A Highly Potent Class of Halogenated Phenazine Antibacterial and Biofilm-Eradicating Agents Accessed Through a Modular Wohl-Aue Synthesis

Abstract

Unlike individual, free-floating planktonic bacteria, biofilms are surface-attached communities of slow- or non-replicating bacteria encased within a protective extracellular polymeric matrix enabling persistent bacterial populations to tolerate high concentrations of antimicrobials. Our current antibacterial arsenal is composed of growth-inhibiting agents that target rapidly-dividing planktonic bacteria but not metabolically dormant biofilm cells. We report the first modular synthesis of a library of 20 halogenated phenazines (HP), utilizing the Wohl-Aue reaction, that targets both planktonic and biofilm cells. New HPs, including 6-substituted analogues, demonstrate potent antibacterial activities against MRSA, MRSE and VRE (MIC = 0.003–0.78 µM). HPs bind metal(II) cations and demonstrate interesting activity profiles when co-treated in a panel of metal(II) cations in MIC assays. HP 1 inhibited RNA and protein biosynthesis while not inhibiting DNA biosynthesis using ³H-radiolabeled precursors in macromolecular synthesis inhibition assays against MRSA. New HPs reported here demonstrate potent eradication activities (MBEC = 0.59–9.38 µM) against MRSA, MRSE and VRE biofilms while showing minimal red blood cell lysis or cytotoxicity against HeLa cells. PEG-carbonate HPs 24 and 25 were found to have potent antibacterial activities with significantly improved water solubility. HP small molecules could have a dramatic impact on persistent, biofilm-associated bacterial infection treatments.

Full text

1
Scientific RepoRts | 7: 2003 | DOI:10.1038/s41598-017-01045-3
www.nature.com/scientificreports
A Highly Potent Class of
Halogenated Phenazine
Antibacterial and Biolm-
Eradicating Agents Accessed
Through a Modular Wohl-Aue
Synthesis
Hongfen Yang1, Yasmeen Abouelhassan1, Gena M. Burch1, Dimitris Kallidas1, Guangtao
Huang2, Hussain Yousaf1, Shouguang Jin2, Hendrik Luesch1 & Robert W. Huigens1
Unlike individual, free-oating planktonic bacteria, biolms are surface-attached communities of
slow- or non-replicating bacteria encased within a protective extracellular polymeric matrix enabling
persistent bacterial populations to tolerate high concentrations of antimicrobials. Our current
antibacterial arsenal is composed of growth-inhibiting agents that target rapidly-dividing planktonic
bacteria but not metabolically dormant biolm cells. We report the rst modular synthesis of a library
of 20 halogenated phenazines (HP), utilizing the Wohl-Aue reaction, that targets both planktonic and
biolm cells. New HPs, including 6-substituted analogues, demonstrate potent antibacterial activities
against MRSA, MRSE and VRE (MIC = 0.003–0.78 µM). HPs bind metal(II) cations and demonstrate
interesting activity proles when co-treated in a panel of metal(II) cations in MIC assays. HP 1 inhibited
RNA and protein biosynthesis while not inhibiting DNA biosynthesis using 3H-radiolabeled precursors in
macromolecular synthesis inhibition assays against MRSA. New HPs reported here demonstrate potent
eradication activities (MBEC = 0.59–9.38 µM) against MRSA, MRSE and VRE biolms while showing
minimal red blood cell lysis or cytotoxicity against HeLa cells. PEG-carbonate HPs 24 and 25 were found
to have potent antibacterial activities with signicantly improved water solubility. HP small molecules
could have a dramatic impact on persistent, biolm-associated bacterial infection treatments.
Pathogenic bacteria present two distinct biomedical challenges, which include (1) antibiotic resistance1–3 and
(2) antibiotic tolerance3–6. Rapidly-dividing, free-oating planktonic bacteria have multiple well-dened mech-
anisms by which they gain, or develop, resistance to antibiotics2, 3. We have learned that planktonic bacteria use
a signaling process known as quorum sensing to communicate and coordinate the simultaneous attachment to a
surface for initial colonization and biolm development (Fig.1)7, 8. Bacterial biolms are surface-attached com-
munities of metabolically dormant (slow- or non-replicating) bacteria encased within a protective extracellular
polymeric matrix of biomolecules that enable them to thrive in hostile environments8–12. Antibiotic tolerance
results from persistent, non-replicating bacteria (i.e., biolms) and are the underlying cause of chronic infections,
which aect 17 million individuals and lead to >500,000 deaths each year13, 14.
Our entire antibiotic arsenal is composed of growth-inhibiting agents that hit bacterial targets critical to
rapidly-dividing bacteria (e.g., cell wall machinery, bacterial ribosomes)3. e large majority of our antibiotic
classes were discovered before 1970, which is ~25 years before bacterial biolms were identied as a biomedical
problem15. Despite the historical signicance of natural products and phenotype screening in antibiotic discovery,
1Department of Medicinal Chemistry, Center for Natural Products Drug Discovery and Development (CNPD3),
College of Pharmacy, University of Florida, 1345 Center Dr., Gainesville, FL, 32610, USA. 2Department of Molecular
Genetics and Microbiology, College of Medicine, University of Florida, 1200 Newell Drive, Gainesville, FL, 32610,
USA. Correspondence and requests for materials should be addressed to R.W.H. (email: rhuigens@cop.u.edu)
Received: 9 February 2017
Accepted: 17 March 2017
Published: xx xx xxxx
OPEN

www.nature.com/scientificreports/
2
Scientific RepoRts | 7: 2003 | DOI:10.1038/s41598-017-01045-3
the paradigm shi in drug discovery to high throughput screening of synthetic libraries in target-based screens
has generated no new chemical entities that have entered the market to date16. To make matters worse, few phar-
maceutical companies continue antibiotic discovery programs17.
Innovative discovery strategies must be implemented to identify compounds that eectively eradicate persis-
tent, antibiotic-tolerant biolms that operate via unconventional, growth-independent mechanisms. Considering
the history of antibiotic discovery, which relies heavily on microbial warfare agents (i.e., penicillin, streptomycin,
vancomycin), it stands to reason that alternative microbial warfare agents exist that include biolm-eradicating
agents we have yet to harness for therapeutic purposes18. Clinically eective biolm-eradicating agents would
signicantly change bacterial treatments and enable the control of persistent biolm infections.
To develop a new anti-biolm strategy, our group has been inspired by the microbial interaction between
Pseudomonas aeruginosa and Staphylococcus aureus in the lungs of young Cystic Fibrosis (CF) patients19. When
CF patients are young, they oen experience S. aureus lung infections; however, as these CF patients age, P. aerug-
inosa co-infects the lung and secretes a series of phenazine antibiotics which help eradicate S. aureus to become
the primary pathogen in the CF patient’s lung. With this microbial interaction in mind, we felt there was a strong
possibility that the eradication of established S. aureus biolms enabled P. aeruginosa to successfully compete
for the CF patient’s lung. We reasoned that phenazine antibiotic metabolites20, 21 may be an ideal starting point
to explore unique antibacterial agents that eradicate bacterial biolms. During our initial studies, we explored
a diverse panel of phenazine small molecules, including several phenazine antibiotics22. From these investiga-
tions, we identied the marine phenazine antibiotic 2-bromo-1-hydroxyphenazine and its brominated analogue,
2,4-dibromo-1-hydroxyphenazine 1, as potent antibacterial agents against S. aureus (MIC = 6.25 and 1.56 µM,
respectively). ese halogenated phenazines demonstrated signicantly higher antibacterial activity compared to
other phenazine antibiotics, incliding pyocyanin.
We have identied a focused series of HP analogues (i.e., 1–3, Fig.1) that potently eradicates planktonic and
biolms of multiple drug-resistant pathogens, including: S. aureus, S. epidermidis and Enterococcus faecium23–25.
Based on our preliminary ndings, HPs bind copper(II) and iron(II) and elicit their potent antibacterial activities
through a metal(II)-dependent mechanism25. Our pursuits have been met with challenges regarding the chemical
synthesis of the HP scaold. Our previous work has enabled the exploration of 7,8-disubstituted HPs along with
2,4-mixed halogenations HPs with the use of o-phenylenediamines building blocks24, 25; however, the availability
of these building blocks does not enable the level of diversity necessary for our campaign. Our goal for this work
is to develop a rapid and convergent synthesis of HP small molecules that enable us to further explore the 6-, 7-,
8- and 9-positions of the HP scaold through the use of diverse aniline building blocks.
Results and Discussion
Chemical Synthesis of New HPs Using the Wohl-Aue Reaction. e Wohl-Aue reaction26, 27 involves
the base-promoted condensation between a nitroarene and an aniline to yield a phenazine. With the potential
to diversify the HP scaold at the 6-, 7-, 8- and 9-positions with an array of substituted aniline building blocks,
we utilized 2-nitroanisole and 4-methyl-2-nitroanisole in Wohl-Aue condensation reactions with a panel of 14
diverse anilines to produce a total of 20 new HPs in three steps enabling rapid and extensive biological evaluations
(Fig.2).
We found the Wohl-Aue reaction to be useful in the convergent synthesis of new 1-methoxyphenazine scaf-
folds; however, we found this reaction to be very low yielding (2–22%, 20 examples; Fig.2). In addition, it is
known that the Wohl-Aue reaction suers from unproductive side reactions, including bond formation between
aniline’s nitrogen and the 4-position of 2-nitroanisole28. We attempted to mitigate this undesired reaction by using
4-methyl-2-nitroanisole; however, reaction yields remained generally low regardless of which 2-nitroanisole
material was used. e 4-methyl-2-nitroanisole condensation reactions with anilines gave a 9% average yield
Figure 1. Illustration of free-swimming planktonic bacteria, persistent surface-attached biolms and
halogenated phenazine small molecules that eectively eradicate both bacterial lifestyles.

