Jemi.microbiology.ubc.ca

Journal of Experimental Microbiology and Immunology (JEMI) Copyright April 2015, M&I UBC Deletion of the Escherichia coli K30 Group I Capsule
Biosynthesis Genes wza, wzb and wzc Confers Capsule-
Independent Resistance to Macrolide Antibiotics
Sandra Botros, Devon Mitchell, Clara Van Ommen
Department of Microbiology and Immunology, University of British Columbia
The Escherichia coli capsule functions to protect bacterial cells from desiccation and environmental stresses. The E.
coli group I capsule is polymerized and transported to the surface of the cells through the action of the wza, wzb
and wzc gene products. It is thought that the presence of a capsule may confer a level of intrinsic antibiotic
resistance. Previous work exploring the role of capsule in antibiotic resistance showed inconsistent results between
different studies, and that the role of capsule in antibiotic resistance may be dependent on antibiotic class. In this
study we sought to examine the role of the E. coli K30 group I capsule in antibiotic resistance across ten different
antibiotic classes. We examined the E. coli K30 strain CWG655Δ[wza-wzb-wzcK30] that has a chromosomal deletion
of three key capsule biosynthesis genes (wza, wzb and wzc) and its isogenic parental strain E69. We quantified the
capsule production of both strains and compared the susceptibility of the strains to ten different antibiotics. In
doing so, we identified macrolide antibiotics as a class of interest and further examined the susceptibility of the
strains to additional macrolides and a ketolide. We observed that CWG655Δ[wza-wzb-wzcK30] exhibited diminished
production of capsular polysaccharides compared to E69 at 21°C, but that both strains produced comparably low
amounts of capsule at 37°C. Contrary to past work on other antibiotic classes, we observed that CWG655Δ[wza-
wzb-wzcK30] was more resistant to macrolide antibiotics, but not ketolides, when compared to E69 at both 21°C and
37°C. From this study, we conclude that a deletion of the capsule biosynthesis genes wza, wzb and wzc confers
resistance to the macrolide family of antibiotics in a mechanism independent of capsule production.
Capsular polysaccharides (CPS) are synthesized, kanamycin resistance (7) and Song et. al. reported that transported and anchored to the surface of the cell by many capsule could interact with tetracycline, providing bacterial species, forming a hydrated layer around the cell resistance via an unknown mechanism (8). Conversely, that protects it from desiccation and environmental stress Parmar et. al found that the capsule did not confer (1). The Escherichia coli K30 group I capsule is assembled resistance to kanamycin or tetracycline, while Drayson et. via the Wyz-dependent biosynthesis system, and al concluded that antibiotic resistance following exposure polymerized and transported via the action of the Wza, to sub-inhibitory antibiotic concentrations was conferred in Wzb and Wzc proteins (3). Wza is found in the outer a capsule-independent fashion (9, 10). It has been membrane and polymerizes to form a channel through suggested by several groups that capsule involvement in which the CPS is translocated (2). Wzc is an integral antibiotic resistance is antibiotic class specific, which may membrane protein of the inner membrane, and participates explain, in part, the varied and contradictory results seen in in the polymerization of CPS through its tyrosine previous work (6-10). autokinase activity. Wzb is found in the cytoplasm and is Each of the many classes of antibiotics has a unique size, the cognate phosphatase of Wzc. (2). Whitfield et. al. structure and bacterial target (11). The macrolide family is developed an E. coli K30 group I mutant strain, characterized by the presence of a large 14, 15, or 16- CWG655Δ[wza-wzb-wzcK30], that has a chromosomal membered lactone ring and attached sugar groups (12). deletion of the wza, wzb and wzc genes resulting in a Different macrolides vary in ring size and in the chemical mutant that exhibits decreased surface assembly of group I groups attached to the ring or sugar moieties (12). CPS when compared to the isogenic parental strain E69 Macrolides of interest in this study include erythromycin, a common representative macrolide, as well as its Previous work suggests that the barrier function of the capsule may confer a level of antibiotic resistance by Additionally, a new sub-group of macrolides called inhibiting access of the antibiotics to the cell (4,5). These ketolides has been recently developed that includes the studies have demonstrated that exposure of E. coli strains antibiotic telithromycin (12). to sub-inhibitory concentrations of antibiotics results in an Macrolides act by binding to the 50s subunit of the increase in CPS production and a corresponding increase bacterial ribosome at the 23s rRNA and inhibit protein in antibiotic resistance (4). However, there has been synthesis by inducing dissociation of peptidyl-tRNA (13). conflicting evidence surrounding the direct role of capsule Four main mechanisms of macrolide resistance have been in mediating antibiotic resistance. For example, Ganal et. previously observed. Firstly, the outer membrane of many al. reported resistance to kanamycin and streptomycin in a Gram-negative bacteria can confer resistance (14). For capsule-dependent fashion (6). In addition, Al Zharani et. example, mutations that impair the barrier function of the al. found that the E. coli capsule was necessary for outer membrane were found to increase susceptibility to Page 1 of 8
