Characterization of antibiotic resistance profiles in Pseudomonas aeruginosa isolates from burn patients

Abstract

Objective:

To investigate the prevalence of Multidrug Drug-Resistant (MDR) Pseudomonas aeruginosa (PA) producing Extended-Spectrum Beta-lactamases (ESBLs) and metallo-betalactamases (MBLs) in burn patients in Algeria.

Methods:

Between April 2016 and October 2019, 47 non-redundant isolates of PA were collected from 47 burn patients admitted to the Department of Burns at the Military Hospital of Algiers in Algeria. Antibiotic susceptibility testing was performed by agar diffusion and the Phoenix automated method. Resistance genes were identified by PCR, and molecular typing of isolates was carried out by enterobacterial repetitive intergenic consensus (ERIC) sequences-polymerase chain reaction (PCR).

Results:

Among the 47 non-redundant MDR PA strains isolated, 59.57% were phenotypically ESBLs-positive, and 100% were phenotypically MBL-positive. The ESBL-positive isolates were subsequently screened for six groups of bla genes encoding ESBL-type enzymes, namely blaCTX-M2, blaPER, blaTEM, blaSHV, blaVEB, and blaGES. Out of the 28 ESBL-producing strains, 23 (82.14%) were blaCTX-M2 positive; 18 (38.29%) were blaPER positive, and 16 (34.04%) were blaTEM positive, while 5 (17.9%) were co-harboring blaCTX-M2, blaTEM, and blaPER genes. The blaSHV, blaVEB, and blaGES genes were not detected in any of the ESBL positive isolates. Since all isolates were MBL-positive, all 47 strains were screened for the blaNDM-1, blaIMP, blaVIM genes that produce MBLs; however, none of these genes were detected. Additional screening for the oprD gene demonstrated that 45 (95.74%) of the isolates were positive for this gene. Finally, ERIC PCR revealed 6 distinct PA clones among the blaCTX-M2 positive strains.

Conclusion:

This is the first study to report the presence of CTX-M2-producing PA in the North Africa region and the first to detect blaCTX-M2-positive and blaPER-positive PA clinical isolates in Algeria, therefore demonstrating the spread of such MDR strains to this part of the world. Identification of bacterial genotypic alterations that confer antibiotic resistance is critical in determining the most effective antimicrobial strategies to be employed. Therefore, our findings could potentially facilitate clinical decision making regarding the antibiotics of choice for the treatment of burn patients that suffer from PA infections in Algeria.

Introduction

Despite significant improvements in burn care, multidrug-resistant (MDR) Pseudomonas aeruginosa (PA) remains one of the most common causes of life-threatening infections in patients suffering from thermal injuries [1]. Rapid antibiotic resistance emergence leaves physicians with limited available effective antibiotics against MDR, PA. Therefore, the outcomes of the affected patients are poor, despite intensive resuscitative and anti-microbial treatments. Infections caused by MDR PA have been mainly implicated in higher morbidity and mortality rates, in increased length of hospital stay, and considerably higher healthcare-related costs following burns [2]. Hence, MDR PA infections pose a substantial threat to the burn patient population. Importantly, this threat is further augmented in developing countries of North Africa, such as Algeria, where the paucity of crucial healthcare-related resources is the norm [3, 4].

Due to its great adaptability, its metabolic versatility, and its ability to acquire antimicrobial resistance traits, PA is considered a model pathogen in the field of antibiotic resistance [5, 6]. It employs an array of mechanisms to protect itself from antimicrobial agents including but not limited to rendering its outer membrane-impermeable, modifying the antibiotic-target site, forming multidrug efflux pumps, and producing beta-lactamases [7, 8]. This latter mechanism also exerts an inherent variability, since the enzymes produced can be either extended-spectrum beta-lactamases (ESBLs), metallo-beta-lactamases (MBLs), or AmpC-beta-lactamases [9]. ESBLs in pathogenic strains of PA are enzymes from Ambler class A (PER, GES [10, 11], VEB [10,12], BEL [10,13,14], and PME [15] families) and from Ambler class B (OXA family) [10,16,12]. A small number of PA isolates also produce three ESBL classes that are mainly found in Enterobacteriaceae, namely TEM, SHV, and CTX-M [11,17]. Furthermore, membrane impermeability and MBL production have been implicated in the observed increase of PA carbapenem resistance. Specifically, membrane impermeability occurs mostly as a result of the loss of the oprD gene, while responsible MBLs include those that belong to the IMP, VIM, SPM, GIM, SIM, AIM-1, FIM-1, and NDM families (Ambler class B) [18, 19, 20]. Importantly, The ESBL enzymes are usually codified by genes of mobile genetic elements, which may be associated with aminoglycoside resistance genes and are therefore a matter of major concern due to their remarkable capacity to disseminate [21]. The production of aminoglycoside modifying enzymes (AME) is considered the most common mechanism of aminoglycoside resistance in PA. These enzymes can phosphorylate (aminoglycoside phosphoryl-transferases [APH]), acetylate (aminoglycoside acetyl-transferases [ACC]) or adenylate (aminoglycoside nucleotidyltransferase [ANT]) aminoglycosides, hence rendering them inactive [21]. Finally, an increasing body of evidence shows a rise in the prevalence of PA strains harboring both ESBL and MBL genes, thus further augmenting the challenge for effective antimicrobial treatments [17].

