1 Background

Multidrug resistance became a great and serious issue attacking the global health and so attracted the attention of most of the scientists worldwide, either in the field of human or veterinary medicine. Antimicrobial resistance is attributed to the extensive and abuse of antimicrobials either in veterinary, human medicine or even agricultural field. The non-supervised practices of using antimicrobials and detergent in particularly veterinary medicine acts as one of the major causes of the antimicrobial resistance widespread, especially in the field of the food animals’ rearing, where the animals act as reservoir of the antimicrobial resistant bacteria, transmitting them to human either through its food products (meat or milk sources) or via direct contact with human [1]. However, the antimicrobial resistance is known to be widely spread in a population of horse, like that in hospitals or housing (stables), through the horizontal transmission of the resistance genes from bacteria to another or even a bacterial population from the infected animal to another animal or even to human and vice versa [2, 3]. The acquisition of the resistance genes is carried out via the transmission of the mobile genetic elements like the insertion sequences, transposons and integrons [1]. There are various mechanisms of drug resistance recorded including: degradation of the drug by the action of an enzyme, alteration of the target site of the drug, alteration of the membrane permeability for the drug and drug efflux pump [4]. Studies conducted on equine infections caused by multidrug resistant bacteria are mostly restricted to the methicillin-resistant Staphylococcus species (MRS) and the extended spectrum extended spectrum β-lactamases (ESBL)-producing Escherichia coli like the work of the following studies [3, 5,6,7,8,9,10]. The extended spectrum beta-lactamases (ESBL) are enzymes that have the ability to hydrolyze penicillins, different generations of cephalosporins (first, second and third ones) and aztreonam, while inhibited by clavulanic acid. There’re different types and classes of beta-lactamases. The largest group of beta-lactamases’ mutants is the group of TEM (was the first one discovered, derived from Temoniera “blood culture” in Greece) and SHV (Sulfhydryl Variable) β-lactamases, this group includes over 150 members [11, 12]. The followed largest one is the CTX-M (cefotaximase) group, is the most frequent one nowadays with 5 subgroups involved about 40 members [12, 13]. Followed by OXA-type β-lactamases and is most frequent in Pseudomonas aeruginosa. Finally, a variety of beta-lactamases as PER, VEB, GES, BES, TLA, SFO, IBC groups. ESBL-producing K. pneumoniae are usually associated with resistance against carbapenems via KPC (carbapenemase-producing K. pneumoniae) and/or resistance against other antibiotic classes such as sulfonamides, fluoroquinolones and aminoglycosides [12].

K. pneumoniae is one of the top six pathogens considered as the most virulent and multidrug resistant bacteria, which are abbreviated as ESKAPE (Enterococcus faecium, Staphylococcus aureus, Klebsiella pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter species) [14]. K. pneumoniae causes respiratory manifestations in the form of hemorrhagic nasal discharge and pneumonia in equine [15] and may induce abortion in pregnant mares [16], additionally, it is one of the most important bacteria causing nosocomial infections [17,18,19]. The most predominant detected drug resistance in K. pneumoniae are the resistance against carbapenems, cephalosporins, fluoroquinolones, and aminoglycosides [20]. Studies concerning drug resistant K. pneumoniae in equine are almost null in Egypt. The aim of this study was to investigate and analyze the distribution pattern of the drug resistance genes among fifteen K. pneumoniae isolates, were the most predominant bacteria in a previous study [21]. These were recovered from dead foals suffered respiratory infection and tested for their sensitivity against different antibiotics recommended by the CLSI 2018. The determinant genes tested were: blaESBL (blaCTX, blaTEM), metallo-beta-lactamases (blaVIM, blaNDM-1), carbapenemase (blaKPC), Quinolones’ target protective proteins “Qnr proteins” (qnrB, qnrS), aminoglycoside adenylyltransferase (aadA1), efflux pump (AcrAB) and porin protein gene mutation/deletion (ompK35).

2 Methods

2.1 Preparation of the bacterial isolates and DNA elution

Bacterial isolates (15 strains of K. pneumoniae) were isolated and preserved at − 70 °C in a previous study [21]. The isolates were thawed and sub-cultured on MacConkey agar at 37 °C/24 h. The chromosomal DNA was eluted using DNA Extraction kits (ABIO pure™ Genomic DNA Extraction kits, AllianceBio, Bothell, USA) according to the manufacturer instruction. The DNA templates were examined for the KP_16S (NM3) for confirmation of the identified isolates biochemically. All the positive isolates were tested for the presence of the drug resistance determinant genes class A β-lactamase enzymes (ESBL determinant genes): blaCTX-M and blaTEM, class B metallo-β-lactamase/carbapenemase enzymes: blaVIM, blaNDM1, carbapenemase: blaKPC, aminoglycoside adenylyltransferase: Aada1, quinolones target’s protective proteins determinant genes: qnrB, qnrS, drug efflux pump: acrAB, porin muted/deleted gene: OmpK35. The primers (Metabion, Germany) used were listed in Table 1.

Table 1 Primers sequences, target genes, amplicon sizes and cycling conditions
Full size table

2.2 PCR amplification

Primers were utilized in a 25 µl reaction volume containing 12.5 µl of PCR Master Mix (EmeraldAmp Max PCR Master Mix, Takara, Japan), 1 µl of each primer of 20 pmol concentration, 5.5 µl of water, and 5 µl of DNA template. The reaction was performed in an Applied biosystem 2720 thermal cycler.

2.3 Analysis of the PCR products

The products of PCR were separated by electrophoresis on 1.5% agarose gel (Applichem, Germany, GmbH) in 1 × TBE buffer at room temperature using gradients of 5 V/cm. For gel analysis, 15 µl of the products was loaded in each gel slot. Gelpilot 100 bp plus DNA ladders (Qiagen, Gmbh, Germany) and generuler100 bp ladder (Fermentas, Germany) were used to determine the fragment sizes. The gel was photographed by a gel documentation system (Alpha Innotech, Biometra) and the data was analyzed through computer software.

2.4 CTX-M sequencing and analysis

Positive PCR products of CTX-M genes were purified using PCR Product extraction kit (Qiagen, Valencia). Sequencing kit (Perkin-Elmer) was used for the sequence reaction and then it was purified using Centrisep spin column. DNA sequences were obtained by Applied Biosystems3130 genetic analyzer (HITACHI, Japan), a BLAST® analysis (Basic Local Alignment Search Tool) [30] was initially performed to establish sequence identity to GenBank accessions. The phylogenetic tree was created by the MegAlign module of Lasergene DNAStar version 12.1 [31] and phylogenetic analysis was done using neighbor-joining method in MEGA6 [32].

2.5 Statistical analysis

The results were analyzed using QI Macros software; has been loaded to the startup directory of Microsoft Office Excel 2019.

3 Results

3.1 PCR results

The phenotypic identification of the fifteen K. pneumoniae isolates was confirmed using the Polymerase Chain Reaction (PCR). All the tested isolates showed positive amplification product of KP_16S (NM3). The fifteen isolates were tested for ten drug resistance gene determinants. The incidence profile of the determinants was analyzed and graphically presented in histogram (Fig. 1). The results showed that the most predominant gene determinants were the extended spectrum β-lactamase (ESBL) blaTEM, efflux pump AcrAB and the muted/deleted porin protein gene OmpK35 (harbored by 100% of the isolates). With less incidences, the determinants of the ESBL blaCTX-M, quinolones’ target protective protein QnrS, aminoglycoside adenylyltransferase AadA1, carbapenemases blaKPC and blaNDM-1 have occurred in rates 86.7, 73.3, 40, 23 and 7.7% respectively. With complete absence of the carbapenemase blaVIM and the quinolones’ target protective protein QnrB.

Fig. 1
figure 1

A Incidence of the multidrug resistance determinants, B Incidence of multidrug resistance categories

Full size image

By investigating the most frequent number of the associated drug resistant determinant genes, it was found that the most frequent number was 5 associated determinants in 40% (6/15) of the isolates, followed by 6 associated determinants 26.7% (4/15), while the frequency of 7 associated genes was 13.3% (2/15) as well as 4 associated ones and finally only 1 isolate (6.7%) harbored 3 associated determinants. Pareto chart was used to explore the most predominant associated determinants (Fig. 3) contributing in the drug resistance of the isolates, where more than one quarter (26.7%) of the isolates harbored the ESBL (TEM and CTX-M), associated with quinolones’ target protective protein gene determinant QnrS, efflux pump AcrAB and the muted/deleted porin protein gene determinant OmpK35. Whilst, the both of the ESBLs, together with the efflux pump (AcrAB) and the muted gene OmpK35 had the same quote (2/15, 13.3% of the isolates) as the association of the ESBLs with AcrAB, QnrS, KPC, muted gene OmpK35 and the association between ESBLs with AcrAB, QnrS, AadA1 and the muted gene OmpK35. On the other hand, four isolate harbored four different drug resistance profiles: TEM, AcrAB, the muted gene OmpK35, TEM, CTX-M, AcrAB, the muted gene OmpK35, QnrS, KPC, AadA1, TEM, CTX-M, AcrAB, the muted gene OmpK35, QnrS, AadA1, NDM-1 and the last one was TEM, AcrAB, AadA1, QnrS and the muted gene OmpK35 (Table 2). The detailed phenotypic and genotypic antibiotic resistance profiles of each isolate is available in Appendix 1 (Table 3), while figures showing the amplification products of the PCR in Appendix 2.

Table 2 Genetic profile of the K. pneumoniae isolates
Full size table

3.2 CTX-M sequencing and analysis

All the isolates showed positive PCR products of CTX-M genes were purified and sequenced. Ten of the sequences showed clear similarities with known genes in the National Center for Biotechnology Information (NCBI) genbank and were added to the genbank data base with accession numbers mentioned in Table 2. The ten sequences were consequently aligned with related sequences in NCBI and phylogenetically analyzed as shown in Fig. 4. The lineages were typed as OXY-5 (50%), CTX-M-15 (40%) and CTX-M-27 (10%).

4 Discussion

The World Health Organization described K. pneumoniae as one of the life-threatening pathogens, that’s for the widespread of its highly resistant strains all-over the world not only against beta-lactams (penicillins, cephems, monobactams…), but also against the last resort treatment carbapenems (imipenem, meropenem…), which were considered as drug of choice [12, 33]. After the emergence of Extended Spectrum beta-lactamase-producing bacteria (ESBL), the situation became more imperil [12]. Unfortunately, drug resistance against certain antibacterial category is multifactorial, making facing such a problem very complex. For example: resistance against beta-lactams may be due to hydrolysis by enzymes as β-lactamases, metallo-β-lactamases, efflux pump or mutation of the outer cell membrane porins or altogether.

Alteration in the outer cell membrane permeability causing increased efflux of antibiotics by the aid of the efflux pumps and/or loss of porin proteins OmpK35 and OmpK36 are other important causes of drug resistance [10, 17, 34]. The efflux pump (AcrAB) is encountered in the resistance against β-lactams, quinolones, chloramphenicol, nalidixic acid, ertapenem, tetracycline, trimethoprim, macrolides, and piperacillin [35,36,37], while the loss of porins is known to be encountered in the resistance against carbapenem, ciprofloxacin, cephalosporines and chloramphenicol [18, 37, 38]. The expression of the porin protein OmpK35 allows efficient penetration of some of the cephems like cefoxitin and cefotaxime and carbapenems [38]. In the current study, AcrAB was present in 100% (15/15) of the strains, whilst OmpK35 was absent in all isolates. These results were complying the results of Wassef et al. [17] mentioned that the loss of Omp35 is detected in 85.7% of the isolates.

This revealed that, the current strains’ resistance against beta-lactams (penicillins, cephems and monobactams) according to the determinant genes was not only attributed to beta-lactamases (TEM and CTX-M) and metallo- beta-lactamase (NDM1), but also to the efflux pump acrAB and the loss of the porin protein ompK35 expression. Indeed, the results of the carbapenemases were correspondingly interesting, where only 3 isolates were positive for the determinant genes (2 positive for KPC and only 1 for NDM-1), on the contrary, the phenotypic resistance profile of 60% of the strains (9/15: K2, K4, K5, K7, K9, K11, K12 and K13) were resistant against meropenem, while the carbapenemases (VIM, NDM1, KPC) were negative. This explained likewise by the positive results of the efflux pump and the absence of the porin protein gene. Similarly, in case of quinolones; the results of the tested Qnr proteins (protect of the quinolones’ targets from inhibition) were not conforming the phenotypic resistance profile, where there were 4 isolates (K1, K3, K5 and K11) showed resistance against quinolones while the determinant genes were absent. These cases were explained by the results of the tested gene ompK35, which was absent in all strains and the presence of AcrAB efflux pump in all isolates conferring the high virulence and drug resistance of the strains against carbapenems and quinolones [10].

Aminoglycosides are very important antibiotic class, usually used as empirical treatment in combination with beta-lactams in case of gram-negative infections. There’re various mechanisms of resistance against aminoglycosides, but the most predominant one is the production of aminoglycoside-modifying enzymes [39]. We used aadA1 (aminoglycoside adenylyltransferase) as a determinant gene in this study. It was found in 6 isolates only (40%), while all the strains were actually resistant to streptomycin in the phenotypic profile that may be due to the coding genes were present on plasmids not chromosome and/or the resistance was attributed to genes other than the tested; further studies are required. These results are completely disagreed with Kim et al. [34], whose result is 100%.

The analyzed data showed in (Fig. 2) was astounding, it was recognized that the most frequent associated number of drug resistance determinant genes was 5 genes in 40% of the strains, followed by 6 associated genes 26.7%. The results addressed a serious alarm for the facilitated spread of the MDR K. pneumoniae and the virulence of the infections caused by such strains and of course reflected how difficult was the dealing with such hard cases. It’s noted that, ESBL producing plasmid encoding genes are usually associated with genes responsible for the resistance against quinolones, aminoglycosides and metallo-beta-lactamases making the alternatives for treatment of such cases very difficult [40,41,42,43]. Nevertheless, the co-resistance mechanism of such associated genes is still unknown [41]. The present study revealed that the most predominant associated drug resistance genes coded on the chromosome were beta-lactamases (blaTEM as sole or with blaCTX-M) and efflux pump (acrAB) with the absence of the OmpK35 which were associated in 100% (15/15) of the strains either associated alone (3/15 of strains) or with other drug resistance determinant genes (12/15). Followed by the resistance gene qnrS, occurred in 73.3% (11/15) of the strains, either associated with beta-lactamases and efflux pump with absent OmpK35 only 26.7% (4/15), or in addition to other resistance genes (illustrated in Table 2 and Fig. 3). These were agreed with that of Almaghrabi et al. [44], whose work is on K. pneumoniae producing carbapenemase, they mentioned that “all KPC producers harbored TEM-1 and SHV-12” and none of the strains harbored NDM and VIM beta-lactamases. Also conformed the results of Hennequin and Robin [45], where the KPC enzymes are frequently associated with other antibiotic resistance like fluoroquinolones and aminoglycosides.

Fig. 2
figure 2

Frequency of association in determinant genes for MDR occurrence within the strain

Full size image
Fig. 3
figure 3

Pareto Chart showing the frequency of the associated MDR determinant genes

Full size image

To the best of our knowledge, there are no previous records about the types of the CTX beta-lactamases in K. pneumoniae concerning the veterinary medicine in Egypt. In the current study, sequencing of the CTX-M gene was carried out for exploring the types of CTX beta-lactamases existed in K. pneumoniae isolated from equine in Egypt. This study may be the first record of the occurrence of OXY-5 in K. pneumoniae in Egypt, either in veterinary medicine or in human medicine. The targeted sequence while choosing the primer was CTX-M, nevertheless, it was surprising to find 50% of the lineages similar to that of OXY-5, by investigation it was found that the “universal CTX-M primer” amplifies the chromosomally located blaOXY and blaCTX-M as well and the only method to differentiate between them is the sequencing [46].

The phylogenetic tree (Fig. 4) illustrated that the sequences typed as CTX-M-15 in the current study (marked with red circles) were closely related to lineages originated from E. coli with human source recovered from stool of a cancer patient in Cairo, Egypt (retrieved from the data reported in NCBI with the accession number KX013145) and from raw milk cheese originated also from Egypt (NCBI data of accession number CP042974) in addition to other K. pneumoniae strains recovered from different sources such as patients in Czech hospitals, Ireland (CP050379), waste water (MH190898) and fish (MT371968). However, the CTX-M-27 sequence (marked with black square) was related to lineages originated entirely from E. coli strains, mostly recovered from human. Moreover, the five lineages typed as OXY-5 in the current study were related to lineages recovered from isolates of K. oxytoca with human origin. The discrimination of the sources of these lineages indicated the widespread of such resistance gene across continents, in different species of either bacteria and/or animals and human being, this’s attributed to the global trading of different types of food animals either as frozen meat or as processed food products [47], in addition to, the close contact between human and animals without commitment to perfect infection control rules, permitting the transmission of the carrier bacteria from and to human and the exchange of the resistance genes inbetween different bacterial species.

Fig. 4
figure 4

Molecular phylogenetic tree of CTX-M gene sequences in 10 strains of K. pneumoniae isolated from dead foals with respiratory infection. The phylogenetic analysis was conducted in MEGA6 software [32] for the linear DNA, partial CDS of 534 bp each, with related genes in CTX-M family using Neighbor-joining method. Accession number is indicated adjacent to each gene. The tree is drawn to scale with branch lengths measured in the number of substitutions per site. Bootstrap values more than 50% (basing on 500 replications) are indicated at nodes. The genes of the current study were labeled with red, black and blue marks according to their relatedness with CTX-M-15, CTX-M-27 and OXY-5 respectively

Full size image

5 Conclusions

The current study represented the first record of the predominance of the resistance genes and their associations among K. pneumoniae strains recovered from dead foals with respiratory infections. The results of the study were alarming a serious problem in the cases infected by K. pneumoniae, where a strain harbored at least three drug resistance genes and the most predominant number of associated drug resistant genes harbored in a particular strain is 5, making the dealing with such cases very difficult. Additionally, the study proceeded in the sequencing of the CTX-M gene in order to identify the predominant type and find its evolutionally lineages with other genes either in Egypt or globally. It may be the first record of OXY-5 existence in K. pneumoniae recovered from either human or animals in Egypt. These findings enforce us to enhance considering the “One Health” approach in all studies concerning MDR bacteria and conducting more applied studies concerning the bacterial drug resistance with consideration of the plasmids in both human and veterinary medicine, exploring the case status of the drug resistant important pathogens “ESKAPE” in Egypt.