Journal of Experimental Microbiology and Immunology (JEMI) Copyright April 2015, M&I UBC azithromycin, clarithromycin and roxithromycin (15). al. (16), with slight modifications. A colony of each cell type was Secondly, modification of the antibiotic target through inoculated in 5 ml of either LB or MH media and grown methylation of the 23s rRNA can confer resistance (14). overnight at either 21oC or 37oC. We conducted experiments Thirdly, resistance can be conferred through an efflux using both LB and MH media because past work by Parmar et. al. identified differences in the production of capsular polysaccharide pump (14). Lastly, macrolides can be inactivated by between strains grown in LB and MH media (9). The following enzymatic activity in the cell, including that of esterases day, optical density readings at 660nm for each culture were and phosphotransferases (14). measured using a Spectronic 20+ spectrophotometer, and 1 ml of Given the conflicting evidence surrounding the role of the same culture was transferred into a sterile microcentrifuge capsule polysaccharides in antibiotic resistance, we tube. Next, the 1ml samples were centrifuged using an Eppendorf examined the role of capsule production on antibiotic 5415D microcentrifuge for 2.5 minutes at 16,100 x g. The resistance to a range of antibiotics. We examined the E. supernatants were discarded and the pellets were washed 3 times coli K30 group I mutant strain CWG655Δ[wza-wzb- with 1ml of 50 mM NaCl. Next, the pellets were re-suspended in 1ml of 50 mM EDTA. The samples were then incubated at 37oC wzcK30] in addition to its wild type (WT) parental strain on a shaker for 30min. After incubation, the samples were E69 (3). We quantified the capsule production of both pelleted at 16,100 x g and the supernatant containing capsular strains and compared the susceptibility of the two strains to polysaccharides was transferred into a sterile microcentrifuge ten different classes of antibiotics. From this, we identified tube. The subsequent capsule quantification was performed with macrolides as a class of interest and further examined the the phenol-sulphuric acid assay (16). A 1.0 mg/ml carbohydrate susceptibility of the strains to additional macrolides and a stock solution containing 0.05% w/v sucrose and 0.05% w/v ketolide. By examining different macrolide antibiotics as fructose was used to prepare the standard curve. For capsule well as a ketolide, we were able to determine if patterns of quantification, 400 uL of supernatant was combined with 400 uL antibiotic susceptibility or resistance were specific to an of 5% phenol and 2 mL of 93% sulphuric acid in a glass test tube. Colour was allowed to develop for 10 min and the absorbance individual antibiotic or if they applied to the larger was measured at 490nm on a Spectronic 20+ spectrophotometer. antibiotic class. Each experiment was done in replicates of three. We observed that CWG655Δ[wza-wzb-wzcK30] produced Capsule Staining. A colony of each cell type was streaked onto
diminished capsule compared to the WT and showed either LB or MH solid media and grown overnight at either 21oC increased resistance to macrolide antibiotics. Overall, our or 37oC. Colonies were taken from the plates using a sterilized results suggest that, for macrolide antibiotics, the E. coli loop and suspended in 250uL of sterile saline. Capsule staining K30 group I capsule does not play a role in antibiotic was performed using a modified version of the Maneval's capsule resistance, and that CWG655Δ[wza-wzb-wzc staining method described by Hughes and Smith (17). First, the cell suspension in sterile saline was mixed with 250µL of Congo resistant to macrolides via a mechanism independent of Red (1% aqueous solution, Sigma Chemical Company C-6767), capsule but related to the absence of the wza, wzb and wzc spread onto a glass microscope slide using a sterilized loop, and air-dried for 5-10 minutes. Next, 150µL of Maneval's solution was then pipetted onto the dried smears (0.047% w/v acid MATERIALS AND METHODS
fuchsin, JT Baker Chemicals, A355-3; 2.8% w/v ferric chloride, Bacterial Strains, Preparation of Media and Growth
Fisher Scientific I-89; 4.8% v/v aqueous glacial acetic acid, Conditions. E. coli K30 strains E69 (serotype: O9a:K30:H12)
Acros, 42322-0025; 3.6% v/v aqueous phenol solution, Invitrogen and CWG655 [wza IS509-037) and allowed to sit for approximately 2 minutes. The 22 min::aadA Δ(wza-wzb-wzc) K30::aphA3 Kmr Spr] were obtained from the laboratory of Dr. Chris Whitfield counterstain was washed off with dH2O and the slides were air- (Department of Molecular and Cellular Biology, University of dried before being viewed using a light microscope at 1000x magnification with oil immersion K30] has a polar aadA insertion in the wza locus corresponding to 22 minutes on the E. coli K12 Disc Diffusion Assay. Disc diffusion assays were performed
lineage map that eliminates expression of this copy of the wza- using a modified version of the Kirby-Bauer method (18). Strains wzb-wzc locus (3). The second locus of wza-wzb-wzc was were grown overnight in liquid culture of LB or MH media at inactivated using PCR amplification and cloning into the suicide 21°C or 37°C. The optical density of the cultures was measured at vector pWQ173, which was used to excise parts of wza and wzc 660nm using a Spectronic 20+ spectrophotometer and the cultures as well as all of wzb (3). In this paper, strain CWG655 is referred were then diluted with sterile broth to 1 optical density unit. LB to as either CWG655 Δ[wza-wzb-wzc] or MH plates were spread plated with 100µL of the diluted liquid K30 or as "mutant strain" while E69 is denoted as "wild type" (WT). All experiments were cultures. Antibiotic discs (7mm diameter) prepared with either performed at either 21°C or 37°C. Liquid cultures were incubated sulfamethoxazole, on a shaker contained in either a 37°C walk-in incubator or at polymyxin, vancomycin, erythromycin, tetracycline, gentamycin, room temperature (approximately 21°C). Plates were incubated in or norfloxacin (AB-biodisk) obtained from the Department of either a 37°C walk-in incubator or at room temperature Microbiology and Immunology at UBC were placed onto the (approximately 21°C). Bacterial cells were grown in either Luria plates using sterilized forceps. For the roxithromycin, Bertani (LB) broth (1.0% w/v tryptone, 0.5% w/v yeast extract, clarithromycin, and telithromycin disc diffusions, stock solutions 0.5% w/v NaCl, pH 7) or Mueller Hinton (MH) broth (0.2% w/v of 10mg/mL roxithromycin, clarithromycin and telithromycin beef extract, 1.75% w/v acid digest of casein, 0.15% starch, pH were obtained from the lab of Dr. Charles Thompson 7.3, not cation-adjusted) for capsule isolation as well as capsule (Department of Microbiology and Immunology, UBC) and 10µL staining. For other capsule staining experiments, as well as for the of each solution was pipetted onto blank discs. Each experiment disc-diffusion assay, bacterial cells were grown on plates made was done is replicates of three, with three or four discs per plate. from either LB (1.5% agar) or MH (1.7% agar) media. The plates were incubated for 18 hours at either 21°C or 37°C Capsule Extraction and Quantification. Capsule extraction
depending on the initial incubation temperature of the liquid and quantification was performed as outlined by Brimacombe et culture, and the diameters of the zones of inhibition were Page 2 of 8
Journal of Experimental Microbiology and Immunology (JEMI) Copyright April 2015, M&I UBC measured in millimetres. An increase in the diameter of the zone of inhibition indicates increased susceptibility and a decrease in the size of the zone of inhibition indicates increased resistance. Statistical Analysis. Statistical analysis was performed for the
disc diffusion assay as well as for the phenol-sulphuric acid assay. Statistical significance was determined using an unpaired, two tailed t-test (p<0.5). For the phenol-sulphuric acid assay, comparisons were made between CWG655Δ[wza-wzb-wzcK30] and the WT strain, at both 21oC and 37oC for LB and MH media. Comparisons were also made between 21°C and 37°C for the WT and CWG655Δ[wza-wzb-wzcK30]. For the disc diffusion assay, comparisons were made between CWG655Δ[wza-wzb-wzcK30] and the WT strain, at both 21oC and 37oC for LB and MH media. Deletion of the wza-wzb-wzc genes decreases capsule
production at 21°C but not at 37°C compared to the
WT strain. To confirm decreased capsule biosynthesis
FIG 1 Differences in capsular polysaccharide produced by the
WT strain and CWG655Δ[
ability of CWG655Δ[ wza-wzb-wzcK30] using the phenol-
wza-wzb-wzcK30] compared to the sulphuric acid capsule quantification method. Strains were
WT strain, we quantified capsular polysaccharide cultured overnight in 21°C or 37°C shaking incubators in LB liquid production of both strains at 21°C and 37°C using the media, and capsule polysaccharide was extracted and quantified using phenol-sulphuric acid assay. Given that past groups have the phenol-sulphuric acid assay. * indicates p<0.05, n.s. indicates not observed decreased capsule production at 37°C, compared to 21°C, we decided to conduct our analysis at both culture, this experiment confirmed that differences in temperatures (9). We expected that CWG655Δ[wza-wzb- capsule production between the strains were also observed wzcK30] would produce less capsular polysaccharide on solid media. Based on the phenol-sulphuric acid assay compared to the WT and that both strains would produce results (Fig. 1), we expected that the WT cells would have more capsular polysaccharide at 21°C, compared to 37°C. a larger visible capsule than the CWG655Δ[wza-wzb- At 21°C, the WT strain exhibited 14-times greater production of capsular polysaccharide compared to K30] cells at 21°C, but not 37°C. Resulting images of stained WT cells (Fig. 2A) and CWG655Δ[wza-wzb- wza-wzb-wzcK30] when grown in LB broth (Fig. 1). At 37°C, we found no significant difference in K30] cells (Fig. 2B) grown at 21°C showed increased capsule size visible around the WT cells, and not the capsular polysaccharide production for the WT compared to CWG655Δ[ K30] cells. Both WT (Fig. 2C) K30] (Fig. 1). Additionally, we cells and CWG655Δ[wza-wzb-wzc observed 12-times greater production of capsular comparable capsule size at 37°C (Fig. 2D). However, we polysaccharide for the WT strain at 21°C compared to observed only minor differences in capsule size between 37°C (Fig. 1). We did not observe a significant difference in capsular polysaccharide production for CWG655Δ[ WT cells grown at 21°C and 37°C (Fig. 2A, 2C). These results are unexpected given that we observed that the WT wzb-wzcK30] between 21°C and 37°C (Fig.1). We produced more capsular polysaccharides at 21°C, replicated these experiments using both strains grown in compared to 37°C (Fig. 1). We suspect that our inability to MH broth and observed a similar trend in which the WT detect large differences in capsule size is due to disparate microscope image quality. Despite our inability to detect wza-wzb-wzcK30] (Supplemental Fig. 1). When large difference in capsule size between WT cells grown at grown in MH media we observed a less pronounced 21°C and 37°C, from these results we conclude that, when difference in polysaccharide production between the two grown on solid media, CWG655Δ[wza-wzb-wzc strains at 21°C, indicating that LB would be a more decreased capsule size compared to the WT at 21°C, and suitable media for further study regarding the effects of the wza-wzb-wzc gene deletion on antibiotic resistance. From K30] and the WT show similar capsule sizes at 37°C. these results, we conclude that the WT strain produces more capsular polysaccharide than CWG655Δ[ exhibits
increased
resistance to erythromycin compared to the WT strain.
wzcK30] at 21°C, but not 37°C. Due to the increase in capsule production observed for the WT cells exhibit increased capsule thickness
compared to CWG655Δ[
WT strain compared to CWG655Δ[wza-wzb-wzc K30] on solid LB
21°C, we hypothesized that an increase in capsular agar. To further confirm that CWG655Δ[wza-wzb-wzcK30]
polysaccharides might influence antibiotic resistance in a was deficient in capsule compared to the WT, we performed capsule staining using Maneval's staining class-dependent manner. In addition, differences in antibiotic susceptibility between the WT strain and procedure, and visualized capsule size using light microscopy. Given that the antibiotic disc diffusion K30] were predicted to be more prominent at 21°C when compared to 37°C, due to the lack experiments were to be performed on solid media but the of differential capsule production between the strains phenol-sulphuric acid assay used cells grown in liquid Page 3 of 8

Journal of Experimental Microbiology and Immunology (JEMI) Copyright April 2015, M&I UBC We observed no significant differences in capsule production at 37°C between CWG655Δ[wza-wzb-wzcK30] and the WT strain (Fig. 1). However, we observed CWG655Δ[wza-wzb-wzcK30] and the WT strain at 37°C (Fig. 3). These results suggest that the differential susceptibility of the strains to erythromycin may not be due to the physical presence of capsule. From this we conclude that although the deletion of the wza-wzb-wzc capsule biosynthesis genes confers increased resistance to erythromycin, this effect may not be due to decreased capsule production. Resistance
erythromycin extends to the macrolides clarithromycin
and
ketolide
FIG 2 Differences in capsule thickness of WT and
telithromycin. We observed differences in antibiotic
susceptibility between the WT strain and CWG655Δ[wza- K30] cells grown on solid LB agar media at
21°C and 37°C. (A) E69 WT cells grown on LB agar at 21°C; (B)
wzb-wzcK30] that varied with antibiotic class (Supplemental CWG655Δ[wza-wzb-wzcK30] cells grown on LB agar at 21°C. (C) E69 Fig. 2). Additionally, we observed that CWG655Δ[wza- WT cells grown on LB agar at 37°C. (D) CWG655Δ[wza-wzb-wzcK30] wzb-wzcK30] exhibited increased resistance to erythromycin cells grown on LB agar at 37°C. Strains were grown overnight on LB agar plates at 21°C, and cell capsules were stained using Maneval's compared to the WT (Fig. 3). In order to determine if the staining protocol and visualized at 1000x magnification. Grey regions resistance conferred by the wza-wzb-wzc gene deletion was indicate cell bodies, and white regions indicate capsule. specific to erythromycin or if it also applied to other antibiotics in the macrolide class, we conducted further observed at 37°C (Fig. 1). To test our hypothesis we conducted a screen of ten antibiotics, each of a different roxithromycin. Additionally, we used telithromycin, which antibiotic class, using an antibiotic disc diffusion assay on is a member of the macrolide sub-group the ketolides. We LB and MH agar media for both strains at 21°C and 37°C. examined the susceptibilities of the WT strain and We observed that CWG655Δ[wza-wzb-wzcK30] exhibited CWG655Δ[wza-wzb-wzcK30] at both 21°C and 37°C. At decreased resistance to some antibiotics, such as 21°C, disc diffusion results showed a 3-fold increase in nitrofurantoin, yet no consistent trends in resistance susceptibility to roxithromycin for the WT strain compared changes to many other antibiotics, when compared to the to CWG655Δ[wza-wzb-wzcK30]. A similar trend of WT (Supplemental Fig. 2). However, we also observed increased susceptibility of the WT strain was seen for that CWG655Δ[wza-wzb-wzcK30] exhibited increased clarithromycin, but these results were not significant at resistance to some antibiotics when compared to the WT, 21°C. (Fig 4). When grown at 37°C, we observed at 10- such as erythromycin and tetracycline (Supplemental Fig. fold increase in susceptibility to roxithromycin for the WT 2). These results suggest that a deletion of the wza-wzb- compared to CWG655Δ[wza-wzb-wzcK30] (Fig. 4). wzc genes can increase, decrease, or have no effect on Similarly, we observed that the WT was susceptible to resistance to antibiotics, depending on the antibiotic tested. clarithromycin with a clear zone of inhibition around the Based on the observed results, we identified erythromycin antibiotic disc, while CWG655Δ[wza-wzb-wzcK30] was as an antibiotic of interest for further study. At 37°C, the resistant with growth up to the edge of the antibiotic disc. WT strain had some degree of susceptibility to At both temperatures, we observed that the WT and erythromycin, as indicated by the presence of a zone of CWG655Δ[wza-wzb-wzcK30] were comparably resistant to the ketolide, telithromycin (Fig. 4). We observed a similar CWG655Δ[wza-wzb-wzcK30] showed no susceptibility, trend in results when disc diffusion assays were replicated growing consistently up to the edge of the disc (Fig. 3A). on MH media (Supplemental Fig. 4). From these results The zones of inhibition surrounding the erythromycin discs we conclude that the wza-wzb-wzc gene deletion confers appeared as a gradient, not a distinct line (Fig. 3A). increased resistance to the macrolides clarithromycin and CWG655Δ[wza-wzb-wzcK30] showed a significant increase roxithromycin, similar to the pattern seen in erythromycin, in resistance to the erythromycin compared to the WT but this does not extend to the ketolide telithromycin. This strain (Fig. 3B). We observed a similar pattern at 21°C, but effect is independent of physical capsule presence. at this temperature results did not reach significance (Fig. 3B). Additionally, we obtained similar results with a disc DISCUSSION
