Antimicrobial Susceptibility and Genetic Basis of Resistance of Klebsiella spp Isolated from Diarrheic and Non-Diarrheic Children at Health Facilities in Mukuru Informal Settlement, Nairobi, Kenya

Antimicrobial resistance (AMR) is a global threat to public health and particularly to children. This study aimed to determine the prevalence of multidrug resistance of fecal Klebsiella spp on selected beta lactam (3rd generation cephalosporins and carbapenems) and fluoroquinolone classes of drugs in four health facilities serving the slum communities of Nairobi city in Kenya. Additionally, determine the genetic basis for the multidrug resistance observed. A cross sectional laboratory based study was undertaken where a total of 1171 children below 16 years were selected, from whom stool samples were collected, tested and analyzed. 395 (33.73%) Klebsiella spp were isolated, consisting of 365 (92.4%) Klebsiella pneumoniae and 30 (7.6%) Klebsiella oxytoca were isolated. The proportion of multi-drug resistance (MDR) K. pneumoniae and MDR K. oxytoca was 64.1% (234/365) and 96.67% (29/30) respectively. Third generation cephalosporins, cefotaxime ceftriaxone and ceftazidime showed the highest resistance of 30.7%, 29.9% and 27.4% respectively, whereas carbapenems including imipenem and meropenem had the least resistance of 1.6%, each, to K. pneumoniae. A significant association was observed in diarrheic children (OR = 1.88; p = 0.01) and those below 50 months (OR = 0.43; p = 0.002) and carrying K. pneumoniae resistance to one or more third generation cephalosporins. Genes associated with resistance included bla TEM 100%, bla CTX-M 95.2%, bla SHV 57.1%, bla OXA-1 66.7%, How to cite this paper: Wairimu, C.W., Odari, E.O., Makobe, C.K. and Kariuki, S. (2021) Antimicrobial Susceptibility and Genetic Basis of Resistance of Klebsiella spp Isolated from Diarrheic and Non-Diarrheic Children at Health Facilities in Mukuru Informal Settlement, Nairobi, Kenya. Advances in Microbiology, 11, 554-578. https://doi.org/10.4236/aim.2021.1110041 Received: September 14, 2021 Accepted: October 22, 2021 Published: October 25, 2021 Copyright © 2021 by author(s) and Scientific Research Publishing Inc. This work is licensed under the Creative Commons Attribution International License (CC BY 4.0). http://creativecommons.org/licenses/by/4.0/ Open Access C. W. Wairimu et al. DOI: 10.4236/aim.2021.1110041 555 Advances in Microbiology qnrS 54.1%, qnrB 47.6% and bla NDM 7.1%. In conclusion, there’s need for more effective infection control measures, antimicrobial stewardship to reduce emergence of antimicrobial resistance, improved drinking water, sanitation and hygiene (WASH) practices.


