Prevalence of 16S rRNA methylase genes among β-lactamase-producing Enterobacteriaceae clinical isolates in Saudi Arabia

Abstract

Background

Co production of 16S rRNA methylases gene and β-Lactamase gene among Enterobacteriaceae isolates conferring resistance to both therapeutic options has serious implications for clinicians worldwide.

Methods

To study co existence of 16S rRNA methylases (armA, rmtA, rmtB, rmtC, rmtD, and npmA) and β-Lactamase (bla_TEM-1, bla_SHV-12, bla_CTX-M-14) genes, we screened all phenotypic positive β-Lactamase producing enterobacteriaceae by polymerase chain reaction (PCR) targeting above genes. A total of 330 enterobacteriaceae strains were collected during study period out of that 218 isolates were identified phenotypically as β-Lactamase producers, which include 50 (22.9%) Escherichia coli; 92 (42.2%) Klebsiella pneumoniae, 44 (20.2%), Citrobactor freundii and 32 (14.7%) Enterobacter spp.

Results

Among this 218, only 188 isolates harbored the resistant gene for β-Lactamase production. Major β-Lactamase producing isolates were bla_TEM-1 type. 122 (56 %) isolates were found to produce any one of the 16S rRNA methylase genes. A total of 116 isolates co produced β-Lactamase and at least one 16S rRNA methylases gene Co production of armA gene was found in 26 isolates with rmtB and in 4 isolates with rmtC. The rmtA and rmtD genes were not detected in any of the tested isolates. Six isolates were positive for a 16S rRNA methylase gene alone.

Conclusion

β-Lactamase producing isolates appears to coexist with 16S rRNA methylase predominantly armA and rmtB genes in the same isolate. We conclude the major β-Lactamase and 16S rRNA methylases co-producer was K. pneumoniae followed by E. coli. We suggest further work on evaluating other β-lactamases types and novel antibiotic resistance mechanisms among Enterobacteriaceae.

• β-lactamase
• Enterobacteriaceae
• 16S rRNA methylase gene

The widespread use of antibacterial drugs has resulted in the emergence of multidrug and pan-drug resistant Gram negative bacterial strains, which pose challenges in an era when new antibiotic choices are limited (1, 2). Multidrug resistant strains are intermediate or resistant to at least three drugs in the classes of β-lactams, carbapenems, aminoglycosides, and fluoroquinolones, whereas pan-drug resistant strains are treatable only with a colistin (3). Pan-drug resistant pathogens with acquired resistance to almost all available antimicrobial agents have emerged in recent years and challenged the treatment of infections (2, 3). Strains in the Enterobacteriaceae family have been well documented for their resistance to clinically used antibiotics (4, 5). Extended spectrum β-lactamase (ESBL) and metallo β-lactamase (MBL) production are the commonly recognized resistance mechanisms in Enterobacteriaceae strains worldwide (6).

Usually, aminoglycosides such as amikacin and tobramycin are used to treat infections by Gram negative pathogens, in combination with β-lactam agents. But increasing resistance to aminoglycosides by novel mechanisms is being reported (7, 8). 16S rRNA methylases have emerged recently as a mechanism of high level aminoglycoside resistance to most of the clinically important aminoglycosides, including arbekacin, amikacin, tobramycin, and gentamicin (4, 9). In 2003, the first 16S rRNA methylase gene, armA, was identified. Since then, eight plasmid-mediated 16S rRNA methylases (armA, rmtA, rmtB, rmtC, rmtD, rmtE, rmtF, and npmA) have been found in clinical isolates of Gram negative bacilli (10–12). Recently, rmtD (which was first identified in a pan-resistant Pseudomonas aeruginosa strain isolated in Brazil and hitherto only found in Latin America), was also detected in a Klebsiella pneumoniae isolate in Brazil (13–15). armA and rmtB have been found in many species of Gram negative bacilli in Asia (4, 16).