diffusion assay performed on MH agar, where In this study, comparison of capsule production between the WT strain and CWG655Δ[wza-wzb- increased erythromycin resistance at 37°C and a similar but less pronounced result at 21°C compared to the WT revealed that CWG655Δ[wza-wzb-wzcK30] exhibited 14-times less CPS production than the WT strain (Supplemental Fig. 3). strain at 21°C . This is consistent with the expected Page 4 of 8

Journal of Experimental Microbiology and Immunology (JEMI) Copyright April 2015, M&I UBC results, due to the chromosomal deletion of the capsule biosynthesis genes wza, wzb and wzc, as described by Whitfield et al. (3). Our results are consistent with the findings of Parmar et al., who examined a strain deficient in only the Wza channel-forming protein necessary for capsule assembly, and demonstrated decreased CPS production by that mutant strain (9). We observed a greater difference in capsule production between the WT strain and CWG655Δ[wza- wzb-wzcK30] at 21°C than at 37°C, and an overall increase in capsule production at 21°C for the WT (Fig. 1). Although the group I E. coli capsule was not previously thought to be thermoregulated (2), studies by Parmar et. al., Drayson et. al, and Stout et. al., and have reported increased capsule production at 21°C compared to 37°C (9, 10, 19). This observation is of interest because some previous studies that have failed to find differences in antibiotic resistance based on the presence or absence of capsule carried out experiments at only 37°C (20). For example, Naimi et. al. examined the role of capsule in streptomycin susceptibility with organisms grown at 37°C and observed no difference in susceptibility between a wza mutant and its isogenic WT strain (20). A possible explanation for their result is that the strains were producing comparable levels of FIG 3 Susceptibility of CWG655Δ[wza-wzb-wzcK30] and the WT
strain to erythromycin via disc diffusion assay. (A, B)
polysaccharides at 37°C. However, other studies have Representative disc diffusion results showing that the WT is examined the same strains at both 21°C and 37°C and susceptible to erythromycin, as seen by a zone of clearance around the also failed to find differences in antibiotic resistance to erythromycin disc indicated by a red dashed circle. CWG655Δ[wza- streptomycin between WT and capsule deficient wzb-wzcK30] is shown to be resistant to erythromycin, as seen by the mutants (9, 10). The previous findings on the topic of lack of inhibition around the erythromycin disc. Scale bars = 7mm; (C) Differences in susceptibility of the WT and CWG655Δ[wza-wzb- capsule-dependent wzcK30] to erythromycin at 21°C and 37°C. Disc diffusion assays were contradictory, however, as there have been other groups carried out using antibiotic discs on LB agar plates. An increase in the that have suggested a link between capsule production diameter of the zone of inhibition indicates an increase in and resistance to certain antibiotics, such as kanamycin, susceptibility. * indicates p<0.05, n.s. indicates non-significant. Dashed line indicates diameter of antibiotic disc. tetracycline, and streptomycin (4, 7, 8). Therefore, we suggest that capsule may influence antibiotic resistance compared to the WT strain (Supplementary Fig. 2). Our for specific classes of antibiotics. results indicate that our understanding of the role of In this study, we used a screen of ten different capsule in antibiotic resistance should be modified to antibiotics to compare the susceptibilities of the WT suggest that an increase in capsule production can either strain and CWG655Δ[wza-wzb-wzcK30] using the disc increase, decrease or have no effect on antibiotic diffusion assay for antibiotic resistance. Due to the resistance depending upon the antibiotic class being limited number of replicates we performed, the majority of antibiotics tested showed no significant difference in Due to the increase in resistance observed by susceptibility between the two strains (Supplementary CWG655Δ[wza-wzb-wzcK30] to erythromycin, we Fig. 2). Indeed, at 21°C we noticed no significant decided to further investigate macrolides as our differences between the antibiotic susceptibilities of antibiotic family of focus. In this study, we either strain for any of the antibiotics tested. demonstrated that CWG655Δ[wza-wzb-wzcK30] showed Nonetheless, the general trend supports the observation increased resistance to the macrolide antibiotics in the literature that the presence of an intact capsule erythromycin, clarithromycin, and roxithromycin, but can increase resistance to a variety of antibiotics (4, 7, not the ketolide antibiotic telithromycin, when 8), or have no effect (9, 10). However, we also compared to the WT strain (Fig. 3, 4). We observed identified another possibility: the absence of wza, wzb, comparable results at 21°C and 37°C. However, and wzc in our mutant may increase antibiotic differences in susceptibility to erythromycin and clairithromycin between the WT and CWG655Δ[wza- CWG655Δ[wza-wzb-wzcK30] exhibited an increase in wzb-wzcK30] only reached significance at 37°C (Fig. 3, resistance to both tetracycline and erythromycin 4). We also observed that there were no significant Page 5 of 8