www.nature.com/scientificreports/
3
Scientific RepoRts | 7: 2003 | DOI:10.1038/s41598-017-01045-3
in the Wohl-Aue reaction (11 examples), while 2-nitroanisole was condensed with anilines to give phenazine
products in an average of 6% yield (9 examples). Despite these low yields, these reactions were carried out on
scales that produced 40–650 milligrams of diverse 1-methoxyphenazines, which was sucient for the nal two
steps of the synthetic sequence. Aer the Wohl-Aue reaction, 1-methoxyphenazines were subjected to boron
tribromide demethylation to aord 1-hydroxyphenazines (33–100% yield, 88% average yield; 20 examples). Final
bromination with N-bromosuccinimide (NBS) produced 2-bromo-4-methyl-HP analogues 4–14 (39–99% yield,
63% average yield, 11 examples) or 2,4-dibromo-HP analogues 15–23 (30–81% yield, 50% average yield, 9 exam-
ples). During our investigations, we were able to prepare 12–220 milligrams of each HP, which enabled sucient
material for biological studies.
Biological Evaluation of HPs. Antibacterial Studies. Following the synthesis of 4–23, these new HPs were
evaluated for their antibacterial activity against methicillin-resistant S. aureus (MRSA), methicillin-resistant S.
epidermidis (MRSE), vancomycin-resistant E. faecium (VRE) and Mycobacterium tuberculosis (MtB). HPs 1–3
(Fig.1) are from our previous studies and were used as positive-controls in our MIC assays. In addition, a panel
of conventional antibiotics (vancomycin29), metal-binding agents (EDTA30, TPEN31, 32) and membrane-active
biolm eradicator (QAC-1033) were used as comparators (Table1).
Our Wohl-Aue derived HPs were tested in MIC assays against MRSA BAA-1707. We found that the
4-methyl-HP analogues 4–14 (MIC = 1.17–18.8 µM; 8 was inactive, MIC > 100 µM; Table1) generally displayed
signicantly reduced antibacterial activities compared to the 2,4-dibromo-HP series 15–23 (MIC = 0.05–3.13 µM;
16 had an MIC = 9.38 µM; 15 was inactive, MIC > 50 µM). To our surprise, 2,4-dibromo-HP analogues 17, 18,
21, 23 are substituted at the 6-position with either a methyl-, ethyl-, chloro- and bromo- groups and demon-
strated potent antibacterial activities against MRSA BAA-1707 (MIC = 0.05–0.30 µM), which is 4- to 16-fold
more potent than parent HP 1 (Fig.3). We have been unable to explore HP analogues containing 6-position
substitutions in previous studies; however, by using the Wohl-Aue reaction described herein, diverse 2-alkylated
Figure 2. Chemical synthesis of a diverse panel of HPs using a convergent Wohl-Aue route. Note:
1,6-Dimethoxylphenazine (Wohl-Aue product) was dibrominated (NBS), then demethylated (BBr3) before nal
dibromination (NBS) to give HP 15. See Supporting Information for full synthesis details.

www.nature.com/scientificreports/
4
Scientific RepoRts | 7: 2003 | DOI:10.1038/s41598-017-01045-3
Compound MRSA
BAA-1707 MRSA
BAA-44 MRSA-1 MRSA-2 MRSE
35984 S. epi
12228 VRE
700221 MtB
H37Ra HeLa Cytotox.
IC50
1 1.17a1.56 0.78 2.35a2.35a2.35a4.69a25 ~100
2 0.78 2.35a0.78 3.13 0.59a1.56 0.78 — —
3 0.20 0.78 2.35a0.78 3.13 3.13 0.78 — —
4 3.13 50 50 3.13 2.35a3.13 3.13 — >100
5 9.38a— 37.5a— 1.17a—>100 — —
6 4.69a— 0.78 — 0.30a— 0.78 — —
7 4.69a— 50 — 18.8a— 18.8a— —
8>100 — >100 — >100 — >100 — —
9 1.17a1.56 1.56 2.35a0.39 2.35a2.35a50 >100
10 18.8a— 50 — 37.5a— 50 — —
11 2.35a3.13 0.39 3.13 1.56 1.56 3.13 — —
12 18.8a— 75a— 9.38a— 12.5 — —
13 1.17a1.17a0.39 2.35a0.78 1.17a1.17a— —
14 9.38a— 75a— 75a— 18.8a— —
15 >50 — >50 >50 >50 — >50 — —
16 9.38a18.8a25 18.8a9.38a18.8a2.35a— —
17 0.30a0.39 0.10b0.30a0.30a0.39 0.59a25 >100
18 0.10b0.59a0.15a0.39 0.10b0.30a0.15a25 >100
19 0.15a9.38a4.69a4.69a0.30a6.25 0.15a—>100
20 2.35a3.13 18.8a2.35a6.25 9.38a2.35a— —
21 0.08a0.59a0.10b0.39 0.30a0.39 1.56 50 >100
22 0.15a0.30a0.15a0.59a18.8a12.5 1.56 6.25 >100
23 0.05 0.59a0.10b0.30a0.78 0.30a1.17a25 >100
24 0.59a1.56 1.17a2.35a1.56 0.78 6.25 25 >100
25 0.003 0.013 0.10b0.10b0.15a0.10b0.59a12.5 >100
EDTA 25 — 12.5 — — — — — —
TPEN 46.9a— 75a50 12.5 — — — —
QAC-10 4.69a— 4.69a3.13 2.35a— 2.35a— —
BAC-16 1.56 — — — 1.56 — 3.13 — —
vancomycin 0.39 0.39 0.39 0.59a0.78 1.17a>100 — —
daptomycin 3.13 18.8a— 4.69a12.5 6.25 125 — —
linezolid 12.5 1.56 4.69a3.13 3.13 1.56 3.13 — —
streptomycin — — — — — — — 1.32 —
Table 1. Summary of antibacterial activities and cytotoxicity against HeLa cells for HP analogues. All values are
reported in micromolar (µM) concentrations. Note: aMidpoint for a 2-fold range in observed values. bLowest
concentration tested. All biological results were generated from three independent experiments.
Figure 3. MIC assay results of potent new HPs (MIC = 0.10 µM) against MRSA BAA-1707.

www.nature.com/scientificreports/
5
Scientific RepoRts | 7: 2003 | DOI:10.1038/s41598-017-01045-3
anilines yield 6-substituted HPs. In addition, HPs 17, 18, 21, 23 along with 8-chloro HP 22 proved to be highly
potent (MIC = 0.05–0.59 µM) against a panel of four MRSA isolates (Table1).
In addition to MRSA isolates, several Wohl-Aue derived HPs demonstrated potent antibacterial activities
against MRSE and VRE strains (Table1). 2-Bromo-4-methyl HP analogues 6, 9 and 13 displayed impressive
antibacterial activities against MRSE 35984 (MIC = 0.30–0.78 µM) and VRE 700221 (MIC = 0.78–2.35 µM)
while 2,4-dibrominated HP analogues 17, 18, 19, 21 and 23 demonstrated more potent antibacterial activi-
ties against MRSE 35984 (MIC = 0.10–0.78 µM) and VRE 700221 (MIC = 0.15–1.56 µM). HPs 18 (6-ethyl HP)
demonstrated the most potent antibacterial activities against S. epidermidis (MIC = 0.10–0.30 µM) and E. faecium
(MIC = 0.15 µM) while 17 (6-methyl HP) gave a similar prole with a slight reduction in antibacterial potency
(S. epidermidis MIC = 0.30–0.39 µM; E. faecium MIC = 0.59 µM). Each of the HPs discussed here, including the
4-methyl HP analogues, demonstrated improved antibacterial activities against MRSE and VRE compared to
parent HP 1 (MRSE MIC = 2.35 µM; VRE MIC = 4.69 µM).
M. tuberculosis (MtB) is the largest bacterial threat to humans worldwide killing over 1.5 million humans
around the globe each year34, 35. MtB is a slow-growing or non-replicating pathogen that requires prolonged antibi-
otic treatments (i.e., ≥6 months), oen with multiple drug combinations due to problems with drug-resistance36–38.
To further compound these problems, currently there is an inadequate antibiotic pipeline for new MtB drugs37.
With active HPs targeting both rapidly-dividing planktonic and persistent biolm cells of Gram-positive path-
ogens, we felt that HPs could demonstrate antibacterial activities against the persistent pathogen, MtB. We
previously reported HPs with good anti-TB activity24, 25 and, during the course of these studies, we evaluated
Wohl-Aue derived HPs for antibacterial activity against MtB H37Ra. We tested a smaller panel of HPs (9, 17, 18,
21–23; Table1) that demonstrated potent antibacterial activities against MRSA, MRSE, and VRE. From these
studies, most HPs reported moderate antibacterial activities against MtB H37Ra (MIC = 12.5–50 µM); however,
8-chloroHP analogue 22 (MIC = 6.25 µM, 2.4 µg/mL) demonstrated the most potent antibacterial activity from
this series against MtB H37Ra. Unlike our previous investigations that reported lead anti-TB HPs not containing
a bromine atom in the 2-position, HP 22 is the most potent 2,4-dibromoHP analogue we have reported to date.
Macromolecular Synthesis Inhibition of HP 1 against MRSA BAA-1707 Cultures. To characterize the antibacterial
mechanism of action for HP small molecules, we investigated the eects HP 1 has against MRSA BAA-1707 on
global biosynthetic pathways in rapidly-dividing (exponentially-growing) planktonic cells through quantifying
the incorporation of various [3H]-labeled precursors into their corresponding macromolecules39. For these exper-
iments, MRSA-1707 cultures were treated with [3H]-thymidine (DNA biosynthesis), [3H]-uridine (RNA biosyn-
thesis) and [3H]-leucine (protein biosynthesis) in the presence of HP 1 and antibiotic controls (4–16 x MIC).
From these initial experiments, we found that HP 1 does not inhibit DNA biosynthesis at 8 x MIC; however, at 4
x MIC, HP 1 inhibits both RNA and protein biosynthesis in MRSA-1707 (Fig.4). Currently, more elaborate mode
of action studies are underway in our labs and aim to understand these results in the context of metal(II)-binding
and biolm eradication.
Mammalian Cytotoxicity Studies. Several new HPs (4, 9, 17–19, 21–23) that displayed potent antibacterial activ-
ities against MRSA, MRSE and VRE were evaluated against HeLa cells to determine mammalian cell cytotoxicity
in lactate dehydrogenase (LDH) release assays40 at 25, 50 and 100 µM. During these investigations, we found
parent HP 1 to have an IC50 value ~100 µM while the remaining HP analogues recorded IC50 values >100 µM
Figure 4. Results of HP 1 against MRSA BAA-1707 in macromolecular synthesis inhibition experiments
with [3H]-labeled precursors (generated from three, or more, independent experiments). HP 1 inhibits RNA
and protein biosynthesis while not inhibiting DNA biosynthesis (p-values ≤ 0.005*; pairwise student t-test
comparing relative treated samples to DMSO vehicle only samples).

