Exploring the Antibiotic Resistance Profile of Clinical Klebsiella pneumoniae Isolates in Portugal

While antibiotic resistance is rising to dangerously high levels, resistance mechanisms are spreading globally among diverse bacterial species. The emergence of antibiotic-resistant Klebsiella pneumoniae, mainly due to the production of antibiotic-inactivating enzymes, is currently responsible for most treatment failures, threatening the effectiveness of classes of antibiotics used for decades. This study assessed the presence of genetic determinants of β-lactam resistance in 102 multi-drug resistant (MDR) K. pneumoniae isolates from patients admitted to two central hospitals in northern Portugal from 2010 to 2020. Antimicrobial susceptibility testing revealed a high rate (>90%) of resistance to most β-lactam antibiotics, except for carbapenems and cephamycins, which showed antimicrobial susceptibility rates in the range of 23.5–34.3% and 40.2–68.6%, respectively. A diverse pool of β-lactam resistance genetic determinants, including carbapenemases- (i.e., blaKPC-like and blaOXA-48-like), extended-spectrum β-lactamases (ESBL; i.e., blaTEM-like, blaCTX-M-like and blaSHV-like), and AmpC β-lactamases-coding genes (i.e., blaCMY-2-like and blaDHA-like) were found in most K. pneumoniae isolates. blaKPC-like (72.5%) and ESBL genes (37.3–74.5%) were the most detected, with approximately 80% of K. pneumoniae isolates presenting two or more resistance genes. As the optimal treatment of β-lactamase-producing K. pneumoniae infections remains problematic, the high co-occurrence of multiple β-lactam resistance genes must be seen as a serious warning of the problem of antimicrobial resistance.


Introduction
Klebsiella pneumoniae is a gram-negative opportunistic pathogen capable of colonizing, invading, and causing infections in diverse anatomical sites of the human body [1,2]. Unfortunately, over the last decades, the indiscriminate use of antibiotics in human health and agriculture has contributed to the emergence of the antibiotic resistance problem [3]. Consequently, the growing resistance of K. pneumoniae infections has become a burden for clinicians. β-lactam antibiotics, including penicillin derivatives, cephalosporins, monobactams, and carbapenems, are the most widely used class of drugs in the treatment of bacterial In total, 102 K. pneumoniae isolates collected by the laboratories from two Portuguese hospitals between 2010 and 2020 were used during this work. Strains were isolated from patients admitted to different areas of health care within the hospitals. Despite the source samples being diverse, the most significant was urine (39.2%), rectal swabs (15.7%), blood cultures (12.7%), and sputum (8.8%). Figure 1 shows the source of all isolates included in this work. To facilitate further analysis, the isolates were divided into five main groups according to sample source: urine (39.2%), rectal swabs (15.7%), respiratory-associated infections (15.7%) (including sputum, bronchial aspirates, and bronchoalveolar aspirates), blood cultures (12.7%), and other samples (16.7%).

Resistance Profiles of K. pneumoniae Isolates to β-Lactam Antibiotics
The susceptibility rates of K. pneumoniae isolates for each β-lactam tested (Table 1) are shown in Figure 2. Ertapenem (ERT) was effective (i.e., susceptible profile) against 24 The susceptibility rates of the other β-lactams were lower or even null for most isolates, with resistance rates >90% for all β-lactams tested other than the carbapenems. Only cephamycins, cefoxitin (CXI), and cefotetan (CTT) exhibited greater effectiveness (40.2% and 68.6%, respectively) against K. pneumoniae isolates. Furthermore, 21 (20.6%) K. pneumoniae isolates showed a complete non-susceptibility profile to all β-lactams tested, including carbapenems. Based on individual antimicrobial profiles, each isolate was classified as a potential producer of CPN, ESBL, or AmpC β-lactamases according to the criteria recommended by EUCAST for the detection of resistance mechanisms and described in the "Materials . Antibiotic resistance portfolio of K. pneumoniae isolates determined according to EUCAST criteria. Cefotetan (CTT) is not included in EUCAST breakpoints guidelines, so we used the CLSI criteria. Note. The percentage of K. pneumoniae considered resistant according to the guidelines was statistically different from the % of isolates classified as intermediate or susceptible (one-way ANOVA with Holm-Sidak's multiple comparisons tests; p < 0.05). The susceptibility rates of the other β-lactams were lower or even null for most isolates, with resistance rates >90% for all β-lactams tested other than the carbapenems. Only cephamycins, cefoxitin (CXI), and cefotetan (CTT) exhibited greater effectiveness (40.2% and 68.6%, respectively) against K. pneumoniae isolates. Furthermore, 21 (20.6%) K. pneumoniae isolates showed a complete non-susceptibility profile to all β-lactams tested, including carbapenems.
