Ciprofloxacin-nitroxide hybrids with potential for biofilm control

Article abstract of DOI:10.1016/j.ejmech.2017.06.058

As bacterial biofilms display extreme tolerance to conventional antibiotic treatments, it has become imperative to develop new antibacterial strategies with alternative mechanisms of action. Herein, we report the synthesis of a series of ciprofloxacin-nitroxide conjugates and their corresponding methoxyamine derivatives in high yield. This was achieved by linking various nitroxides or methoxyamines to the secondary amine of the piperazine ring of ciprofloxacin using amide bond coupling. Biological evaluation of the prepared compounds on preformed P. aeruginosa biofilms in flow cells revealed substantial dispersal with ciprofloxacin-nitroxide hybrid 25, and virtually complete killing and removal (94%) of established biofilms in the presence of ciprofloxacin-nitroxide hybrid 27. Compounds 25–28 were shown to be non-toxic in both human embryonic kidney 293 (HEK 293) cells and human muscle rhabdomyosarcoma (RD) cells at concentrations up to 40 μM. Significantly, these hybrids demonstrate the potential of antimicrobial-nitroxide agents to overcome the resistance of biofilms to antimicrobials via stimulation of biofilm dispersal or through direct cell killing.

Full text of DOI:10.1016/j.ejmech.2017.06.058

Accepted Manuscript
Ciprofloxacin-nitroxide hybrids with potential for biofilm control
Anthony D. Verderosa, César de la Fuente-Núñez, Sarah C. Mansour, Jicong Cao,
Timothy K. Lu, Robert E.W. Hancock, Kathryn E. Fairfull-Smith
PII:
S0223-5234(17)30512-3
10.1016/j.ejmech.2017.06.058
EJMECH 9553
DOI:
Reference:
To appear in: European Journal of Medicinal Chemistry
Received Date: 20 January 2017
Revised Date: 26 June 2017
Accepted Date: 27 June 2017
Please cite this article as: A.D. Verderosa, Cé. de la Fuente-Núñez, S.C. Mansour, J. Cao, T.K. Lu,
R.E.W. Hancock, K.E. Fairfull-Smith, Ciprofloxacin-nitroxide hybrids with potential for biofilm control,
European Journal of Medicinal Chemistry (2017), doi: 10.1016/j.ejmech.2017.06.058.
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.

ACCEPTED MANUSCRIPT
Graphical Abstract
1

ACCEPTED MANUSCRIPT
Ciprofloxacin-Nitroxide Hybrids with Potential for Biofilm Control
1
,¶
3-8,¶
2
3-8
Anthony D. Verderosa, César de la Fuente-Núñez,
Sarah C. Mansour, Jicong Cao,
3
-8
2
*,1
Timothy K. Lu, Robert E. W. Hancock and Kathryn E. Fairfull-Smith
1
ARC Centre of Excellence for Free Radical Chemistry and Biotechnology, Faculty of Science
and Engineering, Queensland University of Technology, Queensland 4001, Australia
2
Centre for Microbial Diseases and Immunity Research, Department of Microbiology and
Immunology, University of British Columbia, Vancouver, BC V6T 1Z4, Canada
3
Synthetic Biology Group, MIT Synthetic Biology Center, Massachusetts Institute of Technology,
Cambridge, Massachusetts, United States of America
4
Research Laboratory of Electronics, Massachusetts Institute of Technology, Cambridge,
Massachusetts, United States of America
5
Department of Biological Engineering, and Department of Electrical Engineering and
Computer Science, Massachusetts Institute of Technology, Cambridge, Massachusetts, United
States of America.
6
Broad Institute of MIT and Harvard, Cambridge, Massachusetts, United States of America.
7
Harvard Biophysics Program, Harvard University, Boston, Massachusetts, United States of
America.
8
The Center for Microbiome Informatics and Therapeutics, Cambridge, Massachusetts, United
States of America.
¶
These authors contributed equally to this work.
Corresponding Author
*
Email: k.fairfull-smith@qut.edu.au. Phone: +61 7 3138 4950. Fax: +61 7 3138 1804.
Author Contributions
All authors contributed to the writing of this manuscript and have given approval to the final
version of the manuscript.
1

ACCEPTED MANUSCRIPT
ABSTRACT
As bacterial biofilms display extreme tolerance to conventional antibiotic treatments, it has
become imperative to develop new antibacterial strategies with alternative mechanisms of action.
Herein, we report the synthesis of a series of ciprofloxacin-nitroxide conjugates and their
corresponding methoxyamine derivatives in high yield. This was achieved by linking various
nitroxides or methoxyamines to the secondary amine of the piperazine ring of ciprofloxacin
using amide bond coupling. Biological evaluation of the prepared compounds on preformed P.
aeruginosa biofilms in flow cells revealed substantial dispersal with ciprofloxacin-nitroxide
hybrid 25, and virtually complete killing and removal (94%) of established biofilms in the
presence of ciprofloxacin-nitroxide hybrid 27. Compounds 25-28 were shown to be non-toxic in
both human embryonic kidney 293 (HEK 293) cells and human muscle rhabdomyosarcoma (RD)
cells at concentrations up to 40 µM. Significantly, these hybrids demonstrate the potential of
antimicrobial-nitroxide agents to overcome the resistance of biofilms to antimicrobials via
stimulation of biofilm dispersal or through direct cell killing.
KEYWORDS
radical; antibiotic; biofilm; nitroxide; ciprofloxacin
1
. INTRODUCTION
The attachment of bacteria to surfaces, and their subsequent ability to aggregate into colonies
called biofilms, is a significant problem in healthcare systems around the world [1-3]. It has been
estimated that biofilms are involved in around 80% of all microbial infections in humans [4],
including those associated with medical devices [5] and chronic wounds [6]. While a variety of
effective antimicrobial strategies exist for the treatment of planktonic bacteria, these approaches
are rarely effective against biofilms [7, 8], which have been reported to be up to one thousand
times more resistant to antibiotic therapies [4, 9, 10]. Accordingly, there is an urgent need to
develop novel strategies for the treatment of established biofilms.
2

ACCEPTED MANUSCRIPT
It is now well recognized that bacteria reside primarily in biofilms but can revert to
planktonic lifestyle by modulating the expression of specific genes [11]. Thus, one approach to
target bacteria in biofilms has involved the development of small molecules with the ability to
inhibit and/or disperse bacterial biofilms through non-microbicidal mechanisms [12, 13]. Nitric
oxide (NO) has been identified to play a central role in biofilm formation and dispersal [14-16]
across a range of biofilm-forming species [17]. When used at low, non-toxic concentrations (in
the pM to low nM range), nitric oxide is capable of dispersing a pre-formed biofilm by triggering
the transitions of cells to the motile, planktonic state [15, 17]. Mechanistically, this effect has
been correlated with a decrease in the intracellular levels of the secondary messenger cyclic di-
GMP, which is involved in biofilm development [11, 18].
The controlled delivery of nitric oxide to biological systems is challenging as it is an
extremely reactive gas with a short half-life of 0.1-5 seconds [19]. Efforts to circumvent the
problems associated with nitric oxide delivery have included the synthesis of NO-donor
molecules [20], and extensive reviews on the dispersal activity of NO-donor in bacterial biofilms
have been written recently [21]. Utilizing the NO-donor concept, a variety of anti-biofilm
compounds have been developed [22]. However, as NO-donor molecules are also often
inherently unstable [23], the use of nitroxides as an alternative for biofilm dispersal have recently
been examined.
Nitroxides are stable free radical species that possess a disubstituted nitrogen atom linked to
a univalent oxygen atom [24]. Both nitroxides and nitric oxide are structurally similar, as both
species possess an unpaired electron, which is delocalized over the nitrogen-oxygen bond
(
Figure 1). Furthermore, the biological effects of nitroxides can be rationalized by their nitric
oxide-mimetic properties, with both compounds known to be efficient scavengers of protein-
derived radicals [25]. In contrast to gaseous nitric oxide, nitroxides have the advantage in that
they are typically air-stable crystalline solids.
Figure 1. The structure of nitric oxide and the general structure of a nitroxide.
3

ACCEPTED MANUSCRIPT
Previously, we have demonstrated that nitroxides can act in a similar manner to nitric oxide
and disperse Pseudomonas aeruginosa biofilms generated in flow cell chambers [26]. When
applied at 20 µM concentrations, nitroxides were able to both inhibit P. aeruginosa biofilm
formation and trigger the dispersal of established P. aeruginosa biofilms. The dispersal ability of
nitroxides has also been documented by others using the less-sensitive crystal violet staining
assay at higher concentrations (in the 5 mM range) [27, 28]. Nitroxides have also recently been
shown to enhance the anti-bacterial activity of silver nanoparticles when coupled together to give
a nitroxide-coated silver nanoparticle [29]. In addition to demonstrating the inhibiting and
dispersal capabilities of nitroxides, we have also reported the potential for biofilm removal when
the biofilm dispersing properties of nitroxides are utilized in combination with an antibiotic
(ciprofloxacin) [30]. The results of this study indicate that the well-known resistance of biofilms
to antimicrobial treatments could be alleviated by employing the dispersal ability of nitroxides.
Furthermore, we have recently shown that combining a nitroxide and an antibiotic within a
single molecule is an effective approach to eradicate mature P. aeruginosa biofilms [31]. These
results demonstrate that the covalent tethering of the antibiotic to the nitroxide positions the
antibiotic near the site of nitroxide-induced biofilm dispersal, and thereby allows the antibiotic to
act directly on the newly dispersed cells before they resume their preferred biofilm mode of
growth. In fact, ciprofloxacin-nitroxide hybrid 1 (Figure 2), which bears the TEMPO nitroxide
moiety, was shown to both induce P. aeruginosa biofilm dispersal and subsequently eradicate
the resulting dispersed cells (up to 95% removal of mature biofilms at 40 uM was observed)
[31].
Figure 2. Ciprofloxacin-nitroxide hybrid 1.
4

ACCEPTED MANUSCRIPT
In our present study, we explored the synthesis of ciprofloxacin-nitroxide conjugate
molecules joined via the secondary amine of the piperazine ring of ciprofloxacin using an amide
linkage. The rationale behind this approach was that the amide functionality may allow access to
ciprofloxacin-nitroxide conjugates which have improved organic (DMSO) solubility to aid in
compound delivery into aqueous biological systems compared to their tertiary amine linked
analogues [31]. Furthermore, the use of an amide bond linkage between the two moieties
expands the variety of carboxylic acid-bearing cyclic nitroxides that can be tethered to the
secondary amine of ciprofloxacin allowing for the effects of nitroxide ring size on anti-biofilm
activity to be explored.
Herein, we report the design and synthesis of the second generation of ciprofloxacin-
nitroxide hybrid molecules together with their biological evaluation as anti-biofilm agents for the
treatment of existing P. aeruginosa biofilms.
2
. RESULTS AND DISCUSSION
2
.1. Chemistry
In line with our previous strategy to generate ciprofloxacin-nitroxide conjugates, we again
chose to exploit the secondary amine of the piperazine ring at the 7-position of the
fluoroquinolone based antibiotic ciprofloxacin 2 (Figure 3) as a useful handle where synthetic
transformations could be performed without significantly altering the antimicrobial properties of
ciprofloxacin [31].
5

ACCEPTED MANUSCRIPT
Figure 3. Ciprofloxacin 2 and N-formyl ciprofloxacin derivative 3.
To generate our second generation of ciprofloxacin-nitroxide hybrids, we tethered nitroxides
to the secondary amine of the piperazine ring of ciprofloxacin 2 using amide bond coupling. The
commercially available cyclic nitroxides 4-carboxy-2,2,6,6-tetramethylpiperidin-1-yloxyl
(
CTEMPO) 6 and 3-carboxy-2,2,5,5-tetramethyl-1-pyrrolidinyloxyl (CPROXYL) 4 as well as
the more rigid isoindoline nitroxide 5-carboxy-1,1,3,3-tetramethylisoindolin-2-yloxyl (CTMIO)
[32] were selected as the nitroxide coupling partners as these systems are resistant to
8
degradation through disproportionation (due to the presence of bis(tert-alkyl) groups on the
carbon atoms α to the nitroxide) [33] and both the piperidine- and isoindoline-based systems
have previously demonstrated dispersal activities in bacterial biofilms [26].
The amide coupling methodology was first optimized using CTEMPO 6. To begin the
synthesis, the carboxylic acid of ciprofloxacin 2 was protected as an ethyl ester using previously
documented procedures to give 16 [34]. The carboxylic acid of CTEMPO 6 was activated by
conversion to the corresponding acid chloride 12 with thionyl chloride and then immediately
reacted with a 1,4-dioxane solution of the protected ciprofloxacin 16 under basic conditions.
o
After heating at 60 C for 1 hour, no starting material 16 remained (TLC analysis) and the desired
ciprofloxacin-nitroxide hybrid 19 was isolated in moderate yield (59%). A second product,
determined to be formamide derivative 3 (Figure 3) by 2D NMR spectroscopy and mass
6

ACCEPTED MANUSCRIPT
spectrometry, was also formed in the reaction (41% yield). N-Formylation of the piperazine
residue of ciprofloxacin has been previously reported to occur in the presence formic acid [35] or
DMF [36] but as neither of these reagents were used directly in the amide synthesis of 19, it was
reasoned the formyl ciprofloxacin 3 may have arisen from an impurity in the commercially
acquired 1,4-dioxane. This theory was confirmed when a 1,4-dioxane solution of protected
o
ciprofloxacin 16 produced the N-formylated analogue 3 after heating at 60 C for 1 hour. The use
of an alternative solvent (DCM) in place of 1,4-dioxane provided a facile solution to avoid
formation of this side product. Reaction of the protected ciprofloxacin 16 and acid chloride 12 in
DCM in the presence of N,N-diisopropylethylamine (Scheme 1) gave the desired ciprofloxacin-
nitroxide hybrid 19 in excellent yield (94%) after stirring at room temperature for 1 hour. These
optimized conditions were then employed to generate conjugate compounds 17 and 21 in
isolated yields of 87% and 98% respectively. Final deprotection of ethyl esters 17, 19 and 21 via
base mediated hydrolysis furnished ciprofloxacin-nitroxide hybrids 23, 25 and 27 in excellent
yield (73-98%). The free carboxylic acid of the fluoroquinolone core is important for
antimicrobial activity as it binds, through magnesium, to the bacterial enzyme DNA gyrase [37,
3
8].
7

ACCEPTED MANUSCRIPT
Scheme 1. Synthetic route to ciprofloxacin-nitroxide hybrids 23, 25 and 27 and their
a
corresponding methoxyamines 24, 26 and 28.
a
Reagents and conditions: (a) SOCl , toluene, pyridine, 0 ºC → rt, 1 h; (b) i-Pr NEt, DCM, rt, 1
2
2
h; (c) 2 M NaOH, MeOH, 50 ºC, 5 h.
In addition to the generation of three novel ciprofloxacin-nitroxide hybrid compounds 23, 25
and 27, methoxyamine derivatives 24, 26 and 28 were also desired as control compounds to
enable a direct comparison of the biofilm dispersal effect of the nitroxide moiety. The
methoxyamine functionality was introduced to the carboxy-functionalized nitroxides 4, 6 and 8
at the beginning of the synthetic sequence such that intermediates could be well characterized by
8

ACCEPTED MANUSCRIPT
NMR spectroscopy (nitroxides are paramagnetic and typically display significantly broadened
NMR signals). Utilising well-known Fenton chemistry [39], the nitroxides 4, 6 and 8 were
treated with methyl radicals generated from hydrogen peroxide, iron(II) sulphate heptahydrate
and DMSO [40] to furnish methoxyamines 5, 7 and 9 in excellent yield (88-92%). Amide
coupling via the corresponding acid chlorides 11, 13 and 15 using the methodology documented
above gave the protected ciprofloxacin-methoxyamine conjugates 18, 20 and 22 in high yield
(83-97%). Subsequent ethyl ester deprotection using base mediated hydrolysis afforded the
desired ciprofloxacin-methoxyamines 24, 26 and 28 in high yield (80-85%).
2.2. Biological Evaluation
Our previous studies have indicated that pre-formed P. aeruginosa biofilms can be dispersed
upon treatment with nitroxides [26]. Furthermore, we have documented the ability of the
nitroxide CTEMPO 6 to almost completely remove mature P. aeruginosa and E. coli biofilms
when used in combination with the antibiotic ciprofloxacin 2 in a flow cell assay [30]. Here, we
employed a similar approach involving pre-formed P. aeruginosa biofilms grown in flow cell
chambers to evaluate the dispersal and biofilm removal properties of the prepared ciprofloxacin-
nitroxide compounds 23-28. P. aeruginosa biofilms were formed in flow cell chambers for 48 h
and then treated with 20 µM solutions of the hybrid compounds 23-28 (dissolved in DMSO and
delivered into BM2 minimal medium supplemented with 0.4% of glucose) for 24 h. This specific
concentration was chosen as it was previously established to be the most effective concentration
for nitroxide-mediated biofilm dispersal [26]. Visualization of the resulting biofilms using the
Live/Dead BacLight bacterial viability kit coupled with confocal microscopy provided the
images shown in Figures 4, 5 and 6. From these images, we calculated the percentages of biofilm
biomass removed by comparing the amount of biomass remaining from 3-day-old untreated
biofilms relative to the remaining adhered biomass from flow cells treated with compounds 23-
2
8). We also calculated the percentage of dead cells present in each biofilm sample to assess the
overall biofilm removal ability of the hybrid compounds. These results are displayed in Table 1.
The results obtained from the hybrid compounds bearing the TEMPO unit (25 and 26) at
concentrations of 20 µM were examined first. The ciprofloxacin-nitroxide 25 displayed biofilm
removal potential (37%) without major killing of biofilm cells (6%) (Table 1 and Figure 4b).
Intriguingly, the corresponding methoxyamine 26 also exhibited some observable removal
9