Dissemination of β-lactamase producing clinical isolates with acquired 16S rRNA methylase genes has become a global concern. Unreasonable use of β-lactams, including cephalosporins, carbapenams, and aminoglycosides for the treatment of Gram-negative nosocomial infections has facilitated the emergence of pan-resistant strains associated with therapeutic failures (3, 4). Coexistence of 16S rRNA methylase genes and β-lactamase genes among Enterobacteriaceae isolates confers resistance to both therapeutic options. This study was undertaken to determine the prevalent mechanisms of resistance among β-lactamase positive Enterobacteriaceae strains and also to investigate by polymerase chain reaction (PCR) the coexistence of β-lactamase and 16S rRNA methylase antibiotic resistance genes in these strains.

Materials and methods

Bacterial strains

Between January 2011 and December 2011, 330 non-duplicate Enterobacteriaceae isolates were consecutively collected from patients at El Iman Hospital, Riyadh: E. coli, K. pneumoniae, C. freundii, and Enterobactor spp (E. aerogenes and E. cloacae). The strains were obtained mostly from urine (72%) and blood (11%), but also from wounds, sputum, and other body fluids. Identification of the clinical isolates was confirmed by using GNI cards on the VITEK-60 system (bioMerieux, Marcy l'Etoile, France) and conventional methods, namely, Gram staining, morphology on differential media (MacConkey agar), and biochemical tests (manitol motility, Mannite, citrate utilization, indole production, triple sugar ion, and methyl red and Voges Proskauer tests).

Antimicrobial susceptibility testing

Antimicrobial susceptibilities were determined by using GNS cards on the VITEK-60 system (bioMerieux, France). Antimicrobial susceptibility to amikacin, tobramycin, and gentamicin was confirmed by the disk diffusion test using commercial disks for cefotaxime, ceftazidime, ciprofloxacin, levofloxacin, cefoxitin, imipenem, tetracycline, and trimethoprim/sulfamethoxazole (Himedia, Mumbai, India) according to the criteria recommended by the Clinical and Laboratory Standards Institute (CLSI) (17). Minimum inhibitory concentrations (MICs) of amikacin, tobramycin, and gentamicin were determined by agar dilution according to CLSI guidelines and interpretative criteria. E. coli ATCC 25922, Enterococci ATCC 29212, and K. pneumoniae ATCC 27853 were used as quality control strains for antimicrobial susceptibility testing.

Phenotypic identification of β-lactamase strains

β-Lactamase-producing strains were identified by resistance to third generation cephalosporins by using a VITEX-60 system (bioMerieux) and by production of β-lactamase utilizing the confirmatory double-disk combination test (17). K. pneumoniae ATCC 700603 was used as the β-lactamase positive control. β-lactamase producers were further characterized for the presence of different β-lactamase genes.

Extraction of total DNA from β-lactamase-producing strains

Whole genomic DNA was extracted from colonies grown overnight (14–16 h) at 37°C on blood agar (Remel, Lenexa, KS) using the QIAamp DNA Mini Kit and the QIAcube instrument (Qiagen, Valencia, CA) according to the manufacturers’ instructions. Total DNA was extracted and stored at −20°C for further assays.

Genotypic characterization of β-lactamase and 16S rRNA methylase genes

bla genes were detected in β-lactamase-producing isolates by PCR using the previously reported oligonucleotide primers for bla_TEM-1, bla_SHV-12, bla_CTX-M-14(18), and the armA, rmtA, rmtB, rmtC, rmtD, and npmA genes by using the following previously described primers (9, 13)(19–21):

Table

Primer	Sequence
armA-f	5′ TATGGGGGTCTTACTATTCTGCCTAT
armA-r	5′ TCTTCCATTCCCTTCTCCTTT
rmtA-f	5′ CTAGCGTCCATCCTTTCCTC
rmtA-r	5′ TTTGCTTCCATGCCCTTGCC
rmtB-f	5′ TCAACGATGCCCTCACCTC
rmtB-r	5′ GCAGGGCAAAGGTAAAATCC
rmtC-f	5′ GCCAAAGTACTCACAAGTGG
rmtC-r	5′ CTCAGATCTGACCCAACAAG
rmtD-f	5′ CTGTTTGAAGCCAGCGGAACGC
rmtD-r	5′ GCGCCTCCATCCATTCGGAATAG
npmA-f	5′ CTCAAAGGAACAAAGACGG
npmA-r	5′ GAAACATGGCCAGAAACTC

Sequencing analysis of multiplex PCR products

Isolates positive for bla or 16S rRNA methylase genes were identified by PCR. The identities of the genes were confirmed by DNA sequencing on an ABI PRISM 3730 automated sequencer (Applied Biosystems, Foster City, CA) following the standard protocol, and the sequences were compared with those in the GenBank nucleotide database at http://www.ncbi.nlm.nih.gov/blast/. Thereafter, positive isolates were used as positive controls for the rest of the respective PCR reactions.

Results

Of the 330 isolated strains of Enterobacteriaceae, 218 isolates were identified as β-lactamase producers by the double-disk combination test: 50 (22.9%) E. coli, 92 (42.2%) K. pneumoniae, 44 (20.2%) C. freundii, and 32 (14.7%) Enterobacter spp. All 330 isolates showed the highest resistance to ciprofloxacin (70%) followed by tobramycin (68%), gentamicin (58%), aztreonam (57%), amikacin (54%), and co-trimoxazole (54%). A significantly higher resistance to nitrofurantoin was seen in K. pneumoniae strains (82%) than in E. coli (24%). Strong resistance was noted for carbapenems: meropenem (49%) and imipenem (38%). Most of the isolates were susceptible to colistin (95%) and tigecycline (93%). MIC values of all 330 isolates were ≥16 µg/ml for gentamicin and tobramycin, and more than 32 µg/ml for amikacin.

β-Lactamase characterization

The major β-lactamase-producing isolates were of the bla _TEM-1 type: at least one bla _TEM-1 gene was present in 83.5% (182/218) of the isolates. All 218 isolates of E. coli and K. pneumoniae possessed one or more β-lactamase gene, whereas 14 of the C. freundi and 16 Enterobacter spp were negative for all tested β-lactamase genes.

In total 70 isolates (32%) harbored only bla _TEM-1: 26 E. coli and 44 K. pneumoniae. We detected the presence bla _TEM-1 in combination with bla _CTX-M-14 in 51% of the isolates (111/218) and in combination with bla _SHV-12 in 39.4% (86/218). Other important combinations were bla _CTX-M-14 +bla _TEM-1+bla _SHV-12 (39%; n=85) and bla _CTX-M-14 +bla _TEM-1 (12%; n=26) (Table 1).

Table 1 Coexistence of different β-lactamase genes in isolates phenotypically positive for β-lactamase

	blaTEM-1	blaTEM-1+CTX-M 14	bla TEM-1+SHV-12 like	blaTEM-1+SHV 12 like+CTX-M 14	bla negative*	Total
E. coli	26	3	1	20	00	50
K. pneumoniae	44	9	0	39	00	92
C. freundi	0	10	0	16	14	44
Enterbacter	0	4	0	10	18	32
Total	70	26	1	85	36**	218

*bla-negative isolates are attributed to presence of bla genes other than those analyzed.

**Six isolates were postive for 16S rRNA methylase genes (armA and rmtB).

16S rRNA methylase characterization of β-lactamase

Among the 218 isolates, 122 isolates (56%) were found to produce 16S rRNA methylase genes, including 24 E. coli, 52 K. pneumoniae, 30 Citrobacter, and 16 Enterobacter spp. armA, rmtB, rmtC, and npmA were found exclusively in 72, 10, 6, and 4 isolates, respectively. armA coexisted with rmtB in 26 isolates and with rmtC in 4 isolates. rmtA and rmtD were not detected in any of the isolates (Table 2).

Table 2 Coexistence of different 16sRNA methylase genes in isolates phenotypically positive for β-lactamase

	armA	rmt B	rmt C	npmA	armA+rmtB	armA+rmtC	arm negative*	Total
E. coli	16	2	2	0	2	2	26	50
K. pneumoniae	34	6	0	2	8	2	40	92
C. freundi	16	2	0	0	12	0	14	44
Enterobacter	6	0	4	2	4	0	16	32
Total	72	10	6	4	26	4	96	218

*Isolates negative for 16S rRNA methylase genes may be attributed to absence of methylase genes or to presence of methylase genes other than those that were analyzed in this study, such as rmtE and rmtF.

Coexistence of β-lactamase and 16S rRNA methylase genes

A total of 116 isolates harbored both β-lactamase and 16S rRNA methylase genes, including 24 E. coli, 52 K. pneumoniae, 26 C. freundi, and 14 Enterobactor spp. Six isolates were positive for 16S rRNA methylase genes alone. Sixty-six β-lactamase-producing isolates did not harbor any methylase genes (Table ^{Table 2} and 4 ^{Table 3}).

Table 3 Coexistence of bla genes with different 16S rRNA methylase genes in isolates phenotypically positive for β-lactamase

Positive for genes	n
β-Lactamase+armA	64
β-Lactamase+armA+rmt B	23
β-Lactamase+rmt B	10
β-Lactamase+rmt C	6
β-Lactamase+armA+rmt C	3
β-Lactamase+npmA	4
β-Lactamase+armA	4
β-Lactamase+armA+rmt C	1
β-Lactamase+armA+rmt B	1
armA only	4
armA+rmt B only	2
Positive for bla gene but no 16S rRNA detected but	66
No bla gene or 16S rRNA methylase gene detected*	30
Total	218

*Negative isolates may attributed to presence of bla or methylase genes that were not analyzed in this study.

Table 4 Coexistence of 16S rRNA genes with different ESBL genes in isolates phenotypically positive for β-lactamase

Positive for gene	n
TEM-1+SHV 12 like+CTX-M 14+16S rRNA methylases	64
TEM-1+CTX-M 14+16S rRNA methylases	23
TEM-1+SHV 12 like+CTX-M 14+16S rRNA methylases	10
TEM-1+SHV 12 like+CTX-M 14+16S rRNA methylases	6
TEM-1+16S rRNA methylases	4
TEM-1+SHV 12 like+CTX-M 14+16S rRNAmethylases	4
TEM-1+CTX-M 14+16S rRNA methylases	3
TEM-1+SHV 12 like	1
TEM-1+16S rRNA methylases	1
TEM-1+SHV 12 like+CTX-M 14+16S rRNA methylases	1
TEM-1 only	65
Positive for 16S rRNA but no bla gene detected	6
No bla or 16S rRNA methylase gene detected*	30
Total	218

*Negative isolates may be attributed to presence of bla or methylase genes that were not analyzed in this study.

Discussion

We noted that almost all (83.5%) Enterobacteriaceae isolates were β-lactamase positive. This rate is even higher than reported in some other developing countries (60−70%) (2, 5) (13, 22). By contrast, the prevalence of ESBL-producing isolates among Enterobacteriaceae in developed nations is much lower:<1% in Sweden (23) and 1.7% in France (24). One review reported that the prevalence of ESBLs in Europe was higher than in the US but much lower than in Asia and South America (25). This issue is evidently a challenge in developing nations, and our study clearly shows that it is a matter of urgency in our region.

The isolates were tested only for the bla genes reported to be prevalent in this region (26) (bla _CTX-M-14 +bla _TEM-1 and bla _SHV), and 83.5% of them were positive for bla _TEM-1. An Indian study (2) reported that 61.1% of MBL producers carried bla(VIM) and 2 (3%) carried bla(IMP). However, in our study we did not examine these two genes but only the major ESBL genes. This is probably the reason why some phenotypically β-lactamase positive isolates were genotypically negative for bla genes.

Previous study showed that the most prevalent microorganism was E. coli, followed by K. pneumonia, and other Enterobacteriaceae (5, 6) (21, 27). A study from France in 2008 reported that the most prevalent species were E. coli (48.5%), Enterobacter aerogenes (23.7%), and K. pneumoniae (14.8%) (24). In our study, the contribution of K. pneumonia to the β-lactamase strains was higher (42.2%, 92/218), whereas that of E. coli was lower (22.9%, (50/218). Noteworthy, most of the isolates (72%) were obtained from urinary tract specimens, and they showed strong resistance to aminoglycosides (amikacin, tobramycin, and gentamicin), thus limiting the treatment options.

β-Lactamase-producing Enterobacteriaceae often harbor other genes (ampC, MBL, and 16S rRNA methylase genes) that confer resistance to other antibiotics, including carbapenams, cephems, and aminoglycosides (3, 4) (9). 16S rRNA methylase genes have been reported worldwide in Enterobacteriaceae, P. aeruginosa and Acinetobacter baumannii (4, 14) (16, 28). A previously published report suggested the spread of armA and rmtB in Enterobacteriaceae isolates globally (29), and these genes were often reported to coexist with other resistance determinants, such as bla and MBL genes (12, 13) (30, 31). Newer 16S ribosomal RNA methylases have been identified recently and added to the database of aminoglycoside resistance mechanisms (4, 14) (16).

Only a few studies are available on the prevalence of 16S rRNA methylase genes in clinical Enterobacteriaceae isolates in our region (4). Therefore, we investigated and characterized the association of 16S rRNA methylase genes with β-lactamase genes among clinical Enterobacteriaceae isolates in a Saudi hospital. We observed that the coexistence of the 16S rRNA methylase genes and β-lactamase was very common: 56% of isolates that were genotypically positive for β-lactamase possessed one of the 16S rRNA methylase genes, predominantly armA. This rate is higher than those reported in a Taiwanese study (21) and a Belgian study (28). But like previous reports (4, 16) (28), we also observed greater prevalence of armA than rmtB. We also report the presence of rmtC in four E. coli and four Enterobacter isolates. It is noteworthy that we identified npmA in two K. pneumoniae and two Enterobacter isolates. This is a novel finding in this region, because rmtC and npmA have been detected in Proteus mirabilis and E. coli, respectively, in Japan (13, 20). We did not detect any rmtA or rmtD gene in our isolates.

In agreement with previous reports (26, 32), β-lactamase positive isolates were more susceptible to colistin and tigecycline, which were the only available therapeutic options for them.

In our study, 35.2% (64/182), 5.5% (10/182), 3.3% (6/182), and 2.2% (4/182) of the isolates harboring bla _TEM-1 also contained armA, rmtB, rmtC, and npmA, respectively, in association with bla _TEM-1+SHV 12 like+CTX-M 14. One isolate produced armA and rmtB along with bla _TEM-1+SHV 12 like +CTX-M 14. This points to horizontal as well as clonal dissemination of these genes in our hospital. We also found strains harboring two different 16S rRNA methylase genes in combinations, including two strains carrying armA+rmtB and one strain carrying armA+rmtC.

In conclusion, β-lactamase-producing isolates of Enterobacteriaceae were common in our hospital and they usually harbored 16S rRNA methylase genes, predominantly armA and rmtB. The major co-producers of β-lactamase and 16S rRNA methylases were K. pneumonia, followed by E. coli. Our findings underline the emerging threat of pan-drug resistant pathogens that produce both β-lactamase and 16S rRNA methylase disseminating in this region. Further work to evaluate other β-lactamase types and novel antibiotic resistance mechanisms among Enterobacteriaceae and other nosocomial pathogens is needed.

Conflict of interest and funding

The authors certify that there is no financial or other conflict of interest regarding the material discussed in the manuscript.

Acknowledgements

The authors extend their sincere appreciation to the Deanship of Scientific Research at King Saud University for funding this research through the Research Group Project No. RGP-VPP-314.