The Complex Interplay of Iron, Biofilm Formation, and Mucoidy Affecting

Pathogens and Disease ISSN 2049-632X

RESEARCH ARTICLE The complex interplay of iron, bioﬁlm formation, and mucoidy affecting antimicrobial resistance of Pseudomonas aeruginosa Amanda G. Oglesby-Sherrouse1,2, Louise Djapgne1, Angela T. Nguyen1, Adriana I. Vasil3 & Michael L. Vasil3

1 Department of Pharmaceutical Sciences, School of Pharmacy, University of Maryland, Baltimore, MD, USA 2 Department of Microbiology and Immunology, School of Medicine, University of Maryland, Baltimore, MD, USA 3 Department of Microbiology, University of Colorado School of Medicine, Aurora, CO, USA

In this study the link between iron acquisition and antimicrobial resistance in Pseudomonas aeruginosa is investigated. The Downloaded from https://academic.oup.com/femspd/article/70/3/307/568079 by guest on 30 September 2021 authors show that iron supplementation (not a lack of iron) enhances the resistance of planktonic P. aeruginosa cells to both tigecycline and tobramycin, albeit by different mechanisms. In P. aeruginosa bioﬁlms, iron decreases the susceptibility to tobramycin, but not to tigecycline, and lack of iron blocks the formation of bioﬁlms normally induced by subinhibitory concentrations of the former antibiotic. These data again highlight that compounds targeting iron metabolism could be interesting leads to develop novel antimicrobial drugs.

Keywords Abstract iron; antibiotic resistance; tobramycin; Pseu- domonas aeruginosa; biofilms; tigecyclin. Pseudomonas aeruginosa is a Gram-negative opportunistic bacterial pathogen that is refractory to a variety of current antimicrobial therapeutic regimens. Correspondence Complicating treatment for such infections is the ability of P. aeruginosa to form Amanda G. Oglesby-Sherrouse, 20 Penn biofilms, as well as several innate and acquired resistance mechanisms. Previous Street, HSF2, Baltimore, MD 21201, USA. studies suggest iron plays a role in resistance to antimicrobial therapy, including Tel.: +1 4107068650 the efficacy of an FDA-approved iron chelator, deferasirox (DSX), or Gallium, an e-mail: [email protected] iron analog, in potentiating antibiotic-dependent killing of P. aeruginosa biofilms. Here, we show that iron-replete conditions enhance resistance of P. aeruginosa Received 14 October 2013; revised 20 nonbiofilm growth against tobramycin and tigecycline. Interestingly, the mecha- December 2013; accepted 2 January 2014. nism of iron-enhanced resistance to each of these antibiotics is distinct. Whereas Final version published online 10 February pyoverdine-mediated iron uptake is important for optimal resistance to tigecycline, 2014. it does not enhance tobramycin resistance. In contrast, heme supplementation results in increased tobramycin resistance, while having no significant effect on doi:10.1111/2049-632X.12132 tigecycline resistance. Thus, nonsiderophore bound iron plays an important role in resistance to tobramycin, while pyoverdine increases the ability of P. aeruginosa Editor: Tom Coenye to resist tigecycline treatment. Lastly, we show that iron increases the minimal concentration of tobramycin, but not tigecycline, required to eradicate P. aeruginosa biofilms. Moreover, iron depletion blocks the previous observed induction of biofilm formation by subinhibitory concentrations of tobramycin, suggesting iron and tobramycin signal through overlapping regulatory pathways to affect biofilm formation. These data further support the role of iron in P. aeruginosa antibiotic resistance, providing yet another compelling case for targeting iron acquisition for future antimicrobial drug development.

Introduction throughout adulthood (FitzSimmons, 1993). Pseudomonas aeruginosa has also been associated with bacteremia in Pseudomonas aeruginosa is a premier opportunistic path- burn victims (Pruitt et al., 1998) and acute ulcerative ogen, particularly among humans with certain underlying keratitis in individuals wearing contaminated contact lenses conditions. It is a leading infectious agent in cancer patients (Lyczak et al., 2000; Willcox, 2007). Complicating treatment with chemotherapy-induced neutropenia (Bendig et al., of individuals infected with this pathogen is the increasing 1987), and most CF patients become infected with P. aeru- ability of P. aeruginosa to resist even the most contempo- ginosa early in life and generally remain chronically infected rary therapeutic agents (Falagas & Bliziotis, 2007). One of

Pathogens and Disease (2014), 70, 307–320, © 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved 307 Iron increases P. aeruginosa antibiotic resistance A.G. Oglesby-Sherrouse et al. the most notable mechanisms contributing to antibiotic (Banin et al., 2005). Furthermore, a recent study showed an resistance of P. aeruginosa is its proclivity to form biofilms FDA-approved iron chelator, deferasirox (DSX), in combi- through the increased production of one or more of three nation with tobramycin, blocks the formation of biofilms on distinct extracellular polysaccharide matrices, designated as CF airway epithelial cells by P. aeruginosa (Moreau-Mar- Pel, Psl, and alginate (Ryder et al., 2007). In this regard, quis et al., 2009). Because of the pervasive role of iron in most P. aeruginosa strains isolated from chronically P. aeruginosa physiology, we postulated that iron might infected cystic fibrosis (CF) patients exhibit the mucoid similarly affect the ability of P. aeruginosa to better resist a phenotype, a consequence of hyperproduction of alginate. broader range of antibiotics. Here, we show iron levels affect This mucoid phenotype was previously thought to increase resistance of P. aeruginosa to two antibiotics: tobramycin, antimicrobial resistance by providing a physical barrier to which is commonly used to control P. aeruginosa lung antibiotic penetration (Hatch & Schiller, 1998; Parad et al., infection in CF patients, and tigecycline, which is used to 1999). However, the conversion of chronic pulmonary treat skin and soft tissue infections by a variety of bacteria, isolates of P. aeruginosa to a mucoid phenotype is corre- although P. aeruginosa is generally considered to be resis- Downloaded from https://academic.oup.com/femspd/article/70/3/307/568079 by guest on 30 September 2021 lated with numerous other physiological changes, such as tant to this antibiotic. While both antibiotics target protein hypermutability, which also contribute to antibiotic resis- synthesis, the mechanisms by which iron enhances resistance (Macia et al., 2005; Waine et al., 2008). Additionally, tance of P. aeruginosa to each vary greatly, indicating the P. aeruginosa can mount a protective response to antibiotic complicated and varied roles that iron uptake and signaling exposure via increased expression of multidrug efflux play in pathogenesis and virulence-associated activities. pumps and ß-lactamases, as well as through the down-reg- Furthermore, our studies, which employ the use of the ulation of outer membrane porins (Driscoll et al., 2007). FDA-approved iron chelator DSX, broaden the potential for Pseudomonas aeruginosa requires an abundance of iron this and other FDA-approved iron chelators to treat individ- during infection (Cox, 1982; Meyer et al., 1996; Takase uals afflicted with P. aeruginosa infections. et al., 2000a, b), but the sequestration of iron by host proteins from potential pathogens creates a substantial Materials and methods barrier to infection. Pseudomonas aeruginosa overcomes iron limitation through a variety of mechanisms, including the Bacterial strains and growth conditions synthesis and secretion of two siderophores, pyoverdine and pyochelin, which can scavenge iron from host proteins Bacterial strains used in this work are listed in Table 1. The and thus contribute to virulence (Cox, 1982; Takase et al., ΔpvdA, ΔpvdD, ΔpvdDΔpchEF, and ΔpchEF mutants were 2000a, b). Pseudomonas aeruginosa can also acquire iron generated by allelic exchange as previously described from heme, an abundant source of host iron, via at least two (Barker et al., 2004). Deletion of each gene in the resulting systems: Phu (Pseudomonas heme uptake) and Has (heme assimilation system; Ochsner et al., 2000). In reducing or anaerobic environments, P. aeruginosa acquires iron via the Table 1 Strains used in this study Feo system, a G-protein-like transporter of ferrous iron Source or (Marlovits et al., 2002; Marshall et al., 2009). Because of Strain Description Reference the potential for iron-accelerated oxidative damage, iron uptake systems are coordinately regulated in response to PAO1 Wild-type P. aeruginosa strain Holloway (1955) iron availability by the Fur (ferric uptake repressor) protein P.S. 3 Mucoid CF clinical isolate Pradeep Singh and an array of sigma factors. As an example, Fur directly P.S. 5 CF clinical isolate Pradeep Singh represses expression of pvdS, encoding a sigma factor that, P.S. 7 Weakly mucoid clinical isolate Pradeep Singh from explanted CF lung in turn, directly activates expression of genes for pyoverdine ΔpvdA PAO1 carrying a deletion in U. Ochsner biosynthesis (pvd) and uptake (fpv), exotoxin A (toxA), and a the pvdA coding sequence secreted protease that can degrade iron-binding proteins ΔpchEF PAO1 carrying a deletion in U. Ochsner (prpL; Cunliffe et al., 1995; Ochsner et al., 1996; Wilson & the pchEF coding sequence Lamont, 2000; Wilderman et al., 2001). Binding of ferri-py- ΔpvdD PAO1 carrying a deletion in U. Ochsner overdine to its outer membrane receptor, FpvA, releases the pvdD coding sequence PvdS, which is normally sequestered at the inner membrane ΔpvdDΔpchEF PAO1 carrying deletions in U. Ochsner by its antisigma factor, FpvR (Lamont et al., 2002; Beare the pchEF and pvdD coding et al., 2003). sequences Several studies demonstrate that high-iron concentrations ΔmucA PAO1 carrying a deletion of Yuta Okkotsu favor the formation of biofilms (Singh et al., 2002; Singh, the mucA coding sequence 2004; Banin et al., 2005; Berlutti et al., 2005; Moreau-Mar- ΔmucB PAO1 carrying a deletion in Yuta Okkotsu quis et al., 2008; Patriquin et al., 2008), and a recent report the mucB coding sequence shows iron released from CF airway epithelial cells fur C6 PAO6261Δanr mutant carrying Barton promotes biofilm formation by this pathogen (Moreau-Mar- a A10?G mutation in the et al. (1996) quis et al., 2008). Pyoverdine-mediated iron uptake contrib- FUR protein utes to the formation of P. aeruginosa biofilms, and ΔprrF1,2 PAO1 carrying a deletion in Wilderman intracellular iron levels are important for biofilm formation the prrF1,2 locus et al. (2004)