ACCEPTED MANUSCRIPT
activity (17%) (Figure 4c), although it was more modest than compound 25. The dispersal of
mature P. aeruginosa biofilms by alkoxyamines has been previously observed by others for an
ethoxyamine derivative in a crystal violet biofilm assay, however in this previous study, the
corresponding nitroxide compound was still more effective at inducing dispersal [27].
Figure 4. Confocal laser scanning microscopy images of 2-day-old pre-formed P. aeruginosa
PA14 biofilms grown in a flow cell at 37°C, treated with (b) 20 µM of 25 and (c) 20 µM of 26
for 24 hours and then visualized with SYTO-9 (stains live cells green) and propidium iodide
(
stains dead cells red). Panel (a) shows an untreated P. aeruginosa PA14 biofilm after 3 days. At
least two replicates were performed per condition. The scale bars represent 40 µm in length for
images (a) and (c), and 50 µm for image (b). Each panel also shows the xy, yz and xz dimensions.
Next we analyzed the flow cell assay results from the ciprofloxacin-nitroxide compounds
bearing PROXYL moieties (23 and 24). Both the nitroxide 23 and the methoxyamine 24 (Figures
5
b and 5c), appeared to remove mature P. aeruginosa biofilms at 20 µM. However, the degree of
removal was greater for the nitroxide 23 (67% biofilm removal) than the corresponding
methoxyamine 24 (24% biofilm removal), and in particular 23, unlike 24, appeared to have
greater dispersal potential (dispersing the large dense bacterial aggregates within the flow cell
(
Figures 5b and 5c). The ciprofloxacin-PROXYL analogue 23 was also tested at 10 µM in the
10

ACCEPTED MANUSCRIPT
flow cell assay against mature P. aeruginosa biofilms. Some dispersal effects along with a 50%
reduction in biofilm mass was observed (Figure 5a), however, the anti-biofilm activity of
compound 23 was more pronounced at 20 µM (67%, Table 1).
Figure 5. Confocal laser scanning microscopy images of 2-day-old pre-formed P. aeruginosa
PA14 biofilms grown in a flow cell at 37°C, treated with (a) 10 µM of 23, (b) 20 µM of 23 and
(c) 20 µM of 24 for 24 hours and then visualized with SYTO-9 (stains live cells green) and
propidium iodide (stains dead cells red). At least two replicates were performed per condition.
The scale bars represent 40 µm in length for images (a)-(c). Each panel also shows the xy, yz and
xz dimensions.
11

ACCEPTED MANUSCRIPT
Lastly, the activity of the conjugate molecules bearing isoindoline moieties (27 and 28) were
examined against mature P. aeruginosa biofilms in flow cell chambers. In the presence of 20 µM
of ciprofloxacin-nitroxide hybrid 27, a substantial reduction in the total biofilm biovolume (85%)
occurred (Figure 6a) and 60% of the remaining biofilm biomass was composed of dead cells.
This represents an improvement over hybrid compound 1 (our previously most active hybrid
conjugate), which reduced total biofilm biovolume by 80% with 50% of the remaining biofilm
biomass containing dead cells [31]. Furthermore, compound 27 was also found to be
substantially more effective at treating P. aeruginosa-based biofilms than the parent antibiotic,
ciprofloxacin (use at its MIC value), which only reduced the total biofilm biovolume by 7% with
very few dead cells (8%) detected in the remaining biofilm biomass [30].
Figure 6. Confocal laser scanning microscopy images of 2-day-old pre-formed P. aeruginosa
PA14 biofilms grown in a flow cell at 37°C, treated with (a) 20 µM of 27 and (b) 20 µM of 28
for 24 hours and then visualized with SYTO-9 (stains live cells green) and propidium iodide
(stains dead cells red). Compound 27 led to cell filamentation, presumably via the active
ciprofloxacin moiety. At least two replicates were performed per condition. The scale bars
represent 40 µm in length for images (a) and (b). Each panel also shows the xy, yz and xz
dimensions.
12

ACCEPTED MANUSCRIPT
Table 1. Total live P. aeruginosa PA14 biofilm biomass eradication and biofilm dead cell values
for 2-day-old biofilms treated with compounds 23-28 at 20 µM relative to the biomass and dead
cell values of untreated 3-day-old biofilms.
Dead Cells in
Remaining
Biomass (%)
Total Live Biofilm
Biomass
Eradication (%)
Remaining
Biomass (%)
Removed
Biomass (%)
Compound
a
Control PA14
--
--
4
0
9
6
16
--
2
2
2
2
2
2
3
4
5
6
7
8
33
76
63
83
15
47
67
24
37
17
85
53
67
31
41
30
94
54
60
2
a
Calculated by adding the amount of dead cells remaining in the biofilm biomass to the initially removed biomass
i.e., compound 27 had initially 85% removed biomass but as 60% of the remaining 15% of biomass was dead, the
(
total live biofilm biomass eradication was 94%).
The observed filamentation and cell death were typical effects in the presence of
ciprofloxacin [41]. Thus, compound 27 led to virtually complete removal of mature biofilms
formed by P. aeruginosa, and treated cells often exhibited filamentous phenotypes (Figure 6a).
Interestingly, the corresponding methoxyamine 28, by comparison, was also able to reduce
biofilm cell density by 53% at the same concentration (Table 1) but fewer cells were actually
killed (2%) (Figure 6b). Of the three ciprofloxacin-nitroxide types examined (23, 25 and 27), the
most effective biofilm removing agent was the isoindoline analogue 27. The TEMPO analogue
2
5 was the least active.
The minimal inhibitory concentrations (MIC) of each prepared compound were also
measured using the broth microdilution method [42, 43]. The results shown in Table 2, revealed
that the prepared compounds 23-28 exhibited minimal antibacterial activity, especially compared
to free ciprofloxacin. The most effective compound with antimicrobial activity against
planktonic bacteria was compound 23, where no bacterial growth was observed at 160 µM. The
obtained values are, however, substantially higher than that of ciprofloxacin 2 alone which gave
4
8
a previously reported MIC of 0.5 µM. Thus, the addition of nitroxide units to the piperazine
ring of ciprofloxacin through amide bonds has resulted in decreased efficacy of the antibiotic.
However, despite the substantially higher MIC of compounds 23-28, ciprofloxacin alone has
been previously shown to be ineffective at biofilm removal [30], whereas the ciprofloxacin-
13

ACCEPTED MANUSCRIPT
nitroxide compounds 23, 25 and 27 prepared in this work display significant anti-biofilm
activities. This is evident with our most active compound (27), which at 20 µM (10.95 µg/mL)
resulted in substantial removal (85%) of existing P. aeruginosa biofilms (Table 1).
Table 2. MIC values of ciprofloxacin-nitroxide hybrids 17-28 against P. aeruginosa PA14
planktonic bacteria.
Compound
MIC (µM)
2
2
2
2
2
2
3
4
5
6
7
8
160
311
312
>303
292 (160)
a
a
>284 (>160)
b
Ciprofloxacin 2
0.5
a
b
More accurate value determined from mechanism of action studies via CFU/mL counts (Figure 7).
Previously reported value measured using the same procedure.
[41]
Additional killing assays of planktonic P. aeruginosa PA14 cultures showed that compound
2
2
7 was more effective at killing P. aeruginosa PA14 planktonic cells compared with compound
8, and led to complete killing of bacterial cultures at the higher concentration used (see
supporting information), thus correlating with its increased ability to directly eradicate and kill
biofilm cells (Table 1). The lower dose required for biofilm killing in flow cell assays compared
to planktonic killing assays is not surprising, as it has been observed before for similar and other
compounds [41].
Next, we aimed to investigate mechanistic insights responsible for the biofilm inhibitory
activity of specific ciprofloxacin-nitroxide hybrid compounds (Figure 7). We chose to examine
the most active and least active nitroxides (27 and 25) from the flow cell assays and their
3
0
corresponding methoxyamines (28 and 26). We leveraged our viable dispersal cell assay
consisting of performing CFU counts from the effluent of flow cell chambers upon nitroxide
treatment. This system allows precise monitoring of viable cell counts over time post-treatment.
We found that compounds 25 and 26 stimulated dispersion of bacteria from biofilms over time
cf. the untreated group (Figure 7a), with compound 25 resulting in more dispersed bacteria as
compared to 26 (Figure 7a). As such, we believe that the increased biofilm removal activity
observed for 25 in comparison with 26 can at least partly be attributed to the increased dispersal
activity of compound 25 (Table 1). On the other hand, no obvious increase in bacterial dispersal
14