Based on individual antimicrobial profiles, each isolate was classified as a potential producer of CPN, ESBL, or AmpC β-lactamases according to the criteria recommended by EUCAST for the detection of resistance mechanisms and described in the "Materials and methods" (see Table S1). Notably, all K. pneumoniae isolates showed a potential ESBLproducing phenotype, while 10 (9.8%) isolates had a characteristic phenotype of both AmpC-and/or ESBL-producing bacteria. In addition, 31 (30.4%) isolates exhibited a resistance phenotype that can be classified as CPN-and/or ESBL-producing isolate, and 50 (49.0%) isolates showed a resistance phenotype that can be classified as CPN-, AmpC-, and/or ESBL-producing isolate.

Screening of Acquired β-Lactam Resistance Genes
Prior to the experiments, four multiplex PCRs were designed and optimized for the detection of the most commonly acquired β-lactam resistance genes using primers described in the literature ( Table 2). Figure S1 shows representative sequences of each of the 11 bla gene families of interest with the binding site of the primer pairs used for PCR detection.
The 102 K. pneumoniae clinical isolates were then tested for the presence of the resistance genes according to the PCR methodology described in "Materials and methods". Figure 3 shows the amplification products of the control strains separated on 1.5% (w/v) agarose gel.  The 102 K. pneumoniae clinical isolates were then tested for the presence of the resistance genes according to the PCR methodology described in "Materials and methods". Figure 3 shows the amplification products of the control strains separated on 1.5% (w/v) agarose gel.   The profile of resistance genes of each K. pneumoniae isolate can be found in Table S1. In total, bla KPC-like was detected in 74/102 (72.5%) K. pneumoniae isolates, while bla OXA-48-like was detected only in three isolates (2.9%). No K. pneumoniae isolates harbored bla OXA-23-like or MBL genes. For the ESBL genes, 76/102 (74.5%) were positive for the presence of bla SHV-like , 38/102 (37.3%) for the presence of bla CTX-M-like , and 47/102 (46.1%) for the presence of bla TEM-like . Among the AmpC genes, 4/102 (3.9%) isolates were positive for bla CMY-2-like , and 4/102 (3.9%) isolates were positive for bla DHA-like . bla KPC-like was most frequently detected in isolates from urine (32.4%), followed by rectal swabs (14.7%), respiratoryassociated infections (8.8%), and blood (7.8%). The isolates positive for bla OXA-48-like were only associated with urine (2.0%), and respiratory (1.0%) isolates. Regarding ESBL genes, the distribution by sample sources is wider. bla TEM-like was detected in the highest percentage in urine, followed by respiratory-associated infections, other samples, blood culture, and rectal swabs. The most prevalent ESBL gene (bla SHV-like ) was most associated with urine, followed by respiratory-associated infections, other samples, rectal swabs, and blood cultures. In turn, bla CTX-M-like was most detected in murine, followed by respiratory-associated infections, blood cultures, other samples, and rectal swabs. bla CMY-2-like and bla DHA-like were detected in 2.0% of urine isolates and 1.0% of respiratory-associated infections and other isolates. This data can be seen in Figure 4A. The profile of resistance genes of each K. pneumoniae isolate can be found in Table S1. In total, blaKPC-like was detected in 74/102 (72.5%) K. pneumoniae isolates, while blaOXA-48-like was detected only in three isolates (2.9%). No K. pneumoniae isolates harbored blaOXA-23-like or MBL genes. For the ESBL genes, 76/102 (74.5%) were positive for the presence of blaSHVlike, 38/102 (37.3%) for the presence of blaCTX-M-like, and 47/102 (46.1%) for the presence of blaTEM-like. Among the AmpC genes, 4/102 (3.9%) isolates were positive for blaCMY-2-like, and 4/102 (3.9%) isolates were positive for blaDHA-like. blaKPC-like was most frequently detected in isolates from urine (32.4%), followed by rectal