Klebsiella pneumoniae is an opportunistic pathogen associated with a variety of infections, being urinary tract infections (UTIs) among the most common worldwide. K. pneumoniae is considered an important uropathogen in ambulatory patients (Foxman, 2010). Although UTIs do not result in a high mortality rate, they represent a significant economic burden to the health care system as they increase treatment costs.

However, K. pneumoniae is not only responsible for UTIs but also for respiratory and bloodstream infections (Podschun and Ullmann, 1998). It is one of the species recognized as part of the ESKAPE group, associated by their characteristic potential to escape or evade the action of antimicrobial agents. This acronym comprises six highly virulent and antibiotic-resistant bacterial pathogens including Enterococcus faecium, Staphylococcus aureus, K. pneumoniae, Acinetobacter baumannii, Pseudomonas aeruginosa, and Enterobacter species. Additionally, the World Health Organization (WHO) lists K. pneumoniae as one of the species of high priority and promotes the research and development of new antibiotics due to the growing global problem of antimicrobial resistance (World Health Organization, 2017).

Uropathogenic bacteria have some characteristics that favor the colonization of human cells, such as the production of adhesins, siderophores, and toxins (Foxman, 2010). In the same way, these bacteria can adhere to medical devices forming biofilms, avoiding the immune system and favoring the antimicrobial therapy failure (Dybowska-Sarapuk et al., 2017). For this reason, the characterization of K. pneumoniae antimicrobial resistance, its ability to form biofilm and its virulence are key to understanding the pathogenicity of this bacteria in the clinical setting and to achieve more efficient antimicrobial therapy.

The aim of this study was to characterize a collection of K. pneumoniae strains isolated from different clinical sources (urine, respiratory tract, or blood) in terms of antimicrobial resistance, the presence of virulence factor genes, serum resistance, hypermucoviscosity, and the ability to form biofilms. The specific characteristics of the uropathogenic strains were studied and compared to K. pneumoniae strains isolated from respiratory and/or blood samples.

K. pneumoniae causes a wide range of infections and uses different virulence factors to colonize and spread in the human body. The increasing antimicrobial resistance of this bacteria in recent years is of great concern to the scientific community.

In this study, we characterized a collection of 127 clinical K. pneumoniae strains isolated from urine, the respiratory tract, and blood and compared these strains according to antimicrobial resistance and the presence of a wide variety of virulence factors.

Overall, we found that more than 40% of isolates were classified as MDR and almost half of these were from urine. The data from the 2017 European Antimicrobial Resistance Surveillance Network report showed that at the European Union and the European economic area level, more than one third of K. pneumoniae isolates were resistant to at least one of the antimicrobial groups under regular surveillance. The report showed resistance percentages for fluoroquinolones (31.5%), followed by third-generation cephalosporins (31.2%), aminoglycosides (24.1%), and carbapenems (7.2%) (European Centre for Disease Prevention and Control (ECDC), 2018) similar to the general percentages observed in this study.

Urine isolates showed the highest overall antibiotic resistance (64.91%), closely followed by strains isolated from blood (63.64%) and the respiratory tract (51.35%). Uropathogenic isolates made up the majority of the ESBL-producing strains. Compared to the multicenter study carried out for Garcia-Fernandez et al., our strains showed a higher prevalence of ESBL and lower carbapenemase production. This previous study reported a prevalence of 23.1% of ESBL-producing Klebsiella spp. and 20% of carbapenemase-producing Klebsiella (García-Fernández et al., 2019). This is of great concern because some antibiotic resistance genes are carried by mobile genetic elements that facilitate horizontal genetic exchange and promote the spread of antimicrobial resistance genes within and between species. Their characterization in future studies could help us to expand our research. On the other hand, multidrug resistance has been associated with increased health care costs, longer hospital stays, and high mortality rates. Moreover, nowadays there are few new alternatives for effective treatments.

Urine strains also showed the highest resistance rates to fosfomycin (24.56%). Fosfomycin trometamol is one of the antibiotics proposed for treating uncomplicated UTIs, and its consumption has increased by 20.7% since 2012 in Spain, according to the Joint Inter-agency Antimicrobial Consumption and Resistance Analysis report (Alonso Herreras et al., 2018). Sorlozano et al. showed that ESBL-producing Klebsiella had low susceptibility rates to fosfomycin ranging between 40 and 78% (Sorlozano et al., 2014). Therefore, it is important to consider this antibiotic in the persistent monitoring and surveillance of antimicrobial resistance.

