Overexpression of the MexXY Multidrug Efflux System Correlates with Deficient Pyoverdine Production in Pseudomonas aeruginosa

Multidrug-resistant Pseudomonas aeruginosa poses a serious problem due to hospital- and healthcare-associated infections. A major drug resistance mechanism of P. aeruginosa involves active efflux via resistance nodulation cell division (RND)-type multidrug efflux pumps of which MexXY is increasingly recognized as a primary determinant of aminoglycoside resistance in P. aeruginosa. MexXY overexpression is often observed in drug-resistant P. aeruginosa clinical isolates. MexXY deficiency increased pyoverdine production in all four P. aeruginosa strains we tested. MexXY-overproducing multidrug-resistant P. aeruginosa PA7 exhibited the greatest effect among the strains. Complementation with a MexXY-expressing plasmid restored low-level pyoverdine production in a MexXY-deficient P. aeruginosa mutant from PA7, indicating that MexXY expression decreases pyoverdine production. Because P. aeruginosa produces pyoverdine to acquire iron, MexXY-deficient mutants might be more susceptible to iron deficiency than MexXY-producing strains or might require extra iron. High-risk clones of multidrug-resistant P. aeruginosa reportedly tend to be MexXY overproducers but defective pyoverdine producers. This study suggests that P. aeruginosa reduces production of a virulence factor after acquiring a drug resistance factor.


Introduction
Pseudomonas aeruginosa is a known opportunistic pathogen and a major threat in hospital and healthcare-associated environments [1]. Infections caused by P. aeruginosa are often difficult to treat; inappropriate chemotherapy readily selects multidrug-resistant P. aeruginosa strains against which very few agents are effective [2,3]. A major factor in the prominence of P. aeruginosa as a pathogen is its intrinsic resistance to various antibacterial agents [2,3]. One of most important chromosomally encoded antimicrobial resistance factors in P. aeruginosa is resistance nodulation cell-division (RND)-type multidrug efflux pumps [2,4]. Among these pumps, the MexXY system is the only significant determinant of efflux-mediated aminoglycoside resistance in P. aeruginosa [5]. In addition to aminoglycosides, MexXY mediates resistance to other clinically relevant drugs such as cefepime, ciprofloxacin, tigecycline, azithromycin, and colistin [5,6].
Worldwide epidemic outbreaks of infection with highly drug-resistant P. aeruginosa are often associated with various so-called international high-risk clones [7], many of which harbor chromosomal mutations that promote drug resistance mechanisms, such as MexXY overproduction [8]. These clones also often produce lower amounts of pyoverdine in vitro, which has been proposed as a potential biomarker [9]. Pyoverdines that facilitate acute infections by pseudomonads include fluorescent siderophores, which specifically chelate Fe 3+ with high affinity [10]. However, the types of gene mutations that contribute to defective pyoverdine production in highly multidrug-resistant P. aeruginosa clinical isolates remain unknown, as no major differences in pyoverdine gene clusters have been identified [11]. In this study, we conducted a detailed examination of the effect of MexXY on the production of pyoverdine in P. aeruginosa.
To confirm that the phenotypic changes due to mexXY deletion were actually due to loss of function of MexXY, we examined pyoverdine production by the complemented strain in which a mexXY-oprA expression plasmid was introduced into PA7 ∆mexXY-oprA, in comparison with the negative control strain (Figure 2). Upon addition of isopropyl β-D-1-thiogalactopyranoside (IPTG) to induce plasmid-driven mexXY-oprA expression, the mexXY-oprA-expressing strain produced 2.7-fold more pyoverdine than the negative control strain (p value: 0.008 [<0.05]). In contrast, without IPTG, no significant difference in pyoverdine production was observed between the strains (p value: 0.30 [>0.05]).    Table 1).
The error ranges shown in the bar graph are the standard error (n = 5). * Indicates p < 0.05. Cultures were supplemented with 5 mM IPTG by addition to Pseudomonas agar F to drive the T7(A1/04/03) promoter expression system fully [12,19].