ACCEPTED MANUSCRIPT
was observed in treatments using 27 or 28 when compared to untreated controls (Figure 7b). This
may imply that dispersed bacteria are immediately removed or, more likely, that 27 and 28 do
not trigger biofilm dispersal but act through direct cell killing. The minimal viable dispersed
bacteria reported in Figure 7b suggests that dispersal is not part of the mechanism of action of
these compounds. However both 27 and 28 are capable of removing biofilms (Table 1), which
indicates they may kill biofilm cells as a major mechanism of action, particularly in the case of
2
7 (Table 1). It is also likely that the compounds killed the dispersed cell population released
from biofilms, as both 27 and 28 are capable of killing planktonic bacteria (see supporting
information).
Figure 7. Mechanism of action studies. (a) Compound 25 stimulated cell dispersal from
biofilms. Viable biofilm dispersal cell assay done using P. aeruginosa PA14 biofilms, which
were grown in flow cells for 2 days and subsequently treated with 20 µM of compounds 25 and
26. Dispersed cells were collected from the effluent of the flow cell chambers upon treatment
with the compounds at the specified times. Collected bacteria were then plated for CFU counts
on Pseudomonas Isolation Agar (PIA) plates and incubated at 37°C. The next day, bacteria were
counted (CFU/mL) and fold-change differences in cell counts were calculated in dispersed cells
of treated samples compared to those from untreated P. aeruginosa PA14 samples. Note: the
fold-change of PA14 viable dispersed cells from compound 26 at 18 h 30 min was virtually 0.
(b) Compound 27 kills planktonic bacteria and does not induce cell dispersal from biofilms.
Biofilm dispersal cell assay was performed as described in (a) except treating biofilms with 20
µM of compounds 27 and 28.
As biofilms are associated with a wide range of infections in humans, the cytotoxicity of a
selection of the prepared compounds was examined in two different human cell lines using the
lactate dehydrogenase (LDH) cytotoxicity assay. Compounds 25-28 were shown to be non-toxic
15

ACCEPTED MANUSCRIPT
in both human embryonic kidney 293 (HEK 293) cells and human muscle rhabdomyosarcoma
(RD) cells at concentrations from 5 µM to 40 µM (Figure 8). The addition of small quantities of
DMSO as the solvent was also shown to have no significant effect on cell viability (see
supporting information).
Figure 8. Cell viability studies for compounds 25-28 as a function of concentration (blue = 5
µM, red = 10 µM, green = 20 µM, purple = 40 µM): (a) HEK 293 cells, (b) RD cells.
3
. CONCLUSION
Three ethyl ester protected ciprofloxacin-nitroxide hybrid compounds 17, 19 and 21 and
their methoxyamine analogues 18, 20 and 22 were prepared using amide bond coupling in high
to excellent yield (83-98%) from the corresponding acid chloride functionalized nitroxides 10,
1
2 and 14 or methoxyamines 11, 13 and 15 and the ethyl ester protected ciprofloxacin 16.
Deprotection of the amide-linked ethyl ester analogues 17-22 with base gave the corresponding
ciprofloxacin-nitroxides 23, 25 and 27 and their methoxyamines 24, 26 and 28 in good to high
yield (73-98 %). The synthesized compounds 23-28 exhibited modest antibacterial activities with
MIC values ranging from 160 to >300 µM. In killing assays, both 27 and 28 exhibited anti-
planktonic activity with 27 showing increased killing. These results suggest that the nitroxide
moiety is key to the effectiveness of analogue 27 against planktonic cells. Evaluation of the
prepared compounds 23-28 for anti-biofilm activity against mature P. aeruginosa biofilms was
16

ACCEPTED MANUSCRIPT
performed in a flow cell assay. Several of the hybrid compounds were found to have the desired
dual-action effect against established biofilms. For instance, treatment with compounds 25 and
2
6 substantially increased dispersal of bacteria from biofilms over time, particularly in the case
of 25, which likely reflected the biofilm removal results (37% for 25 and 17% for 26) with no
major biofilm cell killing events obtained (6% for 25 and 16% for 26). On the other hand,
treatment with compound 27 did not result in increased cell dispersal from biofilms but led to the
highest biofilm removal (85%) and biofilm cell-killing activity (60%) of any of the compounds
tested and also exerted toxicity towards planktonic cells. In addition, the corresponding
methoxyamine 28 also did not stimulate dispersal but disrupted biofilms less than 27 (53%)
without killing biofilm cells (1.5% dead cells). From these experiments, we conclude that the
ability of both 27 and 28 to repress biofilms is independent of dispersal events. We propose that
these compounds likely kill biofilm cells at lower concentrations (~20 µM). The cytotoxicity of a
selection of the prepared compounds (25-28) was also examined in both human embryonic
kidney 293 (HEK 293) cells and human muscle rhabdomyosarcoma (RD) cells and found to be
non-toxic at concentrations up to 40 µM.
The results presented here demonstrate that the combination of an antibiotic and a nitroxide
within a single molecule is an effective approach to facilitate the efficient control of mature
biofilms via stimulation of biofilm dispersion or through direct cell killing, and thereby help
overcome the resistance of biofilms to antimicrobials.
4
. EXPERIMENTAL SECTION
4
.1. General Procedures
Reactions of an air-sensitive nature were carried out under an atmosphere of ultra-high purity
argon. Where anhydrous THF, DMF, DCM or acetonitrile are documented, these solvents were
obtained from the solvent purification system, pure solv micro by Innovative Technologies.
Anhydrous toluene was dried by storage over sodium wire. Triethylamine and i-Pr NEt were
2
stored over potassium hydroxide. All other reagents were purchased from commercial suppliers
1
and used without further purification. All H NMR spectra were recorded at either 400 or 600
MHz on either a Varian Inova 400, a Bruker Avance 400 or a Bruker Avance 600 instrument. All
17

ACCEPTED MANUSCRIPT
1
3
C NMR spectra were recorded at either 100 or 150 on either a Varian Inova 400, a Bruker
Avance 400 or a Bruker Avance 600 instrument. Samples were prepared in CDCl , unless
3
1
otherwise stated, using oven dried glassware. H NMR spectra in CDCl were referenced to the
3
1
3
solvent peak at 7.27 ppm. C NMR spectra run in CDCl
3
were referenced to the solvent peak at
7
7.2 ppm. Coupling constants are reported in Hz. High-resolution ESI mass spectra were
obtained with an Agilent Q-TOF LC high-resolution mass spectrometer, which utilized
electrospray ionization in positive ion mode. The mass-selective detector was optimized by using
calibration standards with reference masses at m/z 121.050873 and 922.009798. Fourier
transform infrared (FTIR) spectra were recorded on a Nicolet 870 Nexus Fourier Transform
Infrared Spectrometer equipped with a DTGS TEC detector and an ATR objective. Melting
points were measured with a Variable Temperature Apparatus by the capillary method and are
uncorrected. Analytical HPLC was carried out on an Agilent Technologies HP 1100 Series
HPLC system using an Agilent C18 column (4.6 × 250 mm, 5 µm) or an Agilent Zorbax RX-SIL
column (4.6 × 250 mm, 5 µm) with a flow rate of 1 mL/min. The purity of all final compounds
was determined to be 95 % or higher using HPLC analysis or qNMR techniques. EPR spectra
were obtained with the aid of a miniscope MS 400 Magnettech EPR spectrometer. Column
chromatography was performed using LC60A 40–63 Micron DAVISIL silica gel. Thin-layer
chromatography (TLC) was performed on Merck Silica Gel 60 F254 plates. TLC plates were
visualised under a UV lamp (254 nm), and/or by development with phosphomolybdic acid
(PMA).
4
.2. Materials
-Carboxy-1,1,3,3-tetramethylisoindolin-2-yloxyl 8 and ethyl 1-cyclopropyl-6-fluoro-4-oxo-7-
piperazin-1-yl)-1,4-dihydroquinoline-3-carboxylate 16 were prepared according to known
procedures [32, 34].
5
(
4
.3. Interference compounds
Nitroxides are recognized quenchers of fluorescent molecules and this effect can occur through
space with optimal separation distances in the range of 0.5-2 nm [44]. Thus, there is potential in
the biofilm flow cell assay for nitroxides to quench the fluorescence of the SYTO-9 and
propidium iodide dyes used during the confocal microscopy analysis if the nitroxides and
18

ACCEPTED MANUSCRIPT
fluorophores are localized within the same region of the bacterial cell. However, as the
fluorescence arising from the nitroxide containing compounds 17, 19, 21, 23, 25 and 27 was
similar to the fluorescence emitted by the methoxyamine control compounds 18, 20, 22, 24, 26
and 28, the potential effect of nitroxide-induced fluorescence quenching on the obtained
biological results is minimal.
4
.4. Biofilm dispersal flow cell assays
P. aeruginosa PA14 biofilms were pre-formed at 37°C over 48 h in flow chambers using
previously established techniques [26]. The biofilms were then exposed for 24 hours to 10 or 20
µM solutions of ciprofloxacin-nitroxide hybrid compounds 17-28 resuspended in DMSO in the
flow cell chambers with channel dimensions of 1 × 4 × 40 mm. Flow chambers were inoculated
with 400 µL of an overnight P. aeruginosa PA14 culture diluted to an OD600 of ~0.05. Next,
chambers were left without flow for 2 h, after which medium was pumped through the system at
a constant rate of 2.4 mL/h. Staining and visualisation of the resulting biofilms was performed
using the Live/Dead BacLight bacterial viability kit and a confocal laser scanning microscope
(Olympus, Fluoview FV1000). Three-dimensional reconstructions and residue biofilm
biovolume calculations were achieved using Imaris software.
4
.5. Bacterial killing experiments
Killing experiments involved performing 1:100 dilutions of overnight cultures of P.
aeruginosa PA14 in the abscense or presence of increasing concentrations of nitroxides (0-160
µM). After 24 h of treatment, 10-fold serial dilutions were performed, bacteria were plated on
LB agar plates and allowed to grow overnight at 37 °C after which colony forming unit (CFU)
counts were recorded.
4
.6. MIC assays
The MIC assays were performed using the broth microdilution method [42, 43] in sterile 96-
well polypropylene microtiter plates. Nitroxides were added to the plate as solutions in DMSO at
5
the desired concentrations, and the bacteria were inoculated at a final concentration of 5 × 10
CFU/mL per well. The plates were incubated at 37 °C for 24 h. The MIC was defined as the
lowest concentration of compound at which no growth was observed.
19

ACCEPTED MANUSCRIPT
4
.7. Cytotoxicity assays
Human embryonic kidney 293 (HEK 293) cells (ATCC, VA) and human muscle
rhabdomyosarcoma (RD) cells (ATCC, VA) were cultured in Dulbecco's Modified Eagle
Medium (DMEM) (ThermoFisher, MA) supplemented with 10% Fetal Bovine Serum (FBS)
(ThermoFisher, MA) at 37°C in 5% CO2. The day before treatment, 50,000 HEK 293 cells or
2
0,000 RD cells were seeded into each well in 96-well plates. The compounds were dispersed in
DMSO at the concentration of 5 mM. Different concentrations of compounds 25-28 were added
into the wells for another 24 hours, and 0.8% of DMSO was added in all wells to eliminate the
effects of DMSO. The release of the lactate dehydrogenase (LDH) was then measured, following
the manual of Pierce LDH Cytotoxicity Assay Kit (Thermofisher, MA). Untreated cells (live
cells) and cells treated with lysis buffer for 3 hours (dead cells) were used as the reference for
normalization. All experiments were performed in triplicate.
4
.8. General procedure for the synthesis of methoxyamine derivatives (
5), (7) and (9)
Iron(II) sulfate heptahydrate (FeSO .7H O, 2.5 equiv) was added to a solution of nitroxide
4
2
compound (1 equiv) in DMSO. The mixture was then cooled to 0 ºC and 35 % aqueous hydrogen
peroxide (4 equiv) was added in a dropwise manner. The resulting mixture was stirred at 0 ºC for
1
0 minutes and then at room temperature for an additional 1.5 hours. The reaction mixture was
diluted with deionized water (40 mL) and adjusted to pH ~3 using aqueous hydrochloric acid (2
M) before being extracted with diethyl ether (3 × 20 mL). The combined organic extracts were
washed with deionized water (200 mL) and dried over anhydrous sodium sulfate. The solvent
was removed in vacuo to yield the desired methoxyamine product.
4
.8.1. 1-Methoxy-2,2,5,5-tetramethylpyrrolidine-3-carboxylic acid (
5)
Reagents: CPROXYL 4 (120 mg, 0.64 mmol, 1 equiv), FeSO .7H O (448 mg, 1.61 mmol, 2.5
4
2
2 2
equiv), 35 % aqueous H O (0.23 mL, 2.56 mmol, 4 equiv) and DMSO (2.5 mL). Data for 5:
-1
white solid (117 mg, 0.58 mmol, 91 %); mp 40-41 ºC. IR (ATR) νmax (cm ) = 3100-2500 (w, br,
1
O-H, COOH) and 1703 (s, C=O, COOH). H NMR (600 MHz, CDCl
3
) δ = 3.63 (s, 1 H,
NOCH ), 2.76 (s, 1 H, C(O)CH), 2.10 (s, 1 H, C(O)CHCH ), 1.74 (dd, J = 12.8, 7.8 Hz, 1 H,
3
2
1
3
C(O)CHCH ), 1.35 (s, 3 H, CH ), 1.23 (s, 1 H, CH ), 1.19 (s, 1 H, CH ), 1.09 (s, 1 H, CH ). C
2
3
3
3
3
20

ACCEPTED MANUSCRIPT
NMR (150 MHz, CDCl
3
) δ = 178.7, 65.0, 61.2, 48.3, 38.6, 33.6, 28.9, 25.8, 16.3. HRMS (ESI):
+
+
m/z calcd for C H NO + H [M+H ]: 202.1428. Found 202.1428. Absolute quantitative NMR:
1
0
19
3
9
8.8% pure.
21

ACCEPTED MANUSCRIPT
4
.8.2. 1-Methoxy-2,2,6,6-tetramethylpiperidine-4-carboxylic acid (7).
Reagents: CTEMPO 6 (120 mg, 0.60 mmol, 1 equiv), FeSO .7H O (448 mg, 1.61 mmol, 2.7
4
2
equiv), 35 % aqueous H O (0.23 mL, 2.56 mmol, 4.3 equiv) and DMSO (2.5 mL). Data for 7:
2
2
-1
clear white solid (113 mg, 0.52 mmol, 88 %); mp 97-99 ºC. IR (ATR) νmax (cm ) = 3100-2500
1
(
w, br, O-H, COOH) and 1693 (s, C=O, COOH). H NMR (600 MHz, CDCl ) δ = 3.62 (s, 3 H,
3
3 2
NOCH ), 2.66 (t, J = 12.7 Hz, 1 H, CH), 1.75 (d, J = 11.8 Hz, 2 H, CH ), 1.66 (t, J = 12.9 Hz, 2
1
3
H, CH ), 1.23 (s, 6 H, 2 × CH ), 1.12 (s, 6 H, CH ). C NMR (150 MHz, CDCl ) δ = 180.7,
2
3
3
3
+
+
6
5.7, 59.4, 41.7, 32.9, 20.3. HRMS (ESI): m/z calcd for C H NO + H [M+H ]: 216.1583.
11 21 3
Found 216.1587. Absolute quantitative NMR: 97.0% pure.
4
.8.3. 2-Methoxy-1,1,3,3-tetramethylisoindoline-5-carboxylic acid (
9)
Reagents: CTMIO 8 (100 mg, 0.43 mmol, 1 equiv), FeSO .7H O (300 mg, 1.08 mmol, 2.5
4
2
equiv), 35 % aqueous H O (0.2 mL, 2.50 mmol, 5.8 equiv) and DMSO (2.5 mL). Data for 9:
2
2
-1
white solid (99 mg, 0.40 mmol, 93 %); mp 181-183 ºC. IR (ATR) νmax (cm ) = 3100-2500 (w,
1
br, O-H, COOH) and 1679 (s, C=O, COOH). H NMR (600 MHz, CDCl ) δ = 8.03 (dd, J = 7.9,
3
1
.6 Hz, 1 H, Ar-H), 7.86 (d, J = 1.2 Hz, 1 H, Ar-H), 7.20 (d, J = 7.9 Hz, 1 H, Ar-H), 3.80 (s, 3 H,
1
3
NOCH ), 1.47 (br s, 12 H, 4 × CH ). C NMR (150 MHz, CDCl ) δ = 172.1, 151.8, 146.0,
3
3
3
1
29.8, 128.7, 123.8, 121.9, 67.5, 67.2, 65.7, 42.9, 30.2, 25.1. HRMS (ESI): m/z calcd for
+
+
C
14 3
H19NO + H [M+H ]: 250.1443. Found 250.1440. HPLC analysis: retention time = 2.992
min; peak area, 95.26 %; eluent A, Methanol; eluent B, H O; isocratic (80:20) over 20 min with
2
-1
a flow rate of 1 mL min and detected at 254 nm; column temperature, rt.
4
.9. General procedure for the synthesis of amide coupled compounds (17-22
)
Pyridine (2 equiv) was added to a solution of carboxylic acid (1 equiv) in anhydrous toluene
under an atmosphere of argon. The resulting solution was cooled to 0 ºC in an ice-water bath and
thionyl chloride (1.5 equiv) was added dropwise. The solution was stirred at room temperature
for 1 hour. The solvent was then removed in vacuo and the resulting residue taken up in
anhydrous dichloromethane (10 mL). This crude product was added to a stirring solution of the
2
amine bearing compound (1.2 equiv) and i-Pr NEt (2 equiv) dissolved in anhydrous
dichloromethane under an atmosphere of argon. The resulting mixture then stirred at room
temperature for 1 hour before water was added to the mixture. The organic phase was separated
22

ACCEPTED MANUSCRIPT
and the aqueous phase was re-extracted with dichloromethane (3 × 20 mL). The combined
extracts were dried over anhydrous sodium sulfate and the solvent removed in vacuo to afford a
crude solid product. Purification was achieved via column chromatography (SiO , gradient
2
elution: 100 % chloroform to 95 % chloroform, 5 % methanol).
4
.9.1. Ethyl 1-cyclopropyl-6-fluoro-7-(4-(2,2,5,5-tetramethyl-1-oxy-pyrrolidine-3-
carbonyl)piperazin-1-yl)-4-oxo-1,4-dihydroquinoline-3-carboxylate (17
)
Reagents for acid chloride formation: CPROXYL 4 (100 mg, 0.54 mmol, 1 equiv), anhydrous
toluene (5 mL) and pyridine (0.1 mL, 1.20 mmol, 2.2 equiv). Reagents for amide coupling: 16
(
230 mg, 0.65 mmol, 1.2 equiv), i-Pr NEt (0.2 mL, 1.08 mmol, 2 equiv) and anhydrous
2
dichloromethane (5 mL). Data for 17: yellow solid (248 mg, 0.47 mmol, 87 %); mp 118-120 ºC.
-1
1
IR (ATR) νmax (cm ) = 3500-3200 (w, br, N-H, amide) and 1688 (m, C=O, amide). H NMR
(
600 MHz, CDCl ) (*note compound is a free-radical, some signals appear broadened and other
3
1
signals are missing) δ = H NMR (600 MHz, CDCl
3
) δ = 8.59 (s, 1 H, NCH=C), 8.11 (d, J =
1
0.2 Hz, 1 H, Ar-H), 7.36 (s, 1 H, Ar-H), 4.46 (q, J = 6.3 Hz, 2 H, OCH CH ), 4.01 (s, 2 H, 2 ×
2 3
2 2 2
NCH ), 3.51 (s, 2 H, 2 × NCH ), 3.36 (s, 2 H, 2 × NCH ), 3.26 (m, 1 H, C=CHNCH), 1.48 (t, J =
1
3
5
.91 Hz, 3 H, OCH CH ), 1.42 (s, 2 H, NCHCH ), 1.23 (s, 2 H, NCHCH ). C NMR (150 MHz,
2
3
2
2
CDCl ) δ = 171.7, 164.3, 152.8, 151.2, 147.0, 142.4, 136.7, 122.4, 112.4, 112.2, 109.3, 103.9,
3
+
+
5
9.7, 51.0, 48.4, 40.0, 33.5, 13.2, 7.1. HRMS (ESI): m/z calcd for C28
4 5
H36FN O + H [M+H ]:
5
28.2758. Found 528.2753. HPLC analysis: retention time = 5.481 min; peak area, 99.52 %;
-1
2
eluent A, Methanol; eluent B, H O; isocratic (70:30) over 25 min with a flow rate of 1 mL min
and detected at 254 nm; column temperature, rt. EPR: g = 1.9975, a = 1.4898 mT.
N
4
.9.2. Ethyl 1-cyclopropyl-6-fluoro-7-(4-(1-methoxy-2,2,5,5-tetramethylpyrrolidine-3-
carbonyl)piperazin-1-yl)-4-oxo-1,4-dihydroquinoline-3-carboxylate (18
)
Reagents for acid chloride formation: 5 (120 mg, 0.60 mmol, 1 equiv), anhydrous toluene (5
mL) and pyridine (0.1 mL, 1.20 mmol, 2 equiv). Reagents amide coupling: 16 (276 mg, 0.77
mmol, 1.2 equiv), i-Pr NEt (0.2 mL, 1.08 mmol, 1.8 equiv) and anhydrous dichloromethane (5
2
mL). Data for 18: pale yellow solid (317 mg, 0.58 mmol, 97 %); mp 132-134 ºC. IR (ATR) νmax
-1
1
(
cm ) = 3500-3200 (w, br, N-H, amide) and 1689 (m, C=O, amide). H NMR (600 MHz, CDCl )
3
δ = 8.50 (s, 1 H, NCH=C), 8.01 (d, J = 13.1 Hz, 1 H, Ar-H), 7.25 (d, J = 7.0 Hz, 1 H, Ar-H), 4.37
23

ACCEPTED MANUSCRIPT
(
q, J = 7.1 Hz, 2 H, OCH
2
CH
3
), 3.89 (s, 2 H, NCH
2
), 3.83 (s, 2 H, NCH
2
3
), 3.66 (s, 3 H, NOCH ),
3
.43 (m, 1 H, C=CHNCH), 3.23 (d, J = 4.9 Hz, 2 H, 2 × NCH ), 3.22 (m, 2 H, 2 × NCH ), 3.10
2
2
(
br, s, 1 H, NC(O)CH), 2.38 (br, s, 1 H, NC(O)CH ), 1.60 (dd, J = 12.6, 7.4 Hz, 1 H,
2
2 2 3 2 3
NC(O)CHCH ), 1.40 (t, J = 7.1 Hz, 3 H, OCH CH ), 1.33 (s, 2 H, NCHCH ), 1.31 (s, 3 H, CH ),
1
.26 (s, 3 H, CH ), (br, s, 3 H, CH ), 1.14 (dd, J = 8.9, 5.9 Hz, 2 H, NCHCH ), 1.06 (br, s, 3 H,
3
3
2
13
CH
3
). C NMR (100 MHz, CDCl
3
) δ = 173.1, 165.8, 154.6, 152.2, 148.3, 144.1, 144.0, 138.1,
1
23.7, 123.6, 113.7, 113.4, 110.6, 105.2, 65.0, 61.0, 50.9, 49.9, 46.4, 42.1, 34.7, 14.6, 8.3.
+
+
HRMS (ESI): m/z calcd for C H FN O + H [M+H ]: 543.2991. Found 543.2990. HPLC
2
9
39
4
5
analysis: retention time = 3.286 min; peak area, 99.70 %; eluent A, Methanol; over 25 min with a
-1
flow rate of 1 mL min and detected at 254 nm; column temperature, rt.
4
.9.3. Ethyl 1-cyclopropyl-6-fluoro-7-(4-(2,2,6,6-tetramethyl-1-oxy-piperidine-4-
carbonyl)piperazin-1-yl)-4-oxo-1,4-dihydroquinoline-3-carboxylate (19
)
Reagents for acid chloride formation: CTEMPO 6 (100 mg, 0.50 mmol, 1 equiv), anhydrous
toluene (5 mL) and pyridine (0.1 mL, 1.20 mmol, 2.4 equiv). Reagents amide coupling: 16 (216
2
mg, 0.60 mmol, 1.2 equiv), i-Pr NEt (0.3 mL, 1.0 mmol, 2 equiv) and anhydrous
dichloromethane (5 mL). Data for 19: light orange powder (255 mg, 0.47 mmol, 94 %); mp 241
-1
ºC decomposed. IR (ATR) νmax (cm ) = 3500-3200 (w, br, N-H, amide) and 1690 (m, C=O,
1
amide). H NMR (600 MHz, CDCl
3
) (*note compound is a free-radical, some signals appear
broadened and other signals are missing) δ = 8.57 (s, 1 H, NCH=C), 8.12 (d, J = 12.1 Hz, 1 H,
Ar-H), 7.33 (br, s, 1 H, Ar-H), 4.43 (q, J = 7.0 Hz, 2 H, OCH CH ), 3.94 (br, s, 4 H, 2 × NCH ),
.46 (br, s, 2 H, NCH ), 3.40 (br, s, 1 H, NCH), 3.32 (br, s, 2 H, NCH ), 1.60-1.48 (br, s, 12 H,
2
3
2
3
2
2
CH ), 1.45 (t, J = 7.0 Hz, 3 H, OCH CH ), 1.37 (br, s, 2 H, NCHCH ), and 1.19 (br, s, 2 H,
3
2
3
2
13
2 3
NCHCH ). C NMR (150 MHz, CDCl ) δ = 172.7, 165.4, 153.8, 152.1, 147.9, 143.5, 123.4,
1
13.4, 110.4, 104.7, 60.7, 50.4, 49.1, 44.7, 42.1, 34.2, 14.1, 7.9. HRMS (ESI): m/z calcd for
+
+
C
29
H
38FN
4
O
5
+ H [M+H ]: 542.2905. Found 542.2902. HPLC analysis: retention time = 6.206
min; peak area, 99.90 %; eluent A, Methanol; eluent B, H O; isocratic (80:20) over 25 min with
2
-1
a flow rate of 1 mL min and detected at 254 nm; column temperature, rt. EPR: g = 1.9989, a =
N
1
.5751 mT.
24