Introduction
The global burden of AMR is increasing alarmingly and the United Nations (UN) General Assembly AMR report estimates that resistance will be responsible for approximately 10 million deaths by 2050 [1], most of which will occur in poor resource setting, mainly, the Sub-Saharan Africa [1]. In the United States of America, for example, it is estimated that more than 2 million people are infected with AMR organisms, annually, with approximately 23,000 deaths [2].
Main causes in the USA are mainly misuse and/or abuse of antibiotics, use of antibiotics in agriculture and increased income [3] whereas in the developing countries, the situation is aggravated due to poor implementation of infection control measures and the availability of counterfeit or low quality drugs [4]. In developing countries, the data is limited due to inadequate surveillance and hence likely to be significantly higher than in developed countries.
Klebsiella spp are common intestinal commensals that obtain, accumulate, and disseminate a variety of antibiotics resistance genes such as bla KPC [5] [6]. Therefore, they serve as a significant reservoir for resistance in the intestinal tract [5] [6] and subsequently increase the risk of nosocomial and community acquired resistant infections [7]. In vivo dissemination of AMR genes from intestinal Klebsiella spp to other bacterial species has been documented [8] [9] [10] [11]. In addition, Klebsiella spp cause diarrheal disease and a myriad of extraintestinal infections especially in severely ill patients [12] [13]. Apart from diarrheal patients [14] [15] [16] multidrug resistant Klebsiella spp has also been documented in apparently healthy patients including children [17] [18].
Multi drug resistance in slums areas ensures faster spread due to high density of humans and livestock living in close proximity, frequent antibiotic misuse and insufficient drinking water, drainage and sanitation infrastructure. These settlements therefore serve as hotspots for AMR transmission [19] [20].
Reports on the emergence and global spread of multidrug-resistant (MDR) and hypervirulent clones of Klebsiella spp especially K. pneumoniae have been increasing in both nosocomial and community-acquired infections [13] [21]. As a result, the treatment of Klebsiella spp infections has become more difficult with the available options being restricted.
Various mechanisms have been implicated in antibiotic resistance including mutation of chromosomal genes and the production of β-lactamases enzymes such as extended-spectrum β-lactamases (ESBLs), cephalosporinases, and carbapenemases [22]. Genes encoding for these enzymes are mostly carried on mobile genetic elements such as conjugative plasmids, integrons, transposons and insertion sequences.
They not only bear resistance genes but also virulence genes which intensify the ability of an organism to colonize and create infection within the host [22].
Colonization precedes infection in pathogenicity of disease [13], therefore understanding colonization dynamics provides a basis for identification of colonized patients and potential establishment of intervention protocols to prevent subsequent infection.

Study Site
Mukuru slum is one of the largest urban settlements in Nairobi. It is located in Nairobi east which has a population of approximately 700,000 people [23]. Mukuru is densely populated and made of temporary structures mostly corrugated metal sheets. Basic services and infrastructure are providing adequate sanitation and clean water. In addition to poverty, a number of factors associated with informal settlements such as overcrowding, substandard housing, unclean and insufficient quantities of water and inadequate sanitation contribute to a high incidence of infectious diseases and increased mortality among children. The immunization coverage for childhood vaccination ranges from 40% -84.9% which is below the WHO recommended rate [24]. Based on unpublished data Mukuru has approximately 5 public schools and 5 health facilities. The collection sites included; Mbagathi hospital (MB), Missionaries of Mary Mukuru kwa Njenga clinic (MMM), Mukuru kwa Reuben clinic (MR) and Municipal city council (MCC).

Study Design
This was a cross sectional laboratory based study which utilized purposive sampling method.

Study Population
Study participants were children and minors under the age of 16 years. Children below 5 years are vulnerable to a myriad of infections due to their under developed immunity while children above 5 years are exposed to lifestyle and behavioral risk factors such as eating habits and WASH challenges and hence included in this study. Included in the study were children and minors below the age of 16 years and who must have been residing in Mukuru slums for at least 3 months prior to the study. For diarrheic cases, participants must have presented with episodes of loose or watery diarrhoea within the last three days.

Ethical Consideration
The study protocol was approved by the Kenya Medical Research Institute Scientific and Ethics Review Unit (SERU) Reference number: KEMRI/RES/7/3/1.

Sample Collection and Specimen Processing
Participants were recruited purposively during regular hospital visits and stool samples collected before initiation of treatment. Up to 5  and Citrate utilization test (Oxoid).

Antibiotic Sensitivity Testing
Kirby-Bauer disc diffusion technique was used on the Klebsiella spp isolates [25]. E. coli ATCC 25922 quality control strains was used as the test quality con- Aztreonam (ATM 30 µg). All discs were obtained from Oxoid, UK.
All the plates were incubated at 37˚C for 18 hours, inhibition zones measured and interpreted according to Clinical Laboratory Standard Institute (CLSI) 2020, guidelines. The standard control strain E. coli (ATCC-25922) was used to assure testing performance of the potency of antibiotics discs and the quality of the media.

