Antimicrobial resistance of Klebsiella pneumoniae strains isolated from COVID-19 patients: genetic analysis and phenotypic characterization

Full Text

Introduction

Klebsiella pneumoniae is one of the most significant opportunistic pathogens causing a wide range of infections in hospitalized patients, including pneumonia, sepsis, and urinary tract infections [1]. This pathogen is characterized by a high capacity to accumulate antimicrobial resistance genes, making it one of the key threats to the healthcare system. Particularly alarming is the fact that K. pneumoniae often acts as a donor of new antimicrobial resistance (AMR) genes, such as blaKPC, blaOXA-48, and blaNDM-1, which were first discovered in this species before spreading to other members of the family Enterobacteriaceae [2].

Traditionally, two main pathotypes of K. pneumoniae are distinguished: classical strains with high levels of antibiotic resistance and hypervirulent (hvKp) strains associated with severe community-acquired infections. Previously, it was believed that these pathotypes had different genetic bases and rarely overlapped, but over the past decade, strains combining multiple drug resistance and increased virulence have emerged. Such strains pose a serious clinical problem, as they combine treatment difficulties with a high ability to cause disseminated infections and are associated with increased mortality [1].

The ability of K. pneumoniae to actively transfer genetic elements horizontally, including plasmids carrying resistance and virulence genes, plays a key role in the formation of new dangerous strains [1, 2]. This highlights the need for comprehensive analysis — both phenotypic and genetic — of strains isolated from COVID-19 patients to assess the level of resistance, identify the mechanisms of its formation, and detect potentially dangerous clones.

The strategic success of such bacteria is determined not only by their resistance to antimicrobials, but also by their ability to effectively redistribute metabolic resources — a phenomenon known as bacterial fitness. This process involves a compromise between maintaining resistance mechanisms and expressing virulence traits, which directly affects the survival of strains in challenging conditions, including antimicrobial therapy and immune system pressure. In recent years, K. pneumoniae has been a leading nosocomial pathogen, causing severe infections with high mortality, especially in sepsis and pneumonia. Its role as a carrier of resistance genes from environmental reservoirs to clinically significant pathogens further increases its danger [4, 5].

The aim of the study is to perform genetic analysis and phenotypic characterization of K. pneumoniae strains isolated from patients with COVID-19 in order to assess their antibiotic resistance profile, identify key resistance genes and analyze possible routes of their spread in the context of the pandemic.

Materials and methods

A total of 102 patients with COVID-19 aged 25–95 years with mild, moderate, or severe disease were examined. The study was conducted with the voluntary informed consent of the patients. The study protocol was approved by the Ethics Committee of Kazan State Medical University (protocol No. 10 dated December 20, 2022).

The study material consisted of nasopharyngeal swabs taken in the emergency room upon admission to large hospitals in Kazan. To isolate bacteria, we used Columbia, chocolate and egg yolk-salt agar cultures, followed by identification of the cultures using the MALDI-ToF mass spectrometry method (Bruker MALDI-biotyper). The study of phenotypic resistance to antimicrobials (amikacin, amoxicillin/clavulanic acid, ampicillin, meropenem, cefepime, cefotaxime, ceftazidime, ciprofloxacin, ertapenem) was performed using the disc diffusion method, in accordance with EUCAST v. 13.0 standards and Russian recommendations "Determination of the susceptibility of microorganisms to antimicrobials. Version 2024-02" [8]. The results were interpreted using the Adagio bacteriological analyzer.

K. pneumoniae strains were isolated from the nasopharynx of 17 patients, with a contamination level of 10³–10⁵ CFU/mL. Among these patients, whose median age was 70 years, 59% had moderate COVID-19, 41% had moderate severity.

Next, an in-depth study was conducted on nine strains of K. pneumoniae with multiple drug resistance. Whole-genome sequencing of nine K. pneumoniae isolates was performed on the GenoLab M platform (Genoanalytics). Libraries were prepared using the SG GM Plus whole genome sequencing kit (Sesana). The procedure included the following steps: enzymatic fragmentation of genomic DNA, adapter ligation, purification of preparations using magnetic particles, and index PCR amplification for multiplexing samples. The resulting libraries were sequenced with paired ends with a read length of 150 bp.