Recently, a dramatic surge in the number of ESBL-producing PA strains isolated from burn patients, has led to significant complications in the treatment of this patient population [22, 23]. Since Algeria has been implicated as one of the countries with the highest antimicrobial resistance rates [3, 20, 24], and given the relevance of PA in burn wound infections, we sought to investigate the prevalence of MBL- and ESBL-related genes among 47 MDR PA strains isolated from burn eschars of patients admitted to the Department of Burns at the Military Hospital of Algiers in Algeria.

Materials and Methods

Bacterial strains

Between April 2016 and October 2019, 47 isolates of PA were collected from 47 burn patients admitted to the Department of Burns at the Military Hospital of Algiers in Algeria that presented with a thermal injury of any degree and subsequently suffered a nosocomial burn-wound infection with PA. If the same isolate was obtained in more than one occasion (in more than one patient), it was included in the study only once; hence our study includes only non-redundant PA clinical isolates. A sample for culture was obtained whenever this was indicated for medical reasons. In particular, a culture sample was taken if there were changes in the odor or color of the wound, if there was cellulitis or graft ghosting, and if the patient was febrile or septic. Our analysis was limited to the first burn-wound PA infection episode for each patient.

According to criteria implemented by the United States Center for Disease Control and Prevention (US CDC), infections that emerged <48 h since admission were not considered as nosocomial infections, and such patients were excluded from our analysis [25].

Bacterial Strain Isolation and identification

All bacterial strains were isolated by burn-wound surface swabs. For routine phenotypical tests usually performed in clinical laboratories, we inoculated burn wound swabs primarily onto several selective media for the isolation of PA, including blood agar, chocolate agar, Mueller-Hinton, and MacConkey agar, and incubated them at 37°C for 24–48 h. All the isolates were identified by conventional biochemical methods that are delineated below and include the colony morphology and pigment production on selective media, followed by the output from the Analytical Profile Index 20E (API 20E) system (bioMerieux, France), the ability of bacteria to ferment lactose, and the cytochrome oxidase activity.

The isolates were identified as truly Pseudomonas species based on the routine lab algorithm that takes into consideration the results from the aforementioned assays [26, 27]. Specifically, the first step of the algorithm looks at the API 20E outcome. If this step determines that the isolate is a Gram-negative rod, then the next step is to determine its ability to ferment lactose. If the isolate does not ferment lactose, then the presence of cytochrome oxidase is assessed. Pseudomonas species are positive for the cytochrome oxidase.

Complementary standard microbiological assays, including nitrate reduction and gelatin liquefaction (assessment of gelatinase presence), were performed as described by Blazevic et al [28] without any modifications. The isolates that had the ability to reduce nitrates and were positive for gelatinase, were deemed to truly belong to the PA species.

Colony morphology and pigment production

To determine whether the colony morphology and the pigment production by the different isolates were consistent with the expected phenotype for PA (i.e., flat, round, spreading colonies with a metallic sheen and a grape-like odor), bacteria were grown on selective cetrimide agar (Pseudosel Agar, BD). Pigment production (green-blue pigment) was evaluated by qualitative observation. Inoculated plates were incubated at 37 °C for 24 h [29, 30].

API 20E system

To determine whether the isolates were Gram negative rods, we used the well-established API 20E system (Biomerieux, France), as per the manufacturer’s instructions [31, 32]. The test for each isolate was repeated twice.

Lactose fermentation

To assess the ability of bacteria to ferment lactose, the isolates were grown in fresh phenol red broth with 1% lactose (Thermo Scientific), as per the manufacturer’s instructions. The inoculated tubes were incubated at 37 °C for 24 h and the color of the broth was then assessed. A control tube that was not inoculated with any bacteria was also incubated along with the test tubes and was assessed for color alterations. The bacteria were deemed to be lactose fermenters if the color of the broth did not change from red (original broth color) to yellow (color alteration in the presence of lactose fermenters) [26].

Oxidase activity

To determine whether bacteria had the cytochrome oxidase enzyme, isolates were grown on nutrient agar (Sigma-Aldrich) and the oxidase activity was determined using the oxidase test (Millipore), as per the manufacturer’s instructions. Briefly, a well-isolated colony was taken using an inoculating loop and was spread on an oxidase disc containing N,N-dimethyl-p-phenylenediamine oxalate and α-naphthol. The reaction was observed within 2 minutes at 25–30°C. Reaction of the N,N-dimethyl-pphenylenediamine oxalate and α-naphthol reacted to indophenol blue, indicated the presence of the enzyme cytochrome oxidase in the tested isolate.

Antimicrobial Susceptibility Testing (AST)

AST was performed using the NMIC/ID-94 Phoenix panel by the automated Phoenix System (BD Diagnostics). The Minimum Inhibitory Concentrations (MICs) of antibiotics were interpreted using the Clinical Laboratory Standards Institute guidelines (CLSI M100-S23) (Table 1), thus meeting the MDR criteria, according to the European Committee on Antimicrobial Susceptibility Testing (EUCAST) guidelines [33, 34]. All tests were confirmed manually by microbiological complementary standard techniques. Sensitivity to all antibiotics listed in Table 1 was determined. Reference strains were included as internal controls in all tests, including Escherichia coli ATCC (25922), Staphylococcus aureus ATCC (25923), and Pseudomonas aeruginosa ATCC (27853). All isolates were cryopreserved at −80°C in Luria Bertani (LB) broth containing 50% glycerol for further analysis.

Table 1:

Phoenix Antibiotic Susceptibility Panel Interpretation.

Antibiotic	MIC (ug/mL)	Interpretation		Number of isolates that are resistant
Amikacin	8–32	<8 S	>32R	38 (80.85%)
Gentamicin	2–8	<2S	>8R	44 (93.61%)
Ertapenem	0.25–4	<0.25S	>4R	47 (100%)
Imipenem	1–8	<1S	>8R	47 (100%)
Meropenem	1–8	<1S	>8R	47 (100%)
Cephalothin	4–16	<4S	>16R	47 (100%)
Cefuroxime	4–16	<4S	>16R	47 (100%)
Cefoxitin	4–16	<4S	>16R	47 (100%)
Ceftazidime	1–16	<1S	>16R	47 (100%)
Ceftriaxone	1–32	<1S	>32R	47 (100%)
Cefepime	1–16	<1S	>16R	47 (100%)
Aztreonam	2–16	<2S	>16R	45 (95.74%)
Ampicillin	4–16	<4S	>16R	47 (100%)
Amoxicillin / Clavulanate	4/2–16/8	<4/2S	>16/8R	47 (100%)
Piperacillin / Tazobactam	4/4–64/4	<4/4S	>64/4R	47 (100%)
Colistin	1–4	<1S	>4R	0 (0%)
Trimethoprim / Sulfamethoxazole	1/19–4/76	<1/19S	>4/76R	47 (100%)
Nitrofurantoin	16–64	<16S	>64R	47 (100%)
Ciprofloxacin	0.5–2	<0.5S	>2R	39 (82.97%)
Levofloxacin	1–4	<1S	>4R	39 (82.97%)
Tigecycline	1–4	<1S	>4R	47 (100%)

S: Sensitive, R: Resistant

Phenotypic detection of production of MBLs

For the phenotypic detection of MBL production, we performed the imipenem-EDTA double-disk synergy (IEDDS) test by using disks of imipenem (10 mg) and EDTA (1.5 mg) spaced at a distance of 20 mm (edge to edge) on Mueller–Hinton agar [34].

Phenotypic detection of production of ESBLs

For the phenotypic detection of ESBL production, we performed the Double-Disc Synergy Test (DDST) [4]. Cefotaxime, ceftazidime, and cefepime disks were placed around a disk of amoxicillin/clavulanic acid at a disk-center to disk-center distance of 20 mm on Mueller–Hinton agar supplemented with cloxacillin (500 mg/mL).

A phenotypic confirmation test on a Mueller–Hinton agar (MHA) plate using discs of ceftazidime (30 μg) and ceftazidime/clavulanic acid (30 μg/10 μg) was performed. Both discs were placed 25 mm apart (center to center) on a lawn culture of the test plate and incubated for 24 hours at 37°C. K. pneumonia ATCC (700603) and E. coli ATCC (25922) were used as positive and negative control strains, respectively [35, 36].