swabs (14.7%), respiratory-associated infections (8.8%), and blood (7.8%). The isolates positive for blaOXA-48-like were only associated with urine (2.0%), and respiratory (1.0%) isolates. Regarding ESBL genes, the distribution by sample sources is wider. blaTEM-like was detected in the highest percentage in urine, followed by respiratory-associated infections, other samples, blood culture, and rectal swabs. The most prevalent ESBL gene (blaSHV-like) was most associated with urine, followed by respiratory-associated infections, other samples, rectal swabs, and blood cultures. In turn, blaCTX-M-like was most detected in murine, followed by respiratory-associated infections, blood cultures, other samples, and rectal swabs. blaCMY-2-like and blaDHA-like were detected in 2.0% of urine isolates and 1.0% of respiratory-associated infections and other isolates. This data can be seen in Figure 4A. Prevalence of acquired β-lactam resistance genes in MDR K. pneumoniae isolates from two Portuguese hospitals. * Prevalence of a specific resistance gene was statistically different when compared with the other genes from the same group (one-way ANOVA; p < 0.05). (B) Relative frequency of acquired β-lactam resistance genes by sample source. The isolates were categorized into five major sample sources: urine, rectal swab, respiratory-associated infections (including sputum, bronchial aspirate, and bronchoalveolar lavage isolates), blood culture, and others (including remaining sample sources). Prevalence of acquired β-lactam resistance genes in MDR K. pneumoniae isolates from two Portuguese hospitals. * Prevalence of a specific resistance gene was statistically different when compared with the other genes from the same group (one-way ANOVA; p < 0.05). (B) Relative frequency of acquired β-lactam resistance genes by sample source. The isolates were categorized into five major sample sources: urine, rectal swab, respiratory-associated infections (including sputum, bronchial aspirate, and bronchoalveolar lavage isolates), blood culture, and others (including remaining sample sources).
Only one (1.0%) isolate did not provide any positive result for all the genes evaluated, with all the other isolates having at least one acquired β-lactam resistance gene. In total, 19/102 (18.6%) revealed the presence of one, 35/102 (34.3%) revealed the presence of two, 32/102 (31.4%) revealed the presence of three, 14/102 (13.7%) revealed the presence of four, and one (1.0%) isolate revealed the presence of five acquired β-lactam resistance genes.

Discussion
K. pneumoniae belongs to the WHO's list of antibiotic-resistant pathogens of critical priority that require the development of new antibiotics to combat them [24]. This alarming superbug is also part of the group of pathogens called ESKAPE (Enterococcus faecium, Staphylococcus aureus, Clostridium difficile, Acinetobacter baumannii, Pseudomonas aeruginosa, and all Enterobacteriaceae) normally described for their ability to "escape" from the general antimicrobial therapy [25]. Furthermore, K. pneumoniae is a well-known nosocomial pathogen that, in recent years, has been established as an MDR and pandrug-resistant problem [26]. Therefore, understanding the genetic determinants involved in AMR in this bacterial species is important from a clinical and epidemiological point of view [27].
The MDR K. pneumoniae isolates used in this study came from a wide range of samples ( Figure 1). The main sample sources were urine (39.2%), rectal swabs (15.7%), respiratoryassociated infections (15.7%; including sputum, bronchial aspirates, and bronchoalveolar aspirates), and blood (12.7%). Urine represents almost 40% of the origin of the isolates, confirming the increasing trend of MDR K. pneumoniae in urinary tract infections (UTIs) [1,28,29]. Furthermore, rectal swabs, respiratory-associated infections, and blood also represented an important source of MDR K. pneumoniae, as already reported in several other studies [1,30].