www.nature.com/scientificreports/
6
Scientific RepoRts | 7: 2003 | DOI:10.1038/s41598-017-01045-3
(Table1). When comparing the HeLa cytotoxicity (IC50) to the antibacterial activities (MIC) of MRSA BAA-1707,
several HPs demonstrated selectivity indexes of >330- to >2,000-fold against MRSA BAA-1707 cells. For exam-
ple, HP 23 reported an IC50 > 100 µM against HeLa cells while reporting potent antibacterial activities with an
MIC = 0.05 µM against MRSA BAA-1707, resulting in a selective index of >2,000-fold towards inhibiting MRSA
planktonic cells. is promising bacterial selectivity prole is crucial for translating HP antibacterial agents into
viable therapeutic options.
PEG Carbonate-HP Synthesis. In an eort to enhance the drug-likeness of our active HP small molecules, we
synthesized two polyethylene glycol (PEG) carbonates as potential prodrugs, which enables: (1) improvement
of water solubility through the installation of a PEG group, (2) enhanced bacterial penetration and release of
HP through possible bacterial esterase processing resulting in active HP, carbon dioxide and non-toxic PEG, (3)
mitigating the metal-binding moiety required for the antibacterial activities of HPs, which would be important in
the development of HPs in more advanced preclinical studies. We selected parent HP 1 (CLogP: 4.68) and active
HP 17 (CLogP: 5.18) to design PEG carbonate-HPs that would have improved water solubility (reduced CLogP
values; calculated using ChemBioDraw Ultra, version 13).
We synthesized PEG carbonate-HPs 24 and 25 through the initial condensation of triphosgene with tetrae-
thyleneglycol monomethyl ether to generate a PEGylated chloroformate intermediate (not shown), which was
immediately condensed with 1 to give 24 in 80% yield and 17 to give 25 in 96% yield (Fig.5A). Both PEG
carbonate-HPs have reduced CLogP values with 24 having a CLogP of 3.56 while 25 has a CLogP of 4.06. ese
carbonates do not directly bind metal(II) cations in UV-Vis experiments (see next section) and are chemically
stable in aqueous formulations for >1 month at room temperature. PEG carbonate-HPs 24 and 25 have main-
tained or enhanced antibacterial and biolm eradication activities compared to their phenolic precursors 1 and
17 suggesting improved bacterial cell entry and possible involvement of bacterial esterase enzymes for cleavage
of the carbonate group to deliver the active HP once inside bacteria. To support this model, 25 reported MICs
of 0.003 and 0.013 µM against MRSA-1707 and MRSA-44 while corresponding HP 17 reported MICs of 0.30
and 0.39 against the same strains correlating to an impressive 100- and 30-fold enhancement of antibacterial
activities for HP carbonate 25 compared to (non-carbonate) 17 (Fig.5B). A further discussion can be found in
the structure-activity relationships section. Similar to non-carbonate HPs, 24 and 25 showed no HeLa cell cyto-
toxicity at 100 µM.
Metal(II) Binding (UV-Vis) and Co-Treatment in Antibacterial Assays with MRSA-1707. We previously reported
that HPs directly binds copper(II) and iron(II) cations in UV-Vis experiments and loses antibacterial activities
when co-treated with these metal(II) cations in antibacterial assays against MRSA-2 (clinical isolate; Shands
Hospital; Gainesville, FL)25. HPs bind metal(II) cations through a chelation event involving the oxygen atom
of the 1-hydroxyl group and the adjacent nitrogen at the 10-position of the HP scaold to form a stable 2:1
HP:metal(II) complex. During these investigations, we evaluated the ability of HPs 1, 17 and 21 to directly bind
copper(II), iron(II), zinc(II) and magnesium(II) in UV-Vis experiments and found these HPs to bind copper(II),
iron(II) and zinc(II), but not magnesium(II) cations. Based on the kinetics of UV-Vis experiments, HP 17
Figure 5. (A) Chemical synthesis of PEG-carbonate HPs 24 and 25. (B) Antibacterial assay of PEG-carbonate
HP 25 alongside non-carbonate version HP 17 demonstrating enhanced antibacterial activities against MRSA.

www.nature.com/scientificreports/
7
Scientific RepoRts | 7: 2003 | DOI:10.1038/s41598-017-01045-3
(6-methyl HP) and 21 (6-chloro HP) chelated copper(II) at a faster rate than HP 1 (Fig.6), which could explain
the enhanced antibacterial activities of 17 and 21 compared to HP 1. Interestingly, HP 25 does not bind any of the
metal(II) cations directly as the carbonate functionality blocks direct metal-chelation.
We found that co-treatment of HPs with metal(II) cations at 200 µM in MIC assays against MRSA BAA-
1707 led to reduced antibacterial activities (i.e., 4- to >20,000-fold elevated MIC values with copper(II) cat-
ion co-treatment; Table2) for 1, 17 and 25 with copper(II) and iron(II); however, it is interesting to note that
co-treatment with zinc(II) increased antibacterial activities (reduced MICs; Table2) of HPs and no changes were
observed in the antibacterial activities when HPs were co-treated with magnesium(II). HP 21 showed reduced
antibacterial activities against MRSA BAA-1707 in the presence of copper(II) and iron(II); however, no changes
in antibacterial activities were observed when 21 was co-treated with zinc(II) or magnesium(II). PEG-carbonate
HP 25 gives the most dramatic antibacterial response to co-treatment with metal(II) cations (>20,000-fold
elevated MIC against MRSA with copper(II) and 67-fold reduction in MIC against MRSA with zinc(II), yet
does not directly bind either metal(II) cation directly) further supporting that PEG-carbonate 25 enters MRSA
Figure 6. UV-Vis analysis of copper(II) binding various halogenated phenazines. e HP:copper(II) complex is
insoluble and precipitates out of solution, thus the disappearance of HP peaks is clear while there is not a strong
appearance of HP:copper(II) complex in the UV-Vis spectrum.
MRSA BAA-1707 (concentration in µM)
Test
Cpd. MIC
Copper(II) Iron(II) Zinc(II) Magnesium(II)
Binding
Y/N MIC w/
Cu2+Binding
Y/N MIC w/
Fe2+Binding
Y/N MIC w/
Zn2+Binding
Y/N MIC w/
Mg2+
1 1.17aY 4.69aY 2.35aY 0.39 N 1.17a
17 0.30aY 75aY 12.5 Y 0.05 N 0.30a
21 0.10bY 6.25 Y 0.59aY 0.15aN 0.15a
25 0.0047aN>100 N 25 N 0.00007aN 0.0047a
8-OHQ 9.38 Y 4.69aY 12.5 Y 12.5 N 6.25
1-OHP 250 — 75 — 250 — >500 — 250
TPEN 46.9a— 500 — 125 — 250 — 62.5
EDTA 25 — 50 — 50 — 50 — 50
Doxy. 0.78 — 0.78 — 3.13 — 0.39 — 0.78
Table 2. UV-Vis and metal(II)-cation co-treatment MIC assays summary for HPs against MRSA BAA-1707.
Note: aMidpoint for a 2-fold range in observed values. bMIC for 21 in these experiments was 0.1 µM. UV-Vis
results are reported as “Binding Y/N” for HPs. Each metal(II) cation was tested at 200 µM in co-treatment MIC
assays. 8-OHQ (8-hydroxyquinoline) and 1-OHP (1-hydroxyphenazine) were used as positive controls and have
metal-binding moieties related to HPs.

