homology among diverse mammalian species (Ralevic and Burnstock 1998) (although lower affinity for human than rat receptors have been reported) (Klotz et al. 1997). A member of the seven transmembrane-spanning G-protein-coupled receptor superfamily (that preferentially couples to inhibitory Gi/o heterotrimeric G-proteins) (Freissmuth et al. 1991; Munshi et al. 1991; Ralevic and Burnstock 1998), the A1 AR inhibits adenylate cyclase and decreases cyclic adenosine monophosphate (Londos et al. 1980; Van Calker et al. 1978). Other second messenger systems may also be coupled to the A1 AR, such as phospholipase C and several types of calcium and potassium channels (Ralevic and Burnstock 1998). A1 ARs are widely distributed in different mammalian species (Ralevic and Burnstock 1998) and ubiquitous within the central nervous system, with the highest density in the cerebral cortex, cerebellum, hippocampus, thalamus, brainstem and spinal cord (Dixon et al. 1996; Reppert et al. 1991). These receptors are localized in the active zone of the presynaptic nerve terminal (Rebola et al. 2003; Swanson et al. 1995), a highly specialized region of the cytoplasm that directly faces the synaptic cleft (Sankaranarayanan and Ryan 2007).

At presynaptic nerve terminals, A1 ARs play a role in the release of neurotransmitters (Dunwiddie 1985). Adenosine acts by inhibiting cholinergic transmission, among other, via A1 ARs (Phillis 1991). The cholinergic system has been associated with a number of cognitive functions, for example, learning and memory as well as emotion (Jackson 2011).

Interaction between the adenosine and cholinergic system is regrettably controversial; on the one hand, administration of a selective A1 AR agonist (e.g. CPA) caused learning and memory deficits, which may be prevented by a selective A1 AR antagonist (e.g. DPCPX) (Normile and Barraco 1991), and on the other hand, the said antagonist also caused learning and memory deficits (Vollert et al. 2013)! Moreover, it was found that chronic treatment with A1 AR modulators result in behavioural effects different from those observed following acute administration; therefore, particular caution is required in the development of adenosine-based strategies for the chronic treatment of neurodegenerative or cognitive disorders (Von Lubitz et al. 1993). Other studies advocated the use of A1 AR antagonists and considered the A1 AR an interesting drug target for the development of cognitive enhancers based on the observation that A1 AR antagonism positively modulated memory process (Maemoto et al. 2004; Normile and Barraco 1991; Pitsikas and Borsini 1997; Suzuki et al. 1993).

Apart from learning and memory, adenosine and its receptors—specifically the A1 AR—also modulate anxiety where blockade of the A1 AR results in anxiolytic actions (Maemoto et al. 2004). The A1 AR agonist R-PIA exacerbated the effects of ethanol withdrawal, whereas the A1 AR antagonist CPT improved these anxiogenic effects in the elevated plus-maze and light/dark test, relevant behavioural animal models of anxiety (Gatch et al. 1999). These results suggest that A1 AR antagonists, at some doses, may be useful for ameliorating the anxiogenic effects produced by ethanol withdrawal, although it does not appear useful for reducing ethanol consumption (Gatch et al. 1999). Furthermore, the widely used anxiolytic agents, benzodiazepines, block adenosine uptake (Noji et al. 2004) and decreased A1 AR binding capacity in vivo at doses of clinical relevance (Kaplan et al. 1992). As with the adenosine system and its effects on learning and memory, there is controversy: DPCPX did not affect the anxiety state of mice in the elevated plus-maze test (Jain et al. 1995) and A1 AR knockout mice showed signs of increased anxiety in the light/dark test (Johansson et al. 2001).

Adenosine and its analogues have been known to cause behavioural despair in animal models relevant to depression. The A2A, rather than the A1, AR is involved in depression; based on evidence from pharmacology and A2A AR knockout mice (Yacoubi et al. 2001). (Additionally, the selective A2A AR antagonist istradefylline not only improves motor fluctuations but also some non-motor symptoms of Parkinson’s disease (PD); such as the mood disorder depression—the most common non-motor symptom of PD (Aarsland et al. 2012; Nagayama et al. 2019) This is yet to be confirmed by a double-blind placebo-controlled trial (Nagayama et al. 2019). Interestingly, the selective A1 AR antagonist DPCPX enhanced the anti-depressant effects of selective serotonin re-uptake inhibitors (SSRIs) imipramine, escitalopram and reboxetine in mice behavioural tests (Szopa et al. 2018); nevertheless, the effects of atypical antidepressants agomelatine and tianeptine were increased, not by DPCPX, but by the selective A2A AR antagonist DMPX (Szopa et al. 2019).

The orally active and brain penetrable pyrazolopyridine derivative FR194921 (1) exhibits potent affinity for the A1 AR (A1 (h) Ki = 2.9 nM) without affinity for the A2A and A3 ARs (Fig. 1). In addition, no species differences among human, rat and mouse were observed in the binding affinity profile of this compound (Maemoto et al. 2004). After oral administration of FR194921 (1), hypolocomotion in rats caused by the A1 AR agonist CPA was bettered demonstrating A1 AR antagonism in vivo (Maemoto et al. 2004). Additionally, scopolamine-induced memory deficits (passive avoidance test) as well as anxiety (social interaction test and elevated plus-maze test) were also bettered by the said compound (1), without influencing general behaviour (Maemoto et al. 2004). FR194921 (1) showed no anti-depressant activity (forced swim test)—even at high doses. It may be said that FR194921 (1) shows promise in the treatment of memory deficits and anxiety (Maemoto et al. 2004).

Fig. 1

The chemical structure and A1Ki (h) value (nM) of FR194921 (1)

The role of A1 AR antagonists in brain functioning is not yet fully understood, and at times, debatable—yet, selective A1 AR antagonists have potential for the treatment of neurological and psychiatric conditions, as seen with the potent and selective A1 AR antagonist FR184921 (1).

While diverse classes of A1 AR antagonists have been documented since the discovery of the receptor and specific radiolabelled ligands, most of these compounds were derived from xanthine analogues which show poor water solubility and oral bioavailability, central nervous system penetration and considerable species differences in terms of receptor binding affinity (Maemoto et al. 1997). Due to these undesirable physiochemical properties of xanthine derivatives, the focus has shifted from xanthine to non-xanthine based heterocycles as A1 AR antagonists (de Lera Ruiz et al. 2013).

The chemical structure 1,3-diphenyl-2-propen-1-one—perhaps better known by the general term “chalcone”—has attracted attention from both chemical (biosynthesis and synthesis) and biological standpoints; due to the wide-ranging pharmacological properties exhibited by these compounds (Mathew et al. 2019; Zhuang et al. 2017). These properties include anticancer, antileishmanial, antimalarial, antimicrobial, antiviral, antifungal, antioxidant, antihypertensive and antidiabetic biological activities—to name a few (Batovska and Todorova 2010; Sahu et al. 2012; Karthikeyan et al. 2015; Singh et al. 2014; Zhou and Xing 2015). Additionally, chalcones show potential in the treatment of neurological conditions by acting as anxiolytics and antidepressants, as well as the modulation of gamma-aminobutyric acid (GABA) receptors and the inhibition of acetylcholinesterase (AChE), butyrylcholinesterase (BChE) and monoamine oxidase (MAO) (Mathew et al. 2019). Of note, the chalcone–coumarin hybrid (2) possesses selective A3 AR affinity (A3Ki (h) = 5.2 µM) (Vazquez‐Rodriguez et al. 2013), the blockade of A3 ARs may be advantageous in brain ischemia (Fig. 2) (Jacobson and Gao 2006).

Fig. 2

The chemical structure and A1, A2A and A3Ki (h) values of 2

The term chalcone generally refers to chemicals with an α,ß-unsaturated carbonyl system; thus, the chalcone family has extensive structural diversity (Zhuang et al. 2017). For example, the chalcone-based benzocycloalkanone derivatives are hybrid chalcones with the core scaffold of 1,3-diaryl-2-propen-1-one (Fig. 3). Some of these compounds structurally related to chalcones include aurone derivatives, which also possess selective A1 affinity [such as hispidol (3)] (Jacobson et al. 2002) and served as inspiration for the structurally related 2-benzylidene-1-tetralone (4–8) (Janse van Rensburg et al. 2017; Legoabe et al. 2018) and 2-benzylidene-1-indanone derivatives (9–12) (Janse van Rensburg et al. 2019a, b), leading to compounds with selective affinity for the A1 AR in the micromolar range.

Fig. 3

The chemical structure and A1Ki (r) values of aurone (3), 2-benzylidene-1-tetralone (4–8) and 2-benzylidene-1-indanone (9–12) derivatives

The chemistry of chalcones is as attractive today as it was years ago; due to the open-chain model and the feature of skeletal modification to produce a new class of organic compounds such as isoxazole-, pyrazole- and indole-based chalcones. Heterocycles occupy a central position in medicinal chemistry and a number of both approved and potential drugs contain at least one heterocyclic nucleus (Khanam 2015; McGrath et al. 2010). Based on the above and in continuation of the efforts to obtain heterocycles as A1 and/or A2A AR antagonists, simple chalcone derivatives (15–36) with either ring A or ring B substitutions as well as a series of novel C3 amino-substituted chalcone derivatives (37–41) with halogen (Br-, Cl- and F-atoms) substitutions on ring B and their regioisomers (42–47) were designed containing the basic chalcone scaffold as structural core. Herein the synthesis, characterization and evaluation, both in vitro and in silico, will be discussed to ascertain the structure–affinity relationships (SAR) that govern A1 and/or A2A AR affinity as well as the future of the said compounds as synthons by which a range of analogues may be designed for future drug discovery.

Results and discussion

Chemistry

The synthesis of the chalcone derivatives 15–36 and 37–41 was done by a Claisen–Schmidt condensation reaction (Claisen and Claparède 1881; Schmidt 1881) of acetophenone and benzaldehyde using either a base or an acid catalyst in a polar solvent (Fig. 4a, b)—this well-known synthetic route is deemed a classical organic chemistry reaction. As stated, the condensing agents are either strong bases or acids (generally, basic conditions are more common in chalcone synthesis) (Gaonkar and Vignesh 2017; Gomes et al. 2017; Rammohan et al. 2020; Zhuang et al. 2017); in the case of base catalysts (15–36), the chalcone is generated from the aldol product via dehydration in an enolate mechanism, while in the case of acid catalysts (37–41), the chalcone is generated via an enol mechanism (Nielsen and Houlihan 2004; Noyce and Pryor 1955). The Claisen–Schmidt condensation reaction is widespread in the literature because of its experimental simplicity and highly efficient formation of the carbon–carbon double bond.

Fig. 4

Synthesis of 15–36, 37–41 and 42–47 (R substituents are identified in Tables 2, 3, 4). Reagents and conditions: a EtOH, KOH (10% (w/v) aqueous solution), room temperature; b MeOH, HCl (32 wt. % in H2O, FCC), 120 °C

Table 2

Ki values for the binding affinity of chalcone derivatives (15–36) against rat (r) A1 and A2A ARs

#	Ring A			Ring B					Ki ± SEM (µM)a [specific binding (%)]b
	2	3	4	2′	3′	4′	5′	6′	rA1c vs [3H]DPCPXe	rA2Ad vs [3H]NECAf
	R1	R2	R3	R1′	R2′	R3′	R4′	R5′
Structural modification of ring A
15	H	H	H	H	H	H	H	H	(35)	(79)
16	OH	H	H	H	H	H	H	H	(65)	(129)
17	H	H	OH	H	H	H	H	H	(56)	(83)
18	H	H	OCH3	H	H	H	H	H	(141)	(106)
19	H	H	Br	H	H	H	H	H	(117)	(117)
Structural modification of ring B
20	H	H	H	H	OH	H	H	H	(32)	(40)
21	H	H	H	H	OCH3	H	H	H	(33)	(106)
22	H	H	H	H	H	OCH3	H	H	(45)	(78)
23	H	H	H	OCH3	H	OCH3	H	H	(30)	(50)
24	H	H	H	OCH3	H	OCH3	OCH3	H	5.3 ± 1	(28)
25	H	H	H	OCH3	H	OCH3	H	OCH3	(27)	(28)
26	H	H	H	Br	H	H	H	H	(24)	(50)
27	H	H	H	H	Br	H	H	H	(28)	(42)
28	H	H	H	H	H	Br	H	H	(110)	(91)
29	H	H	H	Cl	H	H	H	H	6.1 ± 1.1	(42)
30	H	H	H	H	Cl	H	H	H	(23)	(88)
31	H	H	H	H	Cl	Cl	H	H	(114)	(77)
32	H	H	H	H	H	F	H	H	(49)	(98)
33	H	H	H	H	H	CF3	H	H	(80)	(156)
34	H	H	H	H	CN	H	H	H	(52)	(155)
35	H	H	H	–	–	–	–	–	(55)	(75)
36	H	OCH3	H	–	–	–	–	–	(43)	(73)

aAll Ki values were determined in triplicate and expressed as mean ± standard error of the mean (SEM) in µM

bSpecific binding (%) of the radioligand at a maximum tested concentration of 100 µM were determined in duplicate and expressed as the mean in %

cRat receptors were used (rA1: rat whole brain membranes)

dRat receptors were used (rA2A: rat striatal membranes)

e0.1 nM [3H]DPCPX

f4 nM [3H]NECA

Table 3

Ki values for the binding affinity of chalcone derivatives (37–41) against rat (r) A1 and A2A ARs

#	Ring B		Ki ± SEM (µM)a [specific binding (%)]b
	2′	3′	rA1c vs [3H]DPCPXe	rA2Ad vs [3H]NECAf
	R1	R2
37	Br	H	7.1 ± 0.57a	(30)b
38	H	Br	1.6 ± 0.02a	(75)b
39	Cl	H	(26)b	(23)b
40	H	Cl	(30)b	(72)b
41	H	F	(26)b	(71)b

aAll Ki values were determined in triplicate and expressed as mean ± standard error of the mean (SEM) in µM

bSpecific binding (%) of the radioligand at a maximum tested concentration of 100 µM were determined in duplicate and expressed as the mean in %

cRat receptors were used (rA1: rat whole brain membranes)

dRat receptors were used (rA2A: rat striatal membranes)

e0.1 nM [3H]DPCPX

f4 nM [3H]NECA

Table 4

Ki values for the binding affinity of Schiff base derivatives (42–47) against rat (r) A1 and A2A ARs

#	Ring B				Ki ± SEM (µM)a [specific binding (%)]b
	2′	3′	4′	5′	rA1c vs [3H]DPCPXe	rA2Ad vs [3H]NECAf
	R1′	R2′	R3′	R4′
42	OH	H	OCH3	H	(91)b	(197)b
43	OH	H	H	OCH3	(70)b	(75)b
44	OH	Br	H	Cl	(101)b	(95)b
45	OH	H	H	Cl	(44)b	(43)b
46	OH	H	N(CH2CH3)2	H	(74)b	(69)b
47	–	–	–	–	(47)b	(66)b

aAll Ki values were determined in triplicate and expressed as mean ± standard error of the mean (SEM) in µM

bSpecific binding (%) of the radioligand at a maximum tested concentration of 100 µM were determined in duplicate and expressed as the mean in %

cRat receptors were used (rA1: rat whole brain membranes)

dRat receptors were used (rA2A: rat striatal membranes)

e0.1 nM [3H]DPCPX

f4 nM [3H]NECA

The synthesis of the Schiff base or imine derivatives 42–47 by condensation of 3-aminoacetophenone and different substituted benzaldehydes using a base catalyst in a polar solvent (Fig. 4a)—analogous to the reaction conditions of chalcone derivatives 15–36—was not planned. Initially, it was thought that the said reaction conditions would yield compounds 37–41 and not the regioisomers of these C3 amino-substituted chalcone derivatives comprised of two aromatic rings linked via an azomethine (C=N) group. Schiff base or imine compounds are formed by the reaction of a primary amine (such as 3-aminoacetophenone) and either aldehydes (such as the different substituted benzaldehydes) or ketones with the simultaneous removal of water (Furniss et al. 1989). When one, or both, of the reactants are aromatic, the imine is quite stable and usually known as a Schiff base, additionally, in the case of wholly aliphatic reactants the imines tend to decompose or polymerise (Furniss et al. 1989).

Under the adopted synthetic routes, the test compounds 15–36, 37–41 and 42–47 were obtained in fair to good yields, purified by recrystallization from a suitable solvent (EtOH), and in the instance of compounds 37–41 and 42–47, the structure, molecular mass and/or purity of these compounds were verified by 1H NMR, 13C NMR, MS and/or HPLC (Supplementary data). The known chalcone derivatives 15–36 were characterized by 1H NMR only, melting point (DSC) and HPLC and are in accordance with the literature values.

Taking the novel 3-amino-substituted chalcone derivative 38 (R2 = NH2; R2′ = Br) as a representative case, the 1H NMR spectrum displayed two characteristic signals for the ethylenic protons Hα (nearer the carbonyl group) and Hß (next to the Hα proton) at 7.73 and 7.97 ppm, respectively, as doublets (JHα,Hß = 15.7 Hz). Protons on the NH2-group (similar to the proton on the OH-group) are not always visible on a 1H NMR spectrum as protons attached to a N-atom (or O-atom) are acidic, and thus, exchangeable. Also, the 13C NMR spectrum displayed three characteristic signals: a prominent signal for the carbonyl group at 188.36 ppm and two other signals typical of Cα (121.39 ppm) and Cß (142.81 ppm).

From the coupling constant value of the ethylenic protons Hα and Hß (JHα,Hß = 15.7 Hz), it may be deduced that compound 38 exists in the trans (E) configuration (Aksöz and Ertan 2012; Barros et al. 2004; Jung et al. 2008; Opletalova et al. 2000; Rao et al. 2001, 2004). As a result of the diamagnetic anisotropy (deshielding of proton due to local diamagnetic current) of the carbonyl functional group, the ethylenic proton of the (E) isomer gave a signal at a greater chemical shift than the ethylenic proton of the (Z) isomer, as the latter is more remote from the carbonyl functional group (Bayer et al. 1991). Stereochemically, a chalcone might exist as either a cis (Z)- or trans (E)-isomer, having two aromatic rings linked via a three carbon α,ß-unsaturated carbonyl system (Gomes et al. 2017; Hallgas et al. 2005). The cis (Z)-conformer is thermodynamically less stable than the trans (E)-conformer due to the steric effects between the carbonyl group and ring A (Hallgas et al. 2005; Van der Werten et al. 1995). The more stable trans (E)-isomer is the predominant configuration among chalcones (Gomes et al. 2017).

The mass spectrum of compound 38 displayed an intense [M + H]+ molecular ion peak at 302.0164 (79Br isotope) as well as an intense [M + H]+2 peak at 304.0152 (81Br isotope)—in accordance with the formulation depicted (C15H13BrNO). Since Bromine has two isotopes (in a ratio of approximately 1:1), a compound containing a Br-atom (such as compounds 37–38) will have two peaks of similar height in the molecular ion region depending on which bromine isotope the molecular ion contains. Additionally, the two M+ molecular ion peaks for compound 38 (C15H12BrNO) may be seen at 301.0096 (79Br isotope) and 303.0095 (81Br isotope). Of note, chlorine also has two isotopes, namely 35Cl and 37Cl (in a ratio of 3:1) meaning a similar effect may be observed on the mass spectrum of an organic compound containing a Cl-atom (for example, compounds 39–40). The purity of compound 38 determined by HPLC was 98.0549%. The melting point of compound 38 determined by DSC was 176.78 °C.

The word chalcone is derived from the Greek word “chalcos”, meaning “bronze”, which results from the colours of most natural chalcones (Sahu et al. 2012). The bronze colour of chalcones is most likely due to the reactive keto-ethylenic group CO–CH=CH–; a chromophore responsible for the colour of chalcones depending on the presence of other auxochromes (Gaonkar and Vignesh 2017). This bronze colour scheme was also observed with most of the synthesised chalcones, for example, the light brown colour of 38.

As stated, the 3-amino-substituted chalcones and the Schiff base derivatives 42–47 are regioisomers (molecules with the same molecular formula, but different bonding patterns and atomic organisations), and as such, dissimilarities were detected on the 1H NMR and 13C NMR spectra when comparing these compounds. The absence of the ethylenic protons Hα and Hß may be observed on the 1H NMR spectrum and the signals typical of Cα and Cß, on the 13C NMR. In the place of these signals, the proton adjacent to the azomethine group was visible on the 1H NMR spectrum and the carbon of the azomethine group on the 13C NMR spectrum. Additionally, on both the 1H NMR and 13C NMR spectra the methyl group from the acetophenone moiety was clearly visible downfield of all other signals.

Biology

In vitro evaluation

Radioligand binding assays

The degree of binding affinity that the test compounds showed toward rat (r) A1 and A2A ARs were determined via radioligand binding assays in either duplicate [specific binding (%)] or triplicate [inhibition constant (Ki, µM)] and expressed as either mean or mean ± standard error of the mean (SEM), respectively. Only compounds 24, 29, 37 and 38—which displayed specific binding values < 20% at a maximum tested concentration of 100 µM—justified the determination of inhibition constant values (Ki, µM), unlike compounds 15–23, 25–28, 30–36, 39–41 and 42–47 (specific binding values > 20%). The radioligand binding assays were validated with CPA (A1 agonist), DPCPX (A1 antagonist) and istradefylline (A2A antagonist) as reference compounds and results were in accordance with literature values (Table 1).

Table 1

Ki values for the binding affinity of reference compounds against rat (r) A1 and A2A ARs

#	Ki ± SEM (µM)a			SIg	GTP shifth
	rA1b vs [3H]DPCPXd	rA2Ac vs [3H]NECAe	rA1c + GTPf vs [3H]DPCPXd
CPA (A1 agonist)	0.0057 ± 0.0015 (0.0068)i (0.015)j (0.0079)k	0.40 ± 0.17 (0.16)i (0.33)j	0.099 ± 0.015 (0.099)i (0.099)j	70 (24) (22)	17 (15)i (14)j
DPCPX (A1 antagonist)	0.0005 ± 0.00003 (0.0004)i (0.0005)j (0.0003)l	0.23 ± 0.03 (0.55)i (0.53)j (0.34)l	0.0006 ± 0.00003 (0.0004)i (0.0004)j	468 (1362) (1060) (1133)	1.2 (1)i (1.3)j
Istradefylline (A2A antagonist)	0.19 ± 0.01 (0.23)m	0.0014 ± 0.0003 (0.0022)m	0.15 ± 0.02	7.3 (0.0096)	0.79

aAll inhibition constant (Ki) values were determined in triplicate and expressed as mean ± standard error of the mean (SEM) in µM

bRat receptors were used (rA1: rat whole brain membranes)

cRat receptors were used (rA2A: rat striatal membranes)

d0.1 nM [3H]DPCPX

e4 nM [3H]NECA

fAddition of 100 µM GTP to A1 AR radioligand binding assay

gSelectivity index (SI) for the A1 AR isoform calculated as a ratio of A2AKi/A1Ki

hGTP shift calculated by dividing Ki value in the presence of 100 µM GTP by Ki value in the absence of 100 µM GTP

iLiterature value obtained from (Janse van Rensburg et al. 2017)

jLiterature value obtained from (Van der Walt and Terre’Blanche 2015)

kLiterature value obtained from (Bruns et al. 1987)

lLiterature value obtained from (Lohse et al. 1987)

mLiterature value obtained from (Shimada et al. 1997)

Structure–affinity relationships (SAR)

As depicted in Tables 2, 3 and 4, most of the test compounds showed poor A1 and/or A2A AR affinity upon in vitro evaluation—making it difficult to determine SAR; however, some broad conclusions may be drawn from these results. In brief, the test compounds were noticeably more active against the A1 AR than A2A AR and the chalcone derivatives 24, 29, 37 and 38 showed selective A1 AR affinity below 10 µM—of these compounds, 38 had the best A1 AR affinity (A1Ki (r) = 1.6 µM). The type of substitution on ring A of the chalcone scaffold played a key role in affinity, and also, the type and position of the substitution on benzylidene ring B.

As stated, the chalcone derivatives 24 (A1Ki (r) = 5.3 µM) and 29 (A1Ki (r) = 6.1 µM) (Table 2) showed selective A1 AR affinity below 10 µM, and although, the Ki (r) values of these compounds were quite similar the substitutions on ring B were not; 24 was C2′, C4′, C5′-trimethoxy substituted and 29 was C2′-chloro substituted. In the past, the antimicrobial (Karaman et al. 2010; Ramyashree et al. 2017), anticancer (Juvale et al. 2012; Mellado et al. 2018) and/or antioxidant (Shenvi et al. 2013) activities of compounds 24 and 29 were also explored, and now, selective A1 AR affinity may be added.

With regard to compound 24 and A1 AR affinity, it seems that the C2′, C4′, C5′ substitution pattern on ring B was preferred to that of compound 25 ([3H]DPCPX-specific binding (r) = 27%)—interestingly, these compounds differ only at the C5′ (24: –OCH3; 25: –H) and C6′ (24: –H; 25: –OCH3) positions. However, the radioligand-specific binding against the A2A AR of these compounds were the same (24–25: [3H]NECA-specific binding (r) = 28%). Comparison of 24 to the structurally related 2-benzylidene-1-indanone compound 13 (Fig. 5) consisting of a fused 6- and 5-membered ring system (ring A and ring C, respectively) along with C4′-methoxy substitution on ring A had poor A1 and A2A AR affinity (Janse van Rensburg et al. 2019b). The decreased affinity may be due to the increased number of C-, H- and O atoms in the form of the C4 OCH3-group substitution on ring A and the CH2-group present in ring C, increasing the molecular weight from 298.34 g/mol (24) to 340.38 g/mol (13), and perhaps, the resultant steric hindrance (of the CH2-group present in ring C and C2′ OCH3-group on ring B) also decreased binding affinity (Liu et al. 2014; Vásquez-Martínez et al. 2019). Interestingly, according to the physiochemical properties calculated by the free web-tool SwissADME, compound 24 will not be orally bioavailable, but compound 13 will be. Both of these compounds were, unfortunately, not lead-like based on the said parameters. (See “In silico evaluation” for the complete results and discussion).

Fig. 5

The structure–affinity relationships (SAR) of compound 24 versus 13

Comparison of 29 (A1Ki (r) = 6.1 µM) to 30–31 ([3H]DPCPX-specific binding > 20%) demonstrated the importance of the position of the Cl-atom on ring B for A1 AR affinity; the C2′ (ortho) position (29) is preferred to the C3′ (meta) and/or C4′ (para) positions (30–31). Additionally, at the C2′ (ortho) position on ring B the more electronegative Cl-atom (29) is favoured over the less electronegative Br-atom (26). (It must be noted that compounds 26–27 and 30 retained some affinity against the A1 AR based on the specific binding of less than 30%).

These trends were not observed with the C3 amino-substituted chalcone derivatives 37–41 (Table ). Compound 29 paralleled to its counterpart 39 showed decreased A1 AR affinity (39: [3H]DPCPX-specific binding = 26%); this may be credited to the combination of the NH2-group at position C3 on ring A and the Cl-atom at position C2′. Furthermore, C2′ (ortho) Br-group substitution on ring B led to A1 AR affinity below 10 µM; for example, compound 37 (A1Ki (r) = 7.1 µM) versus 39 ([3H]DPCPX-specific binding = 26%). The A1 AR affinity was enhanced almost five-fold when the Br-atom was moved from the C2′ position (37) to the C3′ position (38) on ring B. The importance of the C3 NH2-group substitution on ring A to attain A1 AR affinity was demonstrated by comparison of compound 38 to its unsubstituted counterpart 27 ([3H]DPCPX-specific binding = 28%). Compound 38 had a Ki value of 1.6 µM against the A1 AR—the best of the evaluated compounds Therefore, it may be said that the A1 AR favours C3 NH2-group substitution on ring A in combination with C3′ Br-atom substitution on ring B; based on these chalcones (15–41). The SAR of the chalcone derivatives 27 and 29 and 37–39 are summarized in Fig. .

Fig. 6

The structure–affinity relationships (SAR) of compounds 27 and 29 and 37–39; highlighting the importance of C3 NH2-substitution on ring A in combination with C3′ (meta) Br substitution on ring B for selective rA1 AR affinity

Interestingly, the C3 amino-substituted chalcones 37 ([3H]NECA-specific binding = 30%) and 39 ([3H]NECA-specific binding = 23%) with a C2′ substituent [either a Br- (37) or Cl-atom (39)] on ring B retained some affinity against the A2A AR, unlike compounds 38 ([3H]NECA-specific binding = 75%), 40 ([3H]NECA-specific binding = 72%) and 41 ([3H]NECA-specific binding = 71%) with a C3′ substituent [either a Br- (38), Cl- (40) or F-atom (41)]. (Again, it must be noted that compounds 39–41 retained some affinity against the A1 AR based on the radioligand-specific binding of less than 30%). The importance of not only the type of substituent, but also the position of the substituent on ring B is clear from these comparisons, for example, a compound with selective A1 AR affinity may be obtained when a C3 NH2-group substitution on ring A is combined with a C3′ halogen (Br > Cl > F) substitution on ring B of the chalcone scaffold (e.g. 38, 40–41), whereas a C2′ halogen-substituted ring B could yield a compound with dual A1 and/or A2A AR affinity (37 and 39). Previously, it was found that a coplanar geometry between the two ring systems (namely, ring A and benzylidene ring B) in the flavonols (notably, chalcones may be considered open-chain flavonols) is not required for residual A1 and/or A2A AR affinity; however, the C2′ position on ring B of compounds 37 and 39 was substituted (like the flavonol MRS 1067 with a planar geometry), and thus, there is likely also steric hindrance to free rotation of the benzylidene ring B of these compounds (37 and 39) (Karton et al. 1996). Notably, the structurally related 2-benzylidene-1-indanone 14 (Janse van Rensburg et al. 2019a) that is also C3′ bromo-substituted on benzylidene ring B has selective A2A AR affinity, unlike compound 38 (Fig. 7). Therefore, the benzylidene indanone scaffold with C4 hydroxy-substitution on ring A may be essential to selective A2A AR affinity when compared to the C3 amino-substituted chalcone scaffold. Compound 14 is structurally related to chromone hits with selective A2A AR affinity based on virtual screening, and perhaps, the conformation of these chromones binding to the A2A AR is similar to that of compound 14—leading to selectivity of the A2A AR versus the A1 AR (Langmead et al. 2012).

Fig. 7

The structure–affinity relationships (SAR) of compound 38 versus 14

Additionally, SAR demonstrated the importance of, among other, the phenol functionality in triazine virtual screening hits for selective A2A AR affinity; possibly due to the hydrogen bonding between the phenol functionality and Asn2536.55 (Langmead et al. 2012). This key hydrogen bonding residue (Asn2536.55) is situated centrally and capable of high-quality interactions with diverse heterocyclic compounds (Langmead et al. 2012). Therefore, the phenol functionality (present in compound 14) may also be responsible for the selectivity of the said compound (as seen with the triazine virtual screening hits).

As seen in Table 4 the Schiff base containing compounds 42–47 showed poor A1 and A2A AR affinity. The best radioligand-specific binding at a concentration of 100 µM against A1 and/or A2A AR was demonstrated by compounds 45 and 47; displacing more than 50% of the radioligand at A1 (45 and 47) and A2A (45) ARs. Comparison of 44 to 45 showed that the inclusion of a bromine atom at position C3′ (44), rather than a hydrogen atom (45), in combination with C2′–OH and C5′–Cl substitution on ring B was detrimental to both A1 and A2A AR affinity as the radioligand-specific binding increased more than twofold. The A1 AR seemed to be more tolerable toward the said substitution. Schiff base or imine derivatives are important synthetic intermediates (Furniss et al. 1989); therefore, the synthesised and evaluated compounds 42–47 may be of use in future studies.

GTP shift assays

The type of binding affinity that test compounds 24 and 38 exhibited at the rat A1 AR was determined via a GTP shift assay, as described previously (Lohse et al. 1987; Van der Walt and Terre’Blanche 2015; Van der Werten et al. 1995). GTP shifts were calculated by dividing the Ki values of compounds reported in the presence of GTP by the Ki values obtained in the absence of GTP and the results are summarized in Table 5.

Table 5

A1Ki values (in the absence and presence of GTP) and calculated GTP shifts of 24 and 38

#	Ki ± SEM (µM)a		GTP shifte
	rA1b vs [3H]DPCPXc	rA1b + GTPd vs [3H]DPCPXc
24	5.3 ± 1a	7.8 ± 0.68	1.5
38	1.6 ± 0.02a	1.7 ± 1.3	1.1

aKi values were determined in triplicate and expressed as mean ± standard error of the mean (SEM) in µM

bRat receptors were used (A1: rat whole brain membranes)

c0.1 nM [3H]DPCPX

dAddition of 100 µM GTP to A1 AR radioligand binding assay

eGTP shift calculated by dividing Ki value in the presence of 100 µM GTP by Ki value in the absence of 100 µM GTP

Test compounds 24 and 38 were selected as they possessed the highest A1 AR affinity in the three respective series (Tables 2, 3, 4). The results suggested that compounds 24 and 38 acted as A1 AR antagonists, as the binding curves in the presence of GTP were almost unaffected and the calculated GTP shifts were approximately 1 (Table 5; Fig. 8) (Van der Werten et al. 1995; Gütschow et al. 2012). On account of the structural similarity of the test compounds (15–23, 25–36, 37 and 39–41), it may be supposed that these chalcone derivatives were all A1 AR antagonists.

Fig. 8

The binding curves of compounds 24 (a) and 38 (b) are examples of A1 AR antagonistic action determined via a GTP shift assay performed in triplicate (with and without 100 μM GTP) using rat whole brain membranes expressing A1 ARs with [3H]DPCPX as radioligand. Calculated GTP shift of: 1.5 (24) and 1.1 (38)

In silico evaluation

The physiochemical, pharmacokinetic as well as drug-likeness and medicinal chemistry friendliness of compounds 24, 29, 37 and 38 (the only test compounds with selective A1 AR affinity) were computed via SwissADME and these results are summarized in Tables 6, 7 and 8, respectively.

Table 6

Physiochemical properties of compounds 24, 29, 37 and 38

#	Physiochemical properties								Lipophilicity	Water solubility
	Molecular formula	Molecular weight (g/mol)	Fraction Csp3	Num. rotatable bonds	Num. H-bond acceptors	Num. H-bond donors	Molar refractivity	TPSA (Å2)	Consensus logPo/w	Consensus logS
24	C18H18O4	298.33	0.17	6	4	0	85.72	44.76	3.36	− 4.51
29	C15H13lO	242.70	0	3	1	0	71.26	17.07	3.98	− 4.89
37	C15H12BrNO	302.17	0	3	1	1	78.35	43.09	3.46	− 4.69
38	C15H12BrNO	302.17	0	3	1	1	78.35	43.09	3.53	− 4.86

Table 7

Pharmacokinetic properties of compounds 24, 29, 37 and 38

#	Pharmacokinetic properties
	GI absorption	BBB permeation	Pgp substrate	CYP12 inhibitor	CYP2C19 inhibitor	CYP2C9 inhibitor	CYP2D6 inhibitor	CYP3A4 inhibitor
24	High	Yes	No	Yes	Yes	Yes	Yes	Yes
29	High	Yes	No	Yes	Yes	Yes	No	No
37	High	Yes	No	Yes	Yes	Yes	No	Yes
38	High	Yes	No	Yes	Yes	Yes	No	Yes

Table 8

Drug-likeness and medicinal chemistry friendliness of compounds 24, 29, 37 and 38

#	Drug-likeness					Lead-likeness
	Num. violations
	Lipinskia	Ghoseb	Veberc	Egand	Mueggee	PAINSf	Brenkg	Teagueh
24	0	0	0	0	0	0	1	0
29	0	0	0	0	1	0	1	2
37	0	0	0	0	0	0	2	0
38	0	0	0	0	0	0	2	0

aLipinski: MW < 500, MLOGP < 4.15, N or O < 10, NH or OH < 5 (Lipinski et al. 1997)

bGhose: 160 < MW < 480, − 04 < WLOGP < 5.6, 40 < MR < 130, 20 < atoms < 70 (Ghose et al. 1999)

cVeber: Num. rotatable bonds < 10, TPSA < 140 (Veber et al. 2002)

dEgan: WLOGP < 5.88, TPSA < 131.6 (Egan et al. 2000)

eMuegge: 200 < MW < 600, − 2 < XLOGP < 5, TPSA < 150, num. rings < 7, num. carbon > 4, num. heteroatoms > 1, num. rotatable bonds < 15, num. H-bond acceptors < 10, num. H-bond donors < 5 (Muegge et al. 2001)

fPan assay interference compounds (PAINS) implemented from Baell and Holloway (2010)

gStructural alert implemented from Brenk et al. (2008)

hTeague: 250 < MW < 350, num. rotatable bonds < 7, XLOGP3 < 3.5 (Teague et al. 1999)

The bioavailability radar (Fig. 9) gave a first glance at the drug-likeness of these compounds; compounds 24, 29, 37 and 38 fall within the optimal range for the parameters lipophilicity, size, polarity, solubility and flexibility (see Table 6), except for saturation (Csp3 > 0.25). The fraction of carbon atoms in the sp3 hybridization of compounds 24 (0.17), 29 (0), 37 (0) and 38 (0) were smaller than 0.25 and, thus, out of range. Consequently, compounds 24, 29, 37 and 38 were predicted not orally bioavailable.

Fig. 9

The pink area represents the optimal range for lipophilcity (LIPO: − 0.7 < XLOGP3 < + 5.0), size (SIZE: 150 < MW < 500), polarity (POLAR: 20 < TPSA < 130), solubility (INSOLU: logS < 6), saturation (INSATU: fraction Csp3 > 0.25) and flexibility (FLEX: num. rotatable bonds < 9). The red lines represent the said parameters of compounds 24, 29, 37 and 38. The red lines must fall completely within the pink area for a compound to be considered drug-like; therefore, compounds 24, 29, 37 and 38 are predicted not orally bioavailable

The qualitative solubility classes of compounds 24, 29, 37 and 38 were moderately soluble to soluble in water [the predicted values are the decimal logarithm of the molar solubility in water (logS)]. This was encouraging given the poor solubility of most chalcone-based compounds [most probably the reason for poor in vivo efficacy in preclinical studies (Zhuang et al. 2017)]. Seeing as solubility influences GI absorption (Ottaviani et al. 2010), a soluble molecule will most probably facilitate drug development (Ritchie et al. 2013).

The BOILED-Egg model (Fig. 10) predicted passive human gastrointestinal absorption (HIA) and blood–brain-barrier (BBB) permeation as a function of the position of the compound in the WLOGP-versus-TPSA referential (Daina and Zoete 2016). Compounds 24, 29, 37 and 38 were predicted to have high GI absorption and permeate the BBB (Table 7). Additionally, these compounds were not substrates for the permeability glycoprotein (Pgp) responsible for efflux through biological membranes, for instance from the gastrointestinal wall to the lumen or from the brain (Montanari and Ecker 2015). Pgp also protects the CNS from xenobiotics (Szakács et al. 2008).

Fig. 10

Compounds 24, 29, 37 and 38, which are not a substrate for Pgp (PGP-), is represented by the red circles in the yellow region. The white region is for high probability of passive absorption by the gastrointestinal tract (HIA), and the yellow region (yolk) is for high probability of brain penetration (BBB). White and yellow (yolk) regions are not mutually exclusive

The interaction of a compound with cytochromes P450 (CYP) is important; seeing as these isoenzymes modulate drug elimination through metabolic transformation (Testa and Kraemer 2007); probably causing pharmacokinetics-related drug–drug interactions (Hollenberg 2002; Huang et al. 2008), leading to adverse or even toxic effects due to the lower clearance and accumulation of the drug or its metabolites (Kirchmair et al. 2015). Compounds 24, 29, 37 and 38 were predicted to inhibit the major CYP isoforms (CYP12, CYP2C19, CYP2C9, CYP2D6 and CYP3A4), with the exception of 24 and/or 37–38 that did not inhibit CYP2D6 (29 and 37–38) and CYP3A4 (29).

The drug-likeness of compounds 24, 29, 37 and 38 was assessed by various rule-based filters (Egan et al. 2000; Ghose et al. 1999; Lipinski et al. 1997; Muegge et al. 2001; Veber et al. 2002) (Table 8). Compound 29 violated a rule-based filter from Muegge (Muegge et al. 2001); namely that a compound should contain more than one heteroatom and the only atom other than carbon and hydrogen that 29 contained was a chloro atom. Additionally, compounds 24, 29, 37 and 38 all had a bioavailability score of 0.55 (the probability that a compound will have > 10% bioavailability in rat or measurable Caco-2 permeability) (Martin 2005).

The medicinal chemistry friendliness of compounds 24, 29, 37 and 38 was assessed by the identification of potentially problematic fragments; first, pan assay interference compounds (PAINS) (Baell and Holloway 2010) and, second, structural alerts (Brenk et al. 2008). Compounds 24, 29, 37 and 38 contained no PAINS; however, Brenk violations were present (Table 8). All these compounds contained an α,ß-unsaturated carbonyl system perceived as a potential Michael acceptor. Additionally, compounds 37–38 contained a potentially unwanted aniline group; as aniline groups are perhaps inclined toward problems with genetic toxicity (Clayson 1981).

As compounds 24, 29, 37 and 38 did not have pronounced affinity toward the A1 AR, the present structures need to be optimised, and here, lead-likeness [a molecular entity suitable for optimization, most probably increasing size and lipophilicity (Hann and Keserü 2012)] plays a role (Teague et al. 1999). Accordingly, compounds 24, 37 and 38 (but not compound 29; due to molecular weight value smaller than 250 g/mol and consensus logPo/w value larger than 3.5) may be considered lead-like (Tables 6, 8).

Therefore, the optimization of the physiochemical properties of chalcones and, thus, the pharmacokinetic properties as well as drug-likeness and medicinal chemistry friendliness is of great importance for medicinal chemistry research on chalcone-based compounds.

Experimental

Materials

Unless otherwise noted, all starting materials and solvents were purchased from commercial vendors and used without further purification. Thin layer chromatography on TLC silica gel 60 F254 aluminium sheets from Merck was used to monitor reaction progress. Melting points were determined by means of differential scanning calorimetry (DSC) with a Mettler DSC 3 Star System (Mettler Toledo, Greifensee, Switzerland). Proton (1H) and carbon (13C) nuclear magnetic resonance (NMR) spectra were recorded on a Bruker Avance III 600 spectrometer at frequencies of 600 and 151 MHz, respectively, using either CDCl3 or DMSO-d6 as solvent and TMS as reference. Chemical shifts were reported in parts per million (ppm) in relation to the solvent peak (CDCl3: residual CH at 7.26 ppm and DMSO-d6: residual CH3 at 2.50 ppm for 1H NMR). Spin multiplicities were indicated as follows: singlet (s), doublet (d), triplet (t), quartet (q), doublet of doublets (dd), triplet of doublets (td), double double doublet (ddd) and multiplet (m). Coupling constant (J) values were reported in Hertz (Hz). High-resolution mass spectra (HRMS) were recorded on a Bruker micrOTOF-Q II mass spectrometer in atmospheric chemical ionisation (APCI) mode. High-performance liquid chromatography (HPLC) analyses were done on an Agilent 1100 HPLC system.

Synthesis of 15–36

(2E)-1,3-Diphenylprop-2-en-1-one (15)

A solution of acetophenone (0.50 g, 4.16 mmol) and benzaldehyde (0.44 g, 4.16 mmol) in EtOH (5 mL) was mechanically stirred at room temperature for approximately 5 min before the dropwise addition of KOH (10% (w/v) aqueous solution, 5 mL). The subsequent reaction mixture was mechanically stirred at room temperature and continuously monitored by TLC. Upon completion, the reaction mixture was quenched with crushed ice (15 g) and acidified to pH 2 with HCl (32 wt. % in H2O, FCC). The subsequent precipitate was collected by vacuum filtration, dried (30 °C) and recrystallized from EtOH to yield the title compound 15 as light yellow crystals (0.43 g, 49%): mp: 57.87 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.16 (d, J = 7.7 Hz, 2H), 7.95 (d, J = 15.6 Hz, 1H), 7.89 (d, J = 4.8 Hz, 2H), 7.76 (d, J = 15.6 Hz, 1H), 7.68 (t, J = 7.3 Hz, 1H), 7.58 (t, J = 7.5 Hz, 2H), 7.46 (d, J = 4.4 Hz, 3H) (Xie et al. 2017). Purity (HPLC): 99.8%.

(2E)-1-(2-Hydroxyphenyl)-3-phenylprop-2-en-1-one (16)

Prepared as for 15 from 2′-hydroxyacetophenone (0.50 g, 3.67 mmol) and benzaldehyde (0.39 g, 3.67 mmol) to yield the title compound 16 as dark yellow crystals (0.05 g, 6%): mp: 87.21 °C (EtOH); 1H NMR (600 MHz, CDCl3) δ 12.82 (s, 1H), 7.97–7.90 (m, 2H), 7.71–7.64 (m, 3H), 7.51 (t, J = 7.7 Hz, 1H), 7.45 (d, J = 2.6 Hz, 3H), 7.04 (d, J = 8.3 Hz, 1H), 6.95 (t, J = 7.5 Hz, 1H) (Xie et al. 2017). Purity (HPLC): 99%.

(2E)-1-(4-Hydroxyphenyl)-3-phenylprop-2-en-1-one (17)

Prepared as for 15 from 4′-hydroxyacetophenone (0.50 g, 3.67 mmol) and benzaldehyde (0.39 g, 3.67 mmol) to yield the title compound 17 as beige crystals (0.15 g, 18%): mp: 171.29 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 10.46 (d, J = 7.9 Hz, 1H), 8.10–8.04 (m, 2H), 7.90 (d, J = 15.6 Hz, 1H), 7.86 (dd, J = 7.5, 1.7 Hz, 2H), 7.68 (d, J = 15.6 Hz, 1H), 7.45 (d, J = 6.9 Hz, 3H), 6.93–6.89 (m, 2H) (Xie et al. 2017). Purity (HPLC): 98.9%.

(2E)-1-(4-Methoxyphenyl)-3-phenylprop-2-en-1-one (18)

Prepared as for 15 from 4′-methoxyacetophenone (0.50 g, 3.33 mmol) and benzaldehyde (0.35 g, 3.33 mmol) to yield the title compound 18 as white crystals (0.79 g, 52%): mp: 105.91 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.17 (d, J = 7.7 Hz, 2H), 7.94 (d, J = 15.6 Hz, 1H), 7.88 (d, J = 6.8 Hz, 2H), 7.71 (d, J = 15.6 Hz, 1H), 7.45 (d, J = 5.5 Hz, 3H), 7.09 (d, J = 7.7 Hz, 2H), 3.87 (s, 3H) (Zhang et al. 2015). Purity (HPLC): 97.2%.

(2E)-1-(4-Bromophenyl)-3-phenylprop-2-en-1-one (19)

Prepared as for 15 from 4′-bromoacetophenone (0.50 g, 2.51 mmol) and benzaldehyde (0.27 g, 2.51 mmol) to yield the title compound 19 as light yellow crystals (0.72 g, 63%): mp: 103.26 °C (EtOH); 1H NMR (600 MHz, CDCl3) δ 7.92–7.87 (m, 2H), 7.82 (d, J = 15.7 Hz, 1H), 7.69–7.60 (m, 4H), 7.48 (d, J = 15.7 Hz, 1H), 7.44–7.40 (m, 3H) (Zhou et al. 2016). Purity (HPLC): 98%.

(2E)-3-(3-Hydroxyphenyl)-1-phenylprop-2-en-1-one (20)

Prepared as for 15 from acetophenone (0.50 g, 4.16 mmol) and 3-hydroxybenzaldehyde (0.51 g, 4.16 mmol) to yield the title compound 20 as light yellow crystals (0.51 g, 55%): mp: 163.81 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 9.63 (s, 1H), 8.17–8.09 (m, 2H), 7.83 (d, J = 15.6 Hz, 1H), 7.69–7.63 (m, 2H), 7.57 (t, J = 7.7 Hz, 2H), 7.31 (d, J = 7.7 Hz, 1H), 7.26 (t, J = 7.8 Hz, 1H), 7.23 (s, 1H), 6.88 (dd, J = 7.9, 1.7 Hz, 1H) (Xie et al. 2017). Purity (HPLC): 99.9%.

(2E)-3-(3-Methoxyphenyl)-1-phenylprop-2-en-1-one (21)

Prepared as for 15 from acetophenone (0.50 g, 4.16 mmol) and 3-methoxybenzaldehyde (0.57 g, 4.16 mmol) to yield the title compound 21 as dark yellow crystals (0.19 g, 19%): mp: 60.43 °C (EtOH); 1H NMR (600 MHz, CDCl3) δ 8.14–8.10 (m, 2H), 7.88 (d, J = 15.7 Hz, 1H), 7.69 (t, J = 7.4 Hz, 1H), 7.64–7.58 (m, 3H), 7.44 (t, J = 7.9 Hz, 1H), 7.38–7.32 (m, 1H), 7.26 (d, J = 1.7 Hz, 1H), 7.07 (dd, J = 8.2, 2.4 Hz, 1H), 3.96 (s, 3H) (Unoh et al. 2013). Purity (HPLC): 98.2%.

(2E)-3-(4-Methoxyphenyl)-1-phenylprop-2-en-1-one (22)

Prepared as for 15 from acetophenone (0.50 g, 4.16 mmol) and 4-methoxybenzaldehyde (0.57 g, 4.16 mmol) to yield the title compound 22 as light yellow crystals (0.06 g, 6%): mp: 72.02 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.17–8.08 (m, 2H), 7.88–7.83 (m, 2H), 7.80 (d, J = 15.6 Hz, 1H), 7.72 (d, J = 15.6 Hz, 1H), 7.66 (t, J = 7.4 Hz, 1H), 7.57 (t, J = 7.7 Hz, 2H), 7.06–6.97 (m, 2H), 3.83 (s, 3H) (Zhang et al. 2015). Purity (HPLC): 100%.

(2E)-3-(2,4-Dimethoxyphenyl)-1-phenylprop-2-en-1-one (23)

Prepared as for 15 from acetophenone (0.50 g, 4.16 mmol) and 2,4-dimethoxybenzaldehyde (0.69 g, 4.16 mmol) to yield the title compound 23 as dark yellow crystals (0.18 g, 16%): mp: 69.63 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.08 (d, J = 7.5 Hz, 2H), 7.99 (d, J = 15.7 Hz, 1H), 7.91 (d, J = 8.6 Hz, 1H), 7.74 (d, J = 15.7 Hz, 1H), 7.64 (t, J = 7.3 Hz, 1H), 7.56 (t, J = 7.6 Hz, 2H), 6.67–6.60 (m, 2H), 3.90 (s, 3H), 3.84 (s, 3H) (Suwito et al. 2014). Purity (HPLC): 94.8%.

(2E)-1-Phenyl-3-(2,4,5-trimethoxyphenyl)prop-2-en-1-one (24)

Prepared as for 15 from acetophenone (0.50 g, 4.16 mmol) and 2,4,5-trimethoxybenzaldehyde (0.82 g, 4.16 mmol) to yield the title compound 24 as dark yellow crystals (0.95 g, 79%): mp: 102.97 °C (EtOH); 1H NMR (600 MHz, CDCl3) δ 8.09 (d, J = 15.8 Hz, 1H), 8.03–7.97 (m, 2H), 7.55 (t, J = 7.3 Hz, 1H), 7.47 (dd, J = 20.6, 11.8 Hz, 3H), 7.13 (s, 1H), 6.52 (s, 1H), 3.94 (s, 3H), 3.90 (s, 6H) (Shenvi et al. 2013). Purity (HPLC): 97.6%.

(2E)-1-Phenyl-3-(2,4,6-trimethoxyphenyl)prop-2-en-1-one (25)

Prepared as for 15 from acetophenone (0.50 g, 4.16 mmol) and 2,4,6-trimethoxybenzaldehyde (0.82 g, 4.16 mmol) to yield the title compound 25 as bright yellow crystals (0.43 g, 35%): mp: 109.17 °C (EtOH); 1H NMR (600 MHz, CDCl3) δ 8.26 (d, J = 15.9 Hz, 1H), 8.03–7.98 (m, 2H), 7.88 (d, J = 15.9 Hz, 1H), 7.53 (t, J = 7.3 Hz, 1H), 7.47 (t, J = 7.5 Hz, 2H), 6.14 (s, 2H), 3.90 (s, 6H), 3.86 (s, 3H) (Sawle et al. 2008). Purity (HPLC): 88%.

(2E)-3-(2-Bromophenyl)-1-phenylprop-2-en-1-one (26)

Prepared as for 15 from acetophenone (0.50 g, 4.16 mmol) and 2-bromobenzaldehyde (0.77 g, 4.16 mmol) to yield the title compound 26 as dark yellow crystals (0.90 g, 75%): mp: 47.76 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.24–8.12 (m, 3H), 8.01 (d, J = 15.5 Hz, 1H), 7.95 (d, J = 15.5 Hz, 1H), 7.75 (dd, J = 8.0, 0.9 Hz, 1H), 7.69 (t, J = 7.4 Hz, 1H), 7.59 (t, J = 7.7 Hz, 2H), 7.50 (t, J = 7.5 Hz, 1H), 7.40 (td, J = 7.9, 1.6 Hz, 1H) (Wu et al. 2017). Purity (HPLC): 97.1%.

(2E)-3-(3-Bromophenyl)-1-phenylprop-2-en-1-one (27)

Prepared as for 15 from acetophenone (0.50 g, 4.16 mmol) and 3-bromobenzaldehyde (0.77 g, 4.16 mmol) to yield the title compound 27 as light yellow crystals (0.62 g, 52%): mp: 84.50 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.24–8.16 (m, 3H), 8.03 (d, J = 15.6 Hz, 1H), 7.87 (d, J = 7.8 Hz, 1H), 7.74–7.62 (m, 3H), 7.58 (dd, J = 10.7, 4.7 Hz, 2H), 7.42 (t, J = 7.8 Hz, 1H) (Wang et al. 2019). Purity (HPLC): 86.5%.

(2E)-3-(4-Bromophenyl)-1-phenylprop-2-en-1-one (28)

Prepared as for 15 from acetophenone (0.50 g, 4.16 mmol) and 4-bromobenzaldehyde (0.77 g, 4.16 mmol) to yield the title compound 28 as light yellow crystals (0.58 g, 49%): mp: 119.66 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.19–8.12 (m, 2H), 7.98 (d, J = 15.7 Hz, 1H), 7.87 (d, J = 8.5 Hz, 2H), 7.72 (d, J = 15.7 Hz, 1H), 7.69–7.64 (m, 3H), 7.58 (t, J = 7.7 Hz, 2H) (Wu et al. 2017). Purity (HPLC): 98.3%.

(2E)-3-(2-Chlorophenyl)-1-phenylprop-2-en-1-one (29)

Prepared as for 15 from acetophenone (0.50 g, 4.16 mmol) and 2-chlorobenzaldehyde (0.58 g, 4.16 mmol) to yield the title compound 29 as light yellow crystals (0.32 g, 32%): mp: 50.30 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.23 (dd, J = 7.5, 1.9 Hz, 1H), 8.20–8.15 (m, 2H), 8.03 (q, J = 15.6 Hz, 2H), 7.72–7.66 (m, 1H), 7.62–7.55 (m, 3H), 7.52–7.43 (m, 2H) (Wu et al. 2017). Purity (HPLC): 97.5%.

(2E)-3-(3-Chlorophenyl)-1-phenylprop-2-en-1-one (30)

Prepared as for 15 from acetophenone (0.50 g, 4.16 mmol) and 3-chlorobenzaldehyde (0.58 g, 4.16 mmol) to yield the title compound 30 as light yellow crystals (0.32 g, 32%): mp: 75.30 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.19 (d, J = 7.3 Hz, 2H), 8.09 (s, 1H), 8.04 (d, J = 15.7 Hz, 1H), 7.83 (d, J = 7.2 Hz, 1H), 7.73 (d, J = 15.7 Hz, 1H), 7.69 (t, J = 7.4 Hz, 1H), 7.58 (t, J = 7.7 Hz, 2H), 7.53–7.46 (m, 2H) (Robinson et al. 2015). Purity (HPLC): 96%.

(2E)-3-(3,4-Dichlorophenyl)-1-phenylprop-2-en-1-one (31)

Prepared as for 15 from acetophenone (0.50 g, 4.16 mmol) and 3,4-dichlorobenzaldehyde (0.73 g, 4.16 mmol) to yield the title compound 31 as light yellow crystals (0.35 g, 30%): mp: 108.62 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.30 (d, J = 2.0 Hz, 1H), 8.21–8.17 (m, 2H), 8.07 (d, J = 15.7 Hz, 1H), 7.89 (dd, J = 8.4, 1.9 Hz, 1H), 7.75–7.66 (m, 3H), 7.59 (t, J = 7.7 Hz, 2H) (Choi et al. 2016). Purity (HPLC): 95.9%.

(2E)-3-(4-Fluorophenyl)-1-phenylprop-2-en-1-one (32)

Prepared as for 15 from acetophenone (0.50 g, 4.16 mmol) and 4-fluorobenzaldehyde (0.52 g, 4.16 mmol) to yield the title compound 32 as light yellow crystals (0.32 g, 34%): mp: 86.71 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.15 (dd, J = 8.2, 1.1 Hz, 2H), 8.03–7.94 (m, 2H), 7.91 (d, J = 15.6 Hz, 1H), 7.75 (d, J = 15.7 Hz, 1H), 7.70–7.64 (m, 1H), 7.57 (dd, J = 10.7, 4.7 Hz, 2H), 7.34–7.26 (m, 2H) (Stroba et al. 2009). Purity (HPLC): 100%.

(2E)-1-Phenyl-3-[4-(trifluoromethyl)phenyl]prop-2-en-1-one (33)

Prepared as for 15 from acetophenone (0.50 g, 4.16 mmol) and 4-(trifluoromethyl)benzaldehyde (0.72 g, 4.16 mmol) to yield the title compound 33 as light yellow crystals (0.41 g, 36%): mp: 126.59 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.20–8.16 (m, 2H), 8.11 (dd, J = 21.5, 11.9 Hz, 3H), 7.81 (dd, J = 11.9, 8.8 Hz, 3H), 7.73–7.66 (m, 1H), 7.59 (dd, J = 10.6, 4.8 Hz, 2H) (Downey et al. 2018). Purity (HPLC): 94.4%.

3-[(1E)-3-Oxo-3-phenylprop-1-en-1-yl]benzonitrile (34)

Prepared as for 15 from acetophenone (0.50 g, 4.16 mmol) and 3-cyanobenzaldehyde (0.55 g, 4.16 mmol) to yield the title compound 34 as light yellow crystals (0.63 g, 65%): mp: 113.82 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.51 (s, 1H), 8.20 (dd, J = 8.1, 1.0 Hz, 3H), 8.12 (d, J = 15.7 Hz, 1H), 7.90 (d, J = 7.7 Hz, 1H), 7.77 (d, J = 15.7 Hz, 1H), 7.67 (dd, J = 12.7, 4.7 Hz, 2H), 7.59 (t, J = 7.7 Hz, 2H) (Nagarajan and Shechter 1984). Purity (HPLC): 94.7%.

(2E)-3-[4-(Morpholin-4-yl)phenyl]-1-phenylprop-2-en-1-one (35)

Prepared as for 15 from acetophenone (0.50 g, 4.16 mmol) and 4-(4-morpholinyl)benzaldehyde (0.80 g, 4.16 mmol) to yield the title compound 35 as bright yellow crystals (0.15 g, 13%): mp: 149.99 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.14–8.09 (m, 2H), 7.78–7.61 (m, 5H), 7.55 (t, J = 7.7 Hz, 2H), 6.99 (d, J = 8.9 Hz, 2H), 3.77–3.71 (m, 4H), 3.28–3.22 (m, 4H) (Li et al. 2017). Purity (HPLC): 99.2%.

(2E)-3-(2H-1,3-benzodioxol-5-yl)-1-(3-methoxyphenyl)prop-2-en-1-one (36)

Prepared as for 15 from 3-methoxyacetophenone (0.50 g, 3.33 mmol) and 3,4-(methylenedioxy)benzaldehyde (0.50 g, 3.33 mmol) to yield the title compound 36 as bright yellow crystals (0.92 g, 98%): mp: 77.10 °C (EtOH); 1H NMR (600 MHz, CDCl3) δ 7.73 (d, J = 15.5 Hz, 1H), 7.58 (d, J = 7.6 Hz, 1H), 7.53 (s, 1H), 7.43–7.31 (m, 2H), 7.16 (s, 1H), 7.12 (d, J = 7.9 Hz, 2H), 6.84 (d, J = 7.5 Hz, 1H), 6.02 (s, 2H), 3.88 (s, 3H) (Ruparelia et al. 2018). Purity (HPLC): 99.1%.

