Detection of Carbapenem-Resistance Genes in Klebsiella Species Recovered from Selected Environmental Niches in the Eastern Cape Province, South Africa

Carbapenemase-producing Enterobacteriaceae (CPE) have been heavily linked to hospital acquired infections (HAI) thereby leading to futility of antibiotics in treating infections and this have complicated public health problems. There is little knowledge about carbapenemase-producing Klebsiella spp. (CPK) in South Africa. This study aimed at determining the occurrence of CPK in different samples collected from selected environmental niches (hospitals, wastewater treatment plants, rivers, farms) in three district municipalities located in the Eastern Cape Province, South Africa. Molecular identification and characterization of the presumptive isolates were determined using polymerase chain reaction (PCR) and isolates that exhibited phenotypic carbapenem resistance were further screened for the possibility of harbouring antimicrobial resistance genes. One hundred (43%) of the 234 confirmed Klebsiella spp. isolates harboured carbapenem-resistance genes; 10 isolates harboured blaOXA-48-like; 17 harboured blaKPC; and 73 isolates harboured blaNDM-1. The emergence of blaKPC, blaOXA-48-like, and blaNDM-1 carbapenem-resistance genes in Klebsiella species associated with environmental sources is of great concern to public health.


Introduction
Antimicrobials have been considered to be on the losing side in the therapeutic battle against pathogenic microorganisms, nevertheless, carbapenems (regarded as the drugs of last resort) are the preference for severe infections caused by antimicrobial resistant bacterial strains [1]. Carbapenem-resistance Enterobacteriaceae (CRE), including Klebsiella pneumoniae, generally possess mucoid which is one of the most important known nosogens that causes urinary tract infections, wound infections and septicaemia in immuno-compromised patients. Reports have shown that CRE infections are also associated with patients that have been exposed to high use of antimicrobial agents or have engaged in transplantation of organ/stem cell, mechanical ventilation, and possibly extended stay in hospitals [2]. CRE can cause treatment failure thereby constituting a significant public health concern [3,4].
The resistance against antibiotics represents an important problem because the phenomenon involves different bacterial species which include Pseudomonas aeruginosa [5] and Escherichia coli [6]. Antimicrobial resistance genes (ARGs) can be transferred to other bacterial species and thereafter can lead to the production of various enzymes (such as beta-lactamases) that inactivate antimicrobial

Description of Study Areas and Sample Collection
The sampling sites comprise of hospitals, WWTPs, surface waters (rivers, dams, canals), and farms within villages and towns in Amathole District Municipality (DM), Sarah Baartman DM, and Chris Hani DM, in the Eastern Cape Province (ECP), South Africa. Four WWTPs, three hospitals and two farms were used as sampling sites for this study. Different samples (243 in total) comprising surface water (47), WWTPs final effluents (29), hospital effluents (27), irrigation water (45), soil (60), and vegetables (35) were aseptically collected once-off between the months of January 2018 and September 2018. Water samples were collected in duplicate using sterile 1-L glass bottles and transported in an ice box to the laboratory for microbiological analyses. Vegetable samples including spinach, broccoli, and cabbage were randomly collected from selected farms in the study areas. Figure 1 is a map showing the study areas within the DM in the ECP, South Africa.

Processing of Vegetable Sample
For the processing of vegetable samples, the method of Du Plessis et al. [19] was adopted. Briefly, 10 grams of each vegetable sample was placed in 90 mL of sterile Trypticasein Soy Broth (TSB) in a Stomacher Bag and was macerated with Mixer machine, and 90 mL of the broth was incubated at 37

Processing of Vegetable Sample
For the processing of vegetable samples, the method of Du Plessis et al. [19] was adopted. Briefly, 10 g of each vegetable sample was placed in 90 mL of sterile Trypticasein Soy Broth (TSB) in a Stomacher Bag and was macerated with Mixer machine, and 90 mL of the broth was incubated at 37 • C for 18-24 h in sterilized bottles. Thereafter, a loopful of TSB was streaked on the surface of MacConkey agar plates, incubated for 24 h at 37 • C. From which, four single colonies from the overnight culture were streaked on nutrient agar (NA) plates, incubated for 24 h at 37 • C. Afterwards, glycerol stock was prepared from the NA plates in nutrient broth (NB).