308 Pathogens and Disease (2014), 70, 307–320, © 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved A.G. Oglesby-Sherrouse et al. Iron increases P. aeruginosa antibiotic resistance mutants was confirmed by polymerase chain reaction Purification of pyoverdine (PCR). Additionally, phenotypic validation of the pyoverdine mutants was conducted by chrome azurol S (CAS) and Mini-preparations of pyoverdine were obtained as described chromophore detection assays as described below. Pseu- previously (Meyer et al., 1997) with minor modifications. domonas aeruginosa strains were maintained in brain-heart Briefly, supernatant from a 50-mL culture of wild-type PAO1, infusion (BHI) broth or on BHI agar plates. For high- and grown in DTSB + 0.5 MSG and 1% glycerol for c. 18 h, was low-iron DTSB medium, tryptic soy broth (TSB) was treated applied to a C8 SPE column (Fisher) by gravity. The column with Chelex-100 resin (Bio-Rad) and dialyzed, then supple- was washed with water and pyoverdine eluted with 50% mented with 50 mM monosodium glutamate and 1% glyc- methanol. The resulting eluate was dried by centrifugation in erol. CAS amino acids medium (CAA) was also prepared as an evaporator and resuspended in 1 mL of water. The previously described (Cornelis et al., 1992) for growth concentration was determined by diluting the stock in 0.5 M curves. FeCl3 was added to a concentration of 100 lM. acetic acid–sodium acetate buffer at pH 5.0 and determina-

À1 À1 Downloaded from https://academic.oup.com/femspd/article/70/3/307/568079 by guest on 30 September 2021 Heme was added at a final concentration of 40 lM. Purified tion of the A380 (e = 16 500 M cm at A380, pH 5.0). pyoverdine was added at a final concentration of 2 lM. Deferasirox (DSX, - Novartis) was added to a final concen- Biofilm growth and measurement tration of 330 lM to ensure complete iron chelation in iron-replete conditions without inhibiting growth of PAO1. The MBEC (minimal biofilm eradication concentration) Desferoxamine (DFO) was added at a final concentration of AssayTM physiology and genetics biofilm device (formerly 400 lgmLÀ1 as previously described (Brock et al., 1988). known as the Calgary Biofilm Device, Innovotech) was used Ethylenediamine-di(o-hydroxyphenylacetic acid; EDDHA) to assay biofilm formation and the MBEC of tobramycin and was deferrated as previously described (Rogers, 1973) tigecycline against PAO1, according to the manufacturer’s and added at a final concentration of 45 lM. For low-oxygen instructions and as previously described (Ceri et al., 1999) growth, ETESTâ plates were incubated in GasPakTM (BD with some modifications. Briefly, 96-well MBEC plates TM TM BBL ) containment systems with either a GasPak EZ (for containing DTSB, supplemented with 100 lM FeCl3, were À anaerobic environment) or CampyPakTM EZ (for microaero- inoculated in duplicate with c. 1 9 107 CFU mL 1 bacteria. bic environment, c. 6–16% O2 after 2 h according to The plates were covered with the MBEC peg lids and manufacturer’s insert) pouch added to deplete atmospheric incubated with gentle shaking (65 r.p.m.) at 37 °C for 24 h to oxygen. allow biofilm formation on the pegs. Following incubation, the peg lid was gently washed in PBS to remove nonadherent bacteria and then transferred to a fresh 96-well ‘challenge Antibiotic resistance testing plate’ with DTSB, with or without 100 lM FeCl3, and varying â ETEST antibiotic gradient strips (Biomerieux) were used concentrations of either tobramycin or tigecycline. Biofilms to assess the level of antibiotic resistance in the presence were incubated with tobramycin for 24 h at 37 °C, at which of different concentrations of iron as described in the point the pegs were stained with 0.1% crystal violet solution manufacturer’s instructions with some modifications. for 10 min and then rinsed in water. After drying, the pegs Briefly, stains were grown overnight from freezer stock were destained in 200 lL 30% acetic acid, and the amount on BHI agar. Either a 0.5 (nonmucoid strains) or 1.0 of crystal violet dye released was determined by reading the (mucoid strains) McFarland standard was prepared in wells at OD595 nm in a Biotek plate reader. sterile saline then spread onto 150-mm DTSB agar plates supplemented as indicated in duplicate. Plates were Results allowed to dry for 30 min, and then antibiotic test strips were applied to the agar. Antibiotic resistance was read Iron enhances antibiotic resistance to tobramycin and according to the manufacturer’s instructions after 20-h tigecycline incubation at 37 °C. Iron was previously shown to play a role in resistance of P. aeruginosa grown on CF airway cells to tobramycin Siderophore detection (Moreau-Marquis et al., 2009). To look more broadly at the Total iron chelator production in DTSB culture supernatants potential effects of iron on P. aeruginosa’s ability to survive was quantified as previously described (Schwyn & Neilands, antibiotic treatment, we explored the effects of iron concen- 1987). Briefly, 100 lL of culture supernatant was mixed with tration on resistance of P. aeruginosa to multiple clinically â 100 lL of chrome azurol S (CAS) solution and left to relevant antibiotics. ETEST antibiotic test strips (shown in equilibrate for 30 min at room temperature. Eight hundred Supporting Information Fig. S1) were used to determine the microliter of sterile water was added to the reaction, and the minimal inhibitory concentration of eight different antibiotics absorbance at 410 nm was determined on the spectropho- with varying clinical efficacy against P. aeruginosa (sum- tometer. The pyoverdine chromophore was directly detected marized in Table 2). For these assays, P. aeruginosa strain in culture supernatants by determining the absorbance of PAO1 was grown on chelexed and dialyzed tryptic soy agar, the filtered supernatants at 410 nm. All readings were an iron-depleted medium used extensively for iron regula- normalized to culture density as determined by the absor- tion studies (Ochsner et al., 2002; Oglesby et al., 2008). bance at 600 nm. Three strains isolated from CF patients (P.S. 3, P.S. 5, and

Pathogens and Disease (2014), 70, 307–320, © 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved 309 Iron increases P. aeruginosa antibiotic resistance A.G. Oglesby-Sherrouse et al.

Table 2 Antibiotics used in this study

S/R* (MIC in À Cellular lgmL 1)of Target Class Antibiotic Use P. aeruginosa Cell wall Penicillin Pipericillin (PP) Broad spectrum, often used in combination with beta-lactam inhibitors, S(≤ 64) active against P. aeruginosa, although not used in CF patients Cephalosporins Ceftazidime (TZ) Broad spectrum, active against P. aeruginosa S(≤ 8) Carbapenem Imipenem (IP) and Broad spectrum, resistant to most beta-lactamases S(≤ 4) Meropenem (MP) Protein Aminoglycoside Tobramycin (TM) Gram (À), including P. aeruginosa; preferred over gentamicin for S(≤ 4) synthesis treatment of CF patients due to better lung penetration ≥ Glycylcycline Tigecycline (TGC) Used for tetracycline-resistant cases of MRSA, Haemophilus inﬂuenzae, R( 8) Downloaded from https://academic.oup.com/femspd/article/70/3/307/568079 by guest on 30 September 2021 and Neisseria spp; limited to no activity against P. aeruginosa Macrolide Azithromycin (AZ) Gram (+), although ineffective against MRSA, and some Gram (À); no R(≥ 8) bacteriocidal activity against P. aeruginosa DNA Quinolone Ciproﬂoxacin (CI) Broad spectrum, although many clinical P. aeruginosa strains are R(≥ 4) S (≤ 1) replication resistant *Expected resistance of P. aerugionsa. MIC of resistant (R) or sensitive (S) strains shown in parentheses.

P.S. 7 – Table 2) were included in our preliminary studies to tigecycline. This could be due to the ability of pyoverdine to determine the clinical relevance of our findings. Results of scavenge iron from EDDHA, which has a lower binding these studies are shown in Fig. 1. affinity for iron than DSX (Heinz et al., 1999; Yunta et al., Changing iron levels in the media had no effect on in vitro 2003). Overall, these results suggest that iron levels known resistance to pipericillin, ceftazidime, or azythromycin to be present during chronic CF lung infections (Stites et al., (Fig. 1c–d and g) and caused a slight decrease in carbape- 1998, 1999; Reid et al., 2002, 2004; Gray et al., 2010; nem resistance for two of the more resistant clinical isolates Gifford et al., 2011) could enhance the ability of P. aeru- (Fig. 1a and b). Iron also caused a decrease in resistance to ginosa to resist either tobramycin or tigecycline. ciprofloxacin in one of the clinical isolates, P.S. 3 (Fig. 1h). In contrast, addition of 100 lM ferric chloride to iron-depleted Siderophores exert different effects on tobramycin and media significantly increased resistance of several P. aeru- tigecycline resistance ginosa strains to both tobramycin and tigecycline (Fig. 1e–f). This increase in resistance was most apparent in the case of Because iron-enhanced survival of P. aeruginosa in the tigecycline, resulting in approximately a threefold increase in presence of tobramycin and tigecycline, we next asked MIC for both P.S. 3 and P.S. 5 (Fig. 1f). Most notably, the whether or not siderophore mutants would be more sensitive MICs of tigecycline against all of the strains grown under iron to these antibiotics. Accordingly, we evaluated the antibiotic limitation were below the threshold for susceptibility resistance profiles of mutants defective for pyoverdine (8 lgmLÀ1), especially intriguing considering tigecycline (ΔpvdA, ΔpvdD, and ΔpvdS) or pyochelin (ΔpchEF) biosyn- has been deemed clinically ineffective against P. aeruginosa thesis in comparison with those observed with their wild-type (Petersen et al., 1999). Although not as robust, iron supple- parent. Surprisingly, no significant difference in tobramycin mentation at 100 lM significantly increased the minimal resistance was observed between wild-type PAO1 and the inhibitory concentration of tobramycin against each of the pvd or pch mutants (Fig. 3a, ΔpvdA and ΔpchEF mutants, strains tested by 50% to 200% (Fig. 1e), further supporting the and data not shown). However, addition of DSX, while idea that iron chelation/limitation could enhance the clinical eliminating the effects of iron in both siderophore mutants, efficiency of tobramycin (Moreau-Marquis et al., 2009). caused tobramycin resistance to be significantly increased These results suggest that iron exerts a protective effect in each of the mutants as compared to wild type, particularly on P. aeruginosa when exposed to tobramycin or tigecy- in the absence of iron supplementation (Fig. 3a). The cline. To further examine this effect, we employed iron pyoverdine mutants grew very poorly on plates supple- chelators to sequester available iron during exposure to mented with DSX, and growth curves of these strains these antibiotics. Addition of either deferasirox (marketed as confirmed a growth defect in low-iron DTSB upon addition of Exjadeâ, abbreviated here as DSX) or desferoxamine DSX (Fig. S1). Thus, the increased resistance of the (marketed as Desferalâ, abbreviated here as DFO), both pyoverdine mutants may be due to decreased growth and FDA-approved iron chelators typically used to treat iron protein synthesis, the target of tobramycin, in the presence overload in humans, abrogated the increased resistance of DSX. Supporting this hypothesis, addition of purified afforded by 100 lM FeCl3 to either antimicrobial agent pyoverdine to the ΔpvdA mutant restored growth to wild-type (Fig. 2), demonstrating that elemental iron enhances resis- levels (Fig. S2) and reduced tobramycin resistance of the tance to these antibiotics. In contrast, addition of ethylene- ΔpvdA mutant to wild-type levels (Fig. 3a). This rationale, diamine-N,N’-bis(hydroxyphenylacetic acid (EDDHA) had however, cannot be applied to the ΔpchEF mutant, which no observable effect on resistance to either tobramycin or grows similar to wild type in liquid (Fig. S2) and on solid

310 Pathogens and Disease (2014), 70, 307–320, © 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved A.G. Oglesby-Sherrouse et al. Iron increases P. aeruginosa antibiotic resistance Downloaded from https://academic.oup.com/femspd/article/70/3/307/568079 by guest on 30 September 2021

Fig. 1 Iron has varying effects on resistance of Pseudomonas aeruginosa to different antibiotics. ETESTâs were used as described in the Materials and methods to determine the minimal inhibitory concentration (MIC) of several clinically relevant antibiotics toward the indicated clinical CF isolates growing on DTSB agar medium without (white bars) and with (gray bars) FeCl3 (100 lM) supplementation. Error bars show the standard deviation in MICs from three independent experiments (a–b and e–f). Asterisks indicate the signiﬁcance of increased resistance upon iron supplementation as determined by a two-tailed Student’s t-test: *P < 0.05.

media (data not shown). Alternatively, it is possible that a In contrast to what was observed regarding the influence compensatory increase in pyoverdine biosynthesis allows of iron on tobramycin resistance, the ΔpvdA mutant was the ΔpchEF mutant to acquire iron and survive tobramycin slightly more sensitive to tigecycline than the wild type in the treatment, perhaps even better than the wild-type parent. absence of deferasirox (Fig. 3b). Curiously, this defect was The ΔpchEF mutant does indeed produce higher levels of only significant when these strains were grown in the pyoverdine as compared to the wild-type PAO1 strain, as presence of iron (P < 0.05). The addition of DSX caused a judged by both chromophore absorbance and chrome significant decrease in resistance to tigecycline in the azurol S assay (Fig. S3). However, addition of purified wild-type grown in low iron, and in the ΔpvdA mutant grown pyoverdine to the wild-type strain did not increase tobramy- in either high or low iron (P < 0.05). In contrast to the ΔpvdA cin resistance (Fig. 3a). In fact, the addition of pyoverdine to mutant, the ΔpchEF mutant displayed increased tigecycline the wild-type strain reduced tobramycin resistance, although resistance in the presence of DSX (Fig. 3b), possibly due to this reduction was only significant in the presence of iron compensatory production of pyoverdine. Supplementation and DSX (Fig. 3a). Based on these data, it seems unlikely of the media with purified pyoverdine restored tigecycline that enhanced pyoverdine production by the ΔpchEF mutant resistance to the ΔpvdA mutant to wild-type levels (Fig. 3b), is the source of increased tobramycin resistance in the supporting the idea that increased tigecycline resistance in presence of DSX. While the rationale for increased resis- the ΔpchEF mutant is due to a compensatory increase in tance of the ΔpchEF mutant in the presence of DSX remains pyoverdine production. Notably, iron still enhanced tobra- unclear, our results do indicate that siderophore-mediated mycin resistance in the ΔpvdA mutant, suggesting other iron uptake is not the primary means by which PAO1 methods of iron acquisition may contribute to iron-enhanced mediates iron-enhanced tobramycin resistance. tobramycin resistance. Moreover, addition of pyoverdine

Pathogens and Disease (2014), 70, 307–320, © 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved 311 Iron increases P. aeruginosa antibiotic resistance A.G. Oglesby-Sherrouse et al.

â Fig. 2 Iron increases resistance of Pseudomonas aeruginosa to tobramycin and tigecycline. ETEST s (shown in a) were used as described in the Downloaded from https://academic.oup.com/femspd/article/70/3/307/568079 by guest on 30 September 2021 Materials and methods to determine the minimal inhibitory concentration (MIC) of (b) tobramycin and (c) tigecycline toward PAO1 growing on DTSB agar medium without (white bars) and with (gray bars) FeCl3 (100 lM) supplementation. Error bars show the standard deviation in MICs from at least three independent experiments. Asterisks indicate the signiﬁcance of increased resistance upon iron supplementation as determined by a two-tailed Student’s t-test: *P < 0.05, **P < 0.01, ***P < 0.005, ****P < 0.00005.

eliminated the effect of iron on tigecycline resistance in the absence of DSX (Fig. 3b). Thus, the mere presence of pyoverdine in the media is not sufﬁcient for enhanced antimicrobial resistance to tigecycline and may even exert a negative effect on tigecycline resistance in iron-replete environments. Instead, our data suggest that pyoverdine biosynthesis, even at the low levels present in iron-replete conditions, may contribute to tigecycline resistance. How- ever, our data also indicate that other mechanisms of iron acquisition, independent of pyoverdine, are able to mediate this phenomenon in the absence of pyoverdine.