Synthesis of 37–41

(2E)-1-(3-Aminophenyl)-3-(2-bromophenyl)prop-2-en-1-one (37)

3′-Aminoacetophenone (0.50 g, 3.70 mmol) and 2-bromobenzaldehyde (0.68 g, 3.70 mmol) were suspended in MeOH (4 mL) and HCl (32 wt.% in H2O, FCC, 6 mL). The subsequent reaction mixture was mechanically stirred at 120 °C under reflux while being continuously monitored by TLC. Upon completion, the reaction mixture was cooled to room temperature, ice (15 g) was added and the resulting precipitate was filtered, dried (30 °C) and recrystallized from a suitable solvent to yield the title compound 37 as dark brown powder (1.10 g, 99%): mp: 189.86 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.21–8.12 (m, 2H), 8.01 (d, J = 15.5 Hz, 1H), 7.95 (s, 1H), 7.91 (d, J = 15.5 Hz, 1H), 7.75 (d, J = 7.9 Hz, 1H), 7.64 (t, J = 7.8 Hz, 1H), 7.58 (dd, J = 8.1, 0.8 Hz, 1H), 7.50 (t, J = 7.5 Hz, 1H), 7.44–7.37 (m, 1H); 13C NMR (151 MHz, DMSO) δ 188.30, 141.78, 138.30, 133.80, 133.34, 132.35, 130.19, 128.76, 128.26, 126.55, 126.49, 125.44, 124.70, 121.31. APCI-HRMS m/z calculated for C15H13BrNO [M + H] + : 302.0175, found: 302.0176. Purity (HPLC): 99.8%.

(2E)-1-(3-Aminophenyl)-3-(3-bromophenyl)prop-2-en-1-one (38)

Prepared as for 37 from 3′-aminoacetophenone (0.50 g, 3.70 mmol) and 3-bromobenzaldehyde (0.68 g, 3.70 mmol) to yield the title compound 38 as light brown powder (0.80 g, 72%): mp: 176.78 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.26–8.13 (m, 2H), 7.97 (d, J = 15.7 Hz, 1H), 7.95 (s, 1H), 7.87 (d, J = 7.8 Hz, 1H), 7.73 (d, J = 15.7 Hz, 1H), 7.64 (t, J = 7.8 Hz, 2H), 7.58 (d, J = 7.8 Hz, 1H), 7.43 (t, J = 7.8 Hz, 1H); 13C NMR (151 MHz, DMSO) δ 188.36, 142.82, 138.42, 137.04, 135.23, 133.22, 130.98, 130.91, 130.17, 128.27, 126.72, 126.60, 123.32, 122.40, 121.45. APCI-HRMS m/z calculated for C15H13BrNO [M + H] + : 302.0175, found: 302.0164. Purity (HPLC): 98.1%.

(2E)-1-(3-Aminophenyl)-3-(2-chlorophenyl)prop-2-en-1-one (39)

Prepared as for 37 from 3′-aminoacetophenone (0.50 g, 3.70 mmol) and 2-chlorobenzaldehyde (0.52 g, 3.70 mmol) to yield the title compound 39 as light brown powder (0.94 g, 99%): mp: 188.04 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.23–8.13 (m, 2H), 8.05 (d, J = 15.6 Hz, 1H), 7.95 (d, J = 15.6 Hz, 2H), 7.65 (t, J = 7.8 Hz, 1H), 7.62–7.56 (m, 2H), 7.53–7.43 (m, 2H); 13C NMR (151 MHz, DMSO) δ 188.30, 139.05, 138.31, 134.43, 132.20, 132.12, 130.20, 130.08, 128.61, 127.73, 126.63, 126.58, 124.56, 121.39. APCI-HRMS m/z calculated for C15H13ClNO [M + H] + : 258.0680, found: 258.0673. Purity (HPLC): 99.8%.

(2E)-1-(3-Aminophenyl)-3-(3-chlorophenyl)prop-2-en-1-one (40)

Prepared as for 37 from 3′-aminoacetophenone (0.50 g, 3.70 mmol) and 3-chlorobenzaldehyde (0.52 g, 3.70 mmol) to yield the title compound 40 as light brown powder (0.89 g, 93%): mp: 175.73 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.17 (d, J = 7.6 Hz, 1H), 8.06 (s, 1H), 7.97 (d, J = 15.7 Hz, 1H), 7.95 (s, 1H), 7.83 (d, J = 7.3 Hz, 1H), 7.74 (d, J = 15.7 Hz, 1H), 7.64 (t, J = 7.8 Hz, 1H), 7.58 (d, J = 7.9 Hz, 1H), 7.54–7.45 (m, 2H); 13C NMR (151 MHz, DMSO) δ 188.39, 142.86, 138.41, 136.78, 133.81, 130.73, 130.32, 130.16, 128.05, 127.90, 126.61, 126.56, 123.36, 121.40. APCI-HRMS m/z calculated for C15H13ClNO [M + H] + : 258.0680, found: 258.0658. Purity (HPLC): 98.1%.

(2E)-1-(3-Aminophenyl)-3-(3-fluorophenyl)prop-2-en-1-one (41)

Prepared as for 37 from 3′-aminoacetophenone (0.50 g, 3.70 mmol) and 3-fluorobenzaldehyde (0.46 g, 3.70 mmol) to yield the title compound 41 as dark brown powder (0.19 g, 21%): mp: 206.07 °C (EtOH); 1H NMR (600 MHz, DMSO) δ 8.17 (d, J = 7.7 Hz, 1H), 7.99–7.93 (m, 2H), 7.85 (d, J = 10.3 Hz, 1H), 7.76 (d, J = 15.7 Hz, 1H), 7.70 (d, J = 7.8 Hz, 1H), 7.65 (t, J = 7.8 Hz, 1H), 7.63–7.57 (m, 1H), 7.55–7.47 (m, J = 7.9, 6.3 Hz, 1H), 7.30 (td, J = 8.5, 2.4 Hz, 1H); 13C NMR (151 MHz, DMSO) δ 188.43, 163.28, 161.67, 143.14, 138.43, 137.12, 137.07, 135.17, 130.93, 130.88, 130.19, 126.74, 126.66, 125.61, 123.29, 121.54, 117.54, 117.40, 114.83, 114.69. APCI-HRMS m/z calculated for C15H13FNO [M + H] + : 242.0976, found: 242.0956. Purity (HPLC): 96.4%.

Synthesis of 42–47

1-(3-{(E)-[(2-Hydroxy-4-methoxyphenyl)methylidene]amino}phenyl)ethan-1-one (42)

A solution of 3′-aminoacetophenone (0.50 g, 3.70 mmol) and 2-hydroxy-4-methoxybenzaldehyde (0.56 g, 3.70 mmol) in EtOH (5 mL) was mechanically stirred at room temperature for approximately 5 min before the dropwise addition of KOH (10% (w/v) aqueous solution, 5 mL). The subsequent reaction mixture was mechanically stirred at room temperature and continuously monitored by TLC. Upon completion, the reaction mixture was quenched with crushed ice (15 g) and acidified to pH 2 with HCl (32 wt. % in H2O, FCC). The subsequent precipitate was collected by vacuum filtration, dried (30 °C) and recrystallized from EtOH to yield the title compound 42 as dark green crystals (0.08 g, 8%): mp: 94.63 °C (EtOH); 1H NMR (600 MHz, CDCl3) δ 13.47 (d, J = 2.0 Hz, 1H), 8.58 (s, 1H), 7.83 (s, 2H), 7.54–7.41 (m, 2H), 7.30 (d, J = 8.3 Hz, 1H), 6.51 (d, J = 8.0 Hz, 2H), 3.85 (s, 3H), 2.64 (s, 3H); 13C NMR (151 MHz, CDCl3) δ 197.84, 164.45, 163.95, 162.75, 149.21, 138.47, 133.96, 129.76, 126.32, 126.21, 120.41, 113.11, 107.54, 101.24, 55.65, 26.85. APCI-HRMS m/z calculated for C16H16NO3 [M + H] + : 270.1125, found: 270.1116. Purity (HPLC): 99.8%.

1-(3-{(Z)-[(2-Hydroxy-5-methoxyphenyl)methylidene]amino}phenyl)ethan-1-one (43)

Prepared as for 42 from 3′-aminoacetophenone (0.50 g, 3.70 mmol) and 2-hydroxy-5-methoxybenzaldehyde (0.56 g, 3.70 mmol) to yield compound 43 as dark red crystals (0.59 g, 59%): mp: 92.81 °C (EtOH); 1H NMR (600 MHz, CDCl3) δ 12.53 (s, 1H), 8.63 (s, 1H), 7.86 (d, J = 11.7 Hz, 2H), 7.52 (t, J = 7.6 Hz, 1H), 7.47 (d, J = 7.8 Hz, 1H), 7.02 (d, J = 9.0 Hz, 1H), 6.97 (d, J = 8.9 Hz, 1H), 6.91 (s, 1H), 3.81 (s, 3H), 2.65 (s, 3H); 13C NMR (151 MHz, CDCl3) δ 197.73, 163.63, 155.57, 152.50, 149.20, 138.48, 129.86, 126.84, 126.37, 121.20, 120.44, 118.63, 118.32, 115.52, 56.08, 26.91. APCI-HRMS m/z calculated for C16H16NO3 [M + H] + : 270.1125, found: 270.1112. Purity (HPLC): 94.4%.

1-(3-{(E)-[(3-Bromo-5-chlorophenyl)methylidene]amino}phenyl)ethan-1-one (44)

Prepared as for 42 from 3′-aminoacetophenone (0.50 g, 3.70 mmol) and 3-bromo-5-chlorosalicylaldehyde (0.87 g, 3.70 mmol) to yield the title compound 44 as light orange crystals (0.42 g, 32%): mp: 183.31 °C (EtOH); 1H NMR (600 MHz, CDCl3) δ 14.07 (s, 1H), 8.61 (s, 1H), 7.95–7.84 (m, 2H), 7.64 (s, 1H), 7.56 (t, J = 7.7 Hz, 1H), 7.49 (d, J = 7.8 Hz, 1H), 7.38 (s, 1H), 2.65 (s, 3H); 13C NMR (151 MHz, CDCl3) δ 197.38, 161.58, 156.92, 147.57, 138.65, 136.12, 130.85, 130.11, 127.76, 126.39, 124.20, 120.40, 120.01, 112.03, 26.90. APCI-HRMS m/z calculated for C15H13rClNO2 [M+ H]+: 351.9734, found: 351.9714. Purity (HPLC): 99.7%.