Phenotypic Screening for ESBL-Producing K.pneumoniae
The double disk synergy method was used to detect ESBL-producing K. pneumoniae where 4 antibiotics discs were used including Cefotaxime (CTX) (BD), Cefotaxime/Clavulanic acid (CTX/CLA) (BD, USA), Ceftazidime (CAZ) (BD, USA) and Ceftazidime/Clavulanic acid (CAZ/CLA) (BD, USA). These antibio- isolates that were ESBL positive and were resistant to at least one fluoroquinolone and or carbapenems were then subjected to Minimum Inhibitory Concentration test (MIC) using the vitek 2 machine (bioMerieux, France) using the GN83 card for antibiotic susceptibility testing (AST).
Gel electrophoresis of PCR products was carried out at a voltage of 100 V on a 1.5% agarose gel for 30 minutes and the DNA staining done using SYBR green dye.

Genetic Relatedness of Bacteria Isolates
This was performed using GTG 5

Demographic Characteristics of Children from Mukuru Slums
A total of 1171 children were recruited into this study comprising of 592  (Table 4).

Prevalence of Klebsiella spp Isolated in Children from Mukuru Slums
Of the 1171 participants recruited in the study, prevalence of Klebsiella spp carriage was 33.7% (395/1171). Prevalence of K. pneumoniae was established at 31.2% (365/1171) while that of K. oxytoca was at 2.6% (30/1171). Within Klebsiella spp therefore children were significantly 12 times more likely to be colonized with K. pneumoniae (OR 12.2; p = 0.0001). Although a significant association was statistically derived between Klebsiella intestinal carriage and the residential area, this association could not clearly be concluded due to the number of participants whose villages were not captured (Table 4) ( Figure 1). Further, no significant association was observed between carriage and presentation type (OR 1.2; p = 1.3). All other correlates of carriage included age and gender (Table 5).

Antibiotic Resistance Patterns of K. pneumoniae and K. oxytoca
K. pneumoniae showed highest resistance to ampicillin at 77.5% moderate resistance to one of the most commonly prescribed amoxicillin/clavulanic acid at 37% with low or close to no resistance for imipenem and meropenem each recording percentage resistance of 1.6% ( Figure 2). Generally, K. pneumoniae showed high resistance to 3 rd generation cephalosporins (cefotaxime, ceftriaxone and ceftazidime) compared to fluoroquinolones (nalidixic acid and ciprofloxacin). The least possible resistance from K. pneumoniae was shown for cephamycin (cefoxitin) and carbapenems (imipenem and meropenem).
A similar trend was shown for K. oxytoca that again showed high resistance to ampicillin at 70% with resistance to the most commonly empirically prescribed amoxicillin/clavulanic acid also being relatively high at 56.7%. Moderate resistance was observed for Nalidixic acid and cefotaxime at 33.3% and 26.7% respectively, with low resistance observed for Ciprofloxacin and cefoxitin both at 3.3%. No resistance was observed to the carbapenems (imipenem and meropenem) by K. oxytoca.

Prevalence of Multidrug Resistant (MDR) K. pneumoniae and K. oxytoca and Their Resistance Patterns across Different Antibiotic Panels
Multidrug resistance (MDR) was defined as an isolate non-susceptible to at least        (Table 6).
There was a significant difference in resistance to monobactam (OR = 0.56; p = 0.02), third generation cephalosporins (OR = 1.88; p = 0.01), aminoglycosides (OR = 3.6; p = 0.00) and beta lactam inhibitor (OR = 1.54 p = 0.05) observed in K. pneumoniae isolated from diarrheic children. This means that diarrheic children have a higher chance of colonization with K. pneumoniae resistant to the antibiotics stated above. There was a significant difference noted in resistance to third generation cephalosporins, among K. pneumoniae isolates obtained from children between 1 and 50 months (OR = 0.43; p = 0.002). Children in in this age group have higher odds of carrying K. pneumoniae resistant to third generation cephalosporins. Additionally, a significant difference was observed in resistance to phenicol (OR = 1.81; p = 0.02), tetracycline (OR = 3.14; p = 0.00), aminoglycosides (OR = 4.35; p = 0.000) and folate biosynthesis inhibitor There was no significant difference in resistance to cephamycin from isolates obtained from participants among the various age categories, gender, resident villages and symptoms. There was no significant difference in resistance to third generation cephalosporins among isolates obtained from various resident villages and gender. In addition, no significant difference in resistance to quinolone and fluoroquinolone among isolates obtained from children among various age categories, resident villages and gender. With regard to carbapenems resistance,

Frequency of ESBL Production in the Isolated Klebsiella spp
The proportion of K. pneumoniae Extended Spectrum Beta Lactamase (ESBL) producing isolates was 22.74% (83/365). Out of these ESBLs, 11.50% (42/365) were resistant to at least one fluoroquinolone; while 2.19% (8/365) were resistant to at least one carbapenem and to at least one fluoroquinolone. Comparative analysis showed a significant likelihood with 60% more chance of isolating ESBLs among children aged between 0 -50 months (OR = 0.38; p = 0.001) compared to children 51 -100 months (OR = 0.85; p = 0.66). Again, although an association was observed for ESBLs and residence, this could not effectively be interpreted since majority of the participants did not indicate their areas of residence (Table 11). The prevalence of K. oxytoca ESBLs was 13.33% (4/30).
Comparative analysis showed no difference in age category, gender, health facility and symptoms (Table 11).    Carriage of multiple genes bearing resistance to both 3 rd generation cephalosporins and fluoroquinolones resistance antibiotics was observed in 90.48%) (38/42) isolates (Table 12) while carriage of 3 rd generation cephalosporins and carbapenems resistance genes was observed in 7.14% (3/42) isolates. Carriage of resistance genes against three classes of drugs (Beta lactams, fluoroquinolones and Carbapenems) was only observed in 7.14% (3/42) K. pneumoniae isolates (Table 12).

Phylogenetic Analysis of the Isolated Klebsiella spp
The dendrogram was derived from the 42 K. pneumoniae ESBL producing isolates that were also resistant to fluoroquinolones and/or carbapenems. Within this dendrogram there were Clades (branch that includes a common ancestor and all of its descendants), Clustering groups (descendants in a clade at 40% similarity), Clustering sub groups (descendants in a clade at 100% similarity). Two clades designated A and B, 6 clustering groups designated group1 -6 and 40 subgroups were recorded. Clade A includes group 1 -4 while clade B includes group 5 and 6. There was 100% similarity index in Group 1, 2 and 5 as highlighted in the boxes. Of notice, was an outgroup observed in group 6 of clade B which showed a higher number of bands compared to the rest of the isolates group 6. Out of the 40 Sub groups 37.5% (15/42 showed >80% similarity index with the highest number observed in group 1 while, 62.5% (25/42) showed a similarity index < 80% indicating that these 25 isolates were distantly related. Isolates that carried resistance genes to fluoroquinolones clustered tightly as observed in all the groups, while isolates that carried resistance genes to carbapenems were diverse and did not cluster together as observed in group 1 and 6. From the phylogenetic analysis therefore we state that 83.33% (35/42) of the isolates were diverse and hence disbanding the possibility of clonal spread of MDR strains ( Figure 5).

