Involvement of Outer Membrane Proteins and Peroxide-Sensor Genes in Burkholderia Cepacia Resistance to Isothiazolone

World J Microbiol Biotechnol (2014) 30:1251–1260 DOI 10.1007/s11274-013-1538-3

ORIGINAL PAPER

Involvement of outer membrane proteins and peroxide-sensor genes in Burkholderia cepacia resistance to isothiazolone

Gang Zhou • Qing-shan Shi • You-sheng Ouyang • Yi-ben Chen

Received: 23 June 2013 / Accepted: 22 October 2013 / Published online: 6 November 2013 Ó Springer Science+Business Media Dordrecht 2013

Abstract Isothiazolones are used as preservatives in Keywords Burkholderia cepacia Kathon various modern industrial products. Although microor- preservative Biocide resistance properties Outer ganisms that exhibit resistance towards these biocides have membrane proteins Peroxide-sensor genes been identified, the underlying resistance mechanisms are still unclear. Therefore, we investigated the resistance properties of the following Burkholderia cepacia strains to Kathon (a representative of isothiazolones): a wild-type Introduction (WT) strain; a laboratory resistance strain (BC-IR) induced from WT; and an isolated strain (BC-327) screened from Isothiazolone biocides are widely used as antimicrobial industrial contamination samples. The bacterial cell struc- agents to control microbial growth and biofouling in ture was disrupted by 50 lgml-1 Kathon treatment. BC- industrial applications such as cooling towers, metal- IR and BC-327 did not display resistance in the presence of working fluids, paper systems and cosmetics (Xu et al. 1mlL-1 Tween 80, 1 ml L-1 Triton X-100, 0.1 % 2009; Speksnijder et al. 2010). Isothiazolones possess sodium dodecyl sulfate or 1 mmol L-1 EDTA-2Na. broad-spectrum antimicrobial efficacy against gram-posi- Additionally, BC-IR and BC-327 exhibited lower relative tive and gram-negative bacteria (Pucci et al. 2007, 2011), conductivity from 10 to 180 min. The types as well as the fungi (Pedras and Suchy 2006; Adibpour et al. 2007; Vi- levels of outer-membrane proteins (OMPs) were altered centini et al. 2011), viruses, and algae (Sharmeen et al. among WT, BC-IR and BC-327. Finally, the two Kathon- 2001; Khalaj et al. 2004). resistance strains BC-IR and BC-327 presented higher Isothiazolones were thought to interact oxidatively with resistance capacity to H2O2. We measured the levels of accessible thiols such as glutathione and cysteine within peroxide-sensor genes and observed that the transcriptional the cell (Collier et al. 1990b) resulting in inhibition of activator oxyR, superoxide dismutase sod1, sod2, catalase growth and metabolism and loss of viability (Williams cat1 and cat3 were all up-regulated under oxidative con- 2007). Moreover, one of the most important properties of ditions for all strains. Taken together, OMPs and peroxide- isothiazolone biocides is that they are not specifically tar- sensor genes in B. cepacia contributed to isothiazolone geted to envelope or cytosolic proteins (Collier et al. resistance; However, the laboratory strain BC-IR exhibited 1990a). Several reports demonstrate biocide activity of a different resistance mechanism and properties compared isothiazolone by inhibition of transferases (El Abdellaoui to the isolated strain BC-327. et al. 2006; Furdas et al. 2011; Reddy et al. 2011), oxidoreductases (Collier et al. 1991), isomerases (Cheng et al. 2007; Pucci et al. 2011), ligases (Gedi et al. 2011), G. Zhou Q. Shi (&) Y. Ouyang Y. Chen hydrolases (King et al. 2009), growth factor receptors State Key Laboratory of Applied Microbiology, South China (Kiselyov et al. 2009), and even mouse fibroblast cell (The Ministry-Province Joint Development), Guangdong growth (Adibpour et al. 2010). Institute of Microbiology, Guangzhou 510070, Guangdong, People’s Republic of China To exert antibacterial action, bactericides must penetrate e-mail: [email protected] the cell envelope or accumulated therein at a sufficiently 123 1252 World J Microbiol Biotechnol (2014) 30:1251–1260 high concentration. Adaptation of the microbial cell Materials and methods envelope may contribute to the mechanism allowing for resistance to antimicrobial agents (Cloete 2003). As Bacterial strains and chemicals expected, different outer membrane proteins (OMPs) expression patterns between sensitive and resistant strains B. cepacia ATCC25416 (WT), kindly provided by were observed upon isothiazolones treatment. A 35 kDa Guangdong Culture Collection Center (Guangzhou, OMP was detectable in wild-type (WT) cells of Pseudo- China), was routinely incubated in M9 minimal medium monas aeruginosa but not in cells resistant to