www.nature.com/scientificreports/
8
Scientific RepoRts | 7: 2003 | DOI:10.1038/s41598-017-01045-3
at a high eciency, then is converted to the active, metal-chelating HP 17 which elicits a potent antibacterial
response. Interestingly, structurally related non-halogenated comparators, 1-hydroxyphenazine (1-OHP) and
8-hydroxyquinoline (8-OHQ), demonstrated drastically dierent metal(II) cation proles compared to HPs 1,
17, 21 and 25.
e metal-chelating agents EDTA and TPEN, a membrane-permeable metal-chelating agent, were used as
comparators in metal(II) cation co-treatment assays. EDTA showed only slight reductions in antibacterial assays
(MIC values elevated 2-fold) when co-treated with each metal(II) cation. TPEN showed more dramatic reduc-
tions in antibacterial activities (MIC values elevated 3- to 11-fold) with copper(II), iron(II) and zinc(II); however,
there was not a signicant change in MIC values with TPEN was co-treated with magnesium(II). Doxycycline, a
tetracycline antibiotic that chelates a magnesium(II) ion in the bacterial ribosome during protein synthesis inhi-
bition41, was also tested in these metal(II) cation co-treatment assays against MRSA BAA-1707 and only iron(II)
and zinc(II) modulated antibacterial activities with 4-fold elevated MIC values and 2-fold reduced MIC values,
respectively. ese studies demonstrate unique antibacterial proles for HP small molecules with the enhance-
ment of zinc(II) cations being of particular interest. Future studies are aimed to further understand these results
in the context of biolm viability.
Biolm Eradication Studies. From our previous studies, we have found that biolm eradication activities cor-
relate well to antibacterial activities for HP small molecules24, 25. With this in mind, we advanced a focused set of
Wohl-Aue derived HP analogues to biolm eradication assays using Calgary Biolm Devices (CBD)42, 43 contain-
ing specialized 96-well plates with a lid containing 96 pegs anchored to the lid to provide a surface for biolms to
form and be treated (one peg per microtiter well). Unlike MIC assays which are used to determine the inhibition
of rapidly-dividing planktonic bacteria, CDB assays have three phases to test for biolm eradication, including:
(1) biolm-attachment/establishment to the CBD peg surface (24 hours, media and bacteria only), (2) treating
established biolms on CBD pegs with test compounds (24 hours, media and test compounds only) and (3) recov-
ery, growth, dispersion and planktonic proliferation of viable biolms (24 hours, media only). At the end of bio-
lm eradication assays, turbid microtiter wells in 96-well plates result from pegs that have viable biolms whereas
microtiter wells that results in non-turbid microtiter wells result from pegs containing eradicated biolms. Using
the CBD assay is operationally simple as lid pegs that have attached biolms are rapidly washed and transferred
to new 96-well plates (containing media, with or without test compound) as one moves through each of the
three assay phases. Upon completion of biolm eradication assays, minimum biolm eradication concentration
(MBEC) values are determined as the lowest concentration at which biolms are completely eradicated. Using
the CBD, planktonic-killing activities can also be determined and is useful in our investigations as minimum
bactericidal concentrations (MBC) enable us to assess planktonic-biolm killing dynamics in the same assay
using a single culture. We have found that CBD assays provide superior insights into biolm-eradicating agents
compared to obtaining MIC values and MBEC values using dierent assays.
During these investigations, we evaluated 10 new HP analogues (4, 9, 17–19, 21–25) in biolm eradica-
tion assays against MRSA BAA-1707 and found ve HPs (17–19, 22, 23, 25) to demonstrate potent minimum
biolm eradication concentrations (MBEC) between 4.69 and 37.5 µM (Table3). Halogenated Phenazines 18
(MBEC = 4.69 µM), 17 (MBEC = 6.25 µM), 19 (MBEC = 9.38 µM), 23 (MBEC = 9.38 µM) and PEG-carbonate
25 (MBEC = 9.38 µM) proved to be the most potent MRSA biolm-eradicating HPs from this series (Fig.7A).
Similar to our antibacterial ndings from MIC analysis, HP analogues 17–19 and 23 possess either a methyl
group, ethyl group or bromine atom at the 6-position of the HP scaold and performed with high potency in
biolm eradication assays.
In addition to HPs, we assayed various comparators in biolm eradication assays against MRSA BAA-1707
(Table3), including membrane-lysing agents (quaternary ammonium cation-10, QAC-10, a reported biolm
eradicating agent33; BAC-16), metal-chelating agents (EDTA30, TPEN31, 32) and conventional antibiotics used in
MRSA treatments (i.e., vancomycin, linezolid, daptomycin)44, 45. During these studies, our lead HPs 17–19 were
found to be 10- to 21-fold more potent than QAC-10 (MBEC = 93.8 µM) against MRSA BAA-1707 biolms.
Interestingly, the metal-chelating agents EDTA and TPEN were both found to be inactive (MBEC > 2,000 µM;
Fig.6A) in biolm eradication assays against MRSA BAA-1707 when tested alongside HP small molecules. We
found this result particularly interesting and we feel the lack of biolm eradication activity by EDTA and TPEN
speaks to the unique mechanism that leads to HP-induced biolm killing as HP 18 proves to be >425-fold more
potent than EDTA or TPEN in CBD assays. Vancomycin, daptomycin and linezolid are front-running MRSA
therapies44, 45 and are unable to eradicate MRSA BAA-1707 biolms (MBEC > 2,000 µM) despite being active
against planktonic cells in CBD assays (i.e., vancomycin, MBC = 3.9 µM). ese ndings are illustrative of the
abilities of MRSA biolms to tolerate and thrive in the presence of high concentrations of current antibiotic
therapies.
We previously demonstrated one HP analogue slowly kills MRSA persister cells in kinetic kill experiments of
stationary cultures24, which are known to contain higher populations of non-replicating persister cells46, 47. We
were curious to see if the biolm eradication activities of HPs would show a more potent eect if MRSA biolms
were pulse treated with active HPs (i.e., two subsequent 24 hour compound treatment phases). We tested HPs
17 and 25 alongside BAC-16, a membrane-lysing comparator, in pulse experiments where aer the 24-hour
compound treatment phase, CBD lids were transferred to a second 96-well plate (extending phase 2 of the bio-
lm eradication assay) with the same test compound concentrations to treat MRSA biolms for an additional
24 hours before transferring to the recovery plate (phase 3). e results from the pulse biolm eradication exper-
iments showed only slight improvements in biolm eradication activities for HP 17 (MBEC = 4.69 µM) while
HP 25 reported the same potency as in standard CBD assay conditions (MBEC = 9.38 µM) against MRSA-1707.
BAC-16 reported an MBEC of 25 µM in pulse experiments against MRSA-1707. Following the completion of the
pulse experiments, we removed treated and untreated pegs from the CBD to determine viable biolm cell counts

www.nature.com/scientificreports/
9
Scientific RepoRts | 7: 2003 | DOI:10.1038/s41598-017-01045-3
following sonication and plating to generate the dose-response curve in Fig.7B. At the 6.25 µM, we found that 17
and 25 reduced viable MRSA-1707 biolm cells by 3-logs (99.9% biolm eradication) whereas at 25 µM these HPs
reduced viable MRSA-1707 biolm cells by >4-logs (>99.99% biolm eradication).
In addition to the CBD assays, we tested HP 17 against MRSA-1707 biolms in live/dead staining experi-
ments. Following a 24-hour MRSA-1707 biolm establishment phase, HP 17 was added at 0.1, 1 and 5 µM and
then allowed to incubate at 37 °C for 24 hours. Aer this time, images of the treated and untreated MRSA-1707
biolms were obtained using uorescence microscopy (Fig.7C). As expected, HP 17 demonstrated a potent
biolm eradication and clearance response towards MRSA-1707 biolms, even at the lowest concentration tested
of 0.1 µM.
Our panel of 10 new HP small molecules also demonstrated impressive biolm eradication activities against
methicillin-resistant S. epidermidis (MRSE ATCC 35984) and vancomycin-resistant Enterococcus faecium (VRE
ATCC 700221) biolms in CBD assays. HP 18 the most potent biolm eradication activities against MRSE bio-
lms (MBEC = 2.35 µM) and VRE biolms (MBEC = 0.59 µM). Similar biolm eradication proles were observed
for HPs 17 and 19 as both HPs reported MBECs = 4.69 µM against MRSE and MBECs = 0.59 µM against VRE
(Table3). In addition, we found HPs to be more active against VRE biolms (MBEC = 0.59–9.38 µM) than MRSE
biolms (MBEC = 2.35–75 µM), and similar to our ndings against MRSA-1707 biolms, HPs out-performed
all conventional antibiotics (e.g. vancomycin), metal-chelating agents (e.g., EDTA) and membrane-lysing agents
(e.g., QAC-10, BAC-16). e remarkable biolm eradication activity possessed by these new HPs could lead to
ground-breaking advances in treating persistent, biolm-associated infections.
Red Blood Cell Hemolysis. Following biolm eradication studies, we assayed our HPs for hemolysis activity
against red blood cells (RBCs) at a single high concentration (200 µM). Antimicrobial peptides and mimics
thereof (i.e., quaternary ammonium cationic compounds) eradicate biolm cells through membrane lysis; how-
ever, identifying potent membrane-lysing agents that target bacterial membranes over mammalian membranes
has proven challenging33, 48, 49. Our Wohl-Aue derived HPs demonstrated little, if any, hemolytic activity against
RBCs at 200 µM (≤1 to ≤7% hemolysis, Table3). HP 18, the most potent biolm-eradicating agent from this
collection, demonstrated 1.7% hemolysis at 200 µM while completely eradicating biolms at 85- to 339-fold lower
concentrations (MBEC = 0.59–2.35 µM) compared to membrane-lysing QAC-10 (MBEC = 93.8 µM) that demon-
strated >99% lysis of RBCs at 200 µM when tested alongside HPs. e potent biolm eradication activities of HPs
together with the lack of RBC lysis and HeLa cytotoxicity demonstrates a unique targeting ability of these HP
small molecules which could lead to signicant breakthroughs in the treatment of persistent, biolm-associated
bacterial infections in the clinic.
Structure-Activity Relationships. Based on our previous studies, we have found that the HP scaold 1
requires the 1-hydroxyl group and 2-bromine atom to demonstrate antibacterial activities22, 25. In addition, the
4-position is highly active when containing a bromine atom (i.e., HP 1), but can also tolerate a butyl group (i.e.,
HP 2)25 which motivated these investigations as Wohl-Aue reactions could provide synthetic avenues to both
4-brominated and 4-methylated HPs. During these studies, we found the 4-bromo HPs to be signicantly more
Compound
MRSA
BAA-1707
MIC MRSA BAA-1707
MBC/MBEC
MRSE
35984
MIC MRSE 35984
MBC/MBEC
VRE
700221
MIC VRE 700221
MBC/MBEC Hemolysis at
200 µM (%)
1 1.17a50b/75a2.35a23.5a/250 4.69a18.8a/9.38a≤1
4 3.13 46.9a/93.8a2.35a150a/>200 3.13 12.5b/4.69a≤1
9 1.17a>200/>200 0.39 12.5b/>200 2.35a12.5/9.38a≤1
17 0.30a6.25/6.25 0.30a9.38a/4.69a0.59a1.56b/0.59a6.7
18 0.10c6.25/4.69a0.10c9.38a/2.35a0.15a3.13b/0.59a1.7
19 0.15a4.69a/9.38a0.30a6.25/4.69a0.15a1.56b/0.59b5.3
21 0.08a6.25/75a0.30a18.8a/75a1.56 25/25 ≤1
22 0.15a18.8a/37.5a18.8a50/37.5a1.56 9.38a/2.35a≤1
23 0.05 4.69a/9.38a0.78 4.69a/12.5 1.17a9.38a/9.38a5.1
24 0.59a75a/100 1.56 75b/100 6.25 9.38a/4.69a2.4
25 0.003 9.38a/9.38a0.15a9.38a/4.69a0.59a1.56b/0.59a≤1
QAC-10 4.69a93.8a/93.8a2.35a31.3/31.3 2.35a3.0a/3.0a>99
BAC-16 1.56 31.3b/15.6 1.56 — 3.13 5.9a/3.0a—
TPEN 46.9a375a/>2000 12.5 250/>2000 — — ≤1
EDTA 25 >2000/>2000 — 1000/>2000 — >2000/93.8a≤1
Vancomycin 0.39 3.9/>2000 0.78 3.0b/>2000 >100 >200/150a—
Daptomycin 3.13 125/>2000 12.5 — 125 375/93.8a—
Linezolid 12.5 31.3/>2000 3.13 — 3.13 4.69b/1.56 —
Table 3. Summary of biolm eradication and hemolysis studies with Halogenated Phenazines. Notes:
aMidpoint for a 2-fold range in observed values. bMidpoint value of a 4-fold range in values. cLowest
concentration tested. All biological results summarized in this table were generated from three independent
experiments.

www.nature.com/scientificreports/
10
Scientific RepoRts | 7: 2003 | DOI:10.1038/s41598-017-01045-3
active than 4-methyl HP analogues, which could result from enhanced bacterial membrane permeability or phe-
nolic acidity, which may be critical for metal-chelation. is SAR prole is observed against both planktonic and
biolm cells. To illustrate, the 4-bromo HP analogue, 4-bromo-6,7-dimethyl HP 19 demonstrates 12-fold more
potent antibacterial activities and >21-fold more potent biolm eradication activities against MRSA-1707 com-
pared to it’s 4-methyl HP analogue, 4,6,7-trimethyl HP 7 (Fig.8).
is study is the rst to report HP analogues with substitutions at the 6-position, which proved to signicantly
enhance antibacterial activities of the HP scaold as 6-methyl (17), 6,7-dimethyl (19), 6-ethyl (18), 6-chloro (21)
and 6-bromo (23) all demonstrated potent antibacterial and biolm eradication activities compared to HP 1 (Fig.7).
In addition, we previously reported potent antibacterial activities of 7,8-disubstituted HP analogues24, 25 and here,
we report our investigations of three new 8-monosubstutited HP analogues, including: 8-phenoxyether HP (16),
8-methyl HP (20) and 8-chloro HP (22). Both 8-phenoxyether HP (16) and 8-methyl HP demonstrated a 2- to
24-fold loss in antibacterial activities against staphylococcal pathogens (MRSA, MRSE) compared to HP 1; how-
ever, both 16 and 18 reported a 2-fold increase in antibacterial activities against VRE. Interestingly, 8-chloro HP 22
demonstrated a 5- to 8-fold increase in antibacterial activities against four MRSA isolates (Table1) while losing 5- to
8-fold in antibacterial activities against both S. epidermidis strains compared to HP 1. In addition, HP 22 demon-
strated improved biolm eradication activities against MRSA, MRSE and VRE biolms compared to HP 1 and equi-
potent activities compared to the 7,8-dichloro HP analogue (not shown), which we previously described; however,
8-chloro HP 22 was less potent than the 6-substituted series of HP analogues. Similar to the antibacterial proles of
HP 16 and 18, HP 22 showed an increased potency against VRE (4-fold) compared to HP 1.
Understanding the signicance of each of the functional group substitutions and combinations of multiple
substitutions on positions 6–9 of the HP scaold will require additional investigations, which are underway in
our labs. However, substitutions in these positions clearly control antibacterial activities and one can propose
that these compounds are targeting one or more bacterial metalloproteins that are critical to biolm viability.
We believe that metalloprotein targeting is a viable notion as our HP small molecules generate dierent activity
proles when treated with our panel of metal(II) cations compared to EDTA and TPEN, which are known to
be involved in general metal(II) sequestration. Following this line of thought, the 6-position of the HP scaold
would likely not interfere with a HP-metal binding event in a metalloprotein due to it’s distal location in relation
to the metal-binding moiety of the HP scaold. e observed improvement in antibacterial potency of 6-position
HP analogues could be due to more rapid metal-binding kinetics, which we demonstrated during these inves-
tigations (Fig.6). Similar activities may be explained by substitution at the 7-position and 8-halogenated HPs;
however, 8-methyl HP 23 and 8-phenoxy HP 15 lose activity compared to unsubstituted HP 1, potentially due
to disfavored interactions in their respective target(s) or decreased bacterial membrane penetration. It is also
possible that the halogens in the 6–8 positions of the HP scaold enhance bacterial membrane permeability
Figure 7. (A) Calgary biolm device (CBD) assay of a panel of HPs, TPEN and vancomycin against MRSA-
1707 demonstrating the potent and unique biolm eradication activities of HP small molecules. (B) Dose-
response curve of biolm eradication for HP 17 (MBEC = 4.69 µM), HP 25 (MBEC = 9.38 µM) and BAC-16
(MBEC = 25 µM) against MRSA-1707 (pulse treatment). (C) Live/Dead staining (uorescence images) of
established MRSA-1707 biolms treated with HP 17 aer 24 hours.

www.nature.com/scientificreports/
11
Scientific RepoRts | 7: 2003 | DOI:10.1038/s41598-017-01045-3
compared to alkyl or ether groups, leading to more ecient bacteria entry and metalloprotein targeting. e only
9-substituted HP analogue synthesized during these studies was HP 15, which proved to be inactive against all
strains tested against (MIC > 50 µM). Additional 9-position analogues are needed to conrm this initial result;
however, this position may impede HP-target interactions at a metal(II) center in a metalloprotein. is collective
series of HPs enable us to begin outlining proposals regarding potential bacterial targets and key interactions,
which will guide future developments as we advance our mechanistic studies (probe design) and pre-clinical
evaluations (therapeutic leads). Ultimately, these analogues have enabled us to ask interesting questions about
biolm killing that we are set to answer in future studies.
e synthetic route developed here does not enable for an ideal synthesis of 7- or 9-monosubstituted HP analogues
due to regioisomers that would result from condensing 3-substituted anilines with 2-nitroanisole in the Wohl-Aue reac-
tion. Currently, we are developing synthetic routes to explore such analogues and will report our ndings accordingly.
In addition to our signicant ndings regarding the 6–8 positions of the HP scaold, we report that the phe-
nolic hydroxyl group can tolerate a PEG-carbonate and show enhanced antibacterial activities towards multiple
MRSA isolates and maintain biolm eradication activities against MRSA-1707 compared to their non-carbonate
analogues. In previous studies, we found that some ester groups are well-tolerated while ether- or amine-group
substitution of the phenolic hydroxyl group leads to a complete loss of antibacterial activities22. e structural
requirement for a 1-hydroxyl group (or masked ester/carbonate) on the HP scaold, coupled with the results from
UV-Vis experiments and metal(II) co-treatment MIC proles, leads us to conclude that non-metal(II) binding
carbonates 24 and 25 are prodrugs that release metal(II)-chelators HP 1 and 17, which bind metal(II)-cations and
eradicate planktonic and biolm cells against susceptible, Gram-positive human pathogens. Continued eorts
regarding prodrug investigations are currently underway in our labs and could lead to signicant advances in the
treatment of bacterial infections.
In conclusion, we have utilized a convergent Wohl-Aue reaction to synthesize a diverse class of halogenated
phenazine small molecules that have enabled exploration of the 4-, 6- and 8-positions of the HP scaold using
readily available building blocks (Fig.9). From these studies, we have discovered that 6-substituted HPs demon-
strate enhanced antibacterial and biolm eradication activities against MRSA, MRSE and VRE compared to par-
ent HP 1. We also found that the antibacterial activities of HPs are enhanced with zinc(II) co-treatment, yet are
inhibited with copper(II) and iron(II). is metal(II) co-treatment prole diers from known metal-chelating
agents EDTA and TPEN, which were unable to eradicate MRSA and MRSE biolms during these investigations.
In macromolecular synthesis inhibition experiments, we found HP 1 to inhibit RNA and protein biosynthe-
sis while not inhibiting DNA synthesis in planktonic cultures of MRSA-1707. is is also our rst report of a
Figure 8. New activity proles focused heavily on 4-, 6- and 8-substituted HPs during these studies.

www.nature.com/scientificreports/
12
Scientific RepoRts | 7: 2003 | DOI:10.1038/s41598-017-01045-3
2,4-dibrominated HP (8-chloro HP, 18) demonstrating good antibacterial activity against MtB. In addition, we
synthesized a PEG-carbonate HP with enhanced water solubility that demonstrated 30- to 100-fold enhancement
of antibacterial activities against MRSA strains, likely through a prodrug mechanism, which could prove essential
for translational aspects of HP small molecules. ese ndings, taken together with the low cytotoxicity and lack
of hemolysis activity, give promise that HP small molecules could lead to signicant breakthroughs in the treat-
ment of persistent, biolm-associated bacterial infections.
Methods
General Information. All reagents for chemical synthesis were purchased from commercial sources and
used without further purication. Reagents were purchased at ≥95% purity and commercially available controls
were used in our biological investigations without further purication. All microwave reactions were carried out
in sealed tubes in an Anton Paar Monowave 300 Microwave Synthesis Reactor. A constant power was applied to
ensure reproducibility. Temperature control was automated via IR sensor and all indicated temperatures corre-
spond to the maximal temperature reached during each experiment. Analytical thin layer chromatography (TLC)
was performed using 250 μm Silica Gel 60 F254 pre-coated plates (EMD Chemicals Inc.). Flash column chroma-
tography was performed using 230–400 Mesh 60 Å Silica Gel from Sorbent Technologies. All melting points were
obtained, uncorrected, using a Mel-Temp capillary melting point apparatus from Laboratory Services, Inc.
Bacterial strains used during these investigations include: methicillin-resistant Staphylococcus aureus
(Clinical Isolate from Shands Hospital in Gainesville, FL: MRSA-2; ATCC strains: BAA-1707, BAA-44),
methicillin-resistant Staphylococcus epidermidis (MRSE, ATCC 35984; ATCC 12228), Vancomycin-resistant
Enterococcus (VRE, ATCC 700221) and M. tuberculosis H37Ra (ATCC 25177). Compounds were stored as
DMSO stocks at room temperature in the absence of light when they are stable in DMSO stock without observing
any loss in biological activity for several months at a time. To ensure compound integrity of our DMSO stock
solutions, we did not subject these DMSO stocks of our test compounds to multiple freeze-thaw cycles.
General Chemical Synthesis Procedures. Wohl-Aue reaction. To a 100 mL round-bottom ask was added
4-tert-butylaniline (1.60 mL, 10.0 mmol), 4-methyl-2-nitroanisole (1.53 mL, 11.0 mmol), and potassium hydroxide
(2.80 g, 50.0 mmol) in toluene (16 mL). e reaction was then allowed to reux for 10 hours. Aer the reaction was
complete, the resulting mixture was then transferred to a separatory funnel with brine and extracted with dichlo-
romethane (20 mL × 3). e organic layers were combined, ltered and concentrated in vacuo. e resulting crude solid
was puried via column chromatography using 99:1 to 85:15 hexanes:ethyl acetate to aord yellow solid 8-tert-butyl-4-
methyl-1-methoxyphenazine (612 mg, 22%; compound 30 in supporting information, precursor to HP 8).
Demethylation of 1-methoxyphenazines. To a round bottom ask, 6-methyl-1-methoxyphenazine (376 mg,
1.68 mmol) was dissolved in anhydrous dichloromethane (50 mL) and cooled to −78 °C before dropwise addi-
tion of 1 M boron tribromide solution in dichloromethane (10.0 mL, 10.0 mmol). e reaction was le to stir at
−78 °C for 1 hour, and allowed to reach ambient temperature overnight. e reaction was then heated to reux
for 8 hours until complete (monitored by TLC). Upon completion of the reaction, brine (50 mL) was added to
quench the reaction. e contents of the resulting biphasic mixture were then transferred to a separatory fun-
nel and dichloromethane was used to extract the product. e resulting organic layers were dried with sodium
sulfate, ltered through cotton, and removed in vacuo. e resulting solid was puried via column chromatogra-
phy using dichloromethane to elute 6-methyl-1-hydroxyphenazine as a yellow solid (100%, 350 mg). Note: Some
1-hydroxyphenazines were puried with the addition of 1% acetic acid to 99% dichloromethane via column
chromatography.
Figure 9. Development of a modular synthesis to a diverse array of halogenated phenazines that target both
planktonic and biolms cells.

www.nature.com/scientificreports/
13
Scientific RepoRts | 7: 2003 | DOI:10.1038/s41598-017-01045-3
Bromination of 1-hydroxyphenazines. 6-Methyl-1-hydroxyphenazine (156 mg, 0.742 mmol) and
N-bromosuccinimide (277 mg, 1.56 mmol) were dissolved in dichloromethane (60 mL) and allowed to stir at
room temperature for 4 hours. e reaction contents were washed with brine (60 mL) and extracted with dichlo-
romethane. The extracts were dried with sodium sulfate, filtered, and concentrated in vacuo. The resulting
solid was puried via column chromatography using 99:1 dichloromethane:acetic acid to elute 6-methyl-2,4
-dibromo-1-hydroxyphenazine 17 as a yellow solid. Note: 1.0 Equivalent of N-bromosuccinimide was used to
synthesize 4-methyl HP analogues.
Synthesis of PEG-carbonate HPs. Tetraethyleneglycol monomethyl ether (69 µL 0.33 mmol) was placed in an
oven-dried round-bottomed ask and dissolved in anhydrous dichloromethane (1 mL). e solution was then
cooled to 0 °C. Pyridine (37 µL, 0.47 mmol) and triethylamine (11 µL 0.73 mmol) was then added via syringe,
followed by triphosgene (48.2 mg, 0.16 mmol) dissolved in dichloromethane (1 mL). e resulting mixture was
stirred from 0 °C to room temperature and continued to stir at room temperature for 5 hours. Aer that, then the
reaction was cooled to 0 °C before the addition of solution of 17 (86 mg, 0.23 mmol) and triethylamine (49 µL
0.35 mmol) in anhydrous dichloromethane was added to the reaction in dropwise. e reaction solution was
stirred for 5 min at 0 °C and then reach ambient temperature and stirred at room temperature overnight. Aer
the reaction was complete, the reaction mixture was poured into a separatory funnel containing 1 M ammonium
chloride (20 mL), and the biphasic mixture was shaken vigorously. Upon separation of layers, the aqueous layer
was re-extracted with dichloromethane (2 × 30 mL). Organic extracts were collected, dried over Sodium Sulfate,
ltered, and concentrated under vacuum. e resulting crude material was puried using ash column chro-
matography with 3:1 hexanes:ethyl acetate to 100% ethyl acetate as eluent yield 25 as a yellow oil (135 mg, 96%).
Biology and UV-Vis Experimental Procedures. Minimum Inhibitory Concentration (MIC) Susceptibility
Assays. e minimum inhibitory concentration (MIC) for each test compound was determined by the broth
microdilution method as recommended by the Clinical and Laboratory Standards Institute (CLSI). In a 96-well
plate, eleven two-fold serial dilutions of each compound were made in a nal volume of 100 μL Luria Broth. Each
well was inoculated with ~105 bacterial cells at the initial time of incubation, prepared from a fresh log phase
culture (OD600 of 0.5 to 1.0 depending on bacterial strain). e MIC was dened as the lowest concentration of
compound that prevented bacterial growth aer incubating 16 to 18 hours at 37 °C (MIC values were supported
by spectrophotometric readings at OD600). e concentration range tested for each test compound during this
study was 0.10 to 100 μM. DMSO served as our vehicle and negative control in each microdilution MIC assay.
DMSO was serially diluted with a top concentration of 1% v/v. All compounds were tested in a minimum of three
independent experiments. NOTE: Metal(II) cation studies were performed in a similar setup to the standard
MIC assay, with the addition of 200 µM of the metal(II) cation (i.e., copper(II) sulfate) to the media. All data were
obtained from three independent experiments.
MIC Assay for Mycobacterium tuberculosis. M. tuberculosis H37Ra (ATCC 25177) was inoculated in 10 mL
Middlebrook 7H9 medium and allowed to grow for two weeks. e culture was then diluted with fresh medium
until an OD600 of 0.01 was reached. Aliquots of 200 µL were then added to each well of a 96-well plate starting
from the second column. Test compounds were dissolved in DMSO at nal concentration of 10 mM. 7.5 µL of
each compound along with DMSO (negative control) and streptomycin (positive control-40 mg/ml stock solu-
tion) were added to 1.5 mL of the Mycobacterium diluted cultures, resulting in 50 µM nal concentration of each
halogenated phenazine analogues and 340 µM for streptomycin. e nal DMSO concentration was maintained
at 0.5%. Aliquots of 400 µl were added to wells of the rst column of the 96-well plate and serially diluted two-fold
(200 µl) per well across the plate to obtain nal concentrations that ranges from 0.024 to 50 µM for the test com-
pounds and 0.16 to 340 µM for streptomycin. ree rows were reserved for each compound. e plates were then
incubated at 37 °C for seven days. Minimum inhibitory concentrations are reported as the lowest concentration
at which no bacterial growth was observed. OD600 absorbance was recorded using SpectraMax M5 (Molecular
Devices). Data obtained from three independent experiments were analyzed using Excel.
Macromolecular Synthesis Inhibition Assays. Macromolecular syntheses experiments were carried out in
methicillin-resistant Staphylococcus aureus BAA-1707. An overnight culture (100 µL) of S. aureus BAA-1707 was
sub-cultured into 10 mL of fresh TSBG media which was allowed to grow to exponential phase (OD600 = 0.2–0.3)
before transferring 500 µL to each well in a 24 well-plate. e test compounds and vehicle control (DMSO) were
added to achieve the desired concentrations relative to their MIC values against S. aureus BAA-1707. Treated
cultures were then incubated at 37 °C for 30 minutes before radioactive precursors for DNA ([3H] thymidine
(0.5 µCi)), RNA ([3H] uridine (0.5 µCi)) and protein ([3H] leucine (1 µCi)) were added. Antibiotics with known
modes of action were used as positive controls in these experiments, these included: ciprooxacin (DNA inhibi-
tion), rifampicin (RNA inhibition) and linezolid (protein inhibition). DMSO served as our negative control. DNA
and RNA radiolabeled cultures were then incubated in 37 °C for 15 minutes before being stopped by adding 60 µL
of cold 5% trichloroacetic acid (TCA). e protein synthesis experiment was stopped aer 40 minutes by adding
60 µL cold TCA. ese mixtures were then incubated at 2 °C for at least 30 minutes before the contents of the
plates were transferred onto glass microber lters (24 mm) and washed 5 times with 1 mL of 5% TCA. e lters
are allowed to dry overnight before 3.5 mL of the scintillation uid was added to the scintillation vials containing
the lters and the radiation counts were measured using liquid scintillation LS 6500.
UV-Vis Experiments. e rates of halogenated phenazine-copper(II) complex formation were independently
evaluated via UV-Vis spectrometry following addition of 0.5 equivalents copper(II) sulfate to stirring solutions