Processing of Soil Sample
The method of Da Silva et al. [20] was also adopted for biological analysis of the soil samples. Sixty agricultural soil samples were randomly collected from selected farms in the study areas. A total of 10 g of each soil sample was suspended in 90 mL of sterile TSB and incubated at 37 • C for 18-24 h. Thereafter, a loopful of TSB was streaked on MacConkey agar plates and were incubated for 24 h at 37 • C. Purification process was carried out on NA plates as described above.

Processing of Water Samples
To isolate the enteric bacteria from the water samples, the membrane filter technique [21] was used. From each water sample, 100 mL of the water samples was filtered through a sterile membrane filter of 0.45-µm pore size. Thereafter, the membrane filters were aseptically picked, placed onto the surface of MacConkey agar plate, and incubated at 37 • C for 18-24 h. Four colonies were picked for isolation. After incubation, presumptive Klebsiella spp. isolates were streaked onto NA plates and incubated at 37 • C for 24 h. Thereafter, 25% glycerol stock was prepared from the cultured broths and preserved at −80 • C for further analyses.

DNA Extraction
Bacterial DNA was extracted by performing the boiling method as described by Jackson et al. [22]. Briefly, 4 mL of NB was prepared and presumptive isolates were resuscitated and also extracted from the broth cultures following the method as described in a previous study [23]. The DNA templates were stored at −20 • C for further molecular analyses.

Molecular Identification of Presumptive Isolates
The presumptive isolates were confirmed by polymerase chain reaction techniques (PCR) and PCR products were viewed using agarose gel electrophoresis (AGE). The list of primers used and reaction protocols are listed in Table 1.  Salloum et al. [24] 2.5. Antimicrobial Susceptibility Test of the Confirmed Klebsiella Species All confirmed K. pneumonia and K. oxytoca isolates were subjected to antimicrobial susceptibility test against a panel of four carbapenems following Kirby-Bauer disk diffusion method and results were interpreted according to the Clinical and Laboratory Standards Institute [22] guidelines. The list of carbapenem antibiotics includes doripenem (10 µg), imipenem (10 µg), meropenem (10 µg), and ertapenem (10 µg). Briefly, about 100-200 µL of the bacterial overnight broth was transferred into 5 mL normal saline solution which was adjusted matching 0.5 McFarland standard. Thereafter, 100 µL was spread on Muller-Hinton agar plates with the use of sterile glass spreader, and plates were impregnated with the aforementioned carbapenem antimicrobial discs, incubated aerobically at 37 • C for 24 h. After which the diameters of zone of inhibition were measured and interpreted according to the recommended criteria by the CLSI [25].

Molecular Characterization of the Relevant Carbapenem Resistance Genes
All strains of Klebsiella spp. that exhibited phenotypic resistance against the test carbapenems were further screened for the relevant carbapenem-resistance genes using PCR technique and PCR products were viewed using AGE. The list of primers that were used are listed in Table 2.

Identification and Characterization of Presumptive Klebsiella Spp. Isolates
A total of 352 presumptive Klebsiella spp. isolates were obtained from all the samples analyzed. About 83% (291/352) of the isolates were identified as Klebsiella spp. by PCR. The confirmed Klebsiella spp. isolates were further characterized by PCR into different species using specific primers. Out of 291 Klebsiella spp., a total number of 234 (80%) isolates belonging to the two species identified (K. pneumonia 182, 63%; K. oxytoca 67, 23%) and unidentified species, 57 (20%). Figure 2 is a representation of the PCR products of the PCR-confirmed Klebsiella species while Figure 3 represents the agarose gel electrophoresis of the PCR products of the two Klebsiella spp. (K. pneumonia and K. oxytoca). Table 3 highlights the number of confirmed Klebsiella spp. isolated from the different sample types in this study.

Identification and Characterization of Presumptive Klebsiella Spp. Isolates
A total of 352 presumptive Klebsiella spp. isolates were obtained from all the samples analyzed. About 83% (291/352) of the isolates were identified as Klebsiella spp. by PCR. The confirmed Klebsiella spp. isolates were further characterized by PCR into different species using specific primers. Out of 291 Klebsiella spp., a total number of 234 (80%) isolates belonging to the two species identified (K. pneumonia 182, 63%; K. oxytoca 67, 23%) and unidentified species, 57 (20%). Figure 2 is a representation of the PCR products of the PCR-confirmed Klebsiella species while Figure 3 represents the agarose gel electrophoresis of the PCR products of the two Klebsiella spp. (K. pneumonia and K. oxytoca). Table 3 highlights the number of confirmed Klebsiella spp. isolated from the different sample types in this study.

