1. Background
Antibiotic resistance of enteric bacteria remains a major public health concern worldwide (1). Many mechanisms of resistance have been described in Enterobacteriaceae and gram-negative non-fermentative bacilli (2-4). The oldest and most persistent is the production of ESBL by these bacteria, conferring resistance to beta-lactams mainly against third-generation cephalosporin (5, 6). ESBL are enzymes produced by bacteria that hydrolyze the beta-lactam ring common to beta-lactam antibiotics. At the beginning of these resistances, only some genes were described, namely Tem-1, Tem2, and SHV1 (7, 8). However, many other types have been reported recently. Three genes are mostly involved in this resistance, including Tem, VHS, and CTX-M, which have appeared in the 2000s (5, 9). This resistance usually occurs in Escherichia coli and Klebsiella pneumoniae and rarely in other enteric bacteria such as Enterobacter cloacae (10).
The present study sought to detect cephalosporin resistance in gram-negative bacilli isolated from urinary tract and genital infections in women. Since the advent of molecular biology techniques for the identification of resistance to bacteria including the determination of ESBL resistance genes, no study has been conducted to detect the different variants that circulate in Benin by sequencing these resistance genes. This sequencing is essential in order to differentiate the non-ESBL parental enzymes, which is not possible with the commonly used PCR techniques that do not allow the differentiation of the point mutations generating different variants of the ESBL genes. The present study aimed to identify those variants, in particular Tem1, SHV1, and CTX-M15.
In Benin, several studies have been conducted on beta-lactam resistance (11, 12), but none has examined the sequences of ESBL in order to identify the different types of ESBL that circulate in the country. The present study was initiated to bridge this molecular epidemiology gap of the ESBLs isolated in Southern Benin.
2. Methods
2.1. Bacterial Species Identification and Antibiotics Susceptibility
A total of 154 isolates of enteric bacteria (gram-negative Bacillus, Oxidase -) were recovered from 508 urinary and cervicovaginal fluid samples collected from three hospitals in Southern Benin, namely Bethesda Hospital, Zonal Hospital of Menontin, and the Regional Hospital of Oueme-Plateau from July to September 2015. The identification of the strains was carried out by MALDI-TOF MS (13). The susceptibility of the strains to antibiotics was examined by disk diffusion method on Mueller Hinton-2 agar (14). The antibiotic discs used were amoxicillin (AX 25), fosfomycin (FF50), ciprofloxacin (CIP5), amoxicillin + clavulanic acid (AMC30), ertapenem (ERT10), trimethoprim + sulphamethoxazol (SXT25), imipinem (IMP10), amilkacin (AK30), gentamycin (CN15), ceftriazone (CRO30), cefotaxime (CTX30), ticarcillin + sulphamethoxazol (TIM85), cefoxitin (FOX 30), rifampicin (RA 30), and aztreonam (ATM 30). Inhibition diameters were compared to those recommended by CA-SFM, 2013.
2.2. Resistance Genes Detection
The TEM, SHV, CTX-M genes were sought using real time and conventional polymerase chain reaction (PCR). The primers and probes as well as the positive controls used are shown in Table 1. The reaction medium was composed of 10 µL QuantiTec Master Mix, 1 µL Primer F, 1 µL Primer R, 2 µL DNase free water, 1 µL probe, and 5 µL template DNA for qPCR. The reaction mixture for conventional PCR was made of 12.5 µL of QuantiTec, 0.5 µL Primer F, 0.5 µL Primer R, 6.5 µL DNase free water, and 5 µL DNA.