www.nature.com/scientificreports/
14
Scientific RepoRts | 7: 2003 | DOI:10.1038/s41598-017-01045-3
of HP (10 mM, 4 mL) in dimethyl sulfoxide. Spectral scanning was performed from 200 to 800 nm in 2 nm incre-
ments. e disappearance of HPs 1, 17 and 22 was observed over the indicated time points. e halogenated
phenazine-copper(II) complex formation yielded a loss in absorbance due to precipitation. No change of the
UV-Vis spectra was observed for 25 as a result of no metal(II) binding.
Calgary Biofilm Device (CBD) Experiments to Determine Minimum Bactericidal Concentrations (MBC) and
Minimum Biolm Eradication Concentrations (MBEC). Biolm eradication experiments were performed using
the Calgary Biolm Device to determine MBC/MBEC values for various compounds of interest (Innovotech,
product code: 19111). e Calgary device (96-well plate with lid containing pegs to establish biolms on) was
inoculated with 125 µL of a mid-log phase culture diluted 1,000-fold in tryptic soy broth with 0.5% glucose
(TSBG) to establish bacterial biolms aer incubation at 37 °C for 24 hours. e lid of the Calgary device was then
removed, washed and transferred to another 96-well plate containing 2-fold serial dilutions of the test compounds
(the “challenge plate”). e total volume of media with compound in each well in the challenge plate is 150 µL. e
Calgary device was then incubated at 37 °C for 24 hours. e lid was then removed from the challenge plate and
MBC/MBEC values were determined using dierent experimental pathways. To determine MBC values, 20 µL
of the challenge plate was transferred into a fresh 96-well plate containing 180 µL TSBG and incubated overnight
at 37 °C. e MBC values were determined as the concentration giving a lack of visible bacterial growth (i.e.,
turbidity). For determination of MBEC values, the Calgary device lid (with attached pegs/treated biolms) was
transferred to a new 96-well plate containing 150 µL of fresh TSBG media in each well and incubated for 24 hours
at 37 °C to allow viable biolms to grow and disperse resulting in turbidity aer the incubation period. MBEC
values were determined as the lowest test concentration that resulted in eradicated biolm (i.e., wells that had no
turbidity aer nal incubation period). All data were obtained from a minimum of three independent experi-
ments. Note: Pulse experiments followed a normal CBD assay protocol; however, the compound treatment phase
(the “challenge plate”) consisted of two sequential 24 hour compound treatment plates before the nal recovery
plate. Following this, CBD pegs were removed from the lid, sonicated for 30 minutes in PBS and plated out to
determine biolm cell killing in colony forming units per milliliter (CFU/mL).
Live/Dead staining (Fluorescence Microscopy) of MRSA BAA-1707 Biolms. A mid-log culture of MRSA BAA-
1707 was diluted 1:1,000-fold and 500 µL was transferred to each compartment of a 4 compartment CELLview
dish (Greiner Bio-One 627871). e dish was then incubated for 24 hours at 37 °C. Aer this time, the cultures
were removed and the plate was washed with 0.9% saline. e dish was then treated with the compounds in fresh
media at various concentrations. DMSO was used as our negative control in this assay. e dish was incubated
with the compound for 24 hours at 37 °C. Aer this time, the cultures were removed and the dish was washed
with 0.9% saline for 2 minutes. Saline was then removed and 500 µL of the stain (Live/Dead BacLight Viability Kit,
Invitrogen) were added for 15 minutes and le in the dark. Aer this time, the stain was removed and the dish was
washed twice with 0.9% saline. en the dish was xed with 500 µL 4% paraformaldehyde in PBS for 30 minutes.
Images of remaining MRSA biolms were then taken with a uorescence microscope. All data were analyzed
using ImageJ soware from three independent experiments.
LDH Release Assay for HeLa Cytotoxicity Assessment. HeLa cytotoxicity was assessed using the LDH release
assay described by CytoTox96 (Promega G1780). HeLa cells were grown in Dulbecco’s Modied Eagle Medium
(DMEM; Gibco) supplemented with 10% Fetal Bovine Serum (FBS) at 37 °C with 5% CO2. When the HeLa cul-
tures exhibited 70–80% conuence, halogenated phenazines were then diluted by DMEM (10% FBS) at concen-
trations of 25, 50 and 100 µM and added to HeLa cells. Triton X-100 (at 2% v/v) was used as the positive control
for maximum lactate dehydrogenate (LDH) activity in this assay (i.e., complete cell death) while “medium only”
lanes served as negative control lanes (i.e., no cell death). DMSO was used as our vehicle control. HeLa cells
were treated with compounds for 24 hours and then 50 µL of the supernatant was transferred into a fresh 96-well
plate where 50 µL of the reaction mixture was added to the 96-well plate and incubated at room temperature for
30 minutes. Finally, Stop Solution (50 µL) was added to the incubating plates and the absorbance was measured at
490 nm. Results are on the next page and are from three independent experiments.
Hemolysis Assay with Red Blood Cells. As previously described, freshly drawn human red blood cells (hRBC
with ethylenediaminetetraacetic acid (EDTA) as an anticoagulant) were washed with Tris-buered saline (0.01 M
Tris-base, 0.155 M sodium chloride (NaCl), pH 7.2) and centrifuged for 5 minutes at 3,500 rpm. e washing
was repeated three times with the buer. In 96-well plate, test compounds were added to the buer from DMSO
stocks. en 2% hRBCs (50 µL) in buer were added to test compounds to give a nal concentration of 200 µM.
e plate was then incubated for 1 hour at 37 °C. Aer incubation, the plate was centrifuged for 5 minutes at
3,500 rpm. en 80 µL of the supernatant was transferred to another 96-well plate and the optical density (OD)
was read at 405 nm. DMSO served as our negative control (0% hemolysis) while Triton X served as our positive
control (100% hemolysis). e percent of hemolysis was calculated as (OD405 of the compound- OD405 DMSO)/
(OD405 Triton X- OD405 buer) from three independent experiments.
Associated Content. General synthesis procedures; antibacterial, HeLa cell cytotoxicity, hemolysis assay
protocols; full characterization data reported for all new compounds, including: H/ C NMR spectra, HRMS and
melting points (for solids).

