Hfq mutation confers increased cephalosporin resistance in Klebsiella pneumoniae

Klebsiella pneumoniae (K. pneumoniae), is an opportunistic pathogen raising significant public health concerns owing to its multi-drug resistance. Hfq, one of the main RNA-binding proteins, is a key post-transcriptional regulator. This protein is closely related to virulence and resistance in various pathogenic bacteria. Although the role of hfq in K. pneumoniae virulence has been explored, its influence on resistance remains largely unknown. The aim of this study was to investigate the role of hfq in the resistance of K. pneumoniae to cephalosporins. An hfq mutant was constructed, and its resistance to cephalosporins was investigated. The hfq mutant exhibited over 16-fold higher cephalosporin resistance than that exhibited by the wild type. Time-kill curve analysis showed that the hfq mutant could survive under higher concentrations of cephalosporins than the wild-type strain could. Quantitative RT-PCR showed that expression levels for 8 out of the 9 penicillin-binding proteins, which are the targets of cephalosporins, were downregulated in the hfq mutant. Taken together, contrary to its role in many other bacteria, hfq is involved in a negative regulation of K. pneumoniae resistance to cephalosporins by downregulating the expression of penicillin-binding proteins


INTRODUCTION
Klebsiella pneumoniae, a gram-negative bacterium, is a major opportunistic pathogen, causing a variety of human diseases such as pneumonia [1], urinary tract infection [2], bacterial meningitis [3], septicemia [4], and even liver abscess [5].Infections range from local to life-threatening disseminated diseases, especially at the hospital and community levels.Even with appropriate antibiotic treatment, in vulnerable patients the mortality rate of hospital-acquired pneumonia due to K. pneumoniae exceeds 50% [6].K. pneumoniae is one of the main causes of carbapenem-resistant bacterial infections, which have no effective treatment.This pathogen can resist third-generation cephalo-sporins and carbapenems, which are the last line of defense against severe infections [6].K. pneumoniae poses grave public health risks owing to its multidrugresistance and global reach.In light of the high pathogenicity and dangerous resistance of K. pneumoniae, it is necessary to understand more about the resistance mechanisms of this pathogen.
Hfq, an RNA-binding protein, is an important post-transcriptional regulator in various pathogens, including Pseudomonas aeruginosa [7], Listeria monocytogenes [8], Vibrio cholerae [9], Shigella flexneri [10], Salmonella [11,12], Neisseria gonorrhoeae [13], Neisseria meningitidis [14], Yersinia [15,16], Francisella tularensis [17], Erwinia amylovora [18] and Escherichia coli [19,20].There have been many reports of the effects of Hfq on bacterial growth, virulence and biofilm formation; this protein is a key virulence-and stress-response factor [21].Hfq facilitates the interaction between small non-coding RNAs (sRNAs) and mRNA, which can impact the stability and translation of mRNA.Through its influence on gene expression, Hfq affects multiple cellular processes and physiological reactions, particularly virulence and resistance [21][22][23][24].This vital role in regulating virulence has been described in a variety of bacteria.For K. pneumoniae, Chiang et al. [25] reported that the Hfq protein is a key factor in the regulation of gene expression, being critical for virulence, and is involved in many stress conditions.However, the relationship between hfq and drug resistance in K. pneumoniae remains largely unknown.
Cephalosporins are one of the most commonly used antibiotics, and K. pneumoniae resistance to this type of antibiotic is becoming a public health problem.Several explanations for cephalosporin resistance have been developed over the years.The expression of chromosome-or plasmid-mediated β-lactamases, which would bind and hydrolyze β-lactam antibiotics, including cephalosporins, could provide protection to penicillin-binding proteins (PBPs), the targets of cephalosporin, thereby conferring cephalosporin resistance [26].Alternatively, modifications to cephalosporin targets may influence resistance.The modification of PBPs was mediated by the expression of low-affinity PBPs or acquisition of supplementary cephalosporin-insensitive PBPs [27].Moreover, a number of recent studies have demonstrated that changes in PBP expression levels can impact cephalosporin resistance [28][29][30].
In the present study, we investigated the role of hfq in K. pneumoniae resistance to cephalosporins.We also examined PBP expression levels for indications of regulatory processes, since these proteins can directly influence cephalosporin resistance.