isothiazo- (Howard-Flanders and Theriot 1966) containing 1 g L-1 -1 -1 lone. Therefore, it was proposed that this protein was the NH4Cl, 11 g L Na2HPO47H2O, 3gL KH2PO4, -1 -1 -1 channel for isothiazolone (Bro¨zel and Cloete 1994). Sim- 5gL NaCl, 4 g L glucose, 120 mg L MgSO4, and -1 ilarly, in a previous study, upon examination of OMP 10 mg L CaC12 at 37 °C in a bathing rotator with 120 profiles of methylchloroisothiazolone (MCI)-resistant iso- revolutions per minute (rpm). A representative isothiazo- lates of P. aeruginosa, it was found that each of the isolates lone product Kathon, consisting of a 3:1 (volume:volume) lacked the 42 kDa protein which is believed to be a porin ratio of 5-chloro-2-methyl-4-isothiazolin-3-one (CMIT) known as OprD. This result indicates the role of the outer and 2-methyl-4-isothiazolin-3-one (MIT) at a final con- membrane as a permeability barrier allowing for MCI centration of 14 % total active ingredient, was obtained resistance (Chapman et al. 1998). In addition, a T-OMP in from Guangdong Dimei Biology Technology Co., LTD P. aeruginosa PAO1 was found from biocide-exposed (Guangzhou, China). Meantime, all other chemicals used in preparations but not from resistance-induced cultures, this study were of analytical grade and purchased from which were passaged in the absence of biocide. This result Sigma Chemical Co. (St. Louis, MO, USA) unless other- suggests that the disappearance of T-OMP is associated wise stated. with the onset of resistance to isothiazolones (Winder et al. 2000). Isolation and identification of B. cepacia In general, preservative resistance occurs when a for- from a contaminated sample merly effective preservative system no longer inhibits microbial growth (Chapman et al. 1998). It is important A contaminated sample of shampoo, obtained from to completely understand properties and mechanisms of Guangdong Jiangmen Investment Industry CO., LTD (Ji- industrial biocides that confer resistance for optimizing angmen, China) and contained antimicrobial agents practical application (Williams 2007). However, previous including isothiazolones, was diluted with sterile distill studies of the isothiazolone resistance mechanism were water and placed onto MacConkey agar (Oxoid Ltd., nearly all performed by analyzing an induced laboratory Basingstoke, UK), and then repeatedly cultured at 37 °C. strain, whereas little attention has been devoted to an When the morphology of colony appeared to be homolo- environmental strain isolated from a contaminated sam- gous, pure bacterium was considered to be obtained and ple. We hypothesized that the two strains have different preserved at -80 °C with 30 % glycerol. Subsequently, the resistance mechanisms or properties. Meanwhile because biochemical characteristics of isolated strains were analy- of widespread and long-term use, an increasing number sized using API 20NE system (API-BioMerieux, La Balme of isothiazolone resistance cases have been reported. Les Grottes, France) according to the manufacter’s direc- Recently a resistant strain of Burkholderia cepacia,a tions. Meanwhile, genome of selected strain was extracted major industrial contaminant in cosmetic and pharma- with TIANamp Bacteria DNA Kit (Tiangen, Beijing, ceutical raw materials and even finished products (Jime- China) based on the manusfacter’s instructions. The 16 s nez and Smalls 2000), was isolated and identified from rRNA gene was amplified with a primer pair of 16s-F (50- industrial putrefaction in our laboratory. The new genus AGAGTTTGATCATGGCTCAG-30) and 16s-R (50-TAG- Burkholderia was first assigned in 1992 (Yabuuchi et al. GGTTACCTTGTTACGACTT-30) in an Eppendorf ther- 1992), and it was identified to be resistant to most anti- mocycler 5330 (Eppendorf AG, Hamburg, Germany) using microbial agents (Vermis et al. 2003) including isot- the above genome as templates. The PCR product was hiazolones (Rushton et al. 2013). The objective of this cloned into a pMD18-T vector (TaKaRa, Dalian, China) paper is to elucidate the difference of the resistance and sequenced at Beijing Genomics Institute (Guangzhou, features between induced and isolated strains of B. China). The resultant sequence was aligned using Clustal X cepacia. Our results will contribute to the knowledge of 1.83 with corresponding sequences of representative spe- the bacteria resistance mechanism to isothiazolone bio- cies of the genus of Burkholderia. Phylogenetic trees were cides and inform users about the corrected method of constructed using neighbour-joining algorithms from biocide application. MEGA 4 (Tamura et al. 2007). The identified B. cepacia