Table 1.Primers and Probes Used for the Polymerase Chain Reactions
| Genes (T +) | Type of PCR | Primer or Probe | Oligonucleotide Sequences | References |
|---|---|---|---|---|
| TEM (Kpnasey) | qPCR | ALLTEM-F | TTCTGCTATGTGGTGCGGTA | KJ939560.1 |
| ALLTEM-R | GTCCTCCGATCGTTGTCAGA | |||
| ALLTEM-Probe | AACTCGGTCGCCGCATACACTATTCTCAGA | |||
| PCR-Std | ALLTEM-F | ATGAGTATTCAACATTTCCGTG | ||
| ALLTEM-R | TTACCAATGCTTAATCAGTGAG | |||
| SHV (Kpnasey) | qPCR | SHV-F | TCCCATGATGAGCACCTTTAAA | AF124984.1 |
| SHV-R | TCCTGCTGGCGATAGTGGAT | |||
| SHV-Probe | TGCCGGTGACGAACAGCTGGAG | |||
| PCR-Std | SHV-F | ATTTGTCGCTTCTTTACTCGC | ||
| SHV-R | TTTATGGCGTTACCTTTGACC | |||
| CTX-M (Kpnasey) | qPCR CTX-M A | CTX-M-A-F | CGGGCRATGGCGCARAC | JQ397665.1 |
| CTX-M-A-R | TGCRCCGGTSGTATTGCC | |||
| CTX-M-A-Probe | CCARCGGGCGCAGYTGGTGAC | |||
| qPCR CTX-M A | CTX-M-B-RT-F | ACCGAGCCSACGCTCAA | ||
| CTX-M-B-RT-R | CCGCTGCCGGTTTTATC | |||
| CTX-M- B-Probe | CCCGCGYGATACCACCACGC | |||
| PCR-Std CTX-M 1 | CTX-M-1-F | CCCATGGTTAAAAAATCACTGC | ||
| CTX-M-1-R | CAGCGCTTTTGCCGTCTAAG | |||
| PCR-Std CTX-M 9 | CTX-M-9-F | GCGCATGGTGACAAAGAGAGTGCAA | ||
| CTX-M-9-R | GTTACAGCCCTTCGGCGATGATTC |
2.3. Sequencing
PCR products were purified and BigDye PCR was performed with the same primers. For each sample, the primers were used differently in two reactions. The reaction medium for the BigDye PCR was 3 µL buffer BigDye, 2 µL BigDye, 1 µL primers, and 10 µL DNase free water. The product of the BigDye PCR was then filtered on sephadex and subjected to gene sequencing by the Sanger method by ABI 3730 (Applied Biosystems, Foster City, CA, USA). The obtained sequences were aligned according to the primers and blasted in GenBank NCBI (http://blast.ncbi.nlm.nih.gov/) and Arg-Annot databases.
3. Results
Table 2 exhibits the distribution of the strains identified in MALDI-TOF. Escherichia coli was the most isolated organism in both urinary (48.7%) and cervicovaginal fluid (11.0%) samples, followed by Klebsiella pneumoniae (23.4% in urinary and 3.2% in cervicovaginal fluid samples). Enterobacter cloacae was the third most isolated bacterium found only in urinary samples (9%).
Table 2.Distribution of Isolates Per Specimensa
| Urinary | CVS | Total | |
|---|---|---|---|
| Escherichia coli | 75 (48.7) | 17 (11.0) | 92 (59.7) |
| Klebsiella pneumoniae | 36 (23.4) | 5 (3.2) | 41 (26.6) |
| Enterobacter cloacae | 14 (9.0) | 0 (0) | 14 (9.0) |
| Proteus mirabilis | 4 (2.6) | 0 (0) | 4 (2.6) |
| Enterobacter asburiae | 2 (1.3) | 0 (0) | 2 (1.3) |
| Citrobacter koseri | 1 (0.6) | 0 (0) | 1 (0.6) |
| Total | 132 (85.7) | 6 (14.3) | 154 (100) |
Abbreviation: CVS, Cervico-vaginal secretions.
aValues are expressed as No. (%)
The antibiotic susceptibility pattern of the isolated enteric bacteria is depicted in Table 3. High levels of resistance were observed with beta-lactams. Proteus mirabilis, Citrobacter koseri, and Enterobacter asburiae showed low resistance compared to Escherichia coli, Klebsiella pneumoniae, and Enterobacter cloacae.