Discussion
In this study we report a community prevalence of gastrointestinal K. pneumoniae of 31.16% and of K. oxytoca of 2.56% among the slum dwelling children.
This prevalence noted in the community is higher than what has previously reported among ICU patients [13] [33] of 23% and 19%. K. pneumoniae and K. oxytoca are ubiquitous in nature and are found in various environments including mucosal membranes of humans where they colonize the gastrointestinal tract, the skin and the nasopharyngeal. In the gastrointestinal tract, they occur as normal flora. However, when they cross the gastrointestinal mucosal membrane into other systems of the body, they become opportunistic pathogens, causing infections such as pneumonia, bloodstream infections, meningitis and urinary tract infection. The high prevalence of Klebsiella spp in the community may not have a major impact on the children as the organisms do not cause infection in the gastrointestinal tract. However, these micro-organisms indicate the resistance genes circulating in Mukuru. These genes could be disseminated to other pathogens which pose a challenge in patient management.     The proportion of MDR K. pneumoniae observed in this community can be attributed to selective pressure for certain antibiotics [34]. This indicates antibiotic use/misuse in Mukuru which contributes to emergence and persistence of antibiotic resistance. The therapeutic use of different antibiotics for empirical and prophylactic management of gastrointestinal infections is rampant, in this slum community. Indeed, it has been established that due to high burden of pathogens causing gastrointestinal infections, uncontrolled use of antibiotics to the communities contribute to selective pressure leading to resistance [35]. For example, due to poverty in this slum community and ease of access of antibiotics as over the counter (OTC) medications, dispensing chemists record high purchase of relatively cheap antibiotics such as chloramphenicol, ampicillin and co-trimaxazole. Due to high burden of HIV infections in the community we assessed there is also a high rate of empiric use of antibiotics such as trimethoprim sulfamethoxazole and gentamicin for treatment of gastrointestinal infections. Use of such drugs as first line for treatment of enteric infections or for prophylactic management for prevention of HIV opportunistic infections has been described as a major driver of antibiotic resistance [36]. AMR determinants such as plasmids and insertion sequences containing multiple resistance genes can be present in these microorganisms. These determinants have the ability to transfer resistance genes in vitro [37]. Indeed, other studies describing the MDR patterns in Nairobi have described Mukuru slums as MDR hotspots [38]. The potential for aggravated transmission of MDR genes to the vulnerable populations was demonstrated in this study due to the determination that there was no significant difference in the prevalence of MDR infections in the asymptomatic (non-diarrheic) and the symptomatic (diarrheic) cases. The latter finding demonstrated that both the symptomatic and the asymptomatic play an equally significant role in the carriage of MDR. The proportion of MDR Klebsiella noted in this study (64.1%) is similar to a study done [39] in Kilifi where the proportion of MDR was 63%, although the isolates were from invasive infections. In contrast, a study conducted [40] in rural western Kenya showed a lower proportion MDR Klebsiella of 36.7%. This contrast can be attributed to differences in economic, social and environmental settings. In East Africa, the proportion observed ranges from 80% -95%, which is comparatively higher than that observed in this study [41] [42]. While the prevalence of K. oxytoca was low at 2.56% among children in Mukuru, isolation of MDR K. oxytoca was high at 96.67% (29/30). Though not highly prevalent from children in Mukuru, it's alarming that nearly all the isolates of K. oxytoca are MDR. The latter implies that, if colonization by K. oxytoca proceeds infection, the disease can record high treatment failures particularly among the immune compromised persons. Additionally, K. oxytoca can transfer its resistance genes to other organisms including K. pneumoniae and other enteric bacterial pathogens, leading to a high burden of treatment failure. Unlike K. pneumoniae, horizontal transfer of genes in K. oxytoca is not well documented, although trends of low prevalence of K. oxytoca with high isolation rates of MDR K. oxytoca have been documented in India [43], among adults, and in Iran [44]. Similar studies data are scarce in Africa. Resistance patterns observed in 3rd generation cephalosporins can be attributed to their widespread use and/or misuse in the health facilities in Kenya [45]. The high frequency of the βeta-Lactams resistance genes of bla CTX-M, bla TEM, bla OXA-1 and bla SHV may be due to the presence of mobile genetic elements bearing these genes in this slum environment. Further, various studies in Africa [46] [47] [48] have alluded to the fact that the high economic growth in the recent years has led to ease of accessibility of βeta-Lactams over the counter leading to increased abuse and/or misuse of these antibiotics, hence the predominance of bla TEM and bla CTX-M genes in the environment. Of the 22.7% prevalent ESBL producing K. pneumoniae, the rate of isolation was significantly higher in children below 