123 World J Microbiol Biotechnol (2014) 30:1251–1260 1253 with higher resistance level to Kathon was desiganed as external structure by transmission electron microscopy BC-327. (TEM; H-7650, Hitachi, Japan) at 80 kV.

Inducement of the resistance strain Determination of cell conductivity

Inducement of resistance to Kathon biocide was performed Aliquots of 1 ml overnight culture of WT, BC-IR and BC- according to the method previously reported by Bro¨zel and 327 were pipetted into 50 ml fresh M9 medium respectively, Cloete (1994) with some modifications. Briefly, the WT and then the flasks were placed in a rotary incubator at 37 °C -1 strain was first sub-cultured at 0.5 lgml of Kathon. with 180 rpm shaking until the optical density (OD600) Then the culture was transferred into a fresh M9 medium reached about 2.5. After centrifugation at 5,000 rpm, the supplemented with 1 lgml-1 of Kathon, and the concen- pellets were washed three times with deionized water, tration of Kathon was increased by 0.5 lgml-1 until the transferred into new tubes and resuspended with 5 ml sub-culture could not grow within 48 h. The final culture deionized water. Subsequently the cell conductivity of WT, that grew under the highest Kathon concentration was BC-IR and BC-327 suspensions were measured at 10, 20, 40, considered as the resistance cells and termed BC-IR. 60, 80, 100, 120, 140, 160 and 180 min with a Mettler Toledo SevenMulti Dual pH/Conductivity Meter Kit (Met- Determination of susceptibility to Kathon tler Toledo, Columbus, OH, USA) according to the manual instructions. Finally, the mixtures were boiled for 5 min, and The minimum inhibitory concentrations (MIC) of WT, BC- the final conductivity values were measured to calculate the IR and BC-327 to Kathon were determined according to relative conductivity (conductivity/final conductivity for the method described previously with slight modifications each time point). The above-described treatment was repe- (Chapman and Diehl 1995). Briefly, various amounts of ated three times for each sample at different days. Kathon stock solution were diluted in M9 minimal medium to give a range of appropriate concentrations in increments Determination of OMP profiles of 0.5 lgml-1 according to preliminary test results. Approximately 2 9 107 colony-forming unit (cfu) ml-1 of Another portion of WT, BC-IR and BC-327 cultures, which fresh culture of WT, BC-IR and BC-327 were added into were used in conductivity determination, were centrifuged the above medium in equal volumes. After culturing at at 4 °C at 10,0009g for 5 min, and the supernatants were 37 °C for 48 h, the MIC of WT, BC-IR and BC-327 was discarded. Cell pellets were resuspended in 0.1 M pH 7.0 estimated as the lowest concentration of biocide that phosphate buffer and washed twice. Subsequently, OMPs inhibited visible growth. All the above experiments were was extracted using a Bacteria Membrane Protein Extrac- conducted three times independently for each strain. tion Kit (Bestbio, Shanghai, China) according to the manufacturer’s directions. Finally, an approximately 50 ll Effect of some factors on MIC suspension was obtained and the protein content was determined using a BCA Protein Assay Reagent Kit In order to determine the influence of some surfactants or (BestBio, Shanghai, China). Then, OMPs (about 15 lg) metal chelating agents to the MIC of WT, BC-IR and were separated by SDS-polyacrylamide gel electrophoresis BC-327, 1 ml L-1 Tween 80, 1 ml L-1 Triton X-100, (SDS-PAGE) according to previously reported methods 0.1 % SDS (Aladdin, Shanghai, China) or 1 mmol L-1 (Tattawasart et al. 2000), except that the stacking and EDTA-2Na was added in M9 minimal medium and cul- separating gels contained 5 and 10 % acrylamide (Amre- tured for 48 h at 37 °C respectively. The MIC was esti- sco, OH, USA) respectively. Electrophoresis was carried mated as the lowest concentration of biocide that out at 80 V for about 3 h with an Electrophoresis Power inhibited visible growth. Supply EPS 301 (Amersham Pharmacia Biotech, Piscata- way, NJ, USA). The final gel was stained with 0.1 % Action of Kathon on cell physiology Coomassie Brilliant Blue R-250 (Amresco, OH, USA) and photographed by ImageQuant 350 (GE Healthcare, Buck- Aliquots of 2 ml 108 cfu ml-1 overnight cultures of WT, inghamshire, UK). A standard curve was constructed by BC-IR and BC-327 were transferred into fresh tubes con- plotting the log of molecular weight of marker proteins taining 6 ml M9 liquid medium with 0 (as control), 10, and versus relative mobility (Rf) which was calculated from 50 lgml-1 Kathon. The mixtures were incubated at 37 °C distance migrated by protein/distance migrated by dye. The for 12 h with shaking at 120 rpm. Then, the cultures were molecular weight of separated OMPs was determined using directly loaded on a copper screen for observing their the above standard curve.

123 1254 World J Microbiol Biotechnol (2014) 30:1251–1260

Assay of antioxidative capacity toward H2O2 and qRT- Table 1 Physiological and biochemical characteristics of B. cepacia PCR analysis of the peroxide-sensor genes BC-327 Tests Active ingredients BC-327 The procedures to determine oxidative resistance levels to NO3 Potassium nitrate ? H2O2 (Aladdin, Shanghai, China) of WT, BC-IR, and BC- 327 was followed according to the above-mentioned TRP L-Tryptophane - method for the MIC assay, except that Kathon was replaced GLU D-Glucose ? ADH L-Arginine - by H2O2. For qRT-PCR analysis, 1 ml aliquots of fresh overnight URE Urea - culture of each strain were pipetted into 5 ml normal M9 ESC Esculin ferric citrate ? GEL Gelatin (bovine origin) - medium or adjusted M9 medium supplemented with H2O2 to the ﬁnal concentration of half MIC. After culturing at PNPG 4-Nitrophenyl-b-D-galactopyranoside ? 37 °C for 48 h, total RNAs were extracted from the colo- GLU D-Glucose ? nies of each strain with TRNzol Reagent (Tiangen, Beijing, ARA L-Arabinose ? China) and transcribed into cDNA using the Quantscript MNE D-Mannose ? RT Kit (Tiangen) according to the manufacturer’s MAN D-Mannitol ? instructions. Ten-fold dilutions of WT, BC-IR and BC-327 NAG N-acetyl-glucosamine ? cDNA were used as templates for assessing the transcrip- MAL D-Maltose ? tional levels of seven genes via qRT-PCR with paired GNT Potassium gluconate ? primers and the 16S rRNA (AF097530.1) as an internal CAP Capric acid - standard to normalise the data (Table 2). The selected ADI Adipic acid - genes included transcriptional activator oxyR MLT Malic acid ? (AFQ49200.1), two superoxide dismutase (sod) genes sod1 CIT Trisodium citrate ? (AFQ47444.1) and sod2 (AFQ47335.1), and three catalase PAC Phenylacetic acid - (cat) genes cat1 (AFQ50589.1), cat2 (AFQ46464.1) and ? Positive, - negative cat3 (AFQ49198.1). All qRT-PCR reactions were performed with SuperReal PreMix SYBR Green (Tiangen) on MastercyclerÒ ep realplex (Eppendorf, Hamburg, Ger- many) following the user’s guide. The transcriptional level of each gene in the cDNA was assessed using the 2-DDCt 99 Burkholderia cepacia BC-327 (KF658271) 39 Burkholderia cepacia ATCC 25416 (AF097530.1) method (Livak and Schmittgen 2001). The ratio of levels of Burkholderia seminalis DSM 2351 (AM747631.1) a gene transcript from BC-IR or BC-327 versus that from 13 Burkholderia contaminans LMG 23361 (JX986975.1) WT was deﬁned as the relative transcriptional level for a 49 Burkholderia lata 383 (NC_007509.1) Burkholderia anthina LMG 20980 (JX986972.1) given strain. 37 42 Burkholderia arboris R-24201 (AM747630.1) Burkholderia vietnamiensis LMG 10929 (AF097534.1) Statistical analysis Burkholderia uboniae GTC-P3-415 (AB030584.1) 25 75 Burkholderia cenocepacia LMG 16656 (AF148556.1) Burkholderia multivorans LMG 13010T (Y18703.1) To evaluate the difference among groups, all the results of 97 MIC obtained in this study were recorded and analyzed 30 Burkholderia dolosa LMG 18943 (JX986970.1) 41 Burkholderia latens R-5630T (AM747628.1) with DPS software using Student’s t tests or one-way Burkholderia metallica R-16017T (AM747632.1) analysis of variance (ANOVA) (Tang and Feng 2007). 95 Burkholderia vietnamiensis AMMD (AF043302.1) 90 Burkholderia diffusa R-15930T (AM747629.1)

32 Burkholderia stabilis LMG 14294 (AF097533.1) 78 Results Burkholderia pyrrocinia LMG14191T (U96930.1) Burkholderia oklahomensis C6786 (DQ108388.1) Burkholderia thailandensis E264 (U91838.1) Identiﬁcation of isolated strain 99 Burkholderia mallei ATCC 23344 (AF110188.1) Burkholderia caryophylli ATCC 25418T (AB021423.1) The biochemical properties of isolated strain BC-327 were 0.002 shown in Table 1. To analyze the phylogenetic position of BC-327, 16s rRNA gene sequence was determined and a Fig. 1 Phylogenetic tree of B. cepacia BC-327 based on its 16S dendrogram was constructed (Fig. 1). The phylogenetic rRNA gene sequences. Multiple sequence alignment was performed to show the position of BC-327 and of representatives of analysis indicated that BC-327 is closely related to B. related genera. GenBank accession numbers are indicated in cepacia ATCC25416 (AF097530.1). Based on Bergey’s parentheses 123 World J Microbiol Biotechnol (2014) 30:1251–1260 1255

Fig. 2 Cell structure of B. cepacia WT (A, D, G), BC-IR (B, E, H) and BC-327 (C, F, I) treated with 0 (as control), 10, and 50 lgml-1 of Kathon

manual of systematic bacteriology (Palleroni 2005) and When surfactants or mental chelating agent were added molecular characteristics, BC-327 was identified into B. to the M9 medium, the MICs of each strain decreased cepacia. (Table 3). For 0.1 % SDS treatment, the MICs of WT, BC- IR and BC-327 declined by 42.27, 75.94 and 88.68 % MIC and influential factors respectively; therefore, SDS showed the best additive effect to recover sensitivity to Kathon. Treatment with For parallel comparison, a WT B. cepacia ATCC25416 1mll-1 Tween 80 and 1 ml l-1 Triton X-100, resulted in was chosen as the parental strain and used for induction of similar MIC reduction (Table 3). Moreover, EDTA-2Na laboratory resistance strain to yield the BC-IR strain. The also contributed to sensitivity rescue, especially for BC-IR results showed that the MIC of WT to Kathon was only whose MIC recovered to nearly the same as the WT -1 3.17 lgml whereas that of the induced BC-IR strain was (Student’s t4 = 1.0, p = 0.42). 17.33 lgml-1 (Table 3), a 5.47 folds increase compared to the WT. However, when repeated sub-culture for ten Effect of Kathon on cell physiology times in the absence of Kathon, the MIC of BC-IR was declined to 15.33 ± 0.58 lgml-1. Additionally, a resis- As observed by TEM, the electron micrographs of WT, tant bacterium strain BC-327 was successfully isolated BC-IR and BC-327 cells treated with different concentra- from contaminated samples shampoo, which had added tion of Kathon are shown in Fig. 2. The cells untreated antimicrobial agents including isothiazolones. The resis- with Kathon (control) had smooth and intact surfaces, a tance levels of BC-327 for Kathon were determined, and typical rod shape, and obvious bacterial flagella (Fig. 2A– the MIC increased for 5.57-fold higher than that of WT C). While cells treated with 10 lgml-1 Kathon were rod- (Table 3). Expectedly, the MICs of BC-IR and BC-327 shaped but lacked flagella (Fig. 2D–F), those treated with were significantly enhanced compared with WT 50 lgml-1 Kathon had severely damaged membranes

(F2,6 = 1,233.2, p \ 0.01), therefore, these two strains with large leakage and fragmented areas, especially the were considered as Kathon-resistant strains and were used WT cells (Fig. 2G–I). In addition, these cells lacked bac- for the following tests. terial ﬂagella.

123 1256 World J Microbiol Biotechnol (2014) 30:1251–1260

Table 2 Primers used for qRT-PCR analysis to assess the relative transcription of the seven genes involved in anti-oxidative response Access no. Locus ID Primer name Sequences (50–30)a

AFQ49200.1 GEM_2802 oxyF/oxyR CGTCAAGCAGATGATCC/CTTCAGCAGTTCGAGCAG AFQ47444.1 GEM_0997 sod1F/sod1R GGTTACCTACAACCTGGC/CGTTCGTGTACCTGCAAG AFQ47335.1 GEM_0887 sod2F/sod2R AGACGATCCAGTACCACTAC/CTTCTTCACGATCTCTTCC AFQ50589.1 GEM_4199 cat1F/cat1R ATGCTGACGATGCTGTAC/ATTGACGAACTTGAACGC AFQ46464.1 GEM_0002 cat3F/cat3R TGGATCAAAGGTGCGAAG/GCATCGACGAAGTAGAAC AFQ49198.1 GEM_2800 cat4F/cat4R GCCGAATCAGCTTAACCT/TAGTCGAAATCCTTGTCCAT AF097530.1 – 16sRNAF/16sRNAR GATTCATTTCCTTAGTAACG/ATCGAATTAATCCACATCAT a The primers were designed using Beacon Designer 7.0 (Palo Alto, CA, USA) and synthesized at Beijing Genomics Institute (Guangzhou, China)

Fig. 3 The relative conductivity of B. cepacia WT, BC-IR, and BC- 327 from 10 to 180 min in deionized water. Error bars represent the standard deviation of triplicate measurements

Differences in relative conductivity Fig. 4 OMP profiles of B. cepacia WT (Lane 1), BC-IR (Lane 2), and BC-327 (Lane 3) separated by SDS-PAGE. M: protein markers; Resistance levels may be affected by membrane perme- OMP1-OMP12: different expression proteins among the three strains; number within brackets: protein molecular weight of specific OMP ability. To test this possibility, the relative conductivity of WT, BC-IR, and BC-327 were assayed. In general, the Separation of OMPs by SDS-PAGE relative conductivity values of WT, BC-IR, and BC-327 increased by 30.12, 35.38 and 26.30 % from 10 to Previous work demonstrated that altered OMP profiles 180 min, respectively (Fig. 3). WT cells were the most contributed to isothiazolone resistance in P. aeruginosa permeable throughout the experiment compared to the (Winder et al. 2000). To further analyze resistance in B. other two resistant strains. Conductivity of BC-IR cells was cepacia, the OMPs were extracted and then separated by lower than that of BC-327 cells early in the experiment, but SDS-PAGE. OMP profiles of WT, BC-IR, and BC-327 are the values increased from 80 to 180 min. The average shown in Fig. 4. There were approximately 12 uniquely relative conductivities from 10 to 180 min were expressed OMPs. As shown, most of the proteins (OMP1– 54.64 ± 2.20, 47.15 ± 1.70 and 46.71 ± 2.04 % OMP7) intensively appeared at high protein weight. While (F2,6 = 15.03, p \ 0.01) for WT, BC-IR, and BC-327, OMP1, OMP8, OMP9, OMP11 and OMP12 appeared in respectively. BC-IR or BC-327, this bands were undetected in the WT.

123 World J Microbiol Biotechnol (2014) 30:1251–1260 1257

Table 3 Resistance levels of B. cepacia WT, BC-IR, and BC-327 strains to Kathon preservative Medium MIC (lgml-1) ± SDa WT BC-IR BC-327

M9 3.17 ± 0.29 b 17.33 ± 0.29 a 17.67 ± 0.58 a M9 ? SDS 1.83 ± 0.29 b 4.17 ± 0.29 a 2.00 ± 0.00 b (0.1 %) M9 ? Tween 2.33 ± 0.29 c 7.67 ± 0.58 a 4.50 ± 0.50 b 80 (1 ml L-1) M9 ? Triton 2.00 ± 0.00 c 7.33 ± 0.58 a 4.17 ± 0.29 b X-100 (1 ml L-1) M9 ? EDTA- 2.00 ± 0.00 b 2.17 ± 0.29 b 4.33 ± 0.58 a 2Na (1 mmol L-1) a Different lowercase after the mean values shows the signiﬁcance among each line respectively (Tukey’s HSD; p \ 0.05)

normal conditions (without H2O2) in BC-IR (Fig. 5a). In BC-327, four genes were up-regulated approximately 1.20–1.47 fold except oxyR and cat1 which were down- regulated by 30.7 and 38.4 % respectively (Fig. 5a). Under

the suppression of H2O2, all six genes were up-regulated from 1.08- to 14.05-fold among all cell types (Fig. 5b). Particularly, sod1 in WT, sod1 in BC-IR, oxyR and sod2 in BC-327 were strongly enhanced by 4.79-fold, 14.05-fold, 4.15-fold and 4.14-fold respectively. Futhermore, oxyR was

Fig. 5 The relative transcript levels of the seven selected genes under commonly observed to be involved in H2O2 resistance in non-oxidative (a) and oxidative (b) conditions in B. cepacia WT, BC- all strains. IR, and BC-327. The relative transcription levels of all genes in WT cells under non-oxidative conditions were considered as 1 when calculated with the 2-DDCt method. Error bars represent the standard deviation of triplicate measurements Discussion

In this study, a resistant strain BC-IR was successfully Additionally, OMP9 and OMP11 only appeared in BC-327. induced under laboratory conditions (Table 3), and char- In addition to the expression of different proteins, some of acteristics of its resistance to Kathon were investigated in the common OMPs such as OMP2, OMP3, OMP4, OMP5, parallel with the BC-327 strain, which was isolated from OMP6, OMP7 and OMP10 exhibited differing expression industrial production. We observed various remarkable levels among the three strains. differences between BC-IR and BC-327 strains with respect to OMPs and anti-oxidative genes.