Identification and Characterization of Presumptive Klebsiella Spp. Isolates
A total of 352 presumptive Klebsiella spp. isolates were obtained from all the samples analyzed. About 83% (291/352) of the isolates were identified as Klebsiella spp. by PCR. The confirmed Klebsiella spp. isolates were further characterized by PCR into different species using specific primers. Out of 291 Klebsiella spp., a total number of 234 (80%) isolates belonging to the two species identified (K. pneumonia 182, 63%; K. oxytoca 67, 23%) and unidentified species, 57 (20%). Figure 2 is a representation of the PCR products of the PCR-confirmed Klebsiella species while Figure 3 represents the agarose gel electrophoresis of the PCR products of the two Klebsiella spp. (K. pneumonia and K. oxytoca). Table 3 highlights the number of confirmed Klebsiella spp. isolated from the different sample types in this study.

Molecular Characterization of the Relevant Carbapenem Resistance Genes
Out of 173 Klebsiella spp. that showed resistance against the test antibiotics, 100 (58%) isolates were confirmed by PCR to harbour carbapenem-resistance genes and the result is shown in Table 4 while Table 5 represents the distribution of multiple carbapenem-resistance genes harboured by the isolates. The PCR products of the amplification of blaNDM-1 and blaKPC genes are shown in Figures 5 and  6, respectively.

Molecular Characterization of the Relevant Carbapenem Resistance Genes
Out of 173 Klebsiella spp. that showed resistance against the test antibiotics, 100 (58%) isolates were confirmed by PCR to harbour carbapenem-resistance genes and the result is shown in Table 4 while Table 5 represents the distribution of multiple carbapenem-resistance genes harboured by the isolates. The PCR products of the amplification of bla NDM-1 and bla KPC genes are shown in Figures 5 and 6, respectively.

Percentage Occurrence of the Antimicrobial Resistance Genes in Klebsiella Species
All the Klebsiella spp. isolates that were subjected to the carbapenems used were further screened for the presence of carbapenem-resistance genes, soil samples (26; 23%) from the various farms had the highest percentage occurrence of CRGs. Figure 7 shows the proportions of the confirmed Klebsiella species harbouring ARGs in the different environmental samples collected. Out of 23 hospital effluent isolates that harboured blaNDM-1 gene, 20 were recovered from Chris Hani DM while three of the isolates were recovered from Amathole DM with five of those isolates harbouring both blaNDM-1 and blaKPC genes. Results also showed all the CRE isolates from WWTP (11) were from Chris Hani DM and eight harboured blaNDM-1 gene, three harbored blaKPC gene and one isolate harboured both blaNDM-1 and blaKPC genes. Klebsiella isolates recovered from surface water (19) showed 12 isolates were recovered from Chris Hani with one isolate harbouring the three CRGs, two isolates harbouring blaNDM-1 and blaKPC genes, while seven isolates were recovered from Amathole. Sarah Baartman DM (14, 36%) had the highest overall percentage of CRE isolates from farm samples (i.e. irrigation water, soil and vegetable samples).

Percentage Occurrence of the Antimicrobial Resistance Genes in Klebsiella Species
All the Klebsiella spp. isolates that were subjected to the carbapenems used were further screened for the presence of carbapenem-resistance genes, soil samples (26; 23%) from the various farms had the highest percentage occurrence of CRGs. Figure 7 shows the proportions of the confirmed Klebsiella species harbouring ARGs in the different environmental samples collected. Out of 23 hospital effluent isolates that harboured blaNDM-1 gene, 20 were recovered from Chris Hani DM while three of the isolates were recovered from Amathole DM with five of those isolates harbouring both blaNDM-1 and blaKPC genes. Results also showed all the CRE isolates from WWTP (11) were from Chris Hani DM and eight harboured blaNDM-1 gene, three harbored blaKPC gene and one isolate harboured both blaNDM-1 and blaKPC genes. Klebsiella isolates recovered from surface water (19) showed 12 isolates were recovered from Chris Hani with one isolate harbouring the three CRGs, two isolates harbouring blaNDM-1 and blaKPC genes, while seven isolates were recovered from Amathole. Sarah Baartman DM (14, 36%) had the highest overall percentage of CRE isolates from farm samples (i.e. irrigation water, soil and vegetable samples).