Genomic DNA Extraction

For each isolate, genomic DNA was extracted as described by Feria et al [37]. Briefly, 3 to 4 pure bacterial colonies obtained from a fresh overnight PA culture on LB agar were suspended in 100μl of ultrapure water. The suspension was boiled at 100°C and was then centrifuged at 12,000 rpm for 3 min. The DNA supernatant obtained by centrifugation was used immediately as the DNA template for the Polymerase Chain Reaction (PCR) assay, or was stored at −20°C.

PCR Assay

We subsequently performed PCR to determine the presence or absence of the following ESBL-related resistance genes: blaTEM, blaPER, blaVEB, blaSHV, blaGES, blaCTX-M2, as well as the presence or absence of the following MBL-related resistance genes: blaIMP, blaNDM-1 and blaVIM. We also screened for the aadA, aac6-Ib, and aph3-VI aminoglycoside resistance genes and for the oprD gene. The primers used in this study are listed in Table 2. The amplification was carried out in a thermocycler (Mastercycler Gradient, Eppendorf, Germany). The reaction solution consisted of 10.5 μL of PCR buffer, 12.5 μL of Dream taq Green PCR Master mix (10X) (10 mM each) (Thermo Fisher Scientific), 0.5 μL of each primer (20 pmol/μL), 1μL of template genomic DNA, and 18.5 μL of nuclease-free water (total reaction solution 25 μL). Electrophoresis of the PCR products was performed at 80 v / 380 mA, in 1.5% agarose gel that was stained with ethidium bromide. Visualization was performed under ultraviolet (UV) light using a UV transilluminator (ChemicDocTM Imaging System, Bio-Rad, USA) Biorad, USA).

Table 2:

Primers used for the detection of oprD, ESBL-related, MBL-related genes, and ERIC sequences in the PA isolates.

Target OprD	Primer	Sequence 5′ to 3′	AT °C	Size (bp)	Reference
oprD	oprD-F	GGAACCTCAACTATCGCCAAG	57	1412	30
	oprD-R	GTTGCCTGTCGGTCGATTAC
Target ESBLs	Primer	Sequence 5′ to 3′	AT °C	Size (bp)	Reference
blaTEM	TEM-F	ATGAGTATTCAACATTTCCGTG	55	840	31
	TEM-R	TTACCAATGCTTAATCAGTGAG
blaSHV	SHV-F	TTTATGGCGTTACCTTTGACC	53	1051	32
	SHV-R	ATTTGTCGCTTCTTTACTCGC
blaGES	GES1-F	ATGCGCTTCATTCACGCAC	55	860	33
	GES1-R	CTATTTGTCCGTGCTCAGG
blaPER	PER-F	GTAGTATCAGCCCAATCCCC	55	738	34
	PER-R	CCAATAAAGGCCGTCCATCA
blaVEB	VEB-F	GGAACAACTTTGACGATTGA	57	374	34
	VEB-R	CCCTGTTTTATGAGCAACAA
bla CTXM2	CTX-M- 2-F	ATGATGACTCAGAGCATTCGC	53	749	35
	CTX-M- 2-R	GATATCGTTGGTGGTGCCA
Target MBLs	Primer	Sequence 5′ to 3′	AT °C	Size (bp)	Reference
blaNDM- 1-like	NDM-1- F	GCGAACACACAGCCTGACTTT	57	813	33
	NDM-1- R	CAGCCACCAAAAGCGATGTC
blaIMP	IMP-F	CATACTCGTTGAAGAAGTTAAC GG	53	448	35
	IMP-R	GAGAATTAAGCCACTCTATTGC
blaVIM	VIM-F	TGGTCTACATGACCGCGTCT	53	766	3
	VIM-R	CGACTGAGCGATTTGTGTG
Target AME	Primer	Sequence 5′ to 3′	AT °C	Size (bp)	Reference
aac6’-Ib	aac(6’)Ib-F	TATGAGTGGCTAAATCGAT	49	395	37
	aac(6’)Ib-R	CCCGCTTTCTCGTAGCA
aadA	aadA-F	TTGTACGGCTCCGCAGTG	53	812	38
	aadA-R	CCCAATTTGTGTAGGGCTTA
aph3’-VI	aph(3’)VI-F	CGGAAACAGCGTTTTAGA	39	716	37
	aph(3’)VI-R	TTCCTTTTGTCAGGTC
Target ERIC	Primer	Sequence 5′ to 3′	AT	Size (bp)	Reference
ERIC2	ERIC2	AAGTAAGTGACTGGGGTGACGC	30	-	39

bp: base-pairs

ERIC- PCR typing

For the ERIC (enterobacterial repetitive intergenic consensus) PCR typing, we used repetitive extragenic palindromic (REP) PCR with primers for ERIC 2 sequences (Table 2). The PCR products were separated in 1.2% agarose gels. The isolates were subsequently grouped by comparing their DNA patterns using the PyElph 1.3 software (Creative Commons) which automatically detects the migration lanes and bands, computes the molecular weight of each separated fragment, matches the bands from all samples (based on their migration distance), and computes similarity and distance matrices (using the Dice coefficient, which expresses the similarity level between two DNA patterns and the unweighted pair group method with arithmetic mean - UPGMA). Based on this information, a phylogenetic tree (dendrogram) was then generated using the same software.

Results

Antimicrobial susceptibility testing and phenotypic detection of ESBLs and MBLs.

Table 3 shows the susceptibility patterns of the strains isolated, all of which were classified as MDR as per the EUCAST guidelines [33, 34]. Among the 47 MDR PA isolates initially described as resistant to third-generation cephalosporins, (100%), resistance to ceftazidime and cefepime was confirmed in all of them by the disc diffusion method. Our antimicrobial tests revealed that 100% of the strains were sensitive to Colistin, 19.15% were sensitive to Amikacin, 17.03% were sensitive to Ciprofloxacin, 17.03% were sensitive to Levofloxacin, 6.39% were sensitive to Gentamicin, and 4.26% were sensitive to Aztreonam. All the PA isolates were resistant to the remaining antibiotics (Table 1).

Table 3:

Susceptibility patterns, ESBL and MBL screening results.

Susceptibility Profile of Isolates ( Total= 47)	Positive ESBL ( Total = 28)	Positive MBL (Total=47)
AN, GM, CL (N=3)	2	3
CIP, LVX, CL (N=8)	7	8
AN, CL (N=6)	5	6
ATM, CL (N=2)	1	2
CL (N=28)	13	28

AN=Amikacin, GM=Gentamicin, CL= Colistin, CIP=Ciprofloxacin, LVX=Levofloxacin, ATM=Aztreonam

We subsequently performed tests for the phenotypic detection of ESBL and MBL production. Overall, 28 (59.57%) MDR PA isolates were ESBL-positive, while all 47 (100%) were MBL-positive. 1 PA isolate that was MBL-positive and 2 that were MBL-positive and ESBL-positive were sensitive only to Amikacin, Gentamicin, and Colistin. 1 PA isolate that was MBL-positive and 7 that were MBL-positive and ESBL-positive were sensitive only to Ciprofloxacin, Levofloxacin and Colistin. 1 PA isolate that was MBL-positive and 5 that were MBL-positive and ESBL-positive were sensitive only to Amikacin and Colistin. 1 PA isolate that was MBL-positive and 1 that was MBL-positive and ESBL-positive were sensitive only to Aztreonam and Colistin. Finally, 15 PA isolates that were MBL-positive and 13 that were MBL-positive and ESBL-positive were sensitive only to Colistin (Fig.1).

Figure 1.

Graphical representation of the susceptibility of the P. aeruginosa isolates to antibiotics. Isolates were only to Colistin, Amikacin, Ciprofloxacin, Levofloxacin, Gentamicin, and Aztreonam. All isolates were resistant to all other anti-Pseudomonas antibiotics.

Detection of bla genes encoding ESBLs and MBLs.

We subsequently screened all the phenotypically ESBL-positive isolates for the 5 groups of blagenes encoding ESBL type enzymes, namely blaTEM, blaPER, blaCTX-M-2, blaVEB, blaSHV, and blaGES. The occurrence of all detected bla genes encoding ESBL-type enzymes in relation to the susceptibility profiles of PA isolates is presented in Table 4. Specifically, among the 28 phenotypically ESBL-positive isolates, only blaTEM, blaPER, and blaCTX-M-2 were detected. Specifically, 23 (82.14%) of these isolates were blaCTX-M2 positive, 18 (38.29%) were blaPER positive, and 16 (34.04%) were blaTEM positive. blaSHV, blaVEB, and blaGES genes were not detected in any of the ESBL-positive isolates. Furthermore, we noticed that 8 (17.02%) isolates were positive for the blaCTX-M2, blaTEM, and blaPER genes (Fig.2A, Fig.2B, Fig. 2C). Surprisingly, none of the 47 MDR PA isolates with MBL positive phenotype was carrying the blaIMP, blaVIM, or blaNDM-1 genes.

Table 4:

Associations between P. aeruginosa isolate genotypes and their phenotypic responses to antimicrobials.