ACCEPTED MANUSCRIPT
4
.9.4. Ethyl 1-cyclopropyl-6-fluoro-7-(4-(1-methoxy-2,2,6,6-tetramethylpiperidine-4-
carbonyl)piperazin-1-yl)-4-oxo-1,4-dihydroquinoline-3-carboxylate (20
)
Reagents for acid chloride formation: 7 (100 mg, 0.46 mmol, 1 equiv), anhydrous toluene (5
mL) and pyridine (0.1 mL, 1.20 mmol, 2.6 equiv). Reagents for amide coupling: 16 (216 mg,
0
.60 mmol, 1.3 equiv), i-Pr NEt (0.3 mL, 1.0 mmol, 2.2 equiv) and anhydrous dichloromethane
2
(5 mL). Data for 20: off-white powder (231 mg, 0.41 mmol, 90 %); mp 218-219 ºC. IR (ATR)
-1
1
νmax (cm ) = 3500-3200 (w, br, N-H, amide) and 1645 (m, C=O, amide). H NMR (600 MHz,
CDCl ) δ = 8.50 (s, 1 H, NCH=C), 8.02 (d, J = 13.1 Hz, 1 H, Ar-H), 7.26 (d, J = 7.0 Hz, 1 H, Ar-
3
H), 4.38 (q, J = 7.1 Hz, 2 H, OCH
2
CH
3
), 3.84 (br, s, 2 H, NCH
2
), 3.71 (br, s, 2 H, NCH
2
) 3.62 (s,
3
H, NOCH ), 3.42 (m, 1 H, C=CHNCH), 3.30 (br, s, 2 H, NCH ), 3.21 (br, s, 2 H, NCH ), 2.87
3
2
2
(m, 1 H, NC(O)CH), 1.80 (m, 2 H, CHCH
2 2
), 1.51 (m, 2 H, CHCH ), 1.40 (t, J = 7.1 Hz, 3 H,
OCH CH ), 1.33 (q, J = 6.6 Hz, 2 H, NCHCH ), 1.23 (s, 6 H, 2 × CH ), 1.15 (s, 6 H, 2 × CH ),
2
3
2
3
3
1
3
and 1.14 (m, 2 H, NCHCH ). C NMR (150 MHz, CDCl ) δ = 173.7, 173.2, 154.3, 152.6, 148.4,
2
3
1
3
5
44.2, 144.1, 138.1, 123.7, 113.7, 110.7, 106.2, 66.7, 61.1, 59.5, 50.9, 49.7, 45.5, 42.3, 41.6,
+
+
4.7, 33.0, 31.5, 20.6, 14.6, 8.3. HRMS (ESI): m/z calcd for C H FN O + Na [M+Na ]:
3
0
41
4
5
79.2933. Found 579.2933. HPLC analysis: retention time = 3.355 min; peak area, 99.73 %;
-1
eluent A, Methanol; over 25 min with a flow rate of 1 mL min and detected at 254 nm; column
temperature, rt.
4
.9.5. Ethyl 1-cyclopropyl-6-fluoro-7-(4-(1,1,3,3-tetramethylisoindolin-2-yloxyl-5-
carbonyl)piperazin-1-yl)-4-oxo-1,4-dihydroquinoline-3-carboxylate (21
)
Reagents for acid chloride formation: 9 (120 mg, 0.51 mmol, 1 equiv), anhydrous toluene (5
mL) and pyridine (0.1 mL, 1.20 mmol, 2.4 equiv). Reagents for amide coupling: 16 (218 mg,
0
2
.61 mmol, 1.2 equiv), i-Pr NEt (0.3 mL, 1.0 mmol, 2 equiv) and anhydrous dichloromethane (5
mL). Data for 21: light yellow powder (286 mg, 0.50 mmol, 98 %); mp 126-128 ºC. IR (ATR)
-1
1
ν
max (cm ) = 3500-3200 (w, br, N-H, amide) and 1688 (m, C=O, amide). H NMR (600 MHz,
CDCl ) (*note compound is a free-radical, some signals appear broadened and other signals are
3
missing) δ = 8.57 (s, 1 H, NCH=C), 8.09 (d, J = 12.3 Hz, 1 H, Ar-H), 7.34 (br, s, 1 H, Ar-H),
4
.44 (q, J = 6.5 Hz, 2 H, OCH
2 3 2 2
CH ), 4.12 (br, s, 2 H, NCH ), 3.85 (br, s, 2 H, NCH ), 3.47 (s, 1
H, C=CHNCH), 3.39 (br, s, 4 H, 2 × NCH ), 2.87 (m, 1 H, NC(O)CH ), 1.56 (br, s, 12 H, 4 ×
2
3 2 3 2 2
CH ), 1.46 (t, J = 6.8 Hz, 3 H, OCH CH ), 1.38 (br, s, 2 H, NCHCH ), 1.20 (s, 2 H, NCHCH ).
25

ACCEPTED MANUSCRIPT
1
3
C NMR (150 MHz, CDCl
3
) δ = 172.2, 164.9, 153.4, 151.7, 147.5, 143.2, 137.2, 125.4, 122.9,
+
1
12.7, 109.9, 104.4, 60.2, 33.9, 13.7, 7.5. HRMS (ESI): m/z calcd for C H FN O + H
32 36 4 5
+
[
M+H ]: 576.2717. Found 576.2713. HPLC analysis: retention time = 2.867 min; peak area,
-1
9
9.90 %; eluent A, Methanol; over 20 min with a flow rate of 1 mL min and detected at 254
nm; column temperature, rt. EPR: g = 1.9981, a = 1.4793 mT.
N
4
.9.6. Ethyl 1-cyclopropyl-6-fluoro-7-(4-(2-methoxy-1,1,3,3-tetramethylisoindoline-5-
carbonyl)piperazin-1-yl)-4-oxo-1,4-dihydroquinoline-3-carboxylate (22
)
Reagents for acid chloride formation: 9 (90 mg, 0.36 mmol, 1 equiv), anhydrous toluene (5
mL) and pyridine (0.1 mL, 1.20 mmol, 3.3 equiv). Reagents for amide coupling: 16 (155 mg,
0
2
.43 mmol, 1.2 equiv), i-Pr NEt (0.2 mL, 0.7 mmol, 2 equiv) and anhydrous dichloromethane (5
mL). Data for 22: white foamy solid (178 mg, 0.30 mmol, 83 %); mp 124-126 ºC. IR (ATR) νmax
-1
1
(
cm ) = 3500-3200 (w, br, N-H, amide) and 1689 (m, C=O, amide). H NMR (600 MHz,
CDCl ) δ = 8.53 (s, 1 H, NCH=C), 8.05 (d, J = 13.0 Hz, 1 H, Ar-H), 7.31 (dd, J = 7.7, 1.3 Hz, 1
H, Ar-H), 7.28 (d, J = 7.0 Hz, 1 H, Ar-H), 7.20 (s, 1 H, Ar-H), 7.15 (d, J = 7.7 Hz, 1 H, Ar-H),
3
4
2
.39 (q, J = 7.1 Hz, 2 H, OCH
2 3 2 3
CH ), 4.01 (br, s, 2 H, NCH ), 3.79 (s, 3 H, NOCH ), 3.71 (br, s,
H, NCH ), 3.42 (ddd, J = 10.8, 7.1, 3.9 Hz, 1 H, C=CHNCH), 3.30 (br, s, 2 H, NCH ), 3.24
2
2
(
br, s, 2 H, NCH ), 1.69 (s, 6 H, 2 × CH ), 1.45 (br, s, 6 H, 2 × CH ), 1.41 (t, J = 7.1 Hz, 3 H,
2
3
3
1
3
OCH
2
CH
3
), 1.33 (q, J = 6.6 Hz, 2 H, NCHCH
2
2
), 1.15 (q, J = 6.5 Hz, 2 H, NCHCH ). C NMR
(
150 MHz, CDCl ) δ = 173.3, 171.0, 166.6, 154.4, 152.7, 148.4, 147.7, 146.1, 144.3, 138.2,
3
1
8
34.4, 126.4, 123.9, 123.8, 121.9, 120.9, 113.8, 113.7, 110.8, 105.3, 67.3, 65.7, 61.1, 34.7, 14.6,
+
+
.4. HRMS (ESI): m/z calculated for C H FN O + H [M+H ]: 591.2956. Found 591.2955.
3
3
39
4
5
HPLC analysis: retention time = 3.268 min; peak area, 99.61 %; eluent A, Methanol; over 25
-1
min with a flow rate of 1 mL min and detected at 254 nm; column temperature, rt.
4
.10. General procedure for ester hydrolysis (compounds 23-28
)
2
M aqueous sodium hydroxide (7 equiv) was added to a solution of the specific ethyl ester (1
equiv) in HPLC grade methanol and the resulting solution was stirred at 50 ºC for 5 hours. The
reaction mixture was cooled to room temperature and diluted with deionized water (50 mL). The
pH was adjusted to ~6 using 2 M aqueous hydrochloric acid and the mixture extracted with
26