Strains, plasmids and growth conditions
The K. pneumoniae strains and plasmids used in this study are listed in Table 1.We used K. pneumoniae from the wild-type (WT) isolate K. pneumoniae WJ101, which was a kind gift from Y. F. Wang.All strains were maintained in Luria-Bertani agar (LB agar: 1% tryptone, 0.5% yeast extract, 1% NaCl and 3% agar) plates and stored at 4°C during the experimental period.Mueller-Hinton broth (MHB) was used as the medium for susceptibility testing.

Antibacterial agents
Antibiotics for plasmid maintenance were added into the medium at the following concentrations: kanamycin 60 µg/mL and tetracycline 40 µg/mL.All antimicrobial agent stock solutions were maintained at -20°C.

Construction of K. pneumoniae WJ101 hfqmutant and complemented strain
The genome sequence of K. pneumoniae NTUH-K2044 (GenBank accession number NC_012731.1)was retrieved from the NCBI GenBank database and used for primer design.All primers used in this study were synthesized commercially (Sangon Biotech Co., Ltd., Shanghai, China).These primers are listed in Table 2.
The deletion of the hfq gene from K. pneumoniae WJ101 chromosomal DNA was performed using the Red recombinase system [31].Since the ampicillin resistance marker was invalid for K. pneumoniae, a new recombinant plasmid, pKD46T, was constructed.The tetracycline resistance gene fragment was cut from pBR322 and inserted at the XmnI site of pKD46 [32].Thus, ampicillin resistance was replaced with tetra- cycline resistance.The preparation of K. pneumoniae electrocompetent cells was conducted following the procedure described by Wei et al. [31], with slight modification.Briefly, K. pneumoniae WJ101 was inoculated into a flask containing 100 mL of LB, and incubated on ice for 30 min after the cell density value (OD 600 ) was approximately 0.7.The K. pneumoniae WJ101 cells were collected by centrifugation at 5000×g for 10 min at 4°C, and washed twice using ice-cold sterile ddH 2 O. Next, the cells were resuspended in ice-cold sterile ddH 2 O to a final volume of 0.5 mL.
To construct the Δhfq mutant, a disruption cassette consisting of a kanamycin-resistance marker flanked by hup and hdown (the upstream and downstream fragments of hfq) was constructed by fusion PCR using the primer set KPhfq-N-F/R and KPhfq-C-F/R.Purified pKD46T was electroporated into competent K. pneumoniae cells using a Bio-Rad MicroPulser, followed by transformation of the disruption cassette.The recombinant K. pneumoniae were screened on LB plates supplemented with kanamycin.The pKD46T plasmid was eliminated from K. pneumoniae by culturing this recombinant strain at 42°C.The hfq-complemented strain was constructed according to the method reported by Jayol et al. [33], to demonstrate the specificity of deletion.Amplification of the hfq fragment from the genomic DNA of K. pneumoniae WJ101 was performed using primers hfqF and hfqR, which contained the EcoRI and XbaI cleavage sites, respectively.Next, this hfq fragment was digested with EcoRI and XbaI, and cloned into the pHSG298 plasmid, which contains a kanamycin-resistance gene.The recombinant plasmid pHSG298CH was transferred into K. pneumoniae WJ101 hfq mutant strains, yielding the complemented strains.The recombinant K. pneumoniae WJ101ΔHCH were screened on LB plates supplemented with kanamycin in higher concentrations (125 µg/mL) than was once used.All constructed strains were identified by PCR and further verified by DNA sequencing (Supplementary Figs.1S and 2S).

Antimicrobial susceptibility
Cefazolin sodium (CFZN), cefotaxime (CTX), cefuroxime sodium (CXMN) and ceftazidime (CAZ) were used in this study.These antimicrobial agents (all ≥98% pure) were purchased from the National Institute for the Control of Pharmaceutical and Biological Products, Beijing, China.Sterile water was used to prepare stock solutions of these drugs (1024 µg/mL), which were maintained at -20°C.
The minimum inhibitory concentrations (MICs) of cephalosporins against the K. pneumoniae strains were determined using the standard broth microdilution method, as described by the Clinical and Laboratory Standards Institute (CLSI, formerly the National Committee for Clinical Laboratory Standards).All test isolates were diluted with NaCl 0.85% medium (BIOMERIEUX) to a turbidity of 0.5 McFarland.These dilute samples were further diluted (1:1000) into fresh Mueller-Hinton (MH) medium in the plates, and incubated for 18h at 37°C.The MIC was defined as the lowest concentration of antibacterial agent that showed no visible growth compared to that of the drug-free control.Meanwhile, the minimum bactericidal concentrations (MBCs) were determined using the colony-counting method.The complete content of every dilution was spread on an LB agar plates and incubated at 37°C for 24 h.We counted the number of colonies on each plate.The MBC was defined as the lowest concentration of bactericide with which no colonies appear on the LB agar plate.Each assay was performed in triplicate.

Time-kill curves
Time-kill curves for CFZN and CXMN against WT K. pneumoniae WJ101 and the hfq mutant, respectively, were analyzed.The test isolates were diluted to 0.5 McFarland turbidity using NaCl 0.85% medium (BIOMERIEUX), and then inoculated (1:500) into fresh MH medium in 5-mL test tubes.The tubes contained 0.5×drug concentrations obtained from the previous experiment.We incubated these tubes at 37°C.At the 0, 4, 8, 12, and 24 h time points, a 100-µL aliquot was removed from each test tube and serially diluted 10-fold in sterile water.One hundred mL of each dilution was spread on the LB agar plates and incubated at 37°C for 18 h.Subsequently, we counted the colonies on each plate.Each assay was performed in triplicate.

PCR
We used PCR to confirm the presence of PBP genes in K. pneumoniae WJ101.Genomic DNA from WJ101 was extracted using the TIANamp bacteria DNA kit (TianGen), and this DNA was the PCR template.The reaction system consisted of 12.5 µL 2×Taq Master-Mix, 1 µL of each primer (PBPs-F/R primers, Table 2), 1 µL genomic DNA as template, and ddH 2 O, for a final volume of 25 µL.The PCR was performed using the GeneAmp PCR system 9700 (Applied Biosystem) according to following program: 5 min of predenaturation at 94°C, followed by 35 cycles of denaturation for 30 s at 94°C, primer annealing for 30 s at 55°C and extension for 30s at 72°C.The final step was extension for 5 min at 72°C.Five mL of each reaction mixture was electrophoresed in 1% agarose gel containing ethidium bromide (EB) at 125 V for 35 min.The DNA bands were imaged using Gel Doc TM XR System (Bio-Rad Laboratories).

Quantitative reverse transcription PCR
RNA was isolated from overnight bacterial cultures using TRIzol reagent (Invitrogen).The RNA yield and purity were determined using a NanoDrop ND-1000 (Thermo Scientific).The purified RNA was reverse transcribed to cDNA using an ImProm-II TM Reverse Transcription System (Promega).Quantitative reverse transcription PCR (qRT-PCR) was carried out to detect PBP gene expression levels in WJ101 and the hfq mutant.The primers and their target genes are listed in Table 2. RT-PCR was performed on an IQ5 realtime PCR detection system (Bio-Rad) using the SYBR Premix Ex Taq TM (TliRNaseH Plus) (Takara).Amplification was performed in a 25-μL reaction system consisting of 12.5 μL Ex Taq mixture, 0.6 μL of each primer, 1μL cDNA and ddH 2 O.We used the following program: predenaturation for 30 s at 95°C, followed by 40 cycles of denaturation for 30 s at 95°C and primer annealing for 30 s at 60°C.To acquire melting curves, we performed 71 denaturation-annealing cycles to confirm that this amplification was specific.The analyses of the expression levels were conducted on the following targets: PBP1a, PBP1b, PBP2, PBP3, PBP4, PBP5, PBP6a and PBP7.We normalized measurements to those for the housekeeping gene rpoB [34] and used the 2 -ΔΔt method.Each assay was performed in triplicate.

Statistical analysis
Statistical analyses of qRT-PCR data were performed using the Student's t-test for paired samples.A p value of less than 0.05 was considered significant.Plotting and p value calculation were facilitated by GraphPad Prism 5 (GraphPad Software, Incorporated, CA).

Hfq mutant of K. pneumoniae exhibits increased MIC for cephalosporin
To test the possible roles of the Hfq protein in K. pneumoniae resistance to cephalosporins, a deletion mutant of hfq was constructed.The hfq open reading frame (ORF) was deleted using a lambda red-based replacement system.PCR and DNA sequencing confirmed that the hfq deletion was successful (Supplementary Figs.1S and 2S).Next, the antibacterial activities of cephalosporins against K. pneumoniae and the hfq mutant were assessed in a microtiter plate model.The MIC values of the two strains against four cephalosporin antibiotics are shown in Table 3. MIC values for the WT strain ranged from 0.0625 of CAZ to 16 of CTX.For the hfq mutant, the MIC ranged from 1 of CAZ to 1024 of CFZN.The cephalosporin MIC against the hfq mutant was significantly higher than against WT K. pneumoniae.With respect to the WT MICs, the CTX, CFZN, CXMN and CAZ MICs for the hfq mutant increased by 32-, 1024-, 128-and 16-fold, respectively (Table 3).This effect was impaired by an hfq complementary strain (Table 3).

Increased MBC of the hfq-mutant of K. pneumoniae to cephalosporin
To further confirm the increased resistance of the hfq mutant to cephalosporin antibiotics, the MBCs for the hfq mutant and WT strains were tested.MBCs for CTX, CFZN, CXMN and CAZ against the WT strain were 32, 16, 4 and 0.25 µg/mL, respectively.MBCs for the hfq mutant were 512, 1024, 64 and 1 µg/mL, respectively (Table 4).Compared with those of the WT strain, the MBCs for these antibiotics in the hfq mutant increased between 4-fold (CAZ) and 64-fold (CFZN).These results demonstrated that the deletion of hfq increased the MBCs for K. pneumoniae.

Time-kill curves
To assess the role of hfq in K. pneumoniae cephalosporin insensitivity, the antimicrobial activities of CFZN and CXMN against WJ101 and its hfq mutant, respectively, were determined using the time-kill approach.
The results are shown in Fig. 1.As presented in the figure, although the curves of WJ101and the hfq mutant show a similar trend, the hfq mutant exhibited a slow-growing state when compared with WT WJ101.While CFZN exhibited excellent bactericidal activity against WJ101 at 1/2×hfq-mutant MIC during 12 h, the mutant exhibited only a minor growth inhibition during the same period.There was no apparent impact on the mutant when we treated it with CFZN at 1/2×WT MIC concentration, whereas WT WJ101 exhibited significant growth inhibition.As Fig. 1 illustrates, there was a parallel trend for CXMN against WJ101 and the hfq-mutant at 1/2×hfq-mutant MIC and 1/2×WT MIC.These time-kill curves confirmed that hfq plays an important role in resistance of K. pneumoniae to cephalosporins.