Antimicrobial susceptibility tests revealed a small variety of β-lactam treatment options for the MDR K. pneumoniae isolates from Portuguese hospitals (Figure 2). Since this study started from a pool of antibiotic-resistant clinical isolates, high-resistance rates were expected in the susceptibility tests. Therefore, the resistance percentages reported here should not be seen as the prevalence/frequency rates or compared with other studies but as molecular epidemiology data of nosocomial MDR K. pneumoniae in Portugal. Still, the high diversity of genetic determinants found in MDR K. pneumoniae isolated from two central hospitals confirms that β-lactam efficacy might be seriously threatened [31,32]. Recently, high resistance to β-lactams, including penicillins (60.5%), cephalosporins (66.9%), and carbapenems (60.5%) was also reported among a large collection of K. pneumoniae from Portuguese hospitals [33]. In other countries of the European Union/European Economic Area (EU/EEA), the AMR among K. pneumoniae isolates is also a serious problem. According to the ECDC's annual epidemiological report, the AMR percentage for third-generation cephalosporins and carbapenems was 33.9% and 10.0%, respectively, in 2020 [34]. Furthermore, the carbapenem resistance was almost always combined with resistance to several other key antimicrobial groups, leading to a severely limited range of treatment options for K. pneumoniae. The highest percentages of carbapenem resistance (>10%) were observed in south and south-eastern European countries. In Portugal, there has been an increasing annual trend of carbapenem resistance among K. pneumoniae since 2016 (5.2%% in 2016, 8.6% in 2017, 11.7% in 2018, 10.9% in 2019, and 11.6% in 2020), exceeding the overall prevalence for Europe [34]. This trend is also observed in different studies conducted in other countries, such as China and the United States, with prevalence values reaching close to 15% or 25%, respectively [35,36]. Even so, carbapenems remain the most effective β-lactam antibiotics for most species, so they should be rationally used for the treatment of severe cases involv- ing MDR bacteria, for which traditional β-lactams are not effective [37]. Alternately, the results suggest that cephamycins may be an alternative to carbapenems for the treatment of K. pneumoniae infections. Although cephamycins are very similar to cephalosporins and often included in the 2nd generation classification, they are resistant to degradation by ESBL enzymes, and it seems, to a certain extent, to CPN enzymes (especially CTT) [38]. However, they are reported not to be effective against AmpC cephalosporinases and porin mutations, and as noted, most cephamycin-resistant isolates in this work revealed AmpC genes [39].
The classification of potential ESBL-or AmpC-producing species was based on the resistance phenotype to both cefotaxime (CTA) and ceftazidime (CTZ), plus resistance to CXI in the AmpC-producing species. In this way, a phenotype resistant to CTA and CTZ and susceptible to CXI resulted in a classification of potential ESBL-producing bacteria. However, although a CTA, CTZ, and CXI-resistant phenotype corresponds to a potential AmpC-producing isolate, the possibility that it is also an ESBL producer cannot be discarded. Therefore, a phenotype resistant to CTA, CTZ, and CXI was classified as a potential AmpC-and/or ESBL-producing bacteria. Furthermore, it is important to note that for an isolate to be a potential CPN producer, it does not need to have an MER-resistant phenotype. According to the EUCAST criteria, a phenotypic result to MER <28 mm and PIT-resistant with disk diffusion test can be a potential CPN producer. For this reason, some K. pneumoniae isolates susceptible to MER (≥22 mm with disk diffusion) may be potential carriers of CPN genes. Indeed, 16 K. pneumoniae isolates presented an MER susceptible phenotype ( Figure 2) but were classified as potential CPN-producing isolates (Supplementary Material Table S1). The classification of resistance mechanisms based only on direct evaluation of susceptibility phenotype to individual antimicrobial agents can be a simple way to categorize the resistance of clinical isolates, but it can be ambiguous. Phenotype assessment could be improved by using combinatorial testing of different antimicrobial agents to detect the specific production of β-lactamase classes, for example, by using a combination disk test, double-disk synergy test, or biochemical tests as recommended by EUCAST [15]. Still, PCR screening for acquired antibiotic resistance genes appears to be a more direct way of assessing the resistance potential of clinical isolates.
In general, ESBLs and KPC carbapenemase genes were the most detected (Figure 4), confirming their dissemination in MDR K. pneumoniae isolates as verified in several other studies [33,[40][41][42][43][44][45][46]. KPC enzymes were originally identified in K. pneumoniae which justifies its high prevalence in cases of resistance to carbapenems in this species. However, this type of resistance gene has emerged in other bacterial species, such as E. coli and E. cloacae isolates [46]. In Europe, bla KPC-like (mainly KPC-2 and KPC-3) has already been detected in isolates from Austria, Belgium, the Czech Republic, Germany, Greece, Italy, Romania, and the UK, demonstrating its wide dissemination [47]. In a European survey on carbapenemase-producing Enterobacteriaceae, bla KPC (18.9%) was the major resistance mechanism reported in clinical isolates from Italy, Greece, Portugal, and Israel [48]. In other Portuguese studies, bla KPC genes were also the most detected in clinical K. pneumoniae isolates collected in the same time period of our study [33,45,49]. The highest percentage of bla KPC-like positive isolates were from urine, which confirms the high prevalence of this pathogen in UTIs [1,29]. In turn, OXA-48 is a carbapenemase that is endemic in North Africa, the Middle East, India, and some European countries such as Turkey, Spain, France, Belgium, the Czech Republic, and Malta [46,48]. However, it has also been detected in K. pneumoniae collected from Portuguese hospitals in other studies [33,45]. Recent outbreaks of OXA-48-producing K. pneumoniae in Spain, Turkey, and Greece have revealed a serious risk of these strains to human health since they present higher virulence, while laboratory detection is difficult due to their weak capacity to hydrolyze carbapenems [50][51][52]. Nonetheless, in our study, OXA-48-producing K. pneumoniae was detected in a low percentage of the K. pneumoniae isolates. Regarding MBL, no associated-resistance genes were detected in the isolates. However, the presence of bla VIM-like and bla NDM-1 genes was previously seen in K. pneumoniae clinical isolates in Portugal [53,54]. ESBL genes were the most detected in MDR K. pneumoniae isolates, demonstrating its high dissemination in Portugal, as is verified in the rest of Europe [55]. Moreover, the overall prevalence of bla SHV-like was higher than bla KPC-like . The high percentage of all ESBL types is directly related to the high percentage of resistance to β-lactams other than the carbapenems, including penicillins and cephalosporins. ESBL-producing K. pneumoniae isolates containing bla SHV-like , bla CTX-M-like, and/or bla TEM-like genes were also reported in high percentages in other clinical isolates from Portuguese hospitals, demonstrating their high spread in this bacterial species. However, bla CTX-M-like was found in a higher percentage than bla SHV-like and bla TEM-like , in accordance with epidemiology data worldwide [56]. In our study, there were more samples from around 2010 than in 2020, which is why the dominance of CTX-M-types seen in Europe may not yet be reflected in the results. In fact, the prevalence of bla CTX-M-like has been increasing considerably when compared to other ESBL types, not only in Europe but worldwide [57][58][59]. Regarding acquired AmpC β-lactamases, the low prevalence in K. pneumoniae agrees with other studies in the UK, Ireland, and Norway [60,61]. In Portugal, it was also detected among Klebsiella spp. in clinical settings, although bla DHA in a much higher percentage than bla CMY-2 [62].