H2O2 resistance and qRT-PCR analysis It is well known that the outer membrane of microorganisms is highly sensitive to environmental changes, and In addition to the varying resistance levels to Kathon, WT, the membrane has a physiological ability to respond to this BC-IR, and BC-327 cells also showed different sensitivity change such as excluding harmful molecules from the cell levels to H2O2 which MIC of 29.33 ± 3.06, 40.67 ± 2.31 (Winder et al. 2000). By comparing OMPs of WT and -1 and 46.67 ± 2.31 mmol L respectively (F2,6 = 34.87, resistant cells, Bro¨zel and Cloete (1994) proposed that P. p \ 0.01). In order to detect the possible regulation sites, aeruginosa acquired resistance to isothiazolone by a pro- seven genes related to oxidation resistance were chosen for cess of adaptation where a T-OMP (35 kDa) was sup- qRT-PCR analysis by using 16s RNA as an internal stan- pressed. Similar results were also found by Chapman et al. dard. In comparison to the transcriptional levels of WT, all (1998). From these previous studies, the mass of the sup- the selected genes were slightly up-regulated (1.01–1.59 pressed protein, OprD, was determined to be 42 kDa. fold) except cat1 (down-regulated by 17.7 %) under Specifically, the OprD is a member of the class of OMPs 123 1258 World J Microbiol Biotechnol (2014) 30:1251–1260 called porins, which are water-filled transmembrane dif- EDTA by changing the cell surface tension (Table 3). The fusion channels. A loss of porins is known to cause a drug resistance can be reserved by surfactant (Woodcock reduction in the movement of some molecules across the et al. 1992), which can interact with biological membranes membrane barrier and is associated with resistance to a (Bogman et al. 2003). This result confirms the role of the number of antimicrobial agents, including isothiazolones outer membrane as a permeability barrier involved in (Chapman et al. 1998). Winder et al. (2000) demonstrated biocide resistance. Under higher concentrations of Kathon, that the induced biocide resistance of P. aeruginosa PAO1 the normal cell structures of B. cepacia were disrupted was largely reversible and that the T-OMP was associated (Fig. 2G–I) which suggests that the active sites of this with this resistance development and loss. However, the biocide is the outer membrane; however, the effect was molecular weight and other features of T-OMP were not dose-dependent (Fig. 2). Additional evidence for this analyzed. proposition was the altered relative conductivity, which is Previous studies on isothiazolone resistance in P. aeru- an important index to evaluate membrane permeability. ginosa reported a single change in the OMP profile WT cells exhibited higher conductivity values than the between WT and resistant strain. However, in this study, a other two resistance strains from 10 to 180 min (Fig. 3), series of 12 different OMPs were found among WT, BC-IR which indicated that the structure of the cell envelope of and BC-327 strains (Fig. 4). Surpression of OMPs in BC-IR and BC-327 was changed. It has been suggested that resistant strains was not detected, only additional OMPs an altered cell envelope structure contributes to the per- (OMP1, OMP8, OMP12 in BC-IR, and OMP9 and OMP11 meability of antimicrobial agents (Tattawasart et al. 2000). in BC-327) and OMP over-expression (from OMP2 to As a result, some small molecules, including isothiazolone, OMP7 and OMP10 in BC-IR or BC-327) (Fig. 4) were easily penetrate the cells or localize in the membrane and observed. We hypothesized that our results differ from exert harmful or even lethal functions. previous reports for several reasons. First, the focus of this The generation of reactive oxygen species (ROS) plays study was B. cepacia while the previous researches ana- an important role in mediating apoptosis and necrosis lyzed OMPs in P. aeruginosa. Different mechanism may associated with CMI/MI treatment (Ettorre et al. 2003). be activated in different microorganisms to affect resis- Excessive release of free radicals produced by microbial tance to biocides so that there is different OMPs expres- cells as a normal function of metabolism, or lack of their sion. Alternatively, different isothiazolones were examined control, will disable cells, resulting in death. Compared to in other studies. We used a commercial product, Kathon, the control samples (no biocide), Escherichia coli produced which is a mixture of CMIT and MIT (3:1). A similar more radicals when treated with CMIT or MIT (Williams isothiazolone was used by Bro¨zel and Cloete (1994), but it 2007). When treated with a sub-lethal concentration of was stabilized with MgCl2 and Mg(NO3)2. Pure CMIT or H2O2, the oxyR regulon was up-regulated in all cell types MIT was used to induce resistance in strains analyzed by (Fig. 4b). Previous reports have demonstrated that the Winder et al. (2000). It was shown that different isot- induction of oxyR systems results in the synthesis of hiazolones trigger different responses in microorganisms enzymes such as superoxide dismutase and catalase, as (Collier et al. 1990a; Chapman et al. 1998). Although the well as some other unidentified proteins (Greenberg and entire mechanism of OMPs was not elucidated, we con- Demple 1986). Concomitantly with the increase of oxyR, clude that OMPs play important roles in B. cepacia resis- the associated genes sod1 in WT, sod1, sod2, cat3 in BC- tance to isothiazolone. In addition, since Rushton et al. IR, and sod2 in BC-327 were all enhanced (Fig. 4b). Pro- (2013) has demonstrated that efflux was one of the major tection against biocide-induced killing by a radical defense mechanisms leading adaptive Kathon resistance in B. lata, system may be one of the important roles for radicals in the it was speculated that the OMP changes which have antimicrobial mechanism (Chapman and Diehl 1995). observed in our study may be related to efflux. Whereas, B. However, the signal transduction pathway used with anti- lata is a different B. cepacia complex (Bcc) species and oxidative genes or with other genes are yet to be preservative resistance is an intrinsic feature of the determined. complex. The outer membrane can be permeated and destabilized by treatment with the chelating agent EDTA (Leive 1968). Conclusion EDTA treatment of MCI-resistant Pseudomonas isolates restored MCI sensitivity to resistant isolates (Chapman It has been reported that B. cepacia, an important oppor- et al. 1998). Similarly, we observed that the sensitivities of tunistic pathogen in cystic fibrosis (CF) or chronic granu- BC-IR and BC-327 to Kathon were also rescued by lomatous disease, showed resistance to antimicrobial 1 mmol L-1 EDTA (Table 3). In addition, SDS, Tween 80, agents, e.g. polymyxin B, tetracycline, chloramphenicol, and Triton X-100 also functioned in the same manner as ofloxacin and co-trimoxazole (Lewin et al. 1993; Vermis 123 World J Microbiol Biotechnol (2014) 30:1251–1260 1259 et al. 2003). In this study, we demonstrated that B. cepacia Collier PJ, Ramsey AJ, Austin P, Gilbert P (1990b) Growth inhibitory also showed resistance to isothiazolone under laboratory and biocidal activity of some isothiazolone biocides. J Appl Microbiol 69:569–577 and industrial conditions. Some features of an induced Collier PJ, Austin P, Gilbert P (1991) Isothiazolone biocides: enzyme- resistant strain BC-IR and an isolated resistant strain BC- inhibiting pro-drugs. Int J Pharm 74:195–201 327 were studied and compared with WT, including El Abdellaoui H, Varaprasad CVNS, Barawkar D, Chakravarty S, extracellular structure, OMPs, and oxidative gene expres- Maderna A, Tam R, Chen HM, Allan M, Wu JZ, Appleby T, Yan SQ, Zhang WJ, Lang S, Yao NH, Hamatake R, Hong Z (2006) sion. In summary, upon comparing the two resistant strains, Identification of isothiazole-4-carboxamidines derivatives as a we found that they exhibited different resistance properties novel class of allosteric MEK1 inhibitors. Bioorg Med Chem and mechanisms to Kathon treatment. The most likely Lett 16:5561–5566 explanation for this phenomenon is the following: BC-327 Ettorre A, Andreassi M, Anselmi C, Neri P, Andreassi L, Di Stefano A (2003) Involvement of oxidative stress in apoptosis induced by was isolated from more complex environmental conditions, a mixture of isothiazolinones in normal human keratinocytes. with exposure to various antimicrobial agents in addition to J Invest Dermatol 121:328–336 isothiazolones. Hence, proper steps should be taken to Furdas SD, Shekfeh S, Bissinger EM, Wagner JM, Schlimme S, manage different resistance types. To our knowledge, this Valkov V, Hendzel M, Jung M, Sippl W (2011) Synthesis and biological testing of novel pyridoisothiazolones as histone is the first attempt to report and compare the resistance acetyltransferase inhibitors. Bioorgan Med Chem 19:3678–3689 properties of induced and isolated strains in B. cepacia. GediV,MoonJY,LimWM,LeeMY,LeeSC,KooBS, Further efforts to determine the chemical properties and Govindwar S, Yoon MY (2011) Identification and character- functions of changing OMPs, signal transduction pathways ization of inhibitors of Haemophilus influenzae acetohydr- oxyacid synthase. Enzyme Microb Technol 49:1–5 and the regulation of peroxide-sensor genes in B. cepacia Greenberg JT, Demple B (1986) Glutathione in Escherichia coli is should be undertaken through 2-D electrophoresis and dispensable for resistance to H2O2 and gamma radiation. modern bioinformatics techniques. J Bacteriol 168:1026–1029 Howard-Flanders P, Theriot L (1966) Mutants of Escherichia coli Acknowledgments This work was funded by Scientific and Tech- K-12 defective in DNA repair and in genetic recombination. nological Project of Guangdong Province (No. 2011B010500024), Genetics 53:1137–1150 Strategic Cooperation Projects Guangdong Province and Chinese Jimenez L, Smalls S (2000) Molecular detection of Burkholderia Academy (No. 2011B090300018) and Young Foundation of Guang- cepacia in toiletry, cosmetic, and pharmaceutical raw materials dong Academy of Sciences for Scientific Research (No. qnjj201203). and finished products. J AOAC Int 83:963–966 Khalaj A, Adibpour N, Shahverdi AR, Daneshtalab M (2004) Synthesis and antibacterial activity of 2-(4-substituted phenyl)- 3(2H)-isothiazolones. Eur J Med Chem 39:699–705 King A, Lodola A, Carmi C, Fu J, Mor M, Piomelli D (2009) A References critical cysteine residue in monoacylglycerol lipase is targeted by a new class of isothiazolinone-based enzyme inhibitors. Brit J Adibpour N, Khalaj A, Rezaee S, Daneshtalab M (2007) In vitro Pharmacol 157:974–983 antifungal activity of 2-(4-substituted phenyl)-3(2H)-isothiazol- Kiselyov AS, Semenova M, Semenov VV (2009) 3, 4-Disubstituted ones. Folia Microbiol 52:573–576 isothiazoles: novel potent inhibitors of VEGF receptors 1 and 2. Adibpour N, Khalaj A, Rajabalian S (2010) Synthesis and antibac- Bioorg Med Chem Lett 19:1195–1198 terial activity of isothiazolyl oxazolidinones and analogous Leive L (1968) Studies on the permeability change produced in 3(2H)-isothiazolones. Eur J Med Chem 45:19–24 coliform bacteria by ethylenediaminetetraacetate. J Biol Chem Bogman K, Erne-Brand F, Alsenz J, Drewe J (2003) The role of 243:2373–2380 surfactants in the reversal of active transport mediated by Lewin C, Doherty C, Govan J (1993) In vitro activities of multidrug resistance proteins. J Pharm Sci 92:1250–1261 meropenem, PD 127391, PD 131628, ceftazidime, chloram- Bro¨zel VS, Cloete TE (1994) Resistance of Pseudomonas aeruginosa phenicol, co-trimoxazole, and ciprofloxacin against Pseudomo- to isothiazolone. J Appl Bacteriol 76:576–582 nas cepacia. Antimicrob Agents Chemother 37:123–125 Chapman JS, Diehl MA (1995) Methylchloroisothiazolone-induced Livak KJ, Schmittgen TD (2001) Analysis of relative gene expression growth inhibition and lethality in Escherichia coli. J Appl data using real-time quantitative PCR and the 2-[Delta][Delta] Microbiol 78:134–141 CT method. Methods 25:402–408 Chapman JS, Diehl MA, Fearnside KB (1998) Preservative tolerance Palleroni NJ (2005) Genus I. Burkholderia. In: Garrity GM (ed) and resistance. Int J Cosmet Sci 20:31–39 Bergey’s manual of systematic bacteriology, 2nd edn. Volume 2: Cheng JJ, Thanassi JA, Thoma CL, Bradbury BJ, Deshpande M, The Proteobacteria, Part C: The Alpha-, Beta-, Delta-, and Pucci MJ (2007) Dual targeting of DNA gyrase and topoiso- Epsilonproteobacteria. Springer, New York, pp 575-600 merase IV: target interactions of heteroaryl isothiazolones in Pedras MS, Suchy M (2006) Design, synthesis, and antifungal Staphylococcus aureus. Antimicrob Agents Chemother activity of inhibitors of brassilexin detoxification in the plant 51:2445–2453 pathogenic fungus Leptosphaeria maculans. Bioorg Med Chem Cloete TE (2003) Resistance mechanisms of bacteria to antimicrobial 14:714–723 compounds. Int Biodeterior Biodegradation 51:277–282 Pucci MJ, Cheng JJ, Podos SD, Thoma CL, Thanassi JA, Buechter Collier PJ, Austin P, Gilbert P (1990a) Uptake and distribution of DD, Mushtaq G, Vigliotti GA, Bradbury BJ, Deshpande M some isothiazolone biocides into Escherichia coli ATCC 8739 (2007) In vitro and in vivo antibacterial activities of heteroaryl and Schizosaccharomyces pombe NCYC 1354. Int J Pharm isothiazolones against resistant gram-positive pathogens. Anti- 66:201–206 microb Agents Chemother 51:1259–1267

