S.K. Shrestha et al. / European Journal of Medicinal Chemistry 133 (2017) 309e318
313
excellent antifungal activity (<0.03e0.06
m
g/mL) against C. krusei
determined that the compounds displaying the best overall anti-
fungal activity in this series were 28 and 29, which contained C5
and C6 alkyl chains, respectively. For this reason, we next generated
molecules 30e41 with C5 and C6 alkyl chains and mono-substituted
(with fluoro or chloro) phenyl rings.
ATCC 6258 (strain I) and C. parapsilosis ATCC 22019 (strain J). When
looking at compounds 5e7 with alkyl chains of medium lengths,
we observed that they exerted moderate to excellent antifungal
activity against all the strains tested with the exception of A. flavus
ATCC MYA-3631 (strain K) against which they displayed poor ac-
tivity. We noticed that compound 7 exerted the best antifungal
activity followed by compound 6 and then by compound 5. In
general, compound 7 showed excellent antifungal activity against
10 (strains A, C-J, and L), moderate antifungal activity against two
(strains B and M), and poor activity against A. flavus ATCC MYA-
3631 (strain K) out of the 13 fungal strains tested. Similarly, com-
pound 6 showed excellent antifungal activity against 9 (strains A, C-
G, I-J, and L), and moderate antifungal activity against 4 (strains B,
E, H, and M) out of the 13 fungal strains tested. Finally, compound 5
exhibited excellent antifungal activity against 5 (strains A, G, I, J,
and L), moderate antifungal activity against 4 C. albicans strains (B-
eF) and C. glabrata ATCC 2001 (strain H), and poor antifungal ac-
tivity against two Aspergillus strains (K and M) of the 13 strains
tested. From these observations, we determined that increasing the
length of the alkyl chain substituent results in an overall increase in
antifungal activity. This prompted us to explore compounds 8 and 9
with longer C10 and C12 alkyl moieties. We observed that com-
pounds 8 and 9 displayed poor antifungal activity against 5 (strains
B, D, F, G, and K) and 6 (strains A, B, D, G, I, and K) out of 13 fungal
strains, respectively. Additionally, we observed moderate inhibition
of 5 (strains A, C, E, H, and M) and 4 (strains C, E, F, and M) of the 13
strains tested by compounds 8 and 9, respectively. On the basis of
our in-depth analysis of the MIC data for compounds 1e9, we
concluded that the optimal chain length required for the FLC de-
rivatives to confer maximal antifungal activity were C5-C8.
(ii) Are these optimal C5-C8 chain lengths for our FLC derivatives
the same as that of other families of n-alkylated molecules (e.g.,
aminoglycoside, benzimidazole, and ebsulfur derivatives)? In
recent years, the addition of linear alkyl chains to drug scaffolds has
gained popularity as it has been demonstrated that it improves the
activity of the compounds when compared to the parent non-
alkylated molecules. Aminoglycosides are of the families of anti-
biotics to which alkyl chains have been added. It was shown that
the optimal chain lengths required for maximal antifungal activity
were C8 for kanamycin A [17] and C12-C14 for kanamycin B [4] and
tobramycin [5]. For n-alkylated ebsulfur derivatives, the optimal
chain lengths required to confer antifungal activity were reported
to be C5-C6 [6]. Interestingly, alkyl chains varying in length between
1 and 3 carbons (C1-C3) were found to confer optimal antifungal
activity when attached to a benzimidazole core [16]. From these
data, it is clear that there is no universal alkyl chain length that can
be utilized to confer prime antifungal activity. Consequently, one
should always test a range of alkyl chains to determine the optimal
length(s) for any new drug scaffold to be derivatized.
(iv) For a mono-substituted phenyl ring, which halogen sub-
stituent (fluoro or chloro) is best? After establishing that two chloro
groups were better than two fluoro moieties, we wanted to deter-
mine if the trend would remain when only one halogen substituent
was attached to the phenyl ring. In order to shed light on this, we
synthesized 12 additional compounds, 30e41, and performed
direct pairwise comparisons (30 versus 36, 31 versus 37, 32 versus
38, 33 versus 39, 34 versus 40, 35 versus 41) in terms of their activity
against the fungal strains tested based on our MIC data (Table 1).
