which was then subject to a second SNAr with amino azetidine to give ethyl 1-(6-amino-3,5-difluoropyridin-2-yl)-7-(3-aminoazetidin-1-
yl)-8-chloro-6-fluoro-4-oxo-1,4-dihydroquinoline-3-carboxylate 21. This was subsequently hydrolysed to the desired fluoroquinolone 3
with NaOH.
Compounds (1-3) were evaluated for antimicrobial activity against various wild type (wt) gram positive (S. aureus ATCC 25923, and
E. faecalis ATCC 29212), wt gram negative bacteria (E. coli ATCC 25922, K. pneumoniae ATCC 700603, P. aeruginosa ATCC 27853,
A.baumannii ATCC 19606, N.gonorrhoea ATCC 49226) and N.gonorrhoea NCTC13480 (WHO L)) as outlined in table 1. Values were
determined as the lowest concentration that no growth could be visually observed, and all
MICs performed with a biological repeat to Clinical Laboratories Standards Institute (M7-A9). Ciprofloxacin and delafloxacin were
used as standard drugs in the activity test. Microdilution test results are presented in Table 1. Compounds 1 and 2 have similar
microbiological profiles within experimental error across the gram negative strains compared to delafloxacin, with the exception of E.
coli ATCC 25922, with a 10 fold reduction in MIC value. The
reduction in activity was also observed in the gram-positive bacteria
S. aureus ATCC 25923 and E. faecalis ATCC 29212, both of which
are ~10 fold less susceptible to the novel compounds compared to
delafloxacin.
MIC (µg/mL)
Table 1. Microbiological data of compounds 1, 2, 3 and
Delafloxacin Ciprofloxacin
delafloxacin against an elected panel of bacteria
Straina
1
2
3
N.g
0.01
0.12
64
0.06 0.007
0.06
0.03
8
0.06
4
Compound 3 however was either equipotent or more active
compared to delafloxacin versus all gram-negative strains (within
experimental error). The improved microbiological data of 3 can be
ascribed to the differences in logD7.4. Approved
N.g
(R)
0.12
16
4
0.06
2
A.b
64
2
GN
P.a
2
0.12
1
antibiotics that confer gram-negative activity have
Table 2. Selected physicochemical and biological in vitro assays
a statistical cLogD7.4 of ≤ 020. Amongst other
K.p
8
4
0.25
0.25
0.01
0.5
reasons, this is due to the structural differences in
E.c
0.12
0.12 0.001
0.03 0.003
0.003
the outer membrane/cell wall of GP/GN bacteria. The latter of which
presents a barrier to penetration that may be overcome by
physicochemical attenuation. The measured LogD7.4 of delafloxacin
and compound 3 are 0.7 and -0.3 respectively (Table 2). The
clinically relevant FQ resistant strain of N.gonorrhoea (WHO L) has
multiple point mutations at both DNA gyrase and topoisomerase IV,
an increased expression of MtrCDE efflux pump and reduced
permeability to antibiotics (porB1b mutation)21. Ciprofloxacin is
significantly affected by these mutations with > 66-fold reduction in
activity whereas all other tested analogues were either unaffected or
displayed no more than a 12-fold reduction. This implies that it is the
structural difference of the fluoroquinolone core rather than the
S.a
0.03
0.5
0.003
0.03
0.5
GP
E.f b
0.25
0.12
0.25
aStrain definitions: N.g: N.gonorrhoea ATCC 49226, N.g (R):
N.gonorrhoea (WHO L), A.b: A.baumannii ATCC 19606, P.a:
P.aeruginosa PA01, K.p: K.pneumoniae ATCC 700603, E.c: E.coli ATCC
25922, S.a: S.aureus ATCC29213, E.f: E.faecalis ATCC 29212. b:
Performed at 5DHPG ltd
molecule’s physchem properties that has an impact on activity vs this resistance strain.
Having measured the pharmacological activity, it was then necessary to understand the ADME and physicochemical profiles of the
synthesised molecules. All compounds were assessed for their ability of metabolic turnover. Whilst all compounds displayed equivalent
levels of in vitro human liver microsomal metabolism, including delafloxacin, there was a modest difference in human hepatocytes.
Compounds 1, 2, 3 and delafloxacin had Clint values of 6.8, 9.2, 3 and 3 μL/min/106 cells respectively. This would suggest that although
the structures are not susceptible to phase 1 metabolism, the carbon linked analogues are undergoing phase 2 metabolism at a faster rate
and/or are more intrinsically permeable. Met ID studies could help substantiate this observation.
Cyclobutanol isomers 1 and 2 display different CYP inhibition profiles. Compound 1 inhibited CYP2C9 at 41.7 µM and compound 2
inhibited CYP3A4 at 24.8 µM. As a comparison, delafloxacin did not inhibit any of the CYP isoforms tested up to 50 µM. Compound 3
inhibits CYP2C9 and CYP3A4 at 12.7 and 13.6 μM respectively.
Human plasma protein binding data revealed that all anionic compounds had a similar PPB profile (10.5-14.9 % free) whilst compound
3 was 29.7 % free. This is in keeping with literature precedence of related compounds11. Although compound 3 has a higher free drug
fraction (fu), it does not mean it will improve the in vivo efficacy, as the rate of in vivo clearance typically increases proportionally22.
CYP450 inhibitionc
HPPBd
%Fue
KSOLf
C-7
substituent
Compound ID
Charge
HLMa HHepsb
clogP logD7.4
pKa
IC50 (μM)
μM
1
2
C
C
Anion
Anion
<5
<5
6.8
9.2
>50, >50, 41.7, >50,>50
>50, >50, >50, >50, 24.8
10.5
12.4
59.3
44.8
2.65
2.63
0.3
0.3
5.04
4.96
5.33,
7.72
3
N
Zwitterion
<5
3
>50, >50, 12.7, >50, 13.6
29.7
3.6
0.31
-0.3