Percentage Occurrence of the Antimicrobial Resistance Genes in Klebsiella Species
All the Klebsiella spp. isolates that were subjected to the carbapenems used were further screened for the presence of carbapenem-resistance genes, soil samples (26; 23%) from the various farms had the highest percentage occurrence of CRGs. Figure 7 shows the proportions of the confirmed Klebsiella species harbouring ARGs in the different environmental samples collected. Out of 23 hospital effluent isolates that harboured bla NDM-1 gene, 20 were recovered from Chris Hani DM while three of the isolates were recovered from Amathole DM with five of those isolates harbouring both bla NDM-1 and bla KPC genes. Results also showed all the CRE isolates from WWTP (11) were from Chris Hani DM and eight harboured bla NDM-1 gene, three harbored bla KPC gene and one isolate harboured both bla NDM-1 and bla KPC genes. Klebsiella isolates recovered from surface water (19) showed 12 isolates were recovered from Chris Hani with one isolate harbouring the three CRGs, two isolates harbouring bla NDM-1 and bla KPC genes, while seven isolates were recovered from Amathole. Sarah Baartman DM (14, 36%) had the highest overall percentage of CRE isolates from farm samples (i.e., irrigation water, soil and vegetable samples).

Discussion
Different environmental samples collected from the selected sampling sites within the Eastern Cape Province (ECP), South Africa were analyzed in order to detect and characterize the recovered Klebsiella isolates and to screen the phenotypically resistant Klebsiella spp. for the relevant Figure 7. The proportions of occurrence of Klebsiella spp. isolates with ARGs recovered from three DMs in the Eastern Cape Province, South Africa. The following is the order of the percentage occurrence of carbapenem-resistance genes in the different sample types; (a) hospital effluents (33%), soil (45%) and vegetables (22%); (b) irrigation water (33%), soil (60%) and vegetables (7%); (c) hospital effluents (31%), final effluents (26%), surface water (26%), irrigation water (11%), soil (4%) and vegetables (2%).