Table 3.Antibiotic Susceptibility of the Isolates
| AMX | AMC | TIM | CTX | CRO | FOX | ATM | ERT | IPM | CS | AK | GEN | CIP | FF | SXT | |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Escherichia coli | 78/92 (85%) | 69/92 (74%) | 72/92 (78%) | 23/92 (25%) | 25/92 (27%) | 37/92 (40%) | 26/92 (28%) | 7/92 (8%) | 0/92 (0%) | 0/92 (0%) | 2/92 (2%) | 28/92 (30%) | 47/92 (51%) | 3/92 (3%) | 74/92 (80%) |
| Klebsiella pneumoniae | 41/41 (100%) | 24/41 (59%) | 22/41 (54%) | 11/41 (27%) | 11/41 (27%) | 2/41 (5%) | 13/41 (32%) | 1/41 (2%) | 0/41 (0%) | 0/41 (0%) | 1/41 (2%) | 11/41 (27%) | 12/41 (29%) | 0/41 (0%) | 22/41 (54%) |
| Enterobacter cloacae | 14/14 (100%) | 14/14 (100%) | 11/14 (79%) | 3/14 (21%) | 3/14 (21%) | 14/14 (100%) | 6/14 (43%) | 4/14 (29%) | 1/14 (7%) | 0/14 (0%) | 2/14 (14%) | 4/14 (29%) | 3/14 (21%) | 1/14 (7%) | 3/14 (21%) |
| Proteus mirabilis | 2/4 (50%) | 0/4 (0%) | 0/4 (0%) | 0/4 (0%) | 0/4 (0%) | 2/4 (50%) | 0/4 (0%) | 0/4 (0%) | 0/4 (0%) | 4/4 (100%) | 0/4 (0%) | 0/4 (0%) | 2/4 (50%) | 1/4 (25%) | 3/4 (75%) |
| Enterobacter asburiae | 2/2 (100%) | 2/2 (100%) | 0/2 (0%) | 0/2 (0%) | 0/2 (0%) | 2/2 (100%) | 0/2 (0%) | 2/2 (100%) | 0/2 (0%) | 0/2 (0%) | 0/2 (0%) | 0/2 (0%) | 0/2 (0%) | 0/2 (0%) | 2/2 (100%) |
| Citrobacter koseri | 1/1 (100%) | 0/1 (0%) | 0/1 (0%) | 0/1 (0%) | 0/1 (0%) | 0/1 (0%) | 0/1 (0%) | 0/1 (0%) | 0/1 (0%) | 0/1 (0%) | 0/1 (0%) | 0/1 (0%) | 0/1 (0%) | 0/1 (0%) | 0/1 (0%) |
| Total | 138/154 (90%) | 111/154 (72%) | 105/154 (68%) | 37/154 (24%) | 39/154 (25%) | 57/154 (37%) | 45/154 (29%) | 14/154 (9%) | 1/154 (0.6%) | 0/154 (0%) | 5/154 (3%) | 43/154 (28%) | 79/154 (51%) | 5/154 (3%) | 104/154 (67%) |
Bla-TEM was found in high proportion (57%) followed by bla-CTX-M (16%) and bla-SHV (18%). Only the three bacterial species of Escherichia coli, Klebsiella pneumoniae, and Enterobacter cloacae produced ESBL. Escherichia coli produced more ESBL than the other bacteria and a single strain of Escherichia coli produced bla-SHV. On the other hand, the strains of Klebsiella pneumoniae produced more bla-SHV. About 6% of the strains carried the three ESBL genes. The carriage of CTX-M was most often associated with the presence of bla-TEM gene (Table 4).
Table 4.Repartition of Detected Extended-Spectrum Beta-Lactamase Genes
| ESBL | TEM | SHV | CTX-M | TEM + SHV | TEM + CTX-M | TEM + CTX-M + SHV | |
|---|---|---|---|---|---|---|---|
| Escherichia coli | 68/92 (74%) | 54/92 (59%) | 0/92 (0%) | 1/92 (1.1%) | 0/92 (0%) | 14/92 (15%) | 1/92 (1.1%) |
| Klebsiella pneumoniae | 28/41 (68%) | 4/41 (10%) | 11/41 (27%) | 1/41 (2.5%) | 2/41 (5%) | 3/41 (7.5%) | 8/41 (20%) |
| Enterobacter cloacae | 4/16 (25%) | 1/16 (6%) | 2/16 (12%) | 0/16 (0%) | 1/16 (6%) | 0/16 (0%) | 0/16 (0%) |
| Total | 100/154 (64%) | 59/154 (38%) | 13/154 (8%) | 2/154 (1%) | 3/154 (2%) | 17/154 (11%) | 9/154 (6%) |
The analysis of the sequences obtained shows a predominance of the type bla-TEM-1 for TEM, bla-SHV-1 for SHV, and bla-CTX-M-15 for CTX-M (Table 5).