Journal of Experimental Microbiology and Immunology (JEMI) Copyright April 2015, M&I UBC increased resistance observed in CWG655Δ[wza-wzb-wzcK30] in the context of OM permeability. Whitfield et. al. observed that the CWG655Δ[wza-wzb-wzcK30] strain shows depleted surface assembly of CPS (3). These results were replicated in this study (Fig.1, 2). Whitfield et. al. also observed that CWG655Δ[wza-wzb-wzcK30] produces capsular oligosaccharides with a low degree of polymerization that are attached to the lipid A moiety of LPS and form an alternate glycoform of LPS called K-LPS (3). We speculate that the increased resistance to macrolides exhibited by CWG655Δ[wza-wzb-wzcK30] to macrolides may be related to the formation of K-LPS, and therefore an altered OM structure and permeability (C. Whitfield, personal correspondence). This notion is supported by previous studies that have FIG 4 Differences in susceptibility of the WT strain and
CWG655Δ[
found that modifications in OM permeability alter wza-wzb-wzcK30] to the macrolides clarithromycin,
roxithromycin, and the ketolide telithromycin at 21°C and 37°C.
macrolide susceptibility. Vaara found that mutations Disc diffusion assays were carried out on LB agar plates. An increase that affected OM structure and increased permeability, in the diameter of the zone of inhibition indicates an increase in such as mutations in lipid A synthesis in E. coli, susceptibility. * indicates p<0.05, ** indicates p<0.005, n.s. indicates decreased the MICs of erythromycin, roxithromycin, non-significant. Dashed line indicates diameter of antibiotic disc. clarithromycin and azithromycin (23). Similarly, Buyuk differences in CPS production at 37°C between the WT et. al. found that certain strains of Pseudomonas aeruginosa are more susceptible to macrolides due to together these results suggest that the presence or increased membrane permeability (45). Finally, Farmer absence of capsule does not play a role in antibiotic et. al observed that the MIC of azithromycin was resistance to macrolides for these strains, but that the increased 8 times with the addition of a magnesium absence of the wza, wzb and wzc genes may play a role supplementation that decreased membrane permeability in the increase in resistance of CWG655Δ[wza-wzb- (25). These previous findings lend support to our proposed model wherein the wza-wzb-wzc deletion These results are consistent with past observations made by Drayson et. al which suggested confers macrolide resistance through an alteration of the that antibiotic resistance can be conferred in a capsule- OM structure that causes changes in OM permeability. independent fashion (10). Given that our data suggest The last observation of significance in this study is that the absence of wza, wzb and wzc increases that both the WT strain and CWG655Δ[wza-wzb- antibiotic resistance in a capsule-independent manner, a wzcK30] exhibited comparable levels of resistance to discussion of the potential mechanisms of resistance telithromycin, with but differing levels of susceptibility to macrolides that were close derivatives of A variety of mechanisms for resistance to macrolides erythromycin (Fig. 3, 4). Different macrolide antibiotics have been observed (21). Given that capsule is vary in chemical components that are attached to the associated with the outer membrane (OM) and wza, wzb lactone ring or sugar moieties (11). Clarithromycin is and wzc, are involved in capsule assembly we suggest derived from erythromycin by substituting a methoxy that the mechanism of most relevance to this study is group for the C-6 hydroxyl group of erythromycin (26), the role of the OM as a permeability barrier. Typically, while roxithromycin has an N-oxime side chain macrolide antibiotics are used to treat Gram-positive attached to the lactone ring (27). Telithromycin is a infections because the OM of Gram-negative bacteria member of a macrolide derivative family called can confer a level of resistance that makes clinical use ketolides, which have a further modified structure from of macrolides, particularly erythromycin, challenging typical macrolides in that a keto functional group is for those types of infections (22). This is thought to be substituted for the sugar moiety at C-3 on the lactone due to the hydrophobic nature of the macrolides, which ring (26). A methoxy group replaces the hydroxyl group can prevent them from passing the charged lipid A at C-6 and C-11-12 is cyclized to make a carbamate component of LPS present in the OM (22). group with an imidazo-pyridyl group attachment (26). Because our mutant strain lacks three genes and their We suggest that the difference we see in susceptibility corresponding protein products, we cannot identify a may be due to structural differences between single gene product that, when absent, confers the telithromycin and the macrolides. Most literature in this area suggests that ketolides have increased activity developed a model as a potential explanation for the macrolides; however, limited data is available regarding Page 6 of 8
Journal of Experimental Microbiology and Immunology (JEMI) Copyright April 2015, M&I UBC mechanisms of resistance to ketolides (26). Recent work could also focus on elucidating the mechanism of work has reported that few bacterial strains exhibit resistance to telithromycin used by both E69 and telithromycin resistance (26). In the literature, it appears CWG655Δ[wza-wzb-wzcK30]. Given that the incidence of that those strains that are resistant to ketolides exhibit telithromycin is rare, future studies could attempt to their resistance through mechanisms that are common explain the observed resistance. For example, the strains to those for macrolide resistance. Walsh et al. reported could be examined in an attempt to determine if they two mechanisms of telithromycin resistance: target harbour genes that have been identified in the literature as modification by methylation of the 23s rRNA by the being involved in telithromycin resistance (28). We observed that the WT strain produced 12-times erm(B) methylase gene or efflux mediated by the greater levels