Interestingly, in contrary to what we observed with the dihalo-
genated molecules, we found that, in most cases, the mono-
substituted phenyl rings with a fluoro substituent displayed bet-
ter antifungal activity than its counterpart with a chloro substitu-
ent. Briefly, we found that compound 30 (with a 2-F) repeatedly
showed better antifungal activity against C. albicans strains, non-
albicans Candida strains, and Aspergillus strains when compared to
its counterpart 36 (with a 2-Cl). Compounds 31 (with a 2-F) and 32
(with a 3-F) displayed better antifungal activity against C. albicans
as well as non-albicans Candida strains, whereas their respective
counterpart 37 and 38 showed better activity only against Asper-
gillus strains. Similarly, compound 33 (with a 3-F) exhibited better
antifungal activity against C. albicans strains than compound 39
(with a 3-Cl). Both compounds 33 and 39 showed almost identical
activity against Aspergillus strains. Finally, compound 34 (with a 4-
F) showed better antifungal activity against C. albicans strains,
worse antifungal activity against non-albicans Candida strains, and
equivalent antifungal activity against Aspergillus strains than its
counterpart compound 40 (with a 4-Cl). Likewise, compound 35
(with a 4-F) showed worse antifungal activity against C. albicans as
well as non-albicans Candida strains, but identical antifungal ac-
tivity against Aspergillus strains than its counterpart compound 41
(with a 4-Cl).
(v) For a specific alkyl chain length, which level of substitution
(mono- versus di-) confer the best antifungal activity? As previously
mentioned, based on our MIC data we concluded that compounds
with C5-C6 chain lengths confer the best antifungal activity. We
next decided to explore the effect of the level of substitution
(mono- versus di-) on the phenyl ring for our compounds with C5-
C6 alkyl chain length based by comparing the MIC data obtained for
compound 5 versus compounds 30, 32, and 34; compound 6 versus
compounds 31, 33, and 35; compound 28 versus compounds 36, 38,
and 40; as well as compound 29 versus compounds 37, 39, and 41.
We noticed that the 2,4-difluorinated compounds 5 and 6 displayed
inferior antifungal activity when compared to their mono-2-
substituted counterparts 30 and 31. On the other hand, both com-
pounds 5 and 6 showed better antifungal activity than their mono-
3-substituted counterparts 32 and 33, and than their mono-4-
substituted counterparts 34 and 35. Interestingly, we observed
that the 2,4-dichlorinated compounds 28 and 29 always displayed
better antifungal activity than any of their mono-substituted
counterparts.
(iii) For a given alkyl chain length, would a 2,4-dichlorinated
phenyl ring confer better or worse antifungal activity than the
2,4-difluorinated phenyl ring? Although compound 7 was identi-
fied as the best antifungal in our series based on its MIC values, we
were concerned that the longer alkyl chain (C8) could cause po-
tential toxicity to mammalian cells. Therefore, for the next step in
our SAR study, we decided to keep the length of the alkyl chains
present in compounds 24e29 between 1 and 6 carbons. To explore
the importance of the two fluoro groups on the phenyl ring of
compounds 1e6, we generated their dichlorinated counterparts
24e29. By performing a pairwise comparison of for the MIC values
of compounds 1 versus 24, 2 versus 25, 3 versus 26, 4 versus 27, 5
versus 28, and 6 versus 29, we concluded that the compounds with
the 2,4-dichlorinated phenyl ring always displayed better anti-
fungal activity than those containing the two fluoro groups. We also
(vi) For a given substituent (fluoro of chloro), what is the optimal
position (ortho, meta, or para) for mono-substitution on the phenyl
ring? Finally based on a direct comparison of compounds 30 versus
32 versus 34; compounds 31 versus 33 versus 35; compounds 36
versus 38 versus 40; compounds 37 versus 39 versus 41, we found
that substitution at the ortho position (compounds 30, 31, 36, and
37) confers greater antifungal activity than does that at the para
position (compounds 34, 35, 40, and 41). Substitution at the meta
position (compounds 32, 33, 38, and 39) is the least optimal to