www.nature.com/scientificreports/
15
Scientific RepoRts | 7: 2003 | DOI:10.1038/s41598-017-01045-3
References
1. Brown, E. D. & Wright, G. D. Antibacterial drug discovery in the resistance era. Nature 529, 336–343, doi:10.1038/nature17042
(2016).
2. Blair, J. M. A., Webber, M. A., Baylay, A. J., Ogbolu, D. O. & Piddoc, L. J. V. Molecular mechanisms of antibiotic resistance. Na t.  ev.
Microbiol. 13, 42–51, doi:10.1038/nrmicro3380 (2015).
3. Lewis, . Platforms for antibiotic discovery. Nat. ev. Drug Discov. 12, 371–387, doi:10.1038/nrd3975 (2013).
4. Lewis, . Persister cells. Annu. ev. Microbiol. 64, 357–372, doi:10.1146/annurev.micro.112408.134306 (2010).
5. Balaban, N. Q., Merrin, J., Chait, ., owali, L. & Leibler, S. Bacterial persistence as a phenotypic switch. Science 305, 1622–1625,
doi:10.1126/science.1099390 (2004).
6. Grant, S. S. & Hung, D. T. Persistent bacterial infections, antibiotic tolerance, and the oxidative stress response. Virulence 4, 273–283,
doi:10.4161/viru.23987 (2013).
7. Ng, W. L. & Bassler, B. L. Bacterial quorum-sensing networ architectures. Annu. ev. Genet. 43, 197–222, doi:10.1146/annurev-
genet-102108-134304 (2009).
8. Hall-Stoodley, L., Costerton, J. W. & Stoodley, P. Bacterial biolms: from the natural environment to infectious diseases. Nat. ev.
Microbiol. 2, 95–108, doi:10.1038/nrmicro821 (2004).
9. Donlan, . M. & Costerton, J. W. Biolms: survival mechanisms of clinically relevant microorganisms. Clin. Microbiol. ev. 15,
167–193, doi:10.1128/CM.15.2.167-193.2002 (2002).
10. Davies, D. Understanding biolm resistance to antibacterial agents. Nat. ev. Dug Discov. 2, 114–122, doi:10.1038/nrd1008 (2003).
11. Wood, T. . Combatting bacterial persister cells. Biotechnol. Bioeng. 113, 476–483, doi:10.1002/bit.v113.3 (2016).
12. Mus, D. J. Jr. & Hergenrother, P. J. Chemical countermeasures for the control of bacterial biolms: eective compounds and
promising targets. Curr. Med. Chem. 13, 2163–2177, doi:10.2174/092986706777935212 (2006).
13. Worthington, . J., ichards, J. J. & Melander, C. Small molecule control of bacterial biolms. Org. Biomol. Chem. 10, 7457–7474,
doi:10.1039/c2ob25835h (2012).
14. Wolcott, . & Dowd, S. e role of biolms: are we hitting the right target? Plast. econstr. Surg. 127, Suppl 1, 28S–35S (2011).
15. Høiby, N. A personal history of research on microbial biolms and biolm infections. Pathog. Dis. 70, 205–211, doi:10.1111/
m.2014.70.issue-3 (2014).
16. Payne, D. J., Gwynn, M. N., Holmes, D. J. & Pompliano, D. L. Drugs for bad bugs: confronting the challenges of antibacterial
discovery. Nat. ev. Drug Discov. 6, 29–40, doi:10.1038/nrd2201 (2007).
17. Harbarth, S., euretzbacher, U. & Hacett, J. Antibiotic research and development: Business as usual? J. Antimicrob. Chemother. 70,
1604–1607, doi:10.1093/jac/dv020 (2015).
18. Garrison, A. T. & Huigens III, . W. Eradicating bacterial biolms with natural products and their inspired analogues that operate
through unique mechanisms. Curr. Top. Med. Chem. 17, E-pub ahead of print (2017).
19. Machan, Z. A. et al. Interaction between Pseudomonas aeruginosa and Staphylococcus aureus: description of an anti-staphylococcal
substance. J. Med. Microbiol. 34, 213–217, doi:10.1099/00222615-34-4-213 (1991).
20. Laursen, J. B. & Nielsen, J. Phenazine natural products: biosynthesis, synthetic analogues, and biological activity. Chem. ev. 104,
1663–1685, doi:10.1021/cr020473j (2004).
21. Price-Whelan, A., Dietrich, L. E. P. & Newman, D. . ethining ‘secondary’ metabolism: physiological roles for phenazine
antibiotics. Nat. Chem. Biol. 2, 71–78, doi:10.1038/nchembio764 (2006).
22. Borrero, N. V. et al. Phenazine antibiotic inspired discovery of potent bromophenazine antibacterial agents against Staphylococcus
aureus and Staphylococcus epidermidis. Org. Biomol. Chem. 12, 881–886, doi:10.1039/c3ob42416b (2014).
23. Garrison, A. T. et al. Bromophenazine derivatives with potent inhibition, dispersion and eradication activities against Staphylococcus
aureus biolms. SC Adv. 5, 1120–1124 (2015).
24. Garrison, A. T. et al. Halogenated Phenazines that Potently Eradicate Biolms, MSA Persister Cells in Non-Biolm Cultures, and
Mycobacterium tuberculosis, Halogenated phenazines that potently eradicate biolms, MSA persister cells in non-biolm cultures
and Mycobacterium tuberculosis. Angew. Chemie, Int. Ed. 54, 14819–14823, doi:10.1002/anie.201508155 (2015).
25. Garrison, A. T. et al. Structure-activity relationships of a diverse class of halogenated phenazines that targets persistent, antibiotic-
tolerant bacterial biolms and Mycobacterium tuberculosis. J. Med. Chem. 59, 3808–3825, doi:10.1021/acs.jmedchem.5b02004
(2016).
26. Wohl, A. & Aue, W. Ueber die Einwirung von Nitrobenzol auf Anilin bei Gegenward von Alali. Chem. Ber. 34, 2442–2450,
doi:10.1002/(ISSN)1099-0682 (1901).
27. Pachter, I. J. & loetzel, M. C. The Wohl-Aue reaction. I. Structure of benzo [a] phenazine oxides and synthesis of
1,6-dimethylphenazine and 1,6-dichlorophenazine. J. Am. Chem. Soc. 73, 4958–4961, doi:10.1021/ja01154a144 (1951).
28. Gopalan, B., Gharat, L. A. & hairatar-Joshi, N. Preparation of acridine, phenazine and oxanthrene-1-carboxamides as PDE4
inhibitors for treatment of asthma and chronic pulmonary disease. PCT Int. App., 2006040650 (2006).
29. ahne, D., Leimuhler, C., Lu, W. & Walsh, C. Glycopeptide and lipoglycolpeptide antibiotics. Chem. ev. 105, 425–448,
doi:10.1021/cr030103a (2005).
30. O’Brien, L. et al. M2+●EDTA binding anities: a modern experiment in thermodynamics for the physical chemistry laboratory. J.
Chem. Educ. 92, 1547–1551, doi:10.1021/acs.jchemed.5b00159 (2015).
31. Sigdel, T. ., Easton, J. A. & Crowder, M. W. Transcriptional response of Escherichia coli to TPEN. J. Bacteriol 188, 6709–6713,
doi:10.1128/JB.00680-06 (2006).
32. Cho, Y.-E. et al. Cellular Zn depletion by metal ion chelators (TPEN, DTPA and chelex resin) and its application to osteoblastic
MC3T3-E1 cells. Nutr. es. Pract. 1, 29–35, doi:10.4162/nrp.2007.1.1.29 (2007).
33. Jennings, M. C., Ator, L. E., Pania, T. J., Minbiole, . P. C. & Wuest, W. M. Biolm-eradicating properties of quaternary ammonium
amphiphiles: simple mimics of antimicrobial peptides. ChemBioChem 15, 2211–2215, doi:10.1002/cbic.v15.15 (2014).
34. Young, D. B., Perins, M. D., Duncan, . & Barry, C. E. Confronting the scientic obstacles to global control of tuberculosis. J. Clin.
Invest. 118, 1255–1265, doi:10.1172/JCI34614 (2008).
35. Evangelopoulos, D. & McHugh, T. D. Improving the tuberculosis drug development pipeline. Chem. Biol. Drug Des. 86, 951–960,
doi:10.1111/cbdd.12549 (2015).
36. Pérez-Lago, L. et al. Persistent infection by a Mycobacterium tuberculosis strain that was theorized to have advantageous properties,
as it was responsible for a massive outbrea. J. Clin. Microbiol. 53, 3423–3429, doi:10.1128/JCM.01405-15 (2015).
37. Chetty, S., amesh, M., Singh-Pillay, A. & Soliman, M. E. S. ecent advancements in the development of anti-tuberculosis drugs.
Bioorg. Med. Chem. Lett. 27, 370–386, doi:10.1016/j.bmcl.2016.11.084 (2017).
38. Mullowney, M. W. et al. Diaza-anthracene antibiotics from freshwater-derived actinomycete with selective antibacterial activities
toward Mycobacterium tuberculosis. ACS Infect. Dis. 1, 168–174, doi:10.1021/acsinfecdis.5b00005 (2015).
39. angamani, S., Younis, W. & Seleem, M. N. epurposing ebselen for treatment of multidrug-resistant staphylococcal infections.
Sci. ep. 5, 1–13, doi:10.1038/srep11596 (2015).
40. Weidmann, E. et al. Lactate dehydrogenase assay: a reliable, nonradioactive technique for analysis of cytotoxic lymphocyte-mediated
lytic activity against blasts from acute myelocytic leuemia. Ann. Hematol. 70, 153–158, doi:10.1007/BF01682036 (1995).
41. Brodersen, D. E. et al. e structural basis for the action of the antibiotics tetracycline, pactamycin, and hygromycin B on the 30S
ribosomal subunit. Cell 103, 1143–1154, doi:10.1016/S0092-8674(00)00216-6 (2000).

www.nature.com/scientificreports/
16
Scientific RepoRts | 7: 2003 | DOI:10.1038/s41598-017-01045-3
42. Ceri, H. et al. e Calgary Biolm Device: new technology for rapid determination of antibiotic susceptibilities of bacterial biolms.
J. Clin. Microbiol. 37, 1771–1776 (1999).
43. Harrison, J. J. et al. Microtiter susceptibility testing of microbes growing on PEG lids: a miniaturized biolm model for high-
throughput screening. Nat. Protoc. 5, 1236–1254, doi:10.1038/nprot.2010.71 (2010).
44. L eonard, S. N., Cheung, C. M. & yba, M. J. Activities of ceobiprole, linezolid, vancomycin, and daptomycin against community-
associated and hospital-associated methicillin-resistant Staphylococcus aureus. Antimicrob. Agents Chemother. 52, 2974–2976,
doi:10.1128/AAC.00257-08 (2008).
45. Mél ard, A. et al. Activity of cearoline against extracellular (broth) and intracellular (THP-1 monocytes) forms of methicillin-
resistant Staphylococcus aureus: comparison with vancomycin, linezolid and daptomycin. J. Antimicrob. Chemother. 68, 648–658,
doi:10.1093/jac/ds442 (2013).
46. eren, I., aldalu, N., Spoering, A., Wang, Y. & Lewis, . Persister cells and tolerance to antimicrobials. FEMS Microbiol. Lett. 230,
13–18, doi:10.1016/S0378-1097(03)00856-5 (2006).
47. Lechner, S., Lewis, . & Bertram, . Staphylococcus aureus persisters tolerant to bactericidal antibiotics. J. Mol. Microbiol. Biotechnol.
22, 235–244, doi:10.1159/000342449 (2012).
48. Hoque, J. et al. Membrane active small molecules show selective broad spectrum antibacterial activity with no detectable resistance
and eradicate biolms. J. Med. Chem. 58, 5486–5500, doi:10.1021/acs.jmedchem.5b00443 (2015).
49. De Zoysa, G. H., Cameron, A. J., Hegde, V. V., aghothama, S. & Sarojini, V. Antimicrobial peptides with potential for biolm
eradication: synthesis and structure activity relationship studies of battacin peptides. J. Med. Chem. 58, 625–639, doi:10.1021/
jm501084q (2015).
Acknowledgements
We would like to acknowledge the University of Florida for supporting this work through start-up funds (College
of Pharmacy and Division of Sponsored Research) and through an Opportunity Seed Fund. G.M.B. is a University
of Florida Graduate Fellow. High-resolution mass spectrometry (HRMS) data for all new compounds synthesized
were obtained from the Chemistry Department at the University of Florida.
Author Contributions
R.W.H. III and H.Y. conceived this study. H.Y. carried out all of the chemical synthesis, except for the synthesis of
24, performed all UV-Vis experiments and metal(II) cation co-treatment assays against MRSA-1707 and some
additional antibacterial assays. Y.A. performed the large majority of the antibacterial assays, assisted S.J. with the
macromolecular synthesis inhibition assays, carried out all of the biolm eradication assays (CBD, Live/Dead
staining) against MRSA, MRSE and VRE, and all hemolysis assays. G.M.B. synthesized 24 and performed some
antibacterial assays. D.K. performed all antibacterial assays against MtB. G.H. performed all cytotoxicity against
HeLa cells. H.Y. performed some antibacterial assays. S.J. performed macromolecular synthesis inhibition assays
and supervised HeLa cell cytotoxicity experiments. H.L. supervised MtB antibacterial assays. R.W.H. III led the
study, supervised chemical synthesis/antibacterial/biolm eradication/hemolysis studies, reviewed all data and
wrote this manuscript with the assistance of H.Y., Y.A., S.J. and H.L.
Additional Information
Supplementary information accompanies this paper at doi:10.1038/s41598-017-01045-3
Competing Interests: e authors declare that they have no competing interests.
Publisher's note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional aliations.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Cre-
ative Commons license, and indicate if changes were made. e images or other third party material in this
article are included in the article’s Creative Commons license, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons license and your intended use is not per-
mitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the
copyright holder. To view a copy of this license, visit http://creativecommons.org/licenses/by/4.0/.
© e Author(s) 2017