Genes	Total number of isolates	Sensitivity to antimicrobials
		Colistin	Amikacin	Ciprofloxacin	Levofloxacin	Gentamicin	Aztreonam
aadA	18	18	0	0	0	0	0
acc6’-Ib	17	17	0	0	0	0	0
aph3’VI	7	7	0	0	0	0	0
blaTEM	16	16	5	0	0	2	0
blaPER	18	18	7	3	3	1	0
blaCTXM-2	23	23	1	4	4	0	0
oprD	45	45	0	0	0	0	0

Figure 2.

Agarose gel electrophoresis of the amplified blaTEM gene (1A), blaPER gene (1B), blaCTX-M2 gene (1C) from the ESBL producing P. aeruginosa isolates.

(A): Lane M: 100bp DNA Ladder; Lane C+: Positive Control; Lane C-: Negative Control; Lanes 1, 2, 3, 6, 7, 9, 10 are isolates that were positive for the blaTEM gene (=840bp); Lanes 4, 5, 8 are isolates that were negative for the blaTEM gene.

(2B): Lane M: 100bp DNA Ladder; Lane C+: Positive Control; Lane C-: Negative Control; Lanes 1–10 are isolates that were positive for the blaPER gene (= 738bp).

(2C): Lane M: 100bp DNA Ladder; Lane C+: Positive Control; Lane C-: Negative Control; Lanes 1, 2, 4, 6 are isolates that were positive for the blaCTX-M2 (= 749bp); Lanes 3, 5 are isolates that were negative for the blaCTX-M2 gene. The additional bands could potentially be explained by the presence of different gene isoforms, non-specific amplification, or primer dimers.

Occurrence of the oprD gene

The oprD gene was found in 95.74% of the isolates (n=45). This gene was detected in all but two PA isolates (Fig.3).

Figure 3.

Agarose gel electrophoresis of the amplified oprD gene from the ESBL-producing P. aeruginosa isolates. Lane M: 1Kbp DNA Ladder; Lane C+: positive control; Lane C-: negative control; Lanes 1–8: isolates that were positive for oprD gene;Lanes 9–10: isolates that were negative for the oprD gene(=1412bp).

Occurrence of aminoglycosides modifying enzymes

Since resistance to aminoglycosides in ESBL-positive bacteria is frequent, limiting the clinical use of this antibiotic family against these pathogens [38], we subsequently screened for three different AME, namely aph3’-VI, acc6’-Ib and aadA, by PCR. Among the 28 ESBL-positive PA isolates, 18 (64.28%) were carrying the aadA gene, 17 (60.71%) the aac6’-Ib gene, and 7 (25%) the aph3’-VI gene (Fig.4A, Fig.4B, Fig.4C). The aadA and the acc6’-Ib genes were found to be simultaneously present in the same isolates (n=17), while only one MDR PA was simultaneously carrying the aadA and the aph3’-VI genes. The occurrence of all the detected AME-encoding genes in relation to the susceptibility profiles of the PA isolates is presented in Table 4.

Figure 4.

Agarose gel electrophoresis of the amplified acc6’-Ib gene (4A), aadA gene (4B), and aph3’-VI gene (4C) from the ESBL-producing P. aeruginosa isolates.

(4A): Lane M: 100bp DNA Ladder; Lane C+: Positive Control; Lane C-: Negative Control; Lanes 1, 2, 4–6 are isolates that were positive for the acc6’-Ib gene (=395bp); Lane 3 is an isolate that was negative for the acc6’-Ib gene.

(4B): Lane M: 100bp DNA Ladder; Lane C+: Positive Control; Lane C-: Negative Control; Lanes 1–6 are isolates that were positive for the aadA gene (= 812bp).

(4C): Lane M: 100bp DNA Ladder; Lane C+: Positive Control; Lane C-: Negative Control; Lanes 3, 5, 10 are isolates that were positive for the aph3’-VI gene(=716bp); Lanes 1, 2, 4, 6–8 are isolates that were negative for the aph3’-VI gene.

ERIC-PCR Typing

Since this study is the first to our knowledge to detect CTX-M2 positive PA clinical isolates in North Africa, we sought to investigate whether there is a wide clonal diversity or a predominant clone. Interestingly, ERIC-PCR typing of the 23 CTX-M2 positive strains identified 11 different DNA profiles (Fig.5A, Fig.5B).

Figure 5.