The distribution by sample types was similar to all β-lactamase genes, being detected most often in urinary samples, followed by rectal swabs, respiratory-associated infections, and blood cultures. bla OXA-48-like was only detected in urinary isolates and respiratoryassociated infections, whereas acquired AmpC β-lactamases, bla DHA-like and bla CMY-2-like genes, were detected in isolates from urine, respiratory-associated infections, and pus. However, it is important to notice that the number of isolates for each sample type is very different (ranging from 40 urinary isolates to 13 blood cultures), which means that for sample types with more isolates, the distribution of the genetic determinants is certainly more accurate. An analysis within each sample type revealed that bla KPC-like is highly associated with rectal swab isolates and urine isolates ( Figure 4B). This high association of KPC carbapenemases in rectal swabs has already been described in several studies [63][64][65] since K. pneumoniae is an intestinal pathogen subject to a huge selective pressure by antibiotic use. On the other hand, in samples from respiratory-associated infections, blood cultures, and other samples, the frequency of ESBL genes seemed higher compared to carbapenemases genes, in most cases higher than any carbapenemase gene. For example, bla SHV-like had an almost total frequency (0.94/1), higher than bla KPC-like (0.56/1), as well as bla TEM-like (0.63/1) for isolates from respiratory-associated infections. Still, for all sample types, the frequency of ESBL genes was highest for bla SHV-like , followed by bla TEM-like , and finally, bla CTX-M-like .
Regarding the MDR K. pneumoniae isolate with no resistance genes detected, given the enormous diversity of resistance elements, this isolate might simply carry a less common resistance gene different from those tested or have another resistance mechanism (e.g., efflux pumps or porins modifications) responsible for the phenotypic pattern [66]. All other isolates tested positive for one or more resistance genes corresponding to at least one of the predicted resistance phenotypes (Table S1). Still, not all phenotypes predicted for each isolate have been confirmed by the presence of resistance genes. As mentioned above, resistance to different β-lactams can result from the production of enzymes that directly hydrolyze these or from the combination of enzymes, such as AmpC and/or ESBL, with other cellular mechanisms, such as decreased membrane permeability to antibiotics [67]. It also needs to be considered that the development of resistance depends on the level of gene expression, and thus, it cannot be excluded that resistance genes cannot express in-vitro (susceptibility tests) but show expression in-vivo and/or truncated β-lactamase genes may be detected by molecular methods but not be expressed [68]. Lastly, resistance phenotypes may result from the production of β-lactam resistance genes other than those tested in this study.
In addition, approximately 80% of MDR K. pneumoniae isolates were found to have two or more tested resistance genes (Table S1). This evidence may be associated with the high capacity for clonal expansion and exchange of mobile genetic elements (e.g., β-lactam resistance genes) of this species, which promotes increased resistance to antibiotics [67].
That said, the potential for resistance of this important nosocomial pathogen is alarming, and this work highlights the urge to introduce in the marketplace of new alternative therapies to fight infections due to K. pneumoniae.

Bacterial Isolates
One hundred and two K. pneumoniae isolates with an MDR profile were provided by the microbiology laboratories of the Hospital de Braga and of the Centro Hospitalar De Trás Os-Montes E Alto Douro, E.P.E., both in northern Portugal. Isolates were collected over a period of ten years (2010-2020). This study was performed in line with the principles of the Declaration of Helsinki and the guidelines of Good Clinical Practice. Approval was granted by the Ethics Committee of Life and Health Sciences Research Institute (ICVS), School of Health Sciences, University of Minho, Braga, Portugal. The isolates provided were previously identified using VITEK ® 2 Compact B System (bioMérieux, Marcy-l'Étoile, France). Even so, the bacterial isolates were cultured on MacConkey agar (BioLife, Milan, Italy) plates, and typical K. pneumoniae colonies were sub-cultured onto trypticase soy agar (TSA, Liofilchem, Roseto degli Abruzzi, Italy) for further analysis. Isolates were also preserved in 20% (v/v) glycerol (Riedel-de-Haën, Seelze, Germany) in tryptic soy broth (TSB, Liofilchem) and stored at −80 • C. The sample sources and isolation dates from all isolates were gathered together with the bacterial species information.

Antimicrobial Susceptibility Testing
Antibiograms were obtained by a VITEK 2 compact system available at the hospital laboratories. The antibiotic susceptibility profiles were then confirmed and complemented to meet the recommended list of antimicrobial agents proposed by the European Center for Disease Prevention and Control (ECDC) and the Centers for Disease Control and Prevention (CDC) [67], using the reference disk diffusion method as described by the EUCAST [68]. The list of β-lactam antibiotics tested is shown in Table 1. Quality control (QC) was performed using Escherichia coli ATCC 25922 and EUCAST breakpoint tables, version 12.0, were used in the interpretation of the susceptibility results [69]. Based on the antimicrobial susceptibility profiles, the K. pneumoniae isolates were classified as potential producers of CPN, ESBL, and/or AmpC according to the EUCAST guidelines for the detection of resistance mechanisms of clinical and/or public health importance [15]. In summary, the isolates were considered suspected of CPN-producing Enterobacteriaceae if the result of susceptibility to meropenem with disk diffusion was <28 mm. To be considered a potential ESBL-producing Enterobacteriaceae, the isolates required an intermediate or resistance result to one or both of cefotaxime (CTA) and ceftazidime (CTZ), while AmpC β-lactamase-producing Enterobacteriaceae required a resistant profile for CTA or CTZ, and cefoxitin (CXI).

Detection of Acquired β-Lactam Resistance Genes by Multiplex PCR
Four multiplex PCR assays for the detection of the most common β-lactam resistance genes of CPN, ESBL, or AmpC were designed using primers described in the literature. The different multiplexes were optimized using positive control strains containing the resistance genes of interest (Table 2). Subsequently, the K. pneumoniae isolates were screened for the presence of the acquired β-lactam resistance genes. In brief, a loop (1 µL) of each bacterial culture was homogenized in 500 µL of ultra-pure water and then boiled for 15 min. The suspension was centrifuged (12,000× g) for 5 min, and the supernatant (2 µL) was used as a DNA template. PCR was carried out in a 20 µL reaction mixture containing 10× reaction buffer (NzyTech, Lisbon, Portugal), 1.5 mM MgCl 2 , 0.25 mM of dNTPs (dATP, dTTP, dGTP, and dCTP), five pmol each of forward and reverse primers and 1.25 U of Supreme NZYTaq II DNA polymerase (NzyTech) in an MJ Mini Personal Thermal Cycler (Bio-Rad, CA, USA). The PCR conditions included initial denaturation at 95 • C for 5 min, followed by 30 cycles of denaturation at 94 • C for 30 s, annealing for 30 s at an optimized temperature, and extension at 72 • C for 30 s. A final extension was set at 72 • C for 5 min.
The PCR products were mixed with 6x NZYDNA loading dye (NzyTech), subjected to electrophoresis at 100 V for one h in a 1.5 % (w/v) agarose gel, previously stained with GreenSafe Premium (NzyTech), and finally analyzed under UV light.

Statistical Analysis
All data were analyzed with GraphPad Prism version 7 (La Jolla, CA, USA) using one-way ANOVA with Holm-Sidak's multiple comparisons test. Values with a p < 0.05 were considered statistically significant.

Conclusions
In summary, this work applied four multiplex PCRs to rapidly detect the presence of the most common acquired resistance genes of carbapenemases, ESBL, and AmpC β-lactamases. As a fast, reliable, and low-cost methodology, this can be a valuable tool in epidemiological studies as confirmation of standard phenotypic characterization methodologies. Overall, it was observed a diverse pool of β-lactam resistance genetic determinants disseminated among K. pneumoniae isolates (bla KPC-like , bla OXA-48-like , bla TEM-like , bla CTX-M-like , bla SHV-like , bla CMY-2-like and bla DHA-like ), revealing a diverse threat to the effectiveness of the β-lactam class. In addition, this long-term study confirmed the high spread of multiple β-lactam resistance genes among K. pneumoniae isolates from hospital settings in Portugal, highlighting the need for mitigation measures to prevent the advancement of this problem.