of capsular polysaccharides at 21°C mef(B) gene (28). With our limited data we cannot compared to 37°C. Past publications have reported similar develop any conclusive hypotheses regarding the source results (9). Given that group I capsule production is not known to be thermoregulated (2), future work could examine the potential mechanisms behind these results by mechanisms include exclusion due to its size or charge, examining the expression of genes involved in capsule or the presence of any of the resistance mediating genes assembly at 21°C and 37°C. described above. Finally, future experiments could be conducted to further In this study our aim was to examine the examine the other observed results in our antibiotic screen. CWG655Δ[wza-wzb-wzcK30] mutant strain and its WT That is, higher replicate screens could, and should, be parental strain E69, with particular focus on the role of conducted in order to determine the significance of the K30 group I capsule in antibiotic resistance to observed trends, and to further confirm the results multiple antibiotic families. We observed that observed herein. Moreover, certain antibiotics could be further studied in order to aid in the explanation of the K30] produced less capsular polysaccharide than the WT at 21°C, but not 37°C. We results. For example, tetracycline resistance was increased also observed that CWG655Δ[wza-wzb-wzc in the absence of capsule in this study and further work increased resistance to macrolides, but not the ketolide could focus on elucidating this mechanism. Additionally, subgroup, when compared to the WT at 21°C and 37°C. for norfloxacin at 21°C the WT strain showed increased resistance compared to CWG655Δ[ Taken together, our data suggest that a deletion of the wza-wzb-wzcK30], but this result was reversed at 37°C. Future work could focus wza, wzb and wzc genes confers resistance to macrolide, on attempting to replicate this result and examine the but not ketolide, antibiotics via a capsule-independent mechanism behind it. Indeed, these are but two of several mechanism. Overall, we conclude that the absence of examples of antibiotics that could be further studied in the capsule biosynthesis genes wza, wzb and wzc relation to the role of capsule in antibiotic resistance. confers increased resistance to macrolide antibiotics. FUTURE DIRECTIONS
We would like to thank the Department of Microbiology and Although we have indicated that the absence the wza, wzb, Immunology at the University of British Columbia for their and wzc genes is important for macrolide resistance, and funding and support. We would like the thank Dr. David Oliver, that this phenomenon is capsule-independent, the Jia Wang and Céline Michaels, for their guidance, instruction, mechanism of resistance remains unknown. The clearest and support, as well as the staff of the media room for providing and most pressing direction to be taken from this study is us with equipment and supplies. In addition, we would like to an examination of the effect of single gene mutations in thank the Department of Pharmacology at UBC for gifting the wza, wzb or wzc on resistance to macrolide family Department of Microbiology with many of the antibiotic disks used in our experiments. Further, we would like to thank Dr. antibiotics. Since this study was conducted using a mutant Chris Whitfield at the University of Guelph for providing the E. with a deletion of all three of these capsule assembly coli strains and for his valuable insight into our work. Finally we genes, a causal relationship cannot be established between would like Dr. Charles Thompson at the University of British the absence of any one gene (or a combination) and the Columbia for providing us with the macrolide antibiotics used in observed increase in resistance to macrolides. Further our experiments. study is warranted to determine whether or not the observed effect can be traced to a single gene product, or REFERENCES
whether the effect is the result of a combination of 1. Reid, AN, Szymanski, CM. 2010. Biosynthesis and assembly of
deletions. Additionally, the effect of other outer membrane capsular polysaccharides-Chapter 20, p. 351-373. In Moran, A genes could be studied in order to determine whether this (ed), Microbial Glycobiology: Structure, Relevance and Application. Elsevier Inc, London, UK. phenomenon is unique to the capsule biosynthesis genes 2. Whitfield, C. 2006. Biosynthesis and assembly of capsular
wza, wzb and wzc or whether deletions in other membrane polysaccharides in Escherichia coli. Annu. Rev. Biochem. 75:39-
channels, kinases, etc. are sufficient to induce the observed, macrolide-resistant phenotype. 3. Reid, AN, Whitfield, C. 2005. Functional analysis of conserved
We observed that the WT strain and CWG655Δ[wza- gene products involved in assembly of Escherichia coli capsules and exopolysaccharides: evidence for molecular recognition wzb-wzcK30] exhibited telithromycin resistance. Future Page 7 of 8
Journal of Experimental Microbiology and Immunology (JEMI) Copyright April 2015, M&I UBC between Wza and Wzc for colanic acid biosynthesis. J. Bacteriol. 17. Hughes RB, Smith, A. 2011. Capsule stain protocols. ML
Microbe Library, ASM. Accessed on 9th March 2015 at 4. Lu, E, Trinh, T, Tsang, T, Yeung, J. 2008. Effect of growth in
sublethal levels of kanamycin and streptomycin on capsular 18. Hudzicki, J. 2009. Kirby-Bauer disk diffusion susceptibility test
Escherichia coli B23. J. Exp. Microbiol. Immunol. 12:21-26.
protocol. ML Microbe Library, ASM. Accessed on 8th February 5. Slack, MP, Nichols, WW. 1982. Antibiotic penetration through
bacterial capsules and exopolysaccharides. J. Antimicrob. Chemother. 10:368-372.