Discussion
The results of this study suggest that the MexXY multidrug efflux system decreases pyoverdine production in P. aeruginosa, indicating that MexXY-producing P. aeruginosa cells require less iron than MexXY-deficient cells, because P. aeruginosa requires pyoverdine for survival when iron concentrations become low [20]. The promotion of mexXY expression under conditions of oxidative stress is similar to the case of P. aeruginosa infection of chronically inflamed lungs of cystic fibrosis (CF) patients [21], which also induces siderophore biosynthesis genes [22], possibly due to oxidative inactivation of the Fur-Fe 2+ complex [23]. A further study is thus necessary to elucidate the molecular mechanisms in more detail. A whole genome analyses of the studied P. aeruginosa strains can be one approach to use which might shed light and allow to get a broad picture.
Another possibility is that pyoverdine could be a substrate of MexXY, rather than MexXY decreasing pyoverdine production. The approach used for examining pyover-dine production does not rule out this possibility, since cells are taken from solid media, suspended in solution and then pyoverdine in the solution measured fluorometrically. Pyoverdine exported by cells on the solid media would presumably diffuse into the agar and thus might not be associated with cells collected for the assay although preliminary results from liquid culture supports the agar results in this study ( Figure S1). If MexXY promotes pyoverdine efflux, mexXY deletion mutants could have higher cytosolic pyoverdine concentrations than the parental strains and carry more pyoverdine into the assay. The results would be more convincing if the cells were grown in liquid media and the spent media was assayed for pyoverdine concentration. It is also noteworthy that the buffer used for the pyoverdine assay contain the only contaminated iron. The absorbance spectra of free pyoverdine and iron-loaded pyoverdine are different. The maximum absorbance of iron-loaded pyoverdine is 400 nm (as used for excitation in the assay), but the maximum absorbance for unloaded pyoverdine is lower and 400 nm would only hit the shoulder. Therefore, having the only contaminated iron in the buffer might also impact the results of the assay.
Defective pyoverdine production is a biomarker of high epidemic risk P. aeruginosa clones [9], and multidrug resistant P. aeruginosa clinical isolates [5], including high epidemic risk clones [8], often overexpress mexXY. Decreased pyoverdine production could be, at least in part, due to mexXY overexpression in P. aeruginosa clinical isolates. We cannot rule out the possibility that other factors also contribute to defective pyoverdine production in P. aeruginosa clinical isolates. For example, mexXY expression does not appear to be the primary reason for defective pyoverdine production in P. aeruginosa K2153, a panaminoglycoside-resistant clinical isolate obtained from a CF patient (Figure 1). This study provides experimental evidence that upregulated expression of a drug-resistance factor leads to decreased production of a virulence factor in P. aeruginosa.

Assay of Pyoverdine Production
Pyoverdine production was assayed according to a previously reported method [9]. Bacteria grown overnight at 37 • C were suspended in 0.85% sterilized NaCl using a sterile wooden-axis cotton swab (Eiken Chemical Co., Ltd., Tokyo, Japan) and diluted to a final OD 600 of 0.001. Next, 100 µL of the diluted culture was plated uniformly onto Pseudomonas F agar using a spreading stick and incubated at 37 • C for 14 h. Colonies of P. aeruginosa grown on the agar were suspended in 0.85% sterilized NaCl using a sterile wooden-axis cotton swab. After the cell density (OD 600 ) (A) was measured, the suspension was centrifuged at 23 • C and 4000× g for 10 min. Each of three 200-µL aliquots from the supernatant was transferred into each of three wells of a 96-well plate (Costar ® no lid black, flat bottom, nontreated polystyrene (Corning Inc., Corning, NY, USA). The fluorescence emission for each well was measured (B) at an excitation wavelength of 400 nm and an emission wavelength of 460 nm using a SpectraMax iD3 multi-mode microplate reader (Molecular Devices, LLC, San Jose, CA, USA). Pyoverdine production was calculated by dividing (B) by (A). Each experiment was performed independently at least five times. When necessary, Pseudomonas agar F was supplemented with 5 mM IPTG (FUJIFILM Wako Pure Chemical Corp.). * Co-identical strain is defined as a strain when stocked in the previous lab(s).
Statistical analyses were performed using the R software version 4.0.4 (https://www. r-project.org/) with the Wilcoxon rank test or Steel-Dwass test. A p-value of <0.05 was judged as indicating statistical significance.