Page 3 of 7
ACS Medicinal Chemistry Letters
The aims of our study were to validate compound 1 through
two fluorine atoms instead of two hydrogens linked to the
dioxolane ring, was less active than the starting compound 1.
The anti-T. brucei activity decreased replacing the dioxolane
ring of 1 with a dioxane (compound 20), while it was
maintained in compound 21, bearing a tetrahydrofurane.
Compound 21 presented an EC50 towards T. brucei equal to
3.1 µM, but SI = 8. Overall, six compounds (8, 11-15) showed
a low micromolar EC50 and SI >20. Compound 13, the 3-
pivaloyl derivative of 1, was the most selective among the
novel synthesized molecules.
1
2
3
4
5
6
7
8
Structure Activity Relationship (SAR) studies, discover
follow-up hits and characterize their biological profile for
potential liabilities identifications. The synthetic procedure
followed for the synthesis of the compounds (1-21) is shown
in Scheme 1 and the chemical structures are depicted in Table
1-3. The chalcones (22-34) were synthesized by Claisen-
Schmidt condensation using substituted acetophenones and
benzaldehydes in presence of NaOH as base. The reaction was
carried out in ethanol as previously reported.15 The chalcones
were converted into the corresponding flavonol-like
compounds (1-10, 19-21) using the Flynn-Algar-Oyamada
method for epoxidation and subsequent intramolecular
cyclization of the open-chain structure (Scheme 1A). For the
synthesis of esters (11-15) and carbamate 16, compound 1 was
treated with an excess of acyl chloride in dry DCM and in
presence of triethylamine. The reaction was carried out at
room temperature overnight. For the synthesis of ethers 17 and
18, alkyl halide was added to a solution of compound 1 in dry
DMF and in presence of K2CO3. The reaction was carried out
under microwave irradiation (Scheme 1B).
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The synthesized library was assessed at 10 μM in a panel of
early in vitro ADME-tox assays including cytotoxicity (A549
cell line), mitochondrial toxicity, cytochrome P450 (CYP1A2,
CYP2C9, CYP2C19, CYP2D6 and CYP3A4 isoforms) and
hERG inhibition. The data are reported in Figure 2 using a
traffic light system. Compound 1 and all of its derivatives
exhibited no liability towards hERG and mitochondrial
toxicity. Some compounds were shown to be cytostatic, with
two compounds (9 and 12) being cytotoxic (<0% A549 cell
growth). Most of the compounds displayed varying degrees of
CYP450 liability. The IC50 towards hERG and CYP isoforms
were measured for compound 1. The hERG IC50 (>100 μM)
was over 250-fold higher than the EC50 towards the parasite,
thus in accord with the Target Product Profile (TPP) for hit
prioritization. Compound 1 IC50 towards CYP1A2 and
CYP2D6 were 0.4 and 0.05 μM, respectively, whereas for
CYP2C9, CYP2C19 and CYP3A4 the IC50 were equal to 1.6,
1.5 and 6.0 μM, respectively. Compound 1 was the most
optimal for its anti-trypanosomatid activity and ADME-tox
profile and progressed to in vivo pharmacokinetic studies.
The novel library of flavonol-like compounds (2-21) was
evaluated towards T. brucei bloodstream form. The series was
assessed for cytotoxicity on THP1 macrophage-like cells to
estimate the CC50. For compounds showing a percentage of
parasite growth inhibition higher than 70%, the dose-response
curve (DRC) was performed. The percentages of parasite
growth inhibition at 10 µM are reported in Table S1 of the
Supporting Information.
We started the SAR investigation of this scaffold by
modifying the substituents on ring A (Table 1). Nine
compounds (2-10) were synthesized introducing different
substituents in position 6 and 7 of ring A. Five compounds (2,
4, 8-10) showed a significant activity towards T. brucei with
EC50 lower than 5 µM. When the OCH3 in position 7 of
compound 1 was replaced with a methyl group and a chlorine
or fluorine (8, 9 and 10, respectively), the compounds
maintained a meaningful anti-T. brucei activity. Moving the
methoxy group from position 7 to 6 (compound 3), we
observed a huge drop of the antiparasitic activity. Compound
2, bearing unsubstituted ring A, and compound 4, with a
methyl group in position 6 showed activity towards T. brucei,
while compounds bearing halogen in position 6 (5-bromide; 6-
chlorine; 7-fluorine) did not significantly inhibit T. brucei
cells growth. Compound 8 (EC50 = 0.4 µM) displayed a
potency comparable to that of the starting hit 1, however it
presented a reduced selectivity index (SI = 31).
Following this, our SAR was focused on modifications of the
hydroxyl group in position 3 of the chromen-4-one scaffold
(Table 2). The presence of an ester instead of a hydroxyl group
in position 3 (11-15) led to significant activity on T. brucei
(EC50 <1.1 µM) together with a SI >20. Among the esters, the
3-pivaloyl derivative of compound 1 (13) showed the most
interesting profile with an EC50 towards T. brucei of 1.1 µM
and SI >92. On the contrary, the presence of a carbamate (16)
or an ether (17 and 18) led to inactivity towards T. brucei.
These data suggested that the hydroxyl group in position 3
should be free in order to have a meaningful anti-T. brucei
activity. The activity of esters can be related to an easier
hydrolysis with respect to ethers and carbamates. We enlarged
the SAR study modifying the 1,3-benzodioxole ring of
compound 1 (compounds 19-21, Table 3). Compound 19, with
In vivo bioavailability and half-life were evaluated in BALB/c
mice treated IV with 1 mg/kg and orally 20 mg/kg. Compound
1 displayed a half-life of 19 hrs after iv administration and of
45 hrs after oral (os) administration (Table 4). Both AUC and
Cmax values were similar despite the much higher dose
administered per os. Tmax for IV administration was reached
after 1h this suggesting the possible intravascular aggregation
of compound 1 given its low solubility.
The aggregation behavior of compound 1 in aqueous solution
was investigated spectroscopically and the albumin
sequestration assay performed. As compound 1 concentration
is increased, both the absorption and the emission spectra
show an increase of bands due to aggregates relative to the
monomer bands (Figure 3). The absorption data were well
fitted in terms of a monomer/dimer equilibrium, with a 1.8
(±0.3) x 105 M-1 equilibrium constant at 20 °C (see the
Supporting Information). The fact that the aggregate
absorption band is found at shorter wavelengths and its
emission band at longer wavelengths than the corresponding
bands of the monomeric form indicates the aggregates to be of
H-type (as opposed to a J-type), i.e., with the monomers
stacked on top of each other with a small slip angle.24-25
Subsequent additions of human serum albumin (HSA) caused
a progressive recovery of the monomer absorption band and
the replacement of both aggregate and free monomer emission
bands by a single new band that we assign to a compound
1/HSA complex. Therefore, the latter represents a stable state
with respect to the monomeric and dimeric states. Emission
data analysis provided in the Supporting Information allowed
us to estimate the 1/HSA binding equilibrium constant, 2.5
ACS Paragon Plus Environment