6. Ganal, S, Guadin, C, Roensch, K, Tran, M. 2007. Effects of
19. Stout, V, Gottesman, S. 1990. RcsB and RcsC: a two-
streptomycin and kanamycin on the production of capsular component regulator of capsule synthesis in Escherichia coli. J. polysaccharides in Escherichia coli B23 cells. J. Exp. Microbiol. Bacteriol. 172:659-669.
Immunol. 11:54-99.
20. Naimi, I, Nazer, M, Ong, L, Thong, E. 2009. The role of wza in
7. Al Zahrani, F, Huang, M, Lam, B, Vafaei, R. 2013. Capsule
extracellular capsular polysaccharide levels during exposure to formation is necessary for kanamycin tolerance in Escherichia sublethal doses of streptomycin. J. Exp. Microbiol. Immunol. coli K-12. J. Exp. Microbiol. Immunol. 17:24-28.
Vol. 13:36-40.
8. Song, C, Sun, XF, Xing, SF, Xia, PF, Shi, YJ, Wang, SG.
21. Leclercq, R. 2002. Mechanisms of resistance to macrolides and
2013. Characterization of the interactions between tetracycline lincosamides: nature of the resistance elements and their clinical antibiotics and microbial extracellular polymeric substances with implications. Clin. Infect. Dis. 34:482-492.
spectroscopic approaches. Environ. Sci. Pollut. Res. Int. 21:
22. Poole, K. 2002. Outer membranes and efflux: the path to
multidrug resistance in Gram-negative bacteria. Curr. Pharm. 9. Parmar, S, Rajwani, A, Sekhon, S, Suri, K. 2014. The
Biotechnol. 3:77-98.
Escherichia coli K12 capsule does not confer resistance to either 23. Vaara, M. 1993. Outer membrane permeability barrier to
tetracycline or streptomycin. J. Exp. Microbiol. Immunol.18:76-
azithromycin, clarithromycin, and roxithromycin in gram- negative enteric bacteria. Antimicrob. Agents Chemother. 10. Drayson R, Leggat T, Wood M. 2011. Increased antibiotic
37:354-356.
resistance post-exposure to sub-inhibitory concentrations is 24. Buyck, J, Tulkens, PM, Van Bambeke, F. 2011. Increased
independent of capsular polysaccharide in Escherichia coli. J. susceptibility of Pseudomonas aeruginosa to macrolides in Exp. Microbiol. Immunol. 15:36-42.
biologically-relevant media by modulation of outer membrane 11. Walsh, C. 2003. Antibiotics: actions, origins, resistance.
permeability and of efflux pump expression, p. 17-20. In American Society for Microbiology (ASM). Anonymous 51st Interscience conference on antimicrobial agents 12. Omura, S. 2002. Macrolide antibiotics: chemistry, biology, and
and chemotherapy, Chicago. practice. Academic Press. 25. Farmer, S, Li, ZS, Hancock, RE. 1992. Influence of outer
13. Tenson, T, Lovmar, M, Ehrenberg, M. 2003. The mechanism
membrane mutations on susceptibility of Escherichia coli to the of action of macrolides, lincosamides and streptogramin B dibasic macrolide azithromycin. J. Antimicrob. Chemother. reveals the nascent peptide exit path in the ribosome. J. Mol. 29:27-33.
Biol. 330:1005-1014.
26. Zuckerman, JM. 2004. Macrolides and ketolides: azithromycin,
14. Leclercq, R. 2002. Mechanisms of resistance to macrolides and
clarithromycin, telithromycin. Infect. Dis. Clin. North Am. lincosamides: nature of the resistance elements and their clinical 18:621-649.
implications. Clin. Infect. Dis. 34:482-492.
27. Takashima, H. 2003. Structural consideration of macrolide
15. Vaara, M. 1993. Outer membrane permeability barrier to
antibiotics in relation to the ribosomal interaction and drug azithromycin, clarithromycin, and roxithromycin in gram- design. Curr. Top. Med. Chem. 3:991-999.
negative enteric bacteria. Antimicrob. Agents Chemother. 28. Walsh, F, Willcock, J, Amyes, S. 2003. High-level
37:354-356.
telithromycin resistance in laboratory-generated mutants of 16. Brimacombe CA, Stevens A, Jun D, Mercer R, Lang AS,
Streptococcus pneumoniae. J. Antimicrob. Chemother. 52:345-
Beatty JT. 2013. Quorum-sensing regulation of a capsular
polysaccharide receptor for the Rhodobacter capsulatus gene
transfer agent (RcGTA). Mol. Microbiol. 87:802-817
Page 8 of 8

Source: http://jemi.microbiology.ubc.ca/sites/default/files/Botros%20et%20al..pdf