Heme confers a modest increase in tobramycin resistance in the presence of an iron chelator As heme is potentially an abundant source of iron during infection, we next tested the ability of heme to enhance P. aeruginosa resistance to tobramycin and tigecycline. At a concentration of 40 lM, heme allowed for a slight increase in resistance to tobramycin (Fig. 4a), but it is unknown if this À small increase in MIC (0.4–0.5 lgmL 1) would be clinically significant. Interestingly, heme elicited nearly a twofold increase in resistance to tobramycin in the presence of the iron chelator DSX (Fig. 4a), suggesting increased levels of heme during infection may be able to overcome the effects of iron chelation. In contrast, heme did not lead to increased resistance to tigecycline in either the absence or presence of DSX (Fig. 4b). Because of its hydrophobicity, free heme often aggregates and can even be toxic to cells, which could have complicated analysis of our results. Pseudomonas Fig. 3 Siderophores exert complex effects on iron-enhanced resistance â aeruginosa can also acquire heme from hemoglobin, a more to tigecycline and tobramycin. ETEST s were used as described in the Materials and methods to determine the MIC of tobramycin and soluble source of iron found in the host. Thus, we tested the tigecycline toward PAO1 and isogenic pyoverdine biosynthetic mutant ability of P. aeruginosa to resist antibiotic treatment in l À1 (ΔpvdA) growing on DTSB agar medium in the presence or absence of the presence of hemoglobin. At 64 gmL of hemoglobin, which correlates with c. 4 lM heme, we observed no FeCl3, deferasirox (DSX), or pyoverdine, as indicated in the figures. Error bars show the standard deviation in MICs from three independent significant increase in MIC of either tobramycin (Fig. 4c) or experiments. Asterisks indicate the significance of increased or tigecycline (Fig. 4d) as compared to DTSB with no iron decreased resistance either upon supplementation of pyoverdine or as source, either in the presence or absence of DSX. Thus, otherwise indicated as determined by a two-tailed Student’s t-test: hemoglobin does not appear to induce the same protective *P < 0.05, **P < 0.01. ND, not determined. response during tobramycin treatment as does heme or iron.

312 Pathogens and Disease (2014), 70, 307–320, © 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved A.G. Oglesby-Sherrouse et al. Iron increases P. aeruginosa antibiotic resistance

Fig. 4 Heme enhances tobramycin resistance in the presence of deferasirox. â ETEST s were used as described in the Materials and methods to determine the Downloaded from https://academic.oup.com/femspd/article/70/3/307/568079 by guest on 30 September 2021 MIC of tobramycin and tigecycline toward PAO1 growing on DTSB agar medium with or without heme or hemoglobin (HGB) supplementation, in the presence or absence of the iron chelator DSX, as indicated in the ﬁgures. Error bars show the standard deviation in MICs from three independent experiments. Asterisks indicate a signiﬁcant increase in resistance by addition of heme or hemoglobin as determined by Student’s t-test (*P < 0.05; **P < 0.01).

concerns how oxygen availability, which vastly decreases Iron regulation plays a minor role in iron- and in the lung during late stages of CF disease (Worlitzsch heme-induced tigecycline resistance et al., 2002), affects gene expression and production of Previously, a P. aeruginosa fur mutant was shown to form virulence determinants in P. aeruginosa. Oxygen status can biofilms under low-iron culture conditions, whereas the wild also affect iron availability, as iron is reduced to its ferrous, type was unable to form biofilms under iron depletion, soluble form under anaerobic environments, while remaining indicating a role for iron regulation in biofilm formation largely insoluble as ferric iron under aerobic conditions. (Banin et al., 2005). Thus, we wondered if iron regulation Therefore, we next looked at the ability of iron to confer a might additionally be involved in the observed increase in protective effect in the presence of either tigecycline or antibiotic resistance upon iron supplementation. To test this, tobramycin under microaerobic and anaerobic conditions. we determined the MICs of tobramycin and tigecycline for a While we observed an overall increase in resistance to both C6 fur mutant, which produces a Fur protein defective for tigecycline and tobramycin under anaerobic conditions, DNA binding and is de-regulated for siderophore and potentially due to increased iron solubility and/or decreased exotoxin A synthesis (Barton et al., 1996). While the C6 growth rates, resistance to either antibiotic was enhanced by fur mutation had no effect on iron enhancement of resis- iron supplementation under both microaerobic and anaero- tance to either tobramycin or tigecycline (Fig. 5a–b), this bic conditions in a manner similar to what was observed mutation allowed for heme-induced resistance to tigecycline during aerobic growth (Fig. 6). Also similar to what was (Fig. 5d), while heme had very little effect on tigecycline observed under aerobic conditions, addition of DSX elimi- resistance of wild-type cells (Figs 4b and 5d). Similarly, nated the inducing effects of iron on resistance, indicating deletion of the Fur-repressed PrrF regulatory RNAs allowed oxygen availability does not affect iron-enhanced resistance for a small but significant increase in tigecycline resistance to these antibiotics. Curiously, the feoB gene, encoding for a in the presence of heme (Fig. 5d). These results suggest G-protein-like transporter of ferrous iron, was not required iron homeostasis and/or signaling is important for tigecycline for iron-induced resistance to either tobramycin or tigecy- resistance, while the function of these processes has no cline under microaerobic or anaerobic conditions (Fig. 6). significant effect on resistance to tobramycin. Overall, these results indicate that iron is able to increase resistance to tobramycin and tigecycline regardless of its oxidation state. Iron enhances resistance to tobramycin and tigecycline under low-oxygen conditions in a Feo-independent manner Iron enhances tobramycin resistance of PAO1 biofilms The human body is a diverse set of environments, and Iron is an important factor for P. aeruginosa biofilm devel- nutrient availability in specific sites of P. aeruginosa infec- opment (Singh et al., 2002; Singh, 2004; Banin et al., 2005; tion can vary greatly. One current focus of research Berlutti et al., 2005; Moreau-Marquis et al., 2008; Patriquin

Pathogens and Disease (2014), 70, 307–320, © 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved 313 Iron increases P. aeruginosa antibiotic resistance A.G. Oglesby-Sherrouse et al.

Fig. 5 Iron regulation does not play a Downloaded from https://academic.oup.com/femspd/article/70/3/307/568079 by guest on 30 September 2021 major role in iron-dependent enhancement â of antibiotic resistance. ETEST s were used as described in the Materials and methods to determine the MIC of tobramycin and tigecycline toward PAO1 and the indicated iron regulatory mutants growing on DTSB agar medium without

(white bars) and with (gray bars) FeCl3 or heme supplementation. Error bars show the standard deviation in MICs from three independent experiments. Asterisks (*) indicate a signiﬁcant increase in resistance as compared to the wild type as determined by Student’s t-test (P < 0.01).

Fig. 6 Iron and DSX exert similar effects on tigecycline and tobramycin resistance under microaerobic and anaerobic â conditions. ETEST s were used as described in the Materials and methods to determine the minimal inhibitory concentration (MIC) of tobramycin and tigecycline toward PAO1 and a Δfeo ferrous iron uptake mutant growing on DTSB agar medium without (white bars) and with (gray

bars) FeCl3 (100 lM) supplementation. Plates were incubated in jars with anaerobic or Campy Paks as described in the Materials and methods. Error bars show the standard deviation in MICs from three independent experiments. *P < 0.05, **P < 0.005, ***P < 0.0005. et al., 2008), a phenomenon that further complicates anti- Table 3 MICs of planktonic PAO1 biotic treatment for recalcitrant infections. Therefore, we next sought to determine the effects of iron, heme, and DSX MIC* on resistance of PAO1 biofilms to tobramycin and tigecy- ÀDSX +DSX cline. For these studies, we employed the MBEC AssayTM Antibiotic ÀIron +Iron ÀIron +Iron (Innovotech), which utilizes a 96-well plate with a peg lid on which biofilms can form. Using this MBEC device, the Tobramycin 1.0 Æ 0.0 2.7 Æ 1.2 0.6 Æ 0.2 0.8 Æ 0.3 minimal biofilm eradication concentration (MBEC) can Tigecycline 53.3 Æ 18.5 64.0 Æ 0.0 8.0 Æ 0.0 8.0 Æ 0.0 quickly be determined for a large number of compounds *MIC = minimal inhibitory concentration of antibiotic required to produce and conditions. Wild-type PAO1 biofilms were allowed to clearing in MBEC plate wells, as described in Materials and methods.

314 Pathogens and Disease (2014), 70, 307–320, © 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved A.G. Oglesby-Sherrouse et al. Iron increases P. aeruginosa antibiotic resistance Downloaded from https://academic.oup.com/femspd/article/70/3/307/568079 by guest on 30 September 2021 Fig. 7 Iron protects PAO1 biofilms from eradication by tobramycin and tigecycline. PAO1 biofilms were allowed to form on MBEC device pegs for 24 h in DTSB supplemented with 100 lM FeCl3.The biofilms were then washed in saline and then challenged for 24 h with varying concentrations of tobramycin (a) or tigecycline (b) in the presence or absence of 100 lM FeCl3 and/or 330 lM DSX as indicated. Biofilm formation was then determined as described in the Materials and methods. Error bars show the standard deviation of three independent experiments. Asterisks indicate the significance of increased resistance upon iron supplementation as determined by a two-tailed Student’s t-test: *P < 0.05, **P < 0.005, ***P < 0.005. form on the MBEC device pegs in iron-replete media, then tigecycline (Fig. 7). Curiously, planktonic PAO1 cells also gently washed, and exposed to decreasing concentration of showed extremely high resistance (MIC = 64 lgmLÀ1)in either tobramycin or tigecycline, in the presence or absence either high- or low-iron conditions (Table 3). This finding was of iron. Similar to what was observed with the ETESTâ, iron particularly surprising, as the MIC as determined by ETESTâ, enhanced the ability of planktonic cells in the MBEC device which measures resistance of bacteria growing on agar, was to survive antibiotic treatment, and addition of DSX elimi- much lower (c. 10 and 5 lgmLÀ1 in high- and low-iron media, nated this effect (Table 3). Moreover, iron increased the respectively – Fig. 2b). In spite of the high resistance of MBEC of the PAO1 biofilms from 2 lgmLÀ1 in low iron to planktonic cells to tigecycline in this assay, addition of DSX 8 lgmLÀ1 in high iron (Fig. 7a). Previous studies demon- greatly reduced the MIC of tigecycline for planktonic PAO1 (to À strated that subinhibitory concentrations of aminoglyco- 8 lgmL 1 – Table 3). These data demonstrate the complex sides, including tobramycin, induced biofilm formation by relationship between iron and P. aeruginosa antimicrobial P. aeruginosa via increased synthesis of the cylic di-GMP resistance, and further delineate the apparent mechanisms (c-di-GMP) second messenger (Hoffman et al., 2005). In by which iron enhances resistance to tobramycin and tigecyclin. agreement with these earlier studies, we also observed a large induction in biofilm formation at subinhibitory concen- À Mucoidy per se has no effect on antibiotic resistance of trations of tobramycin (< 2 lgmL 1 – Fig. 7a). Interestingly, P. aeruginosa induction of biofilm formation by tobramycin was eliminated in the absence of iron (Fig. 7a), and addition of the DSX iron A hallmark of chronic P. aeruginosa infection of CF patients chelator eliminated tobramycin-induced biofilm formation in is the appearance of a mucoid phenotype, often due to the iron-replete cultures (Fig. 7). Combined, these data suggest mutation of one of the muc genes resulting in overproduc- a link between iron regulation and c-di-GMP signaling in tion of the polysaccharide alginate (Fyfe & Govan, 1980; Gill mediating resistance of PAO1 biofilms to tobramycin. et al., 1987; Boucher et al., 1997). It has been proposed In contrast to what was observed with tobramycin, PAO1 that excessive alginate production by these strains contrib- biofilms were highly resistant to tigecycline regardless of iron utes to antibiotic resistance by providing a physical barrier to supplementation (Fig. 7b), and addition of DSX exerted no antibiotic penetration (Hatch & Schiller, 1998), but this significant effect on biofilm formation in the presence of hypothesis has not yet been directly tested. Analysis of

Pathogens and Disease (2014), 70, 307–320, © 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved 315 Iron increases P. aeruginosa antibiotic resistance A.G. Oglesby-Sherrouse et al.

results indicate that increased antibiotic resistance of mucoid clinical CF isolates is largely due to other charac- teristics of mucoid P. aeruginosa strains, perhaps hypermutability, rather than increased secretion of alginate. Combined with the observed effects of iron chelators on mucoid clinical isolates from CF patients, as well as data on non-CF isolates from previous studies in other laboratories (Moreau-Marquis et al., 2009), these results suggest FDA-approved iron chelators, such as DSX, may potentiate the activity of tobramycin during CF lung infections.

Discussion Downloaded from https://academic.oup.com/femspd/article/70/3/307/568079 by guest on 30 September 2021 Antimicrobial resistance is an ever-increasing problem, and the requirement for iron by pathogenic bacteria provides an ideal target for antimicrobial agents. Yet the time required to develop novel, effective, and safe therapies is a hurdle to antibiotic drug discovery. As such, repurposing already-approved therapies to improve the efficacy of available antibacterials is a favorable option for extending the current antimicrobial arsenal. Iron chelation therapy has been used for more than 40 years to treat iron overload disorders (Cohen, 2006). Because of their proven efficacy and relative safety in humans, both DFO and DSX are ideal candidates for enhancing current antibiotic therapies. Previous studies demonstrated that concurrent treatment with the FDA-approved iron chelator DSX, sold in the US under the brand â name Exjade , may increase the effectiveness of tobramycin in controlling P. aeruginosa infections in the CF lung. DSX was previously shown to block biofilm formation by P. aeruginosa when used in combination with tobramycin (Moreau-Marquis et al., 2009). In this study, we demonstrate that iron chelation increases susceptibility of P. aeruginosa growing in iron-replete conditions to certain antibiotics, namely tobramycin and tigecycline. Additionally, Fig. 8 Mucoidy per se does not increase resistance to tobramycin or â our study further corroborates previous studies showing that tigecycline. ETEST s were used as described in the Materials and iron chelation enhances activity of tobramycin against methods to determine the MIC of tobramycin and tigecycline toward P. aeruginosa biofilms. Moreover, we show that iron chela- PAO1 and the indicated deletion mutants growing on DTSB agar tion eliminates the induction of P. aeruginosa biofilm forma- medium without (white bars) and with (gray bars) FeCl supplementa- 3 tion by subinhibitory concentrations of tobramycin, a tion. Error bars show the standard deviation in MICs from three previously reported phenomenon that likely complicates independent experiments. tobramycin treatment for chronic CF lung infections (Hoff- man et al., 2005). Combined with previous reports, this antibiotic resistance of our clinical isolates showed no study strengthens the argument for development of correlation between mucoidy and antibiotic resistance for FDA-approved iron chelators for the treatment for multidrug any of the antibiotics we tested (Fig. 1 – P.S. 3 and P.S. 7 resistant microbial infections, including P. aeruginosa infec- are mucoid; P.S. 5 is not). This was by no means an tions in the CF lung. exhaustive collection of CF clinical isolates, though. Fur- Iron concentrations in the CF lung vary and measure thermore, numerous factors drive the evolution of P. aeru- anywhere between 2 and 130 lM (Mateos et al., 1998; ginosa during decades-long chronic CF infections, which Stites et al., 1998, 1999; Reid et al., 2004; Gray et al., can result in large variations in the ability of infecting strains 2010; Gifford et al., 2011) – and are correlated with the to resist different antibiotics. We therefore sought to directly amount of inflammation and damage to lung tissues. determine the effects of alginate overproduction on antibiotic Although P. aeruginosa encounters differing iron concen- resistance by testing isogenic muc mutants of PAO1. As trations both temporally and spatially during pathogenesis, shown in Fig. 8, the ΔmucA and ΔmucB mutants are no the importance of iron acquisition and iron-dependent more resistant than the wild-type strain to either tobramycin regulation for a successful infection is well established or tigecycline. Moreover, the effects of iron and DSX on (Cox, 1982; Meyer et al., 1996; Takase et al., 2000a, b). these strains are similar to what was observed for the This study further establishes the importance of iron during wild-type PAO1 parent strain (Fig. 8). Together, these pathogenesis, as higher iron concentrations allow for

316 Pathogens and Disease (2014), 70, 307–320, © 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved A.G. Oglesby-Sherrouse et al. Iron increases P. aeruginosa antibiotic resistance increased resistance to certain antibiotics. The specific ferrous iron. In fact, the PvdS sigma factor required for mechanism(s) for iron-induced antibiotic resistance pyoverdine biosynthesis is repressed in low-oxygen condi- observed here is unclear, but our data suggest the mech- tions (Ochsner et al., 1996). Recent studies further support anisms for iron-induced resistance to tigecycline and tobra- this notion, showing an inverse relationship between CF mycin are distinct. Interestingly, heme was unable to induce lung function and ferrous iron content in the lungs of CF the same increase in tobramycin and tigecycline resistance patients (Hunter et al., 2013). The study from Hunter et al., as free iron, even though heme can serve as an iron source further showed that DSX, which chelates ferric iron, may for P. aeruginosa, suggesting the method in which P. aeru- not eliminate biofilms as adequately in the microaerobic ginosa acquires iron is important for enhancing antibiotic environment of the CF lung. Thus, the development of resistance. While pyoverdine-mediated iron uptake allowed ferrous iron chelators should also be considered for for some increased resistance to tigecycline (Fig. 3), our treatment of CF lung infections, as well as other persistent analysis of several iron uptake mutants did not reveal an infections involving biofilms. Our study does indicate, iron uptake system that was responsible for iron-induced however, that DSX could potentiate tobramycin killing in Downloaded from https://academic.oup.com/femspd/article/70/3/307/568079 by guest on 30 September 2021 tobramycin resistance. P. aeruginosa possesses a number low-oxygen environments. More work, particularly in animal of iron acquisition systems, however; thus, it is possible that models, is clearly needed to define how iron chelation a system not analyzed in our study is responsible for therapy may be developed for treatment of P. aeruginosa iron-induced tobramycin resistance. Alternatively, iron could infections. be playing a more complex role, either through signaling or It is interesting that DFO blocked enhanced resistance to interaction with the antibiotics themselves. More studies will both tobramycin and tigecycline, because it can be used as be necessary to understand the mechanism(s) underlying an iron source by P. aeruginosa (Brock et al., 1988). This this phenomenon. suggests the receptor for DFO is not present under the We were surprised to find that, while deletion of the conditions used in this study, similar to one of the receptors genes for pyoverdine biosynthesis led to a decrease in for ferripyoverdine, FpvB, which is only evident in the tigecycline resistance, addition of purified pyoverdine to presence of certain carbon sources (Ghysels et al., 2004). the ΔpvdA mutant did not complement this defect. Growth curves suggest that DFO, unlike DSX, does not Assuming a 100% yield of pyoverdine during our purifica- inhibit growth in either DTSB or CAA (Fig. S4). However, the tion, the concentration of pyoverdine obtained from the effects of iron supplementation on growth in these media supernatant of an 18-h DTSB low-iron culture of PAO1 differed and indicated that CAA is a much more iron- should be c. 0.04 lM. However, it is unlikely that we depleted media that DTSB. Previous data indicate that actually achieved a 100% yield during our purification of DFO is less efficient than DSX at blocking biofilm formation pyoverdine. Further, this amount does not take into in combination with tobramycin by P. aeruginosa account how much pyoverdine was taken up by PAO1 (Moreau-Marquis et al., 2009). Combined with the lower as a means of iron acquisition prior to purification. Based solubility and more difficult delivery of DFO (Manning et al., on these caveats, we supplemented the DTSB agar plates 2009), these findings suggest DSX is a more appropriate with a higher level of purified pyoverdine (2 lM) than what candidate for supplementary antibacterial therapy. Another was calculated to ensure the levels would be ample curious observation is that addition of EDDHA has no effect enough to mediate iron acquisition over the 20-h course of on resistance of P. aeruginosa to antibiotic resistance the ETEST experiments. It is possible, however, that the (Fig. 2). Each of these chelators can complex with other levels of pyoverdine we added were not high enough to cations to varying lesser degrees than iron, however, and complement the defect of the ΔpvdA mutant. Alternatively, the effects these interactions may have on iron and other our data may suggest that other factors, such as biosyn- metal availability in our experiments are unclear. Thus, an thesis of pyoverdine, contribute to tigecycline resistance. investigation into the role that other metals play in antibiotic These data are interesting to consider in the light of the tolerance may also yield interesting results. fact that some strains of P. aeruginosa lose the ability to This study also distinguished between the roles of synthesize pyoverdine during CF lung infection (De Vos mucoidy and biofilm formation in the ability of P. aeruginosa et al., 2001). It has been hypothesized that pyoverdine to resist antibiotic treatment. While biofilm formation expec- mutants may benefit from pyoverdine-producing strains tantly increased resistance of PAO1 to both tigecycline and also present in the CF lung (Visca et al., 2007), but our tobramycin (Fig. 7), inducing mucoidy through genetic data suggest that these pyoverdine-deficient ‘cheaters’ deletion of the mucA or mucB gene caused no increase in lack some of the signaling pathways present in pyover- antibiotic resistance (Fig. 8). These data further support dine-producing strains of P. aeruginosa. However, our previous findings that biofilms (Hoiby et al., 2010) and finding that pyoverdine has a negative effect on resistance mucoidy in the CF lung (Waine et al., 2008) impart charac- to tobramycin may provide some additional insight into the teristics to P. aeruginosa that contribute to antibiotic resis- evolutionary pressures that might result in loss of pyover- tance, distinct from a physical barrier generated by dine, as CF lung-infecting strains of P. aeruginosa are hypersecretion of polysaccharide. Especially interesting in frequently exposed to tobramycin. this study, however, was the finding that iron was required Loss of pyoverdine production by P. aeruginosa in CF for induction of biofilm formation by subinhibitory concen- lung may also be due to decreased oxygen concentrations, trations of tobramycin. Previous work nicely demonstrated resulting in the conversion of insoluble ferric iron to soluble the role of the c-di-GMP second messenger in this