Regarding virulence, hypermucoviscous pathotypes, genes for adhesins, iron uptake and transport, toxins, and biofilm formation ability are some of the factors that K. pneumoniae employs in pathogenesis. Although many studies define hypervirulent K. pneumoniae (hvKp) as string test positive, it has been demonstrated that not all hvKp strains are hypermucoviscous (Fu et al., 2018). The study carried out by Russo et al. proposed the rmpA and iucA genes as biomarkers for distinguishing between hypervirulent (hvKp) and the classical (cKp) K. pneumoniae pathotypes (Russo et al., 2011). rmpA is a plasmid-located virulence factor gene that regulates the synthesis of capsular polysaccharides  (Wang et al., 2020), and iucA is involved in the biosynthesis of the siderophore aerobactin (De Lorenzo and Neilands, 1986). These genes were significantly associated with the hvKp-phenotype in our study, in which seven of the 17 (41.18%) hypermucoviscous strains carried both genes. On the other hand, two more strains were found, icuA+, rmpA+ and HMV-, five were classified as icuA+, rmpA- and HMV- and ten were classified as icuA-, rmpA- and HMV+. This suggests that other regulatory mechanisms may be involved in the expression of the hypermucoviscous phenotype. Thus, considering both the string test and the hypervirulence-related genes iucA and rmpA, 24 strains (18.9%) were classified as hypervirulent in the present work. In terms of antimicrobial resistance, hvKp have been classified as susceptible or as ESBL-producer strains (Harada and Doi, 2018). We found 13 strains classified as susceptible and nine as ESBL-producers, similar to the studies previously described. Of the other two strains with an icuA-, rmpA- and HMV+ phenotype, one was classified as resistant to fosfomycin and the other as resistant to more than three categories of antibiotics.

Other hypervirulence-associated genes, such as iroBCDN, iutA, iuacABCD, ybt and the clb locus, may be encoded in the chromosome as part of integrative conjugative elements (Arato et al., 2021). The hypervirulent phenotype confers important properties to bacteria for pathogenesis, such as capsule polysaccharide production, phagocytosis resistance, dissemination and systemic infection in the host (Xu et al., 2021). Early identification of hvKp strains would allow better understanding of the infection and the implementation of more effective antimicrobial treatment. Likewise, it is mandatory to avoid the possible dissemination of these pathotypes by implementing barriers between patients, with precautions based on virulence likely being as important as those based on antimicrobial resistance.

Among siderophores, enterobactin (encoded by ent genes), aerobactin (encoded by iuc genes), salmochelin (encoded by iro genes), and yersiniabactin (encoded by ybt genes, regulated by irp genes, and the receptor is encoded by the fyuA gene) are the most representative in K. pneumoniae. Siderophores are indispensable to obtain the iron necessary for the growth and replication of bacteria. They scavenge iron from host transport proteins. Enterobactin is produced by all K. pneumoniae, but its mechanism may be disabled by human lipocalin-2 (Lcn2), which has an affinity for enterobactin and elicits an inflammatory response. In contrast, yersiniabactin eludes binding to Lcn2, thereby avoiding the inflammatory response and promoting bacterial growth (Lam et al., 2018). As expected, we found the entB gene in all the strains tested in agreement with another study (Fatima et al., 2021). Besides, the higher prevalence of the siderophore-encoded ybtS and fyuA genes in the respiratory tract isolates compared to the urinary and blood strains confirmed that yersiniabactin is an important virulence factor for the pulmonary infection, as reported by previous studies (Yan et al., 2016).

The capsule-associated genes wabG, (encoding capsule), uge (encoding capsule lipoprotein) and ycfM (promoting external membrane protein) are highly involved in virulence by promoting infection through resistance to phagocytosis. We observed the presence of the uge (uridine diphosphate galacturonate-4 epimerase) gene in the majority of the isolates from urine, showing a statistically significant difference compared to the other clinical isolates. It has been demonstrated that K. pneumoniae strains lacking the uge gene are less virulent and less capable of causing UTI, pneumonia or sepsis. Mutations in the uge gene reduce the colonization ability of K. pneumoniae in experimentally induced urinary infections (Regué et al., 2004; Candan and Aksöz, 2015; Remya et al., 2019). However, the occurrence of the uge gene in K. pneumoniae has varied widely in different studies. Remya et al. found a prevalence of 48.6% harboring the uge gene. Its distribution in different clinical specimens was blood (50%), urine (48.8%), exudates (48.4%), and respiratory secretions (47.3%). El Fertas-Aissani et al. observed the presence of the uge gene in 84.6% of urinary tract strains and all the blood and respiratory isolates (El Fertas-Aissani et al., 2013). Although the uge gene has previously been described as essential for K. pneumoniae virulence, future research may provide further evidence for its specific association with UTIs.

