and JE2 (EC50 = ~6 µM), with ~20-fold increase in potency in the
presence of CFX, suggesting on-target activity.
EC50 (μM)
Compound Compound +
Collectively, these data demonstrated that CFX derivatives
were more active than corresponding PA derivatives as single
agents or when potentiated by CFX (with the noted exception of
9), consistent with the hypothesis of improved intercalation
increasing potency. Analysis of available literature SAR data for
the PA series[25] with Topliss-guided analysis indicated improved
activity correlated with increased electron-withdrawing ability of
the substituent on the thiourea-linked aryl moiety. Potentiation by
CFX demonstrated favorability of para- (4 and 10) over meta- (1
and 7) trifluoromethyl substituent in both the PA (~5-fold
increase) and CFX (~25-fold increase) series (Table 1).
Compound 10 comprising a CFX core and a para-substituted
trifluoromethyl arylthiourea moiety was the most potent
analogue, and exhibited a ~650 increase in potency against JE2 in
the presence of CFX (EC50 = 5.9 ± 0.6 nM). This represents
~160-fold increase in potency compared to 1
Compound Linker
only
CFX (9.4 μM)
16
17
18
3-methylpiperazine3.9 ± 0.4
3-aminopyrrolidine53 ± 13
ethylenediamine 9.7 ± 1.0
0.018 ± 0.003
3.2 ± 0.3
2.5 ± 0.3
Table 2. Potentiation of Ciprofloxacin by derivatives of 10 in
MRSA (USA300 JE2). Compounds were tested as single agents
and for Ciprofloxacin (CFX) potentiation in MRSA. Data
represent mean ± SEM (n = 3 biological replicates).
The use of SAR optimization strategies based on the
development of quinolone antibiotics, and loss of single agent
potency against JE2 compared to SH1000, prompted
consideration that compounds may be acting through quinolone
targets, rather than AddAB. The presented panel of inhibitors
contain PA or CFX cores, both known inhibitors of DNA gyrase
and topoisomerase IV. DNA gyrase produces supercoiled DNA
from relaxed and open-coiled DNA, which can be separated via
agarose gel electrophoresis. Inhibition of pBR322 supercoiling
by DNA gyrase was assessed for hit compound 1, best
performing compound 10, alongside the potentially on-target
derivative 5, inactive control 15 and positive control CFX.
(Figure 1A). CFX inhibited gyrase (IC50 = 0.28 ± 0.03 μM),
whereas 1, 5, 10, and 15 showed minimal activity against gyrase
inhibited by 1, 10, or 15 at 12.5 µM (Figure S1). This
(EC50 = 1.1 ± 0.1 μM) for potentiation against JE2 by CFX.
The increased activity of electron-withdrawing arylthiourea
substituents in the CFX series led to the synthesis of nitro
derivatives at the meta- (13) and para- (14) positions, following
the Topliss decision tree and empirical antibacterial activities
linked to the nitroaromatic group.[30,37] Para-substitution resulted
in a ~10-fold increase in potency compared to meta. However,
metabolic toxicities, particularly hepatotoxicity, are associated
with nitroaromatic groups, as a result of their abilities to interfere
with cellular redox via nitro reduction,[38] and the lack of
improved bioactivity for 13 and 14 compared to 10 (Table 1)
implied no justification for their continued investigation.
concentration is higher than the cellular EC50 for growth
inhibition for all active compounds and provides initial evidence
that DNA gyrase and topoisomerase IV are not direct targets of
this series of compounds. Indeed, N-acylation is a known
mechanism for loss of antibiotic activity in fluoroquinolones.[40]
To support the proposal that compound potentiation by CFX
occurred by inhibition of AddAB-mediated DNA repair, PA urea
derivative 15 was prepared as this compound does not inhibit the
activity of purified AddAB/RecBCD.[25] Consistent with the
postulated mechanism, the potency of 15 was unaffected by the
presence of CFX (Table 1), suggesting potentiation by CFX may
be mediated by AddAB inhibition. The analogue panel was
further tested for potentiation of alternative DNA-damaging
agents Mitomycin-C (MMC) and hydrogen peroxide (Table S2),
which function via alkylation and oxidization DNA bases,
respectively. No potentiation of MMC or hydrogen peroxide was
observed, consistent with repair of these forms of DNA damage
typically occurring through base excision or mismatch repair.
To further probe if compounds functioned through a different
mechanism to CFX, synergy between compounds and CFX was
analyzed. Drug synergy is the amplification of the effect of two
drugs, such that their effect when co-administered is greater than
the sum of their individual effects. Combination indexes (CIs)
determined using the Chou-Talalay method quantitatively
demonstrate synergism (CI < 1), additive effects (CI = 1), and
antagonism (CI > 1). Constant combination ratio titrations were
prepared as two-fold serial dilutions of CFX with test compounds
in the ratio of their EC50 values for JE2 growth inhibition, such
that both drugs exert approximately equal effects in the
As a result of the successful scaffold hop to the CFX core,
further SAR optimization strategies that improve quinolone
substituent antibiotic activity were investigated. The quinolone
C-7 modulates factors such as antibacterial spectrum,
bioavailability and side-effects.[39] Derivatives of the most potent
analogue 10, were prepared replacing the central piperazine
moiety with 3-methylpiperazine (16), 3-aminopyrrolidine (17), or
ethylenediamine (18). 16-18 were tested in cellular growth
inhibition assay either as single agents against E. coli or S.
aureus SH1000 (Table S3), or in combination with CFX against
JE2 (Table 2). None of these derivatives were more effectively
potentiated by CFX than 10, indicating deviation from the
piperazine ring of 10 is not favored (Table 2). The substantial
decrease in bioactivity of ethylenediamine-derivative 18, is likely
due to increased entropic penalty on binding as a result of
increased flexibility.
combination.[41] A more precise method to determine EC50 values
for individual compounds and combinations was used to increase
accuracy for CI calculations. The OD600 was recorded over a 17 h
period at 37 °C and the rate of the exponential growth phase used
for dose-response analysis, rather than endpoint OD600
measurement, which resulted in decreased error in EC50 values
(Figure S2). In combination, 10 exhibited synergy with CFX
(CI = 0.70) and a dose-reduction index (DRI) of 2.8 (Figure 1C-
E), agreeing with the observed compound potentiation by CFX in
the JE2 cellular assays (Table 1). By contrast, 1 (CI = 1.0) and 5
(CI = 0.96) demonstrated an apparent additive relationship or
weak synergy with CFX, respectively, consistent with their
degree of potentiation (Table 1). These findings indicated that 10,
which we have given the synonym IMP-1700, is a highly active
potentiator of bacterial DNA damage by CFX.
2.3 Demonstration of synergistic effects with Ciprofloxacin
2.4 Investigation of the mechanism of action of IMP-1700
Having demonstrated a synergistic effect of IMP-1700 (10)
with CFX for killing of MRSA, further evidence was sought to