Pathogens and Disease (2014), 70, 307–320, © 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved 317 Iron increases P. aeruginosa antibiotic resistance A.G. Oglesby-Sherrouse et al. phenomenon (Hoffman et al., 2005), and our data indicate Banin E, Vasil ML & Greenberg EP (2005) Iron and Pseudomonas iron regulation is also involved in tobramycin-induced biofilm aeruginosa biofilm formation. P Natl Acad Sci USA 102: 11076– formation. Iron regulation has already been shown to play a 11081. role in PAO1 biofilm formation (Banin et al., 2005), thus Barker AP, Vasil AI, Filloux A, Ball G, Wilderman PJ & Vasil ML presenting the intriguing hypothesis that iron and c-di-GMP (2004) A novel extracellular phospholipase C of Pseudomonas aeruginosa is required for phospholipid chemotaxis. Mol Microbiol regulatory pathways overlap. A more recent study supports 53: 1089–1098. this idea, showing that c-di-GMP levels are positively Barton HA, Johnson Z, Cox CD, Vasil AI & Vasil ML (1996) Ferric correlated with pyoverdine and pyochelin production (Fran- uptake regulator mutants of Pseudomonas aeruginosa with gipani et al., 2013). More studies will be necessary to distinct alterations in the iron-dependent repression of exotoxin elucidate how these regulatory pathway(s) intersect. A and siderophores in aerobic and microaerobic environments. Tigecycline is a glycylcycline antibiotic, a class of antibi- Mol Microbiol 21: 1001–1017. otics derived from tetracycline to overcome resistance Beare PA, For RJ, Martin LW & Lamont IL (2003) Siderophore-med- mediated by efflux pumps and/or ribosomal protection iated cell signaling in Pseudomonas aeruginosa: divergent Downloaded from https://academic.oup.com/femspd/article/70/3/307/568079 by guest on 30 September 2021 (Sum & Petersen, 1999). Our data indicate co-treatment pathways regulate virulence factor production and siderophore – with iron chelators reduces the MIC of tigecycline for receptor synthesis. Mol Microbiol 47: 195 207. À Bendig JW, Kyle PW, Giangrande PL, Samson DM & Azadian BS P. aeruginosa well below the reported levels of 8 lgmL 1 (1987) Two neutropenic patients with multiple resistant (Petersen et al., 1999). However, this effect was not Pseudomonas aeruginosa septicaemia treated with ciprofloxacin. observed when P. aeruginosa was grown as a biofilm, a J R Soc Med 80: 316–317. critical factor in many P. aeruginosa infections. Tigecycline Berlutti F, Morea C, Battistoni A, Sarli S, Cipriani P, Superti F, is often used to treat Acinetobacter baumanii, an opportu- Ammendolia MG & Valenti P (2005) Iron availability influences nistic pathogen causing infections of the urinary tract, aggregation, biofilm, adhesion and invasion of Pseudomonas bloodstream and other sites with increasing resistance to aeruginosa and Burkholderia cenocepacia. Int J Immunopathol multiple antibiotics (Mihu & Martinez, 2011). In spite of its Pharmacol 18: 661–670. efficacy for treating many infections, resistance rates and Boucher JC, Yu H, Mudd MH & Deretic V (1997) Mucoid MICs for tigecycline are increasing among multidrug resis- Pseudomonas aeruginosa in cystic fibrosis: characterization of muc mutations in clinical isolates and analysis of clearance in a tant strains of A. baumanii (Navon-Venezia et al., 2007). mouse model of respiratory infection. Infect Immun 65: 3838– Acinetobacter baumanii is closely related to P. aeruginosa, 3846. raising the possibility that iron and iron chelation may have Brock JH, Liceaga J & Kontoghiorghes GJ (1988) The effect of the same effects in this emerging pathogen. Many of the synthetic iron chelators on bacterial growth in human serum. proposed novel therapies for A. baumanii, P. aeruginosa, FEMS Microbiol Immunol 1: 55–60. and other highly antibiotic resistant bacteria remain years Ceri H, Olson ME, Stremick C, Read RR, Morck D & Buret A (1999) away from clinical use (Allahverdiyev et al., 2011; Falagas The Calgary Biofilm Device: new technology for rapid determina- et al., 2011; Mihu & Martinez, 2011; Moradpour & Gha- tion of antibiotic susceptibilities of bacterial biofilms. J Clin – semian, 2011; Wiener-Kronish & Pittet, 2011), whereas iron Microbiol 37: 1771 1776. chelators have a long history of clinical efficacy and could be Cohen AR (2006) New advances in iron chelation therapy. Hematology/the Education Program of the American Society of implemented at the bedside in much less time (Cohen, Hematology American Society of Hematology Education Program 2006; Manning et al., 2009). As such, iron chelators may be 42–47. a promising avenue for curbing the growing problem of Cornelis P, Anjaiah V, Koedam N, Delfosse P, Jacques P, Thonart antimicrobial resistance. P & Neirinckx L (1992) Stability, frequency and multiplicity of transposon insertions in the pyoverdine region in the chromo- somes of different fluorescent pseudomonads. J Gen Microbiol Acknowledgments 138: 1337–1343. â Cox CD (1982) Effect of pyochelin on the virulence of Pseudomonas We would like to thank Novartis for supplying DSX (Exjade ) aeruginosa. Infect Immun 36: 17–23. for use in this study (Valerie Schule, Material Transfer Cunliffe HE, Merriman TR & Lamont IL (1995) Cloning and Manager, Novartis Pharma AG, WSJ-386.12.06, 4002 characterization of pvdS, a gene required for pyoverdine synthe- Basel, Switzerland). We also thank Pradeep Singh for sis in Pseudomonas aeruginosa: PvdS is probably an alternative supplying clinical strains of P. aeruginosa, as well as Yuta sigma factor. J Bacteriol 177: 2744–2750. Okkotsu and Michael Schurr for sharing their ΔmucA and De Vos D, De Chial M, Cochez C, Jansen S, Tummler B, Meyer ΔmucB mutants. This work was supported by National JM & Cornelis P (2001) Study of pyoverdine type and production Institutes of Health grant R37-AI015940-33 (to MLV) and by Pseudomonas aeruginosa isolated from cystic fibrosis K22-AI089776-01 and laboratory start-up funds from the patients: prevalence of type II pyoverdine isolates and accumu- University of Maryland School of Pharmacy (to AGOS). lation of pyoverdine-negative mutations. Arch Microbiol 175: 384–388. Driscoll JA, Brody SL & Kollef MH (2007) The epidemiology, References pathogenesis and treatment of Pseudomonas aeruginosa infections. Drugs 67: 351–368. Allahverdiyev AM, Kon KV, Abamor ES, Bagirova M & Rafailovich M Falagas ME & Bliziotis IA (2007) Pandrug-resistant gram-negative (2011) Coping with antibiotic resistance: combining nanoparticles bacteria: the dawn of the post-antibiotic era? Int J Antimicrob with antibiotics and other antimicrobial agents. Expert Rev Agents 29: 630–636. Anti-Infect Ther 9: 1035–1052.

318 Pathogens and Disease (2014), 70, 307–320, © 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved A.G. Oglesby-Sherrouse et al. Iron increases P. aeruginosa antibiotic resistance

Falagas ME, Karageorgopoulos DE & Nordmann P (2011) Thera- Marlovits TC, Haase W, Herrmann C, Aller SG & Unger VM (2002) peutic options for infections with Enterobacteriaceae producing The membrane protein FeoB contains an intramolecular G protein carbapenem-hydrolyzing enzymes. Future Microbiol 6: 653–666. essential for Fe(II) uptake in bacteria. P Natl Acad Sci USA 99: FitzSimmons SC (1993) The changing epidemiology of cystic 16243–16248. ﬁbrosis. J Pediatr 122: 1–9. Marshall B, Stintzi A, Gilmour C, Meyer JM & Poole K (2009) Frangipani E, Visaggio D, Heeb S, Kaever V, Camara M, Visca P & Citrate-mediated iron uptake in Pseudomonas aeruginosa: Imperi F (2013) The Gac/Rsm and cyclic-di-GMP signalling involvement of the citrate-inducible FecA receptor and the FeoB networks coordinately regulate iron uptake in Pseudomonas ferrous iron transporter. Microbiology 155: 305–315. aeruginosa. Environ Microbiol DOI: 10.1111/1462-2920.12164. Mateos F, Brock JH & Perez-Arellano JL (1998) Iron metabolism in Fyfe JA & Govan JR (1980) Alginate synthesis in mucoid Pseudo- the lower respiratory tract. Thorax 53: 594–600. monas aeruginosa: a chromosomal locus involved in control. J Meyer J, Neely A, Stintzi A, Georges C & Holder I (1996) Pyoverdin Gen Microbiol 119: 443–450. is essential for virulence of Pseudomonas aeruginosa. Infect Ghysels B, Dieu BT, Beatson SA, Pirnay JP, Ochsner UA, Vasil ML Immun 64: 518–523.

& Cornelis P (2004) FpvB, an alternative type I ferripyoverdine Meyer JM, Stintzi A, De Vos D, Cornelis P, Tappe R, Taraz K & Downloaded from https://academic.oup.com/femspd/article/70/3/307/568079 by guest on 30 September 2021 receptor of Pseudomonas aeruginosa. Microbiology 150: 1671– Budzikiewicz H (1997) Use of siderophores to type pseudomo- 1680. nads: the three Pseudomonas aeruginosa pyoverdine systems. Gifford AH, Miller SD, Jackson BP, Hampton TH, O’Toole GA, Microbiology 143(Pt 1): 35–43. Stanton BA & Parker HW (2011) Iron and CF-related anemia: Mihu MR & Martinez LR (2011) Novel therapies for treatment of expanding clinical and biochemical relationships. Pediatr multi-drug resistant Acinetobacter baumannii skin infections. Pulmonol 46: 160–165. Virulence 2: 97–102. Gill JF, Deretic V & Chakrabarty AM (1987) Alginate production by Moradpour Z & Ghasemian A (2011) Modified phages: novel the mucoid Pseudomonas aeruginosa associated with cystic antimicrobial agents to combat infectious diseases. Biotechnol fibrosis. Microbiol Sci 4: 296–299. Adv 29: 732–738. Gray RD, Duncan A, Noble D, Imrie M, O’Reilly DS, Innes JA, Moreau-Marquis S, Bomberger JM, Anderson GG, Swiate- Porteous DJ, Greening AP & Boyd AC (2010) Sputum trace cka-Urban A, Ye S, O’Toole GA & Stanton BA (2008) The metals are biomarkers of inflammatory and suppurative lung DeltaF508-CFTR mutation results in increased biofilm formation disease. Chest 137: 635–641. by Pseudomonas aeruginosa by increasing iron availability. Am J Hatch RA & Schiller NL (1998) Alginate lyase promotes diffusion of Physiol Lung Cell Mol Physiol 295: L25–L37. aminoglycosides through the extracellular polysaccharide of Moreau-Marquis S, O’Toole GA & Stanton BA (2009) Tobramycin mucoid Pseudomonas aeruginosa. Antimicrob Agents Chemother and FDA-approved iron chelators eliminate Pseudomonas aeru- 42: 974–977. ginosa biofilms on cystic fibrosis cells. Am J Respir Cell Mol Biol Heinz U, Hegetschweiler K, Acklin P, Faller B, Lattmann R & 41: 305–313. Schnebli HP (1999) 4-[3,5-Bis(2-hydroxyphenyl)-1,2, 4-tria- Navon-Venezia S, Leavitt A & Carmeli Y (2007) High tigecycline zol-1-yl]-benzoic acid: a novel efficient and selective iron resistance in multidrug-resistant Acinetobacter baumannii. J (iii) complexing agent. Angew Chem Int Ed Engl 38: 2568– Antimicrob Chemother 59: 772–774. 2570. Ochsner UA, Johnson Z, Lamont IL, Cunliffe HE & Vasil ML (1996) Hoffman LR, D’Argenio DA, MacCoss MJ, Zhang Z, Jones RA & Exotoxin A production in Pseudomonas aeruginosa requires the Miller SI (2005) Aminoglycoside antibiotics induce bacterial iron-regulated pvdS gene encoding an alternative sigma factor. biofilm formation. Nature 436: 1171–1175. Mol Microbiol 21: 1019–1028. Hoiby N, Bjarnsholt T, Givskov M, Molin S & Ciofu O (2010) Ochsner UA, Johnson Z & Vasil ML (2000) Genetics and regulation Antibiotic resistance of bacterial biofilms. Int J Antimicrob Agents of two distinct haem-uptake systems, phu and has,inPseudo- 35: 322–332. monas aeruginosa. Microbiology 146(Pt 1): 185–198. Holloway BW (1955) Genetic recombination in Pseudomonas Ochsner UA, Wilderman PJ, Vasil AI & Vasil ML (2002) GeneChip aeruginosa. J Gen Microbiol 13(3): 572–581. expression analysis of the iron starvation response in Pseudo- Hunter RC, Asfour F, Dingemans J, Osuna BL, Samad T, Malfroot monas aeruginosa: identification of novel pyoverdine biosynthesis A, Cornelis P & Newman DK (2013) Ferrous iron is a significant genes. Mol Microbiol 45: 1277–1287. component of bioavailable iron in cystic fibrosis airways. MBio Oglesby AG, Farrow JM, Lee JH, Tomaras AP, Greenberg EP, 4(4): e00557–13. doi: 10.1128/mBio.00557-13. Pesci EC & Vasil ML (2008) The influence of iron on Pseudomo- Lamont IL, Beare PA, Ochsner U, Vasil AI & Vasil ML (2002) nas aeruginosa physiology: a regulatory link between iron and Siderophore-mediated signaling regulates virulence factor pro- quorum sensing. J Biol Chem 283: 15558–15567. duction in Pseudomonas aeruginosa. P Natl Acad Sci USA 99: Parad RB, Gerard CJ, Zurakowski D, Nichols DP & Pier GB (1999) 7072–7077. Pulmonary outcome in cystic fibrosis is influenced primarily by Lyczak JB, Cannon CL & Pier GB (2000) Establishment of mucoid Pseudomonas aeruginosa infection and immune status Pseudomonas aeruginosa infection: lessons from a versatile and only modestly by genotype. Infect Immun 67: 4744–4750. opportunist. Microbes Infect 2: 1051–1060. Patriquin GM, Banin E, Gilmour C, Tuchman R, Greenberg EP & Macia MD, Blanquer D, Togores B, Sauleda J, Perez JL & Oliver A Poole K (2008) Influence of quorum sensing and iron on twitching (2005) Hypermutation is a key factor in development of motility and biofilm formation in Pseudomonas aeruginosa. J multiple-antimicrobial resistance in Pseudomonas aeruginosa Bacteriol 190: 662–671. strains causing chronic lung infections. Antimicrob Agents Che- Petersen PJ, Jacobus NV, Weiss WJ, Sum PE & Testa RT (1999) In mother 49: 3382–3386. vitro and in vivo antibacterial activities of a novel glycylcycline, the Manning T, Kean G, Thomas J, Thomas K, Corbitt M, Gosnell D, 9-t-butylglycylamido derivative of minocycline (GAR-936). Anti- Ware R, Fulp S, Jarrard J & Phillips D (2009) Iron chelators in microb Agents Chemother 43: 738–744. medicinal applications – chemical equilibrium considerations in Pruitt JBA, McManus AT, Kim SH & Goodwin CW (1998) Burn pharmaceutical activity. Curr Med Chem 16: 2416–2429. wound infections: current status. World J Surg 22: 135–145.

Pathogens and Disease (2014), 70, 307–320, © 2014 Federation of European Microbiological Societies. Published by John Wiley & Sons Ltd. All rights reserved 319 Iron increases P. aeruginosa antibiotic resistance A.G. Oglesby-Sherrouse et al.

Reid DW, Withers NJ, Francis L, Wilson JW & Kotsimbos TC (2002) variables, mucoid phenotype, and antimicrobial resistance in Iron deficiency in cystic fibrosis: relationship to lung disease Pseudomonas aeruginosa. J Clin Microbiol 46: 3491–3493. severity and chronic Pseudomonas aeruginosa infection. Chest Wiener-Kronish JP & Pittet JF (2011) Therapies against virulence 121: 48–54. products of Staphylococcus aureus and Pseudomonas aerugin- Reid DW, Lam QT, Schneider H & Walters EH (2004) Airway iron osa. Semin Respir Crit Care Med 32: 228–235. and iron-regulatory cytokines in cystic fibrosis. Eur Respir J 24: Wilderman PJ, Vasil AI, Johnson Z, Wilson MJ, Cunliffe HE, Lamont 286–291. IL & Vasil ML (2001) Characterization of an endoprotease (PrpL) Rogers HJ (1973) Iron-binding catechols and virulence in Escher- encoded by a PvdS-regulated gene in Pseudomonas aeruginosa. ichia coli. Infect Immun 7: 438–444. Infect Immun 69: 5385–5394. Ryder C, Byrd M & Wozniak DJ (2007) Role of polysaccharides Wilderman PJ et al. (2004) Identification of tandem duplicate in Pseudomonas aeruginosa biofilm development. Curr Opin regulatory small RNAs in Pseudomonas aeruginosa involved in Microbiol 10: 644–648. iron homeostasis. P Natl Acad Sci USA 101(26): 9792–9797. Schwyn B & Neilands JB (1987) Universal chemical assay for the Willcox MD (2007) Pseudomonas aeruginosa infection and inflam-

detection and determination of siderophores. Anal Biochem 160: mation during contact lens wear: a review. Optom Vis Sci 84: Downloaded from https://academic.oup.com/femspd/article/70/3/307/568079 by guest on 30 September 2021 47–56. 273–278. Singh PK (2004) Iron sequestration by human lactoferrin stimulates Wilson MJ & Lamont IL (2000) Characterization of an ECF sigma P. aeruginosa surface motility and blocks biofilm formation. factor protein from Pseudomonas aeruginosa. Biochem Biophys Biometals 17: 267–270. Res Commun 273: 578–583. Singh PK, Parsek MR, Greenberg EP & Welsh MJ (2002) A Worlitzsch D, Tarran R, Ulrich M et al. (2002) Effects of reduced component of innate immunity prevents bacterial biofilm devel- mucus oxygen concentration in airway Pseudomonas infections opment. Nature 417: 552–555. of cystic fibrosis patients. J Clin Invest 109: 317–325. Stites SW, Walters B, O’Brien-Ladner AR, Bailey K & Wesselius LJ Yunta F, Garcia-Marco S, Lucena JJ, Gomez-Gallego M, Alcazar R (1998) Increased iron and ferritin content of sputum from patients & Sierra MA (2003) Chelating agents related to ethylenediamine with cystic fibrosis or chronic bronchitis. Chest 114: 814–819. bis(2-hydroxyphenyl)acetic acid (EDDHA): synthesis, character- + Stites SW, Plautz MW, Bailey K, O’Brien-Ladner AR & Wesselius LJ ization, and equilibrium studies of the free ligands and their Mg2 , + + + (1999) Increased concentrations of iron and isoferritins in the Ca2 ,Cu2 , and Fe3 chelates. Inorg Chem 42: 5412–5421. lower respiratory tract of patients with stable cystic fibrosis. Am J Respir Crit Care Med 160: 796–801. Sum PE & Petersen P (1999) Synthesis and structure-activity Supporting Information relationship of novel glycylcycline derivatives leading to the Additional Supporting Information may be found in the online – discovery of GAR-936. Bioorg Med Chem Lett 9: 1459 1462. version of this article: Takase H, Nitanai H, Hoshino K & Otani T (2000a) Impact of siderophore production on Pseudomonas aeruginosa infections in Fig. S1. Iron increases resistance of P. aeruginosa to immunosuppressed mice. Infect Immun 68: 1834–1839. tobramycin. Takase H, Nitanai H, Hoshino K & Otani T (2000b) Requirement of Fig. S2. Pyoverdine allows for optimal growth in the the Pseudomonas aeruginosa tonB gene for high-affinity iron presence of Exjade. acquisition and infection. Infect Immun 68: 4498–4504. Fig. S3. The ΔpchEF mutant displays a compensatory Visca P, Imperi F & Lamont IL (2007) Pyoverdine siderophores: from biogenesis to biosignificance. Trends Microbiol 15: 22–30. increase in pyoverdine production. Waine DJ, Honeybourne D, Smith EG, Whitehouse JL & Dowson Fig. S4. DFO and DSX have different effects on growth in CG (2008) Association between hypermutator phenotype, clinical different media.