Among the toxins tested, two of the strains were positive for the clbA and clbQ genes, two of the genes encoding colibactin. Colibactin is a recently described toxin that has been associated with colorectal cancer in humans because it causes DNA damage in host cells. Colibactin has been found in E. coli, but its occurrence in other species is steadily increasing. Acquisition of the pks locus, which synthesizes this toxin, has been linked to intestinal colonization and mucosal invasion by K. pneumoniae (Strakova et al., 2021).

Interestingly, one of the strains isolated from the respiratory tract was positive for all the virulence genes tested. Two other strains were positive for almost all the virulence genes except those encoding colibactin. These three strains were susceptible to all antimicrobial agents, and were classified as serotype K1, non-biofilm-forming, or weak biofilm-forming strains and showed high serum resistance, which protects the bacteria from phagocytosis (Hennequin and Robin, 2016). Although more research is needed on the influence of the acquisition of antibiotic resistance mechanisms by the bacteria on fitness cost and the relation with virulence, we found a scenario in which the greater the virulence, the lesser the antimicrobial resistance. Indeed, we found a correlation between the presence of some virulence factor genes and susceptibility to most of the antibiotics tested.

The genes encoding ESBL are carried by plasmids that may also harbor other genes, including some virulence factor genes. In previous studies, three virulence factors (adhesin CF29K, aerobactin or mucoid phenotype) were found in ESBL-producing K. pneumoniae strains. (Hennequin and Robin, 2016). In our study, we found the same association between fyuA, irp2 or ybtS genes and ESBL-producing strains. The ybt carried by the FIB_K plasmid is a novel mechanism for ybt mobilization, Currently, many FIB_K plasmids also contain antimicrobial resistance transposons, which may indicate the coexistence of virulence and resistance genes in the same mobile genetic element (Lam et al., 2018).

Regarding biofilm, almost half of biofilm-forming strains, including weak, moderate and strong, were isolated from urine and the difference from the other clinical sources was statistically significant. K. pneumoniae has been identified as the second most prevalent pathogen in UTIs (Liu et al., 2020), and the ability to form biofilms on indwelling urethral catheters is one of the most common conditions, affecting thousands of people worldwide each year. Recurrence of catheter urinary tract infections after antibiotic treatment could be caused by re-colonization by bacteria that survived in the biofilm (Stahlhut et al., 2012). Besides, biofilms present higher resistance to phagocytosis and higher antimicrobial resistance compared to planktonic bacteria (Trautner and Darouiche, 2004). Nonetheless, truly effective strategies to prevent or counteract biofilm-related infections remain elusive.

We found that isolates from urine were more resistant to antimicrobials and were more ESBL-producers, and more biofilm-formers compared to respiratory and blood strains. Our study also correlated the presence of the uge gene with UTIs and the presence of yersiniabactin with respiratory tract infections. In addition, although we found an inverse relation between antimicrobial resistance and virulence, the acquisition of mobile genetic elements could promote not only the spread of antimicrobial resistance genes but also that of virulence genes that evolve toward pathotypes considered to be more virulent. The increasing coexistence of these two conditions is of particular concern as it can lead to untreatable and invasive K. pneumoniae infections. Active surveillance not only for antimicrobial resistance but also for virulence determinants is imperative to avoid the transmission and spread of hypervirulent or extensively resistant strains.

The original contributions presented in the study are included in the article/Supplementary Material. Further inquiries can be directed to the corresponding author.

Conceptualization, SS and VB. Methodology, VB, YG, MT, RO, and CR. Statistical analysis, VB. Writing-original draft preparation, VB and SS. Writing—review and editing, VB, CR, YG, and SS. All authors contributed to the article and approved the submitted version.

This work was funded by Planes Nacionales de I+D+iI2008‐2011/2013‐2016 and Instituto de Salud Carlos III (PI19/00478), Subdirección General de Redes y Centros de Investigación Cooperativa, Ministerio de Economía y Competitividad, Spanish Network for Research in Infectious Diseases (REIPI RD12/0015/0013 and REIPI RD16/0016/0010) co‐financed by European Development Regional Fund “A way to achieve Europe” and operative program Intelligent Growth 2014‐2020. ISGlobal is a CERCA center from the Generalitat of Catalunya and a Severo Ochoa Center (Spanish Ministry of Science, Innovations, and Universities). VB has a grant from Ministerio de Ciencia, Tecnología e Innovación (Colombia).

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.