Discussion
Different environmental samples collected from the selected sampling sites within the Eastern Cape Province (ECP), South Africa were analyzed in order to detect and characterize the recovered Klebsiella isolates and to screen the phenotypically resistant Klebsiella spp. for the relevant carbapenem-resistance genes (CRGs). Among the total number (352) of presumptive Klebsiella spp. isolates, 291 (83%) were confirmed positive by molecular techniques, a very high percentage occurrence which is in conflict with the report of Rahman et al. [26], where only 14% were confirmed Klebsiella spp. from clinical isolates. This can be as a result of the number of isolates recovered from different niches in our study and also considering the moderately high prevalence of K. oxytoca among the confirmed species, more so, studies by Jacob et al. [27] and Trivedi et al. [28] also reported the presence of K. oxytoca from environmental isolates. In our present study, among the confirmed Klebsiella isolates, Klebsiella pneumonia (182, 63%) had the higher number of isolates than K. oxytoca (67, 23%) and this is in accordance with a study by Siri et al. [29] (2011). Other studies like Sharma et al. [30] and Samanta et al. [31] also reported the occurrence of Klebsiella species in isolates from environmental samples.
The number of reservoirs for CRE is increasing, not only in hospitals, but also in the community, among animals (e.g., livestock, companion animals, and wildlife) and in the environment. In this study, Amathole and Sarah Baartman DMs were observed to have elevated percentage occurrence of carbapenem-resistant Klebsiella spp. isolates recovered from farm soil samples. Generally, the presence of CRKP obtained from the different sampling sites of the study area can be of serious health concern. In this study, the highest percentage of resistance (69%) was against meropenem which is in contrary to the study by Peneş et al. [32], who reported slightly lower resistance against meropenem and imipenem. Another study by Ben Tanfous et al. [33] reports ertapenem showing high activity against Klebsiella spp. and that report is in corroboration with our findings where the Klebsiella isolates showed the lowest resistance against ertapenem (74%). Some other studies [34,35] reported Klebsiella pneumonia with high percentage of resistance against doripenem and imipenem and these are also in accordance with our present study with both antibiotics having resistance percentages of 89% and 86%, respectively.
Enterobacteriaceae members such as Enterobacter spp., Escherichia coli, and Klebsiella spp. have been reported to be associated with exceedingly high levels of antimicrobial resistance (AR) and K. pneumonia is among the group to have exhibited high resistance against carbapenems [36]. Among the various members of CRE, CRKP have received the most attention because they have the greatest potential to contribute to the overall problem of AR [37,38]. In a study conducted in Egypt, ElMahallawy et al. [39] reported the detection of OXA-48-like positive CRE in clinical isolates, which emphasized the prevalence of CRE in sub-Sahara Africa and this calls for more studies to be carried out in the region. In South Africa, reports on the presence of ARGs (specifically carbapenemase-producing genes) in isolates from different environmental niches are scarce, however, various studies on the high occurrence of ESBL have been previously reported [40][41][42].
Some of the positive isolates were phenotypically susceptible to one or more of the selected antibiotics but appeared to possess ARGs, suggesting the possibility of one silent gene. Our findings are in agreement with the findings of Lamba and Ahammad [43] and another report of Ng et al. [44] revealing high prevalence of clinically relevant ARGs in wastewaters. However, isolates that exhibited phenotypic resistance but lacked ARGs and vice versa can be as a result of antimicrobial resistance mediated by non-described or included genes and the limitations involved in the use of PCR to detect these genes. Interestingly, ESBL-producing Enterobacteriaceae in vegetables, soil, and water of the farm environment was reported in study by Said et al. [45], however there is a dearth of information on the presence of CRE harbouring bla NDM , bla KPC, and bla OXA-48-like genes in isolates recovered from wastewater, hospital sewage, lakes, and farm samples within the ECP, South Africa.
To the best of our knowledge, this present study is the first report investigating environmental bacterial isolates of Enterobacteriaceae family harbouring significant CRGs such as bla NDM , bla KPC , and bla OXA-48-like within major DMs in the ECP, South Africa. Our findings show that soil samples generally had the most strains of Klebsiella spp. harbouring ARGs (89%) in comparison with other sample types Antibiotics 2020, 9, 425 9 of 12 from all DMs. This may be as a result of the link between hospitals, WWTPs and receiving surface water used for agricultural purposes, a common practice in rural areas [46]. However, among the bacterial isolates that exhibited multiple resistance against carbapenems, our results show a low percentage of the isolates harbouring the bla OXA-48-like gene (6%) and our findings are in accordance with another report by Mairi et al. [47]. Among the Klebsiella spp. isolates screened for the presence of ARGs, only one K. pneumonia isolate recovered from river water in Chris Hani DM showed the presence of all three CRGs (bla NDM , bla KPC , bla OXA-48-like ) while 15 Klebsiella spp. isolates (from 8 hospital effluents, 3 farm soils, 2 rivers, 1 final effluent, 1 vegetable) were positive for genes of two different CRGs (12 bla NDM and bla KPC ; 3 bla NDM and bla OXA-48-like ). In our study, 17 isolates that exhibited multiple resistance were recovered from Chris Hani DM while only three isolates from farm soil were recovered from Sarah Baartman DM. This may be as a result of industrial activities in the different geographic areas with Chris Hani DM having the highest number of hospitals and industries. Among the few recent studies carried out on the prevalence of CRE, Hoang et al. [48] reported similar distributions of NDM and KPC-producing Klebsiella spp. in hospital and aquatic isolates in Vietnam which supports our present study. Bacterial isolates recovered from surface water and irrigation water had relatively high percentages of ARGs unlike hospital effluent and WWTPs final effluents, which could be as a result of the sample size in this study. Vegetable samples had the lowest percentage of isolates that harboured ARGs in the selected DMs. Our findings showed Enterobacteriaceae group (particularly Klebsiella spp.) perhaps already effectively resistant against carbapenems.
In conclusion, this study evidences the risk of prevalence of potential carbapenem resistant Klebsiella spp. in the environments of the Eastern Cape Province. The detection of carbapenem-resistance genes in pathogenic isolates of Klebsiella spp. recovered from hospital effluents, WWTPs, and river samples indicates that it could be as a result of improper treatment of final effluent before discharge into receiving aquatic environment. Furthermore, the occurrence of ARB in isolates recovered from farm samples may be associated with the possibility of transmission of ARGs due to the use of manure in farming and receiving watershed as a source of irrigation water and this is a concern to public health. Our findings strongly point out the need for surveillance, control, and management of antimicrobial agents use as well as calls for more studies in this area. Also, rapid and reliable detection of CP-CRE is critical for preventing their further spread in these settings.