Table 5.Repartition of the Detected Extended-Spectrum Beta-Lactamase Genes
| Types | No. (%) | E. coli | K. pneumoniae | E. cloacea | |
|---|---|---|---|---|---|
| Bla Tem | Tem 1 | 83 (94) | 65/88 (72%) | 17/88 (20%) | 1/88 (1%) |
| Tem 2 | 3 (3) | 3/88 (3%) | 0/88 (0%) | 0/88 (0%) | |
| Tem 54 | 2 (2) | 1/88 (1%) | 0/88 (0%) | 1/88 (1%) | |
| Bla CTX-M | CTX-M 15 | 24 (89) | 12/28 (43%) | 12/28 (43%) | 0/28 (0%) |
| Others | 4 (11) | 4/28 (14%) | 0/28 (0%) | 0/28 (0%) | |
| Bla SHV | SHV-1 | 15 (60) | 0/25 (0%) | 11/25 (44%) | 2/25 (8%) |
| SHV-12 | 8 (32) | 1/25 (4%) | 8/25 (32%) | 1/25 (4%) | |
| SHV-2 | 2 (8) | 0/25 (0%) | 2/25 (8%) | 0/25 (0%) |
4. Discussion
The production of ESBL by enteric bacteria is one of the most widespread forms of antibiotic resistance in the world. Out of 154 enteric bacteria isolated in three hospitals in Southern Benin, identification with MALDI-TOF revealed a high presence of Escherichia coli in both urinary (48.7%) and cervicovaginal fluid (11 %) samples. Several studies have reported the strong involvement of Escherichia coli in urinary (15-17) and cervicovaginal infections (18-20).
Klebsiella pneumoniae was the second most isolated bacterium in our specimens, which has also been reported in urinary and vaginal infections (21-23). A high resistance to cephalosporins was recorded in the present study. Nevertheless, a very low proportion of carbapenem resistance was observed. In the aminoglycoside family, we observed a high proportion of gentamicin resistance compared to a low resistance to amikacin. This discrepancy in the aminoglycoside family is attributable to the fact that amikacin has a higher minimum inhibitory concentration (MIC) and therefore, it is less used (24). This demonstrates the implication of inappropriate and non-moderate use of antibiotics in the emergence of resistances (25).
No resistance was noted to colistine, which is the last resort against infections caused by Pseudomona aeruginosa, Acinetobacter baumanii, and multidrug-resistant Enterobacteriaceae (26). It is therefore necessary to monitor the use of this antibiotic in both human and veterinary health in order to avoid resistance to this antibiotic. The bacteria that showed higher levels of resistance were Escherichia coli, Klebsiella pneumoniae, and Enterobacter cloacae. The same observation was previously made in Burkina Faso (27, 28).
About 65% of the enteric bacterial strains carried an ESBL gene. This prevalence of ESBL is so far the highest reported in Benin and shows high dissemination of this resistance in the country. Previous studies in Benin were based on the ESBL phenotype, which has shown deficiencies over time because bacteria that produce β-lactamases are regarded as susceptible to 3rd and 4th-generation cephalosporins (29). In addition, most of these studies are biased with the poor quality of antibiotic discs marketed in Benin (30).
This study also revealed the presence of bla-TEM in Escherichia coli and bla-SHV in Klebsiella pneumoniae. Anago et al. (11) also reported a high presence of bla-Tem in Escherichia coli strains isolated from nosocomial infections in Southern Benin. Hou et al. (31) reported high proportions of bla-SHV in Klebsiella pneumoniae strains compared to bla-Tem in China. The natural production of penicillinase by Klebsiella pneumoniae strains encoding a gene with SHV, LEN, and OKP variants explains the presence of SHV in Klebsiella pneumoniae strains (32). Kamga et al. (33) observed a strong presence of TRI (inhibitor-resistant TEM) in Escherichia coli strains, which may justify the high presence of TEM in the Escherichia coli strains isolated in the present study. Zongo et al. (28) in Burkina Faso, Salah (34), and Diagbouga et al. (35) in Togo also found that the bla-Tem gene was the most widespread in ESBL-producing Escherichia coli strains, whereas Bla-SHV was the predominant gene in ESBL-producing Klebsiella pneumonia strains. However, the high proportion of bla-CTX-M in Togo (95.7%) and Burkina Faso (65.49%) is not the same as in our study because this gene came second after bla-Tem and represented only 18%.
The molecular typing of the various resistance genes revealed the presence of three types of Bla-TEM including bla-TEM1 (83/88), bla-TEM2 (3/88), and bla-TEM54 (2/88). Bla-CTX-M15 was the most represented type of bla-CTX-M. The strong presence of this type has been reported in many studies in Africa and worldwide (27, 36-41). SHV-1 was the predominant VHS in our study followed by SHV-12 and SHV-2. These types have also been reported in (42-45) as well.
4.1. Conclusion
The present study revealed the strong presence of ESBL-producing Enterobacteriaceae in Benin. It remains the most dominant resistance to beta-lactam and evolves towards resistance to carbapenems. The types of ESBL gene reported in the study are widely replicated in Africa and are implicated in many infectious pathologies. There is an urgent need to devise policies towards reducing antimicrobial resistance in Africa in order to reduce mortality and the high cost of treatment associated with ESBL-producing Enterobacteriaceae.