Decreased expression level of PBP genes in the hfq mutant
Bacterial resistance to cephalosporins is usually mediated by PBPs.Decreased expression of PBPs results in increased resistance to the antibiotics.To test whether the hfq deletion influenced the expression of PBPs, the distribution of PBPs and their expression were determined.Firstly, K. pneumoniae WJ101 was detected by PCR to confirm the presence of genes that encode PBPs.As shown in Fig. 2(a), all 9 genes encoding PBPs  were successfully detected in the WJ101 strain, indicating that these genes do exist in the strain and it was feasible to analyze their expression.To test the influence of hfq on PBPs, gene expression of was analyzed by qRT-PCR in WT and the hfq-mutant strain as described in the Materials and Methods.As shown in Fig. 2(b), compared with that in the WT strain, the expressions of most of these genes, including 1a, 1b, 2, 3, 5, 6a, 6b and 7, were decreased.The expression levels of these genes were inhibited from 0.3-to 5-fold.
Among these, PBP1b, PBP3, PBP6b and PBP7 showed the greatest decrease.The only exception was PBP4, which showed increased expression.

DISCUSSION
Cephalosporins, which have a broad spectrum of antimicrobial activity against a large variety of pathogens, are widely used as first-line drugs for both the preven-  tion and the treatment of bacterial infection.However, recent studies have shown that they have a reduced effect against many drug-resistant pathogens, such as K. pneumoniae.In the present work, we have investigated the role of hfq in the regulation of K. pneumoniae resistance to cephalosporins.One interesting finding here was that an hfq mutant exhibited increased resistance to cephalosporins in K. pneumoniae.
There have been many reports describing the specific mechanism of the antimicrobial effect of cephalosporins.In brief, this kind of antibiotic can inhibit the activity of the penicillin-binding proteins, which is necessary for the biosynthesis of peptidoglycan, the main component in the negative-bacteria cell wall.Hence, cephalosporins can influence cell-wall integrity, resulting in damage to the pathogen cell.As an important factor contributing to the multifactorial mechanism of cephalosporin resistance, the changes in PBP level have been found to be involved in drug susceptibility in some bacteria.Another interesting finding here is that the inactivation of hfq leads to altered expression levels of PBPs, the targets of cephalosporins in K. pneumoniae.This finding implies that the increased resistance might be mediated by an altered expression of PBPs, the main mediators of resistance.The reduction or loss of one or more PBPs has been identified as a reason for the increased resistance to β-lactam in many reports.Examples include PBP3 for clinical ceftazidime-resistance in Burkholderia pseudomallei [30] and cefotaxime-resistance in Streptococcus pneumoniae [29], and PBP4 for triggering overproduction of AmpC in Enterococcus faecium [35].In addition, the similar phenomenon where increased resistance accompanies a decreased expression level of most PBPs was also observed by Vashist et al. [36] in Acinetobacter baumannii.Contrary to the above results, Mottl et al. [37] and Sanders et al. [38] reported respectively that the loss of PBP4 would result in diminishing the induction of AmpC in Escherichia coli, which is in line with the increased expression of PBP4 in the cephalosporin-resistant hfq mutant.
In conclusion, one key finding in this investigation is that hfq negatively regulates the sensitivity of K. pneumoniae to cephalosporins.Inactivation of hfq leads to reduced expression levels of most of the PBPs, which is the main mechanism of resistance to cephalosporin.Therefore, hfq mediated the sensitivity of K. pneumoniae to cephalosporins probably by decreasing the expression of PBPs.Because the roles of the Hfq protein are mainly mediated by sRNA, our future work is to identify sRNA involved in expression regulation of PBPs.

Table1. Strains and plasmids used in this study. Strains or plasmids Description a
R − resistance to ampicillin; Km R − resistance to kanamycin; T R − resistance to tetracycline.

Table 2 .
Oligonucleotides used for PCR in this study.
Italicized residues are the cleavage sites.

Table 3 .
MIC fold increases for hfq mutant over wild-type K. pneumoniae

Table 4 .
MBC fold increases for hfq mutant over wild-type K. pneumoniae.