(5A): ERIC-PCR fingerprints of blaCTX-M2 positive ESBL-producing P. aeruginosa isolates; Lane M: 1Kbp DNA Ladder; Lane B: Blanc; Lanes 1–11 are isolates that were positive for the ERIC-2 sequence. (5B): Cluster analysis based on the ERIC-PCR fingerprints of P. aeruginosa isolates that were blaCTX-M2-positive. Clustering analysis was performed using the PyElph 1.3 software and was based on the Dice similarity coefficient and the unweighted pair group method with arithmetic mean (UPGMA). 11 major clusters were identified from groups of closely related strains sharing genotype similarities.

Discussion

To the best of our knowledge, this study is the first to identify blaCTX-M2-positive PA clinical isolates in North Africa and the first to detect blaCTX-M2-positive and blaPER positive PA clinical isolates in Algeria, therefore demonstrating the spread of such MDR strains to this part of the world (Fig.6) [22, 39, 40, 41–55]. Such findings of phenotypic and genotypic PA stain surveillance are of tremendous importance since they can determine the strategies needed for the control and treatment ofPA-related nosocomial infections in this geographical region.

Figure 6.

World map showing the spread of the blaCTX-M2- and/or blaPER-positive P. aeruginosa (generated using mapchart.net at https://mapchart.net/world.html).

There is a limited number of studies on PA resistance genes in Algeria. Drissi et al have previously reported blaTEM-110-positive PA clinical isolates in Algeria [24], while in the broader region of North Africa, Ktari et al have identified blaPER-positive isolates in Tunisia. [41]. To the best of our knowledge, blaCXT-M2-positive and blaPER-positive PA isolates have never been reported in Algeria before.

PA strains that are positive for the blaCXT-M2 and blaPER genes have been identified in a few additional parts of the world (Fig.6). In particular, blaPER-positive PA strains have been identified in 5 European countries, namely France, Italy, Greece, Poland, and Hungary [22, 40, 42–45, 47]. Furthermore, blaPER-positive PA strains have been isolated in Latin America, including Brazil, Bolivia, and Uruguay [39, 51]. In Asia, Turkey, Iran, India, China, and Japan have also reported blaPER-positive PA isolates, while Tunisia in North Africa is also among the countries where such strains have spread to [46, 48–50, 52]. The blaCXT-M2-positive PA strains have so far spread to a relatively smaller number of countries, including Brazil and Bolivia in Latin America, Iran and China in Asia, as well as Poland in Europe [39, 40, 50, 52–55].

Notably, our study has not identified any PA clinical isolates that carry the blaNDM-1, blaIMP, blaVIM, blaSHV, blaGES, or blaVEB resistance genes. The blaVEB, blaGES, and blaSHV genotypes are prevalent in Asian countries [56]. None of these genes has previously been reported in Algeria except for the gene blaVIM, which was recently identified in Algerian burn patients for the first time by Meradji et al [57]. Specifically, this study reported that 46.7% of the PA strains isolated from burn patients were MBL producers and contained the blaVIM-2 and blaVIM-4 genes [57]. In the present work, MBL genes were not detected in any of our isolates, which could potentially be explained by the overproduction of cephalosporinase AmpC and/or non-enzymatic mechanisms such as the loss of porin OprD and overproduction of the active efflux system MexAB-OprM [24]. The oprD gene was detected in all but two strains included in our study. This could potentially be secondary to its deletion in these two isolates, as observed in a previous study [24].

Importantly, our ERIC-PCR analysis revealed a genetic diversity among blaCTX-M2 positive strains, with different clones coexisting in our burn care unit. This observation indicates a polyclonal dissemination, which could potentially be explained by the presence of multiple reservoirs, such as carrier patients and environmental sources, or alternatively by the diffusion of mobile genetic elements. Whatever the exact mechanism may be, this finding suggests clonal emergence of CTX-M2 producing strains, which could possibly be promoted by cross-transmission between PA and other bacterial species.

The therapeutic implications of our findings are unfortunately grim. All the isolates included in the present study were multidrug resistant strains, with high rates of resistance to all the commercially available anti-Pseudomonas antibiotics. The isolates were sensitive only to a small number of antimicrobial agents, among which Aztreonam, Gentamicin, Ciprofloxacin, Levofloxacin, Amikacin, and Colistin. Our data shows that less than 1 out of every 10 patients would benefit from treatment with Aztreonam, or Gentamicin (4.26% and 6.39% sensitivity rates respectively; Table 1). Additionally, less than 2 out of every 10 patients would benefit from treatment with Ciprofloxacin, Levofloxacin, or Amikacin (17.03%, 17. 03%, and 19.15% sensitivity rates respectively; Table 1). All the isolates were sensitive to Colistin but were 100% resistant to all the remaining available anti-Pseudomonas antimicrobial agents, including carbapenems that represent a strong weapon in the anti-Pseudomonas armamentarium. These considerations render the treatment of such infections extremely complicated and challenging. In this near dead-end context, Colistin represents the most reliable therapeutic agent. Combinations of this drug with aminoglycosides (Amikacin, Gentamicin), or with fluoroquinolones (Ciprofloxacin, Levofloxacin) could aid toward antibiotic stewardship.

Our study is limited by the fact that our data derive only from burn patients in the Military Hospital of Algiers, Department of Burns. Therefore, our results capture the prevalence of antibiotic resistance genes in this limited patient population from one hospital in Algeria, and it is possible that the gene prevalence rates reported here are not generalizable to the total Algerian patient population, even though relevant for the burn patient population. Furthermore, our results cannot eliminate the possibility that the blaNDM-1, blaIMP, blaVIM, blaSHV, blaGES, and blaVEB genes are present in other PA isolates that derive from different patient populations and other hospitals in Algeria. Hence, the absence of the blaNDM-1, blaIMP, blaVIM, blaSHV, blaGES, and blaVEB genes in our population is not generalizable to the total Algerian patient population. Future studies isolating PA strains from different patient populations and more Algerian hospitals would be necessary for a more comprehensive report of the prevalence of PA resistance-related genes in the country. Moreover, these studies, would confer more confidence in excluding the presence of the blaNDM-1, blaIMP, blaVIM, blaSHV, blaGES, and blaVEB genes in Algerian patients affected by PA infections. Such surveillance studies would aid in employing appropriate strategies for the treatment and control of MDR PA infections in the Algerian patient population.

In summary, our data indicate a high prevalence of ESBL-producing isolates among PA strains causing nosocomial infections in Algerian burn patients. Notably, these isolates seem to harbor a diverse group of ESBL-related enzymes to exert antibiotic resistance. Spreading dissemination of ESBL-producing strains is a concern, as it leads to limitations in the use of available antimicrobials for optimal treatment of patients. The findings of the study add to the increasing identification of ESBLs and emphasize the need for enhanced surveillance of different ESBLs in PA infections.

Conclusion

Taken together, our findings indicate the emergence of resistance of PA in North Africa and, more specifically in Algeria, by harboring new resistance genes. More than half of thePA strains in our burn unit harbor beta-lactamase-encoding genes, and therefore exert resistance to a wide range of antibiotics. Our results reveal the presence of three major ESBL genotypes in the clinical PA strains isolated from burn patients in the military hospital of Algiers, namely blaCTX-M2, blaPER, and blaTEM, two of which (blaCTX-M2 and blaPER) were detected for the first time in PA isolates in Algeria. High ESBL prevalence, diversity of resistance-gene patterns, and co-existence of different resistance-conferring genotypes in the bacterial isolates are alarming. They are very likely to impact patient outcomes adversely. Conceivably, identification of bacterial genotypic alterations that render pathogens resistant to multiple antibiotics is crucial in determining the most effective antimicrobial strategies to be employed. Therefore, our findings could potentially facilitate clinical decision making regarding the antibiotics of choice for the treatment of burn patients that suffer from PA infections in Algeria.

Highlights.

• Prevalence of multidrug-resistant (MDR) Pseudomonas aeruginosa (PA) in burn wounds of thermally injured patients at the Military Hospital of Algiers in Algeria.

• Prevalence of Extended-Spectrum Beta-lactamase (ESBL)-producing PA in burn wounds of thermally injured patients at the Military Hospital of Algiers in Algeria.

• This is the first study to report the presence of CTX-M2-producing PA in the North Africa region and the first to detect blaCTX-M2-positive and blaPER-positive PA clinical isolates in Algeria.

Abbreviations:

MDR

Multidrug resistant

PA

Pseudomonas aeruginosa

ESBL

Extended-spectrum beta-lactamase

MBLs

Metallo-beta-lactamases

ERIC

Enterobacterial repetitive intergenic consensus

PCR

Polymerase chain reaction

APH

Aminoglycoside phosphoryl-transferase

ACC

Aminoglycoside acetyl-transferase

ANT

Aminoglycoside nucleotidyltransferase

AST

Antimicrobial susceptibility testing

MIC

Minimum inhibitory concentration

EUCAST

European Committee on Antimicrobial Susceptibility Testing

LB

Luria Bertani

S

Sensitive

R

Resistant

IEDDS

Imipenem-EDTA double-disk synergy

DDST

Double-disc synergy test

REP

Repetitive extragenic palindromic