123 1260 World J Microbiol Biotechnol (2014) 30:1251–1260

Pucci MJ, Podos SD, Thanassi JA, Leggio MJ, Bradbury BJ, Vermis K, Vandamme P, Nelis HJ (2003) Burkholderia cepacia Deshpande M (2011) In vitro and in vivo proﬁles of ACH-702, complex genomovars: utilization of carbon sources, susceptibil- an isothiazoloquinolone, against bacterial pathogens. Antimicrob ity to antimicrobial agents and growth on selective media. J Appl Agents Chemother 55:2860–2871 Microbiol 95:1191–1199 Reddy BM, Tanneeru K, Meetei PA, Guruprasad L (2011) 3D-QSAR Vicentini CB, Romagnoli C, Manfredini S, Rossi D, Mares D (2011) and molecular docking studies on substituted isothiazole analogs as Pyrazolo [3, 4-c]isothiazole and isothiazolo [4, 3-d]isoxazole inhibitors against MEK-1 kinase. Chem Biol Drug Des 79:84–91 derivatives as antifungal agents. Pharm Biol 49:545–552 Rushton L, Sass A, Baldwin A, Dowson CG, Donoghue D, Williams TM (2007) The mechanism of action of isothiazolone Mahenthiralingam E (2013) Key role for efﬂux in the biocide. Power Plant Chem 9:14–22 preservative susceptibility and adaptive resistance of Burkholde- Winder CL, Al-Adham IS, Abdel Malek SM, Buultjens TE, Horrocks ria cepacia complex bacteria. Antimicrob Agents Chemother AJ, Collier PJ (2000) Outer membrane protein shifts in biocide- 57:2972–2980 resistant Pseudomonas aeruginosa PAO1. J Appl Microbiol Sharmeen L, McQuade T, Heldsinger A, Gogliotti R, Domagala J, 89:289–295 Gracheck S (2001) Inhibition of the early phase of HIV Woodcock D, Linsenmeyer M, Chojnowski G, Kriegler A, Nink V, replication by an isothiazolone, PD 161374. Antiviral Res Webster L, Sawyer W (1992) Reversal of multidrug resistance 49:101–114 by surfactants. Br J Cancer 66:62 Speksnijder P, van Ravestijn J, de Voogt P (2010) Trace analysis of Xu FL, Lin Q, Hou BR (2009) Synthesis and bioactivity of novel isothiazolinones in water samples by large-volume direct benzisothiazolone derivatives as potential microbiocides. J Het- injection liquid chromatography tandem mass spectrometry. erocycl Chem 46:320–323 J Chromatogr A 1217:5184–5189 Yabuuchi E, Kosako Y, Oyaizu H, Yano I, Hotta H, Hashimoto Y, Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: molecular Ezaki T, Arakawa M (1992) Proposal of Burkholderia gen. nov; evolutionary genetics analysis (MEGA) software version 4.0. and transfer of seven species of the Pseudomonas homology Mol Biol Evol 24:1596–1599 group II to the new genus, with the type species Burkholderia Tang QY, Feng MG (2007) DPS data processing system: experimental cepacia (Palleroni and Holmes 1981) comb. nov. Microbiol design, statistical analysis and data mining. Science Press, Beijing Immunol 36:1251–1275 Tattawasart U, Maillard JY, Furr JR, Russell AD (2000) Outer membrane changes in Pseudomonas stutzeri resistant to chlorh- exidine diacetate and cetylpyridinium chloride. Int J Antimicrob Agents 16:233–238

123