50 months, potentially attributed to their underdeveloped immune system or possible nutritional deficiencies due to their residential environment. This ESBL prevalence however appears lower compared to other studies done in Kenya ranging between 44% [39] and 71% [46]. It is however noted that the prevalence of ESBL at 71% was established among K. pneumoniae isolated from urine samples. Resistance to Quinolone and Fluoroquinolone from this study was generally low at 18.36%, indicating low selective pressure for these antibiotics. The predominant qnr genes (which is plasmid mediated; PMQR) were qnrS, qnrB, indicating possible horizontal transfer of these genes can occur to other organisms including pathogens. Therefore, exacerbating fluoroquinolone resistance which is the choice of treatment for a variety of infections. In addition to PMQR, fluoroquinolone resistance can be mediated by chromosomal mutations especially in DNA gyrase and topoisomerase encoding genes such parE and parC genes., which were also detected during this study at 38% and 48% respectively Although these genes were observed in this study in relatively high proportions, the mutations can only be observed after performing DNA sequencing, which was a limitation in this study. Notably, isolates that carried qnr genes were all resistant to nalidixic acid, however some isolates exhibited partial reduction of ciprofloxacin efficacy to K. pneumoniae as opposed conferring complete resistance to the antibiotic (0.25 -0.5 μg/mL). This indicates that qnr genes confer complete resistance to quinolones and partial resistance to fluoroquinolones. The low rate of resistance can be due to the low prescription of Ciprofloxacin and its high cost despite being widely available. Similar findings where qnrS and qnrB genes have been found to be most prevalent in Africa [49] [50] have been documented. Low resistance to carbapenems of 3.3% was noted in this study. The prevalence of carbapenem resistance gene bla NDM-1 was also low at 7.1%. The low resistance can be due to their limited use and availability in the market in Kenya [51]. Indeed, a similar study (Poirel et al., 2010) conducted in Nairobi only observed one bla NDM positive isolate, with similar study conducted in Kilifi, Kenya [39] observed no bla NDM isolates. The study in Kilifi however, documented a plasmid with a genetic architecture of a known bla NDM carrying plasmid in a total of 25 isolates.
In this study qnr B and S genes were found to co-exist with bla CTX-M ESBLs.
Co-carriage of ESBLs with fluoroquinolones can be attributed to the presence of plasmids containing a plethora of resistance determinants such as the qnr genes which encode for qnr protective proteins. According to literature, plasmid mediated resistance to quinolone is often associated with ESBLs [52] [53]. Isolates that carried resistance genes to the 3 classes of drugs that were of interest to this study (3 rd generation cephalosporins, fluoroquinolones and carbapenemes), were  [57].
Phylogenetic relatedness analysis showed a high number of <80% similarity index amounting to 62.5%, which is indicative of the high diversity among the isolates, ruling out the possibility of clonal spread of MDR strains. Isolates that showed >80% similarity index, amounting to 37.5% were closely related. Those that showed 100% similarity index were considered completely related and amounted to 15%. The findings showed a high genetic diversity of Klebsiella strains circulating. Other studies conducted in Kenya have also observed high genetic diversity among K. pnemoniae isolates [39] [40].
If colonization precedes infection, and there's high concordance between colonizing and infecting isolates [13] then MDR K. pneumoniae such as those carrying AMR genes for 3 rd generation cephalosporins (ESBLs), fluoroquinolones and or carbapenems pose a great risk to the community. Therefore, identification of colonizing strains can inform on patient care interventions. Indeed, multidrug resistance is a problem in Mukuru slums and there is urgent need curb this menace. Various measures can be taken to reduce the emergence and spread of resistance. Creating awareness on antibiotic resistance and how it affects their well-being; Improvement of sanitation, provision of clean water and treatment of sewage waste; Antibiotic stewardship that allows for prudent use of antibiotics; Prioritization of research on antibiotics alternatives and development of AMR diagnostic tools [58].

Conclusion
The high proportion of MDR K. pneumoniae and MDR K. oxytoca and the carriage rates of resistance genes observed in the gastrointestinal tract of participants present a threat to community spread of MDR resistant Klebsiella. It accentuates the need for more effective infection control measures, proper implementation of public health policies, prioritization of AMR intervention development, surveillance of AMR circulating genes and mapping of MDR Klebsiella spp especially in the informal settlements. It also shows empirically that the gut is an important reservoir of a plethora of resistance genes especially in asymptomatic individuals who can disseminate to the vulnerable persons in the community. Such asymptomatic individuals provide key target populations for intervention. More studies are required therefore to further understand the gut resistome and transmission dynamics of AMR genes in informal settlements of low resource countries.