ACCEPTED MANUSCRIPT
dichloromethane (3 × 20 mL). The combined organic extracts were dried over anhydrous sodium
sulfate and the solvent was removed in vacuo to afford the pure solid product.
4
.10.1. 1-Cyclopropyl-6-fluoro-7-(4-(2,2,5,5-tetramethyl-1-oxy-pyrrolidine-3-
carbonyl)piperazin-1-yl)-4-oxo-1,4-dihydroquinoline-3-carboxylic acid (23
)
Reagents: 17 (57 mg, 0.10 mmol, 1 equiv), 2 M aqueous NaOH (0.35 mL, 0.70 mmol, 7 equiv)
and HPLC grade methanol (2.5 mL). Data for 23: pale yellow powder (47 mg, 0.09 mmol, 87
-1
%
); mp 229 ºC decomposes. IR (ATR) νmax (cm ) = 3500-3200 (w, br, N-H, amide), 3100-2500
1
(
w, br, O-H, COOH) and 1649 (m, C=O, amide). H NMR (600 MHz, CDCl ) (*note compound
3
is a free-radical, some signals appear broadened and other signals are missing) δ = 14.87 (s, 1
H, COOH), 8.81 (s, 1 H, NCH=C), 8.10 (d, J = 11.7 Hz, 1 H, Ar-H), 7.41 (br, s 1 H, Ar-H), 3.98
(br, s, 4 H, 2 × NCH
2 2
), 3.57 (br, s, 1 H, C=CHNCH), 3.27 (m, 4 H, 2 x NCH ), 1.44 (m, 2 H,
13
NCHCH ), 1.26 (br, d, J = 17.3, 2 H, NCHCH ). C NMR (150 MHz, CDCl ) δ = 177.1, 166.8,
2
2
3
+
1
47.7, 139.0, 108.4, 105.2, 49.4, 35.3, 8.4. HRMS (ESI): m/z calcd for C H FN O + H
26 32 4 5
+
[
M+H ]: 500.2440. Found 500.2440. HPLC analysis: retention time = 4.392 min; peak area,
9
9.90 %; eluent A, DCM; eluent B, THF; isocratic (70:30) over 20 min with a flow rate of 1 mL
-1
min and detected at 254 nm; column temperature, rt. EPR: g = 1.9992, a
N
= 1.5024 mT.
4
.10.2. 1-Cyclopropyl-6-fluoro-7-(4-(1-methoxy-2,2,5,5-tetramethylpyrrolidine-3-
carbonyl)piperazin-1-yl)-4-oxo-1,4-dihydroquinoline-3-carboxylic acid (24
)
Reagents: 18 (78 mg, 0.15 mmol, 1 equiv), 2 M aqueous NaOH (0.5 mL, 1.05 mmol, 7 equiv)
and HPLC grade methanol (3 mL). Data for 24: pale yellow powder (60 mg, 0.12 mmol, 80 %);
-1
mp 196-198 ºC. IR (ATR) νmax (cm ) = 3500-3200 (w, br, N-H, amide), 3100-2500 (w, br, O-H,
1
COOH) and 1626 (m, C=O, amide). H NMR (600 MHz, CDCl ) δ = 14.90 (s, 1 H, COOH),
3
8
.76 (s, 1 H, NCH=C), 8.03 (d, J = 12.4 Hz, 1 H, Ar-H), 7.37 (d, J = 5.1 Hz, 1 H, Ar-H), 3.92 (s,
2
H, NCH ), 3.86 (s, 2 H, NCH ), 3.62 (s, 3 H, NOCH ), 3.56 (br, s, 1 H, C=CHNCH), 3.38 (s, 2
2
2
3
H, NCH
2
2 2
), 3.31 (s, 2 H, NCH ), 3.11 (br, s, 1 H, NC(O)CH), 2.39 (br, s, 1 H, NC(O)CH ), 1.61
(
dd, J = 16.9, 11.7 Hz, 1 H, NC(O)CHCH ), 1.41 (s, 3 H, CH ), 1.33 (s, 2 H, NCHCH ), 1.27 (s,
2 3 2
1
3
3
H, CH ), 1.26 (s, 3 H, CH ), 1.22 (s, 2 H, NCHCH ), 1.07 (br, s, 3 H, CH ). C NMR (150
3 3 2 3
MHz, CDCl
3
) δ = 177.0, 166.8, 154.9, 152.4, 147.6, 145.5, 145.4, 139.1, 120.2, 120.1, 112.6,
1
12.3, 108.1, 105.2, 65.0, 50.5, 49.6, 46.2, 41.9, 35.5, 29.8, 8.4. HRMS (ESI): m/z calcd for
+
+
27 4 5
C H35FN O + H [M+H ]: 515.2651. Found 515.2651. HPLC analysis: retention time = 3.711
27

ACCEPTED MANUSCRIPT
min; peak area, 99.90 %; eluent A, DCM; eluent B, THF; isocratic (70:30) over 20 min with a
-1
flow rate of 1 mL min and detected at 254 nm; column temperature, rt.
4
1
.10.3. 1-Cyclopropyl-6-fluoro-7-(4-(2,2,6,6-tetramethyl-1-oxy-piperidine-4-carbonyl)piperazin-
-yl)-4-oxo-1,4-dihydroquinoline-3-carboxylic acid (25)
Reagents: 19 (84 mg, 0.16 mmol, 1 equiv), 2 M aqueous NaOH (0.6 mL, 1.12 mmol, 7 equiv)
and HPLC grade methanol (3 mL). Data for 25: orange powder (78 mg, 0.15 mmol, 98 %); mp
-1
2
45 ºC decomposes. IR (ATR) νmax (cm ) = 3500-3200 (w, br, N-H, amide), 3100-2500 (w, br,
1
O-H, COOH) and 1624 (m, C=O, amide). H NMR (600 MHz, CDCl ) (*note compound is a
3
free-radical, some signals appear broadened and other signals are missing) δ = 14.92 (s, 1 H,
COOH), 8.78 (s, 1 H, NCH=C), 8.06 (d, J = 10.9 Hz, 1 H, Ar-H), 7.45 (br, s 1 H, Ar-H), 4.00
(br, s, 4 H, 2 × NCH
2
), 3.63 (s, 2 H, NCH
2 2
), 3.54 (s, 1 H, NCH), 3.44 (s, 2 H, NCH ), 1.49 (br, s,
1
3
2
1
7
H, NCHCH ), and 1.29 (br, s, 2 H, NCHCH ). C NMR (150 MHz, CDCl ) δ = 176.0, 165.8,
2
2
3
53.4, 151.8, 146.6, 144.3, 138.0, 119.3, 111.7, 111.6, 107.2, 104.1, 49.4, 48.5, 44.0, 40.5, 34.6,
+
+
4 5
.6. HRMS (ESI): m/z calcd for C27H34FN O + H [M+H ]: 514.2575. Found 514.2587. HPLC
analysis: retention time = 4.283 min; peak area, 99.90 %; eluent A, DCM; eluent B, THF;
-1
isocratic (70:30) over 20 min with a flow rate of 1 mL min and detected at 254 nm; column
temperature, rt. EPR: g = 1.9995, a = 1.6108 mT.
N
4
.10.4. 1-Cyclopropyl-6-fluoro-7-(4-(1-methoxy-2,2,6,6-tetramethylpiperidine-4-
carbonyl)piperazin-1-yl)-4-oxo-1,4-dihydroquinoline-3-carboxylic acid (26
)
Reagents: 20, (78 mg, 0.14 mmol, 1 equiv), 2 M aqueous NaOH (0.4 mL, 1.00 mmol, 7 equiv)
and HPLC grade methanol (3 mL). Data for 26: White powder (61 mg, 0.12 mmol, 82 %); mp
-1
2
59 ºC decomposes. IR (ATR) νmax (cm ) = 3500-3200 (w, br, N-H, amide), 3100-2500 (w, br,
1
O-H, COOH) and 1627 (m, C=O, amide). H NMR (600 MHz, CDCl
3
) δ = 14.89 (s, 1 H,
COOH), 8.77 (s, 1 H, NCH=C), 8.04 (d, J = 12.9 Hz, 1 H, Ar-H), 7.38 (d, J = 6.9 Hz, 1 H, Ar-
H), 3.87 (br, s, 2 H, NCH ), 3.75 (br, s, 2 H, NCH ) 3.63 (s, 3 H, NOCH ), 3.56 (m, 1 H,
C=CHNCH), 3.40 (br, s, 2 H, NCH ), 3.30 (br, s, 2 H, NCH ), 2.88 (t, J = 12.5 Hz, 1 H,
2
2
3
2
2
NC(O)CH ), 1.82 (s, 2 H, CHCH ), 1.52 (d, J = 12.9 Hz, 2 H, CHCH ), 1.41 (d, J = 6.5 Hz, 2 H,
2
2
13
NCHCH
2
), 1.24 (s, 6 H, 2 x CH
3
), 1.22 (m, 2 H, NCHCH
2
3
), 1.16 (s, 6 H, 2 × CH ). C NMR
(
100 MHz, CDCl ) δ = 177.2, 177.1, 166.9, 155.0, 152.5, 147.7, 145.5, 139.2, 120.4, 112.8,
3
1
12.5, 108.3, 105.3, 65.8, 59.6, 49.5, 45.3, 41.4, 35.5, 31.5, 20.7, 8.4. HRMS (ESI): m/z calcd for
28