1 Induction of the MarA/soxS regulon
2 The multiple antibiotic resistance gene (marA) and the superoxide resistance regulator
3 (soxS) were found to be two of the most up-regulated genes in response to PMX 10070 treatment
4 according to the microarray analysis (Fig S5). Activation of marA is known to cause low-level
5 resistance to a wide range of antibiotics, superoxides and organic solvents (reviewed in Alekshun
6 and Levy, 1997)(1) via the induction of efflux pumps, such as the AcrAB multidrug efflux pump.
7 soxS is known to be induced by generation of radical oxidative species (ROS) as well as changes
8 in the redox potential of the cytosol and induces the oxidative stress response regulon(2).
9 Transcriptional profiling by RT-PCR confirmed a transient induction of marA (up to 14 fold in
10 polymyxin B treatment and 7.5 fold in PMX 10070 treatment; Fig S5A) and soxS (up to 75 fold
11 in polymyxin B treatment and 24 fold in PMX 10070 treatment; Supp Fig S5B). The
12 transcription of soxS is normally repressed by soxR, which binds to the polymerase binding
13 region of the soxS gene, preventing transcription. Structural changes induced by the oxidation of
14 the Fe-S cluster of soxR as a result of ROS generation or changes in the redox state of the
15 cytoplasm cause it to dissociate from the promoter, thereby initiating transcription of soxS (3).
16 Direct measurement of ROS generation in E. coli D31 was performed by using a non-
17 fluorescent, membrane permeable derivative of fluorescein, hydroxyphenyl fluorescein (HPF).
18 Once inside the cytoplasm, the phenyl group of HPF can be cleaved by reactions with hydroxyl
19 or peroxynitrite radicals (if present) to give fluorescein, which is membrane impermeable and is
20 therefore trapped in the cytoplasm. Therefore, the fluorescence emission of fluorescein can serve
21 as a reporter of ROS generation. ROS generation was measured after 3 hours of treatment with
22 MIC concentrations of PMX 10070, PMX 10072 or polymyxin B with norfloxacin (a quinilone
23 known to cause ROS generation) used as a positive control. Treatment with arylamides and
24 polymyxin showed virtually no ROS generation compared to the controls as measured by flow
25 cytometry (Supp Fig 5B). This indicates that the induction of marA and soxS is unlikely to be
26 caused by ROS generation. However, a down-regulation of the rsx regulon and rseC genes,
27 which have been shown to be involved in keeping the Fe-S cluster of soxR in a reduced state,
28 was observed in the microarray experiment. A decrease in rsx transcript abundance was
29 observed, which could explain the oxidation of soxR and subsequent induction of soxS. (Table
30 S1). Taken together, these results indicate that the activation of the oxidative-stress response
31 pathway is part of a coordinated response to the presence of antibiotic treatment.
32 Supplementary Methods:
33 ROS generation assay. A culture of E. coli D31 was grown in LB to OD600 = 0.1 and incubated
34 with 5 µM 3-(p-hydroxyphenyl) fluorescein (HPF; invitrogen) for 1 hour to allow passive
35 diffusion into cells. The cultures were then treated with appropriate concentrations of polymyxin,
36 PMX10070, PMX10072 and 250 ng/ml of norfloxacin and incubated with shaking at 37 ºC for 3
37 hours. Aliquots were then removed, pelleted and washed with PBS buffer twice and resuspended.
38 A BD FACSsort cell sorter flow-cytometer was used to measure the fluorescence intensity of
39 approximately 105 cells at 515 nm. Data analysis was performed in the BD CellQuest Pro
42 Fig. S1: Arylamides permeabilized the inner-membrane slowly and less-extensively. Early log
43 phase E. coli D31 cultures were co-incubated with ONPG substrate and indicated concentrations
44 of antimicrobials. A) Polymyxin B treatment causes immediate and extensive permeabilization at
45 concentrations > 1.56 µg/mL. B) PMX 10070 treatment does not cause increased permeability of
46 ONPG substrate. C) PMX 10072 treatment causes increased ONPG permeability at high
47 treatment concentrations 40 mins after treatment.
49 Fig. S2: Genes down-regulated by PMX 10070 and polymyxin B treatment. Genes related to
50 protein synthesis, cell metabolism and transport are down-regulated by both treatments. More
51 genes involved in lipopolysaccharide biosynthesis and cell-cycle regulators are down-regulated
52 by PMX 10070 treatment than polymyxin B treatment.
54 Fig. S3: Changes in morphology caused by treatment with PMX 10070. Cultures of E. coli D31
55 were treated with 62.5 µg/mL (5X MIC) of PMX 10070 for 1 min (A), 2 min (B), 4 min (C), 8
56 min (D), 20 min (E), 40 min (F) and 60 min (G) and visualized by TEM. PMX 10070 treatment
57 shows similar changes of normal cellular morphology (increased cytoplasmic staining,
58 appearance of diffuse halo, non-uniform cellular membrane) with a slower progression than
59 treatment with PMX 10072.
61 Fig. S4: rcs phosphorelay is induced by arylamide treatment. CpsB-lacZ reporter strains were
62 used to measure rcs induction in response to treatment with indicated concentrations of PMX
63 10070 and PMX 10072. Polymyxin B treatment (0.39 µg/mL) was used as positive control. A)
64 Treatment with PMX 10070 induces cpsB at concentrations ≥10 µg/mL. B) Similarly, treatment
65 with PMX 10072 induces cpsB at concentrations ≥6 µg/mL. Growth attenuation caused by these
66 treatments is shown.
68 Fig. S5: SoxS induction is not caused by ROS generation. E. coli D31 were treated with PMX
69 10070 (8.75 µg/mL, 0.7x MIC) and polymyxin B (0.39 µg/mL, 1x MIC). RT-PCR shows the
70 time-course of the up-regulation of marA (A), and soxS (B) in response to PMX 10070 (blue)
71 and polymyxin B (red). C) There is no increase in ROS generation as a result of treatment by
72 MIC concentrations of PMX 10070, PMX 10072 or polymyxin B. Norfloxacin was used as a
73 positive control.
75 1. Alekshun, M. N., and S. B. Levy. 1999. The mar regulon: multiple resistance to
76 antibiotics and other toxic chemicals. Trends Microbiol 7:410-3.
77 2. Demple, B. 1996. Redox signaling and gene control in the Escherichia coli soxRS
78 oxidative stress regulon--a review. Gene 179:53-7.
79 3. Koo, M. S., J. H. Lee, S. Y. Rah, W. S. Yeo, J. W. Lee, K. L. Lee, Y. S. Koh, S. O.
80 Kang, and J. H. Roe. 2003. A reducing system of the superoxide sensor SoxR in
81 Escherichia coli. EMBO J 22:2614-22.
85 Table S1: Down-regulation of the rsx regulon.
Polymyxin B PMX 10070
Gene 20 min 60 min 20 min 60 min
rsxA -1.58 -1.41 -3.37 -2.03
rsxB -1.13 1.06 -1.60 -1.49
rsxC -1.33 1.06 -1.91 -1.96
rsxD -1.22 -1.22 -1.56 -2.53
rsxE -1.16 -1.44 -1.51 -1.96
rsxG -1.22 -1.84 -1.94 -2.76
rseC -1.08 1.05 1.18 1.62