1-(3-{(E)-[(5-Chloro-2-hydroxyphenyl)methylidene]amino}phenyl)ethan-1-one (45)

Prepared as for 42 from 3′-aminoacetophenone (0.50 g, 3.70 mmol) and 5-chlorosalicylaldehyde (0.58 g, 3.70 mmol) to yield the title compound 45 as light orange crystals (0.18 g, 18%): mp: 118.02 °C (EtOH); 1H NMR (600 MHz, CDCl3) δ 12.97 (s, 1H), 8.61 (s, 1H), 7.92–7.82 (m, 2H), 7.54 (t, J = 7.7 Hz, 1H), 7.48 (d, J = 7.8 Hz, 1H), 7.39 (s, 1H), 7.34 (d, J = 8.8 Hz, 1H), 6.99 (d, J = 8.8 Hz, 1H), 2.65 (s, 3H); 13C NMR (151 MHz, CDCl3) δ 197.56, 162.59, 159.84, 148.66, 138.57, 133.51, 131.60, 129.94, 127.23, 126.22, 124.01, 120.58, 119.86, 119.08, 26.91. APCI-HRMS m/z calculated for C15H13ClNO2 [M+ H]+: 274.0629, found: 274.0629. Purity (HPLC): 100%.

1-{3-[(E)-{[4-(Diethylamino)-2-hydroxyphenyl]methylidene}amino]phenyl}ethan-1-one (46)

Prepared as for 42 from 3′-aminoacetophenone (0.50 g, 3.70 mmol) and 4-(diethylamino)salicylaldehyde (0.71 g, 3.70 mmol) to yield the title compound 46 as gold crystals (0.46 g, 42%): mp: 117.85 °C (EtOH); 1H NMR (600 MHz, CDCl3) δ 13.52 (s, 1H), 8.47 (s, 1H), 7.84–7.73 (m, 2H), 7.50–7.38 (m, 2H), 7.18 (d, J = 8.7 Hz, 1H), 6.26 (d, J = 8.8 Hz, 1H), 6.20 (s, 1H), 3.41 (q, J = 7.0 Hz, 4H), 2.63 (s, 3H), 1.21 (t, J = 7.0 Hz, 6H); 13C NMR (151 MHz, CDCl3) δ 197.89, 163.87, 161.38, 152.00, 149.49, 138.15, 133.95, 129.39, 125.85, 125.17, 120.03, 108.90, 103.90, 97.61, 44.55, 26.72, 12.63. APCI-HRMS m/z calculated for C19H23N2O2 [M + H] + : 311.1754, found: 311.1746. Purity (HPLC): 100%.

1-(3-{(E)-[(2H-1,3-benzodioxol-5-yl)methylidene]amino}phenyl)ethan-1-one (47)

Prepared as for 42 from 3′-aminoacetophenone (0.50 g, 3.70 mmol) and 3,4-(methylenedioxy)benzaldehyde (0.56 g, 3.70 mmol) to yield the title compound 47 as light green crystals (0.98 g, 99%): mp: 127.64 °C (EtOH); 1H NMR (600 MHz, CDCl3) δ 7.81 (d, J = 15.6 Hz, 1H), 7.49–7.32 (m, 4H), 7.21 (d, J = 7.9 Hz, 1H), 6.99 (d, J = 8.0 Hz, 1H), 6.94 (d, J = 7.7 Hz, 1H), 6.13 (s, 2H), 3.93 (s, 2H), 3.82 (q, J = 6.9 Hz, 1H); 13C NMR (151 MHz, CDCl3) δ 190.69, 149.97, 148.53, 146.93, 144.50, 139.67, 129.59, 129.54, 125.28, 120.54, 119.41, 118.91, 114.52, 108.79, 106.79, 101.75, 58.61, 18.58. APCI-HRMS m/z calculated for C16H14NO3 [M+ H]+: 268.0968, found: 268.0951. Purity (HPLC): 90.4%.

All reagents were commercially available and purchased from various manufacturers. Radioligands [3H]DPCPX (specific activity 120 Ci/mmol) and [3H]NECA (specific activity 21.1 Ci/mmol) were obtained from PerkinElmer. Radioactivity was counted by a Packard Tri-CARB 2810 liquid scintillation counter.

Ethics

The collection of tissue samples for the A1 and A2A AR radioligand binding assays were approved by the Health Sciences Ethics Office for Research, Training and Support, North-West University (application number NWU-00260-17-A5) and were performed in accordance with the guidelines of the South African National Standard (SANS) document (The care and use of animals for scientific purposes).

Tissue samples

Male Sprague–Dawley rats were dissected to attain rat whole brain membranes (excluding cerebellum and brain stem) and rat striatal membranes for the A1 and A2A AR radioligand binding assays, respectively. Tissue samples were prepared and stored as described in literature (Van der Walt and Terre’Blanche 2015). Protein concentrations were determined according to the Bradford assay, using bovine serum albumin as reference standard (Bradford 1976).

Adenosine A1 and A2A receptor radioligand binding assays

The degree of binding affinity the test compounds possess towards rat A1 and A2A ARs were determined via radioligand binding assays, as described previously (Bradford 1976; Bruns 1987; Bruns and Watson 2012; Van der Walt and Terre’Blanche 2015). The A1 AR radioligand binding assay used rat whole brain membranes (expressing A1 ARs) and 0.1 nM 1,3-[3H]-dipropyl-8-cyclopentylxanthine ([3H]DPCPX) as radioligand (Bruns 1987) and, in turn, the A2A AR radioligand binding assay used rat striatal membranes (expressing A2A ARs) and 4 nM 5′-N-ethylcarboxamido[3H]adenosine ([3H]NECA) as radioligand (Bruns et al. 1986). N-Cyclopentyladenosine (CPA) was also added to the A2A AR radioligand binding assay (in order to eliminate the A1 component of binding exhibited by the non-selective [3H]NECA) as well as MgCl2 (in order to increase radioligand binding and decrease non-specific binding) (Bruns et al. 1986). The final volume of all incubations contained 1 mL 50 mM Tris.HCl buffer and 1% DMSO (Van der Walt and Terre’Blanche 2015). Non-specific binding of [3H]DPCPX and [3H]NECA for the radioligand binding assays were defined as binding in the presence of 100 µM CPA (Bruns et al. 1986; Van der Walt and Terre’Blanche 2015). Specific binding was defined as the total binding minus the non-specific binding (Van der Walt and Terre’Blanche 2015).

Guanosine triphosphate (GTP) shift assays

The type of binding affinity test compounds exhibited at the rat A1 AR was determined via a guanosine triphosphate (GTP) shift assay, as described previously (Lohse et al. 1987; Van der Walt and Terre’Blanche 2015; Van der Werten et al. 1995). A GTP shift assay resembles the A1 AR radioligand binding assay. However, it requires the addition of 100 µM GTP. Non-specific binding was defined as binding in the presence of 10 µM DPCPX (Van der Werten et al. 1995; Lohse et al. 1984).

Statistical data analyses

In short, all statistical data analyses were done using Microsoft Excel and GraphPad Prism Software. Sigmoidal dose response curves, from which IC50 values were calculated, were obtained by plotting the specific binding against the logarithm of the test compounds’ concentrations. Subsequently, the IC50 values were used to calculate the Ki values for the competitive inhibition of [3H]DPCPX (Kd = 0.36 nM) against rat whole brain membranes and [3H]NECA (Kd = 15.3 nM) against rat striatal membranes by the test compounds by means of the Cheng–Prusoff equation. All calculated Ki values were determined in triplicate and given as mean ± standard error of the mean (SEM). Specific binding (%) of the radioligand at a maximum tested concentration of 100 µM were determined in duplicate and expressed as the mean in %. SI index values were calculated as a ratio of the A1 and A2A AR Ki values of test compounds. GTP shifts were calculated by dividing the Ki values of compounds reported in the presence of GTP by the Ki values obtained in the absence of GTP.

SwissADME (https://www.swissadme.ch), a free web tool, was used to evaluate key parameters of small molecules; such as pharmacokinetics, drug-likeness and medicinal chemistry friendliness. Pharmacokinetic parameters (absorption, distribution, metabolism and excretion), among others, were predicted from molecular structures using the most relevant computational methods, and thus, do not focus on just one specific property or model (Daina et al. 2017).

Conclusions

A total of 33 chalcones (15–36 and 37–41) and structurally related compounds (42–47) with different substitutions on ring A and/or benzylidene ring B were synthesised, characterized and evaluated in vitro to determine the degree and type of A1 and A2A receptor binding. While most compounds were not novel, the application of all compounds is original. The chalcone derivatives 24, 29, 37 and 38 possessed selective A1 affinity below 10 µM; compound 38 was the most potent selective A1 AR antagonist (Ki (r) = 1.6 µM). Most of the test compounds showed poor A1 and/or A2A AR affinity upon in vitro evaluation—making it difficult to determine SARs; however, some broad conclusions were drawn from these results. The type of substitution on ring A of the chalcone scaffold played a key role in affinity, and the position of the substitution on benzylidene ring B; the NH2-group at position C3 of ring A of the chalcone scaffold, and the Br-atom at position C3′ on benzylidene ring B were essential to selective A1 AR affinity (exemplified by compound 27 versus 38 and 37 versus 38). Selective A1 AR affinity and antagonistic activity may be added to the known biological activities of compounds 24 and 29. The physiochemical and pharmacokinetic properties (based on in silico evaluation) proved the chalcone derivatives 24, 29, 37 and 38, although drug-like, not at all lead-like (29). Thus far, the chalcone chemical structure was a promising scaffold in medicinal chemistry; due to the diverse chemistry and biology of this α,ß-unsaturated carbonyl system. The said system may be troublesome (it is perceived as a potential Michael acceptor); however, it easily allows structural modification and introduction of a heterocyclic ring system (such as isoxazole, pyrazole, pyrrole etcetera) which may just be beneficial to A1 and/or A2A AR binding affinity; thus, leading to the discovery of new pharmacophores from this classes of compounds.

Below is the link to the electronic supplementary material.

Supplementary file1 (DOCX 9118 KB)