Bioinformatics analysis included de novo genome assembly using SPAdes v. 3.13.0 and Unicycler v. 0.5.0 software. Gene annotation was performed using the NCBI Prokaryotic Genome Annotation Pipeline tool. Antimicrobial resistance and virulence genes were identified using ResFinder (http://genepi.food.dtu.dk/resfinder) and Kleborate v. 2.3.2. Multilocus typing and K- and O-type determination were performed using the Kleborate tool. Annotated genomic sequences are deposited in the GenBank database (SCPM-O-B-14103, SCPM-O-B-14104, SCPM-O-B-14105, SCPM-O-B-14106, SCPM-O-B-14107, SCPM-O-B-14108, SCPM-O-B-14109, SCPM-O-B-14110, SCPM-O-B-14111). In this study, the virulence score (VS) was used to quantitatively assess the virulence potential of K. pneumoniae strains. This parameter is implemented in the Kleborate v. 2.4.0 software package. VS is a summary index calculated based on the presence and combination of known virulence genes associated with hypervirulent K. pneumoniae lines. The Kleborate program automatically annotates genomic sequences and assigns each strain a numerical VS according to a predefined gene weighting scheme based on literature data and epidemiological observations. The VS used in the study is a virulence index specific to the Kleborate program and reflects the relative virulence potential of the analyzed isolates.

Statistical analysis was performed using SPSS Statistics (IBM). Quantitative data with indicators different from normal distribution are presented as medians (Me). To assess the statistical significance of differences between groups, we used the Kruskal–Wallis test with Bonferroni correction for multiple comparisons, as well as the Mann–Whitney test to compare two groups of indicators. Differences were considered significant at p ≤ 0.05. Correlation analysis was used to identify the relationship between quantitative indicators, analyze its direction, strength (closeness) and statistical significance. The normality of the distribution was checked using the Shapiro–Wilk test (p ≤ 0.05).

Results

Study of genetic determinants of resistance to antimicrobials in K. pneumoniae

Analysis of clinical isolates of K. pneumoniae from COVID-19 patients revealed an extensive arsenal of genetic determinants associated with resistance to major classes of antimicrobials.

Resistance to the aminopenicillin group in the population of isolated strains was mainly determined by the production of various β-lactamases. The dominant genetic determinants of extended-spectrum β-lactamases (ESBLs) in K. pneumoniae strains were lactamases encoded by the blaCTX-M-15 (66.6%) and blaOXA-1 (66.6%) genes, as well as blaCTX-M-3 (11.1%). The population also included genetic determinants encoding serine β-lactamases blaTEM-1 (44.4%), blaSHV-1 (33.3%), blaSHV-11 (66.9%), as well as carbapenemases blaOXA-48 (22.2%), blaNDM-1 (44.4%), and plasmid AmpC-β-lactamases blaLAP-2 (11.1%). All strains were found to have genes responsible for the modification of OmpK37 porin channels and disruption of the global regulators marA, H-NS and CRP. The detection of a combination of BLRS genes (blaCTX-M-15, blaCTX-M-3) and non-enzymatic mechanisms, such as OmpK37 porin inactivation, in the studied strains may provide resistance to cephalosporins, including cefepime. A critical finding is the discovery of a combination of carbapenemase and BLRS genes among the isolates. Thus, two strains of K. pneumoniae were found to carry OXA-48-type carbapenemase (blaOXA-48) together with blaOXA-1 and blaCTX-M-15. A distinctive feature of all isolates studied was the presence of genetic determinants providing complex mechanisms of resistance to β-lactams, including β-lactamase production, reduced membrane permeability and activation of efflux systems.

Among the isolated strains, a high level of resistance to aminoglycosides was detected, mediated by a variety of modifier enzymes. Genes encoding fluoroquinolone-acetylating aminoglycoside-(6)-N-acetyltransferases (AAC(6')-Ib10; 66.6%), aminoglycoside kinases (APH(3')-Via; 44.4%), and aminoglycoside-3"-adenylyltransferase (aadA; 22.2%) have been identified. In three strains of K. pneumoniae, the armA gene encoding 16S rRNA methyltransferase was identified, which may cause resistance to virtually all aminoglycosides.

Mutations in genes encoding the main targets of fluoroquinolones, DNA gyrase (gyrA) and topoisomerase IV (parC), play a key role in resistance to fluoroquinolones. In 66.6% of K. pneumoniae strains, characteristic amino acid substitutions were identified in the regions determining resistance to quinolones: Ser83→Ile in GyrA and Ser80→Ile in ParC. Variants of the QnrB1 and QnrS1 genes of the qnrB and qnrS families were also found in the isolated cultures. The oqxA and oqxB genes, which regulate active efflux systems such as oqxAB, were identified in the genome of all isolates, which may indicate additional mechanisms of fluoroquinolone resistance.

The active transport genes KpnG and KpnH were identified in all studied K. pneumoniae strains, which may indicate their potential role as global regulators of resistance. The detection of the regulatory genes marA and H-NS may highlight their key role in the coordinated activation of multiple resistance systems in circulating strains.

The analysis showed that the predominant mechanism of resistance in the studied K. pneumoniae strains is antimicrobial inactivation (42.4%), which emphasizes the importance of enzymatic inactivation as the main mechanism of resistance. A significant proportion of genetic determinants are related to efflux pumps (36.3%), which may also indicate their importance in ensuring resistance to a wide range of antimicrobials. A smaller proportion of genes are associated with antimicrobial inactivation (9.1%), protection of the antibiotic target, reduction of its availability to DNA gyrase and topoisomerase, and protection of the ribosome through ATP-dependent allosteric displacement (6.1%), and reduction of cell permeability (6.1%), however, their combined presence indicates the complex nature of antimicrobial resistance in K. pneumoniae. The data obtained reflect the high potential of circulating strains to develop multiple drug resistance.

Phenotypic resistance of K. pneumoniae strains

Phenotypic resistance to antimicrobials used, except for ampicillin, in the conditions of the infectious hospital in Kazan, for these strains is presented in Table 1. Testing for resistance to ampicillin was performed to control for the natural resistance of K. pneumoniae.

 

Table 1. Characteristics of susceptibility to antimicrobial strains of K. pneumoniae

Strain	Phenotypic expression of resistance to antimicrobials
	amikacin	amoxicillin/clavulanic acid	ampicillin	meropenem	cefepime	cefotaxime	ceftazidime	ciprofloxacin	ertapenem
SCPM-O-B-14110	R	R	R	–	R	R	R	R	R
SCPM-O-B-14111	R	R	R	–	R	R	R	R	R
SCPM-O-B-14104	S	R	R	–	R	R	R	S	S
SCPM-O-B-14107	R	R	–	R	R	R	R	R	–
SCPM-O-B-14103	R	R	–	R	R	R	R	R	–
SCPM-O-B-14105	S	S	R	S	S	S	S	R	–
SCPM-O-B-14109	S	S	R	S	S	S	S	R
SCPM-O-B-14106	R	R	R	–	R	R	R	R	R
SCPM-O-B-14108	R	R	R	–	R	R	R	R	R

Note. R — resistant; S — susceptible; «–» — unidentified.

 

The results of the analysis showed that 6 (66.7%) strains were resistant to amikacin, 7 (77.8%) to amoxicillin/clavulanic acid, ampicillin — 7 (100%) strains out of 7 tested, meropenem — 2 (50%) strains out of 4 tested, cefepime — 7 (77.8%), cefotaxime — 7 (77.8%), ceftazidime — 7 (77.8%), ciprofloxacin — 8 (88.9%), ertapenem — 4 (80%) out of 5 tested. Phenotypically, all strains showed resistance to at least 2 antimicrobials, with 66.7% of the isolates studied being resistant to all tested antimicrobials.

The phenotypic manifestation of resistance to antimicrobial generally reflects genetic determination, but the presence of resistance genes does not always clearly predict its phenotypic manifestation. The BSCPM-O-B-14104 strain was found to be phenotypically susceptible to ertapenem, but this strain has a set of genes capable of neutralizing the effect of this antimicrobial on bacterial cells, namely: blaCTX-M-3, H-NS, KpnE, KpnF, KpnG, KpnH, Ompk37, marA, MdtQ, blaSHV-1.

Virulence factors and resistance of K. pneumoniae strains

Nine K. pneumoniae isolates were analyzed to determine K and O types, as well as virulence (Table 2).

 

Table 2. Description of K- and O-types, final virulence score of isolated strains

Strain	wzi	K-locus	Identity, %	О-locus	О-type	Identity, %	VS
1-SCPM-O-B-14106	wzi77	KL57	98,76	O1/O2v2	O2afg	98,63	1 (ybt)
SCPM-O-B-14103	wzi108	KL137	95,96	O1/O2v1	O2a	98,76	4 (ybt,iuc)
SCPM-O-B-14104	wzi77	KL57	98,76	O3/O3a	O3/O3a	98,38	1 (ybt)
SCPM-O-B-14107	wzi1	KL1	100,00	O1/O2v2	O1	99,11	1 (ybt)
SCPM-O-B-14105	wzi1	KL1	100,00	O1/O2v2	O1	99,16	5 (clb, iuc, iro, ybt)
SCPM-O-B-14108	wzi64	KL107	90,86	O1/O2v1	O2a	98,57	4 (iuc, ybt, rmpA2)
SCPM-O-B-14109	wzi64	KL64	99,99	O1/O2v1	O2a	98,62	5 (clb, iuc, iro, ybt)
SCPM-O-B-14111	wzi64	KL64	99,99	O1/O2v1	O2a	98,62	1 (ybt)
SCPM-O-B-14110	wzi64	KL107	90,23	O1/O2v1	O2a	98,58	4 (iuc, ybt)

 

The most common were wzi64 (n = 4) and wzi1 (n = 2). Strains SCPM-O-B-14109 and SCPM-O-B-14111 had identical wzi64 and KL64 (99.99% identity), but differed in their virulence profile: the SCPM-O-B-14109 strain had a complete set of genes (clb, iuc, iro, ybt, VS 5), while the SCPM-O-B-14111 strain had only ybt (VS 1)

Strains SCPM-O-B-14107 and SCPM-O-B-14105 belonged to the wzi1/KL1 type, which is associated with hypervirulence. With 100% identity at the K locus, their virulence profiles differed: only ybt (VS 1) was detected in strain SCPM-O-B-14107, while all 4 key operons (VS 5) were detected in SCPM-O-B-14105.

O-type O2a was the most common (n = 5), predominantly in strains with wzi64 or wzi108. Strains with O1/O2v2 had more diverse K-types (KL57, KL1).

High virulence VS (≥ 4) was observed in 5 strains, including two with a maximum value of 5 (SCPM-O-B-14105 and SCPM-O-B-14109), which is characteristic of hypervirulent isolates (hvKP). The remaining 4 strains had low virulence potential (VS 1), which is characteristic of less virulent variants.

The total number of identified genetic determinants encoding virulence factors varied depending on the strain, with the maximum value observed in strain SCPM-O-B-14109 (151 genes) and the minimum in strain SCPM-O-B-14107 (123 genes). With regard to adhesion, most strains had a comparable number of genes determining adhesion factors (22–23 genes). The number of genes encoding anti-phagocytic factors ranged from 3 to 6, with strains SCPM-O-B-14107 and SCPM-O-B-14111 showing the highest number (6 genes). All strains had the same number of genes associated with efflux systems (2). More pronounced differences were observed in virulence genes associated with immune evasion, iron acquisition, and iron uptake. The SCPM-O-B-14109 strain showed the highest number of genetic determinants for immune evasion (15), while the SCPM-O-B-14105 and SCPM-O-B-14108 strains showed the lowest (1 each). Strains SCPM-O-B-14105 and SCPM-O-B-14109 had the highest number of genes associated with iron acquisition (37), while in the other strains their number ranged from 18 to 25. Nine genes encoding the iron uptake factor were identified in most strains, with the exception of two strains, SCPM-O-B-14105 and SCPM-O-B-14109, in which this virulence factor was encoded by one gene. As for secretion systems, most strains showed the presence of genes in the range of 33–35, with the exception of strain SCPM-O-B-14105, which had significantly fewer (17). The number of genes encoding serum resistance in all strains ranged from 3 to 10, and genes encoding trophic factors and toxin synthesis were present only in strains SCPM-O-B-14105 and B-SCPM-O-B-14109 and were absent in the other strains. The number of genes belonging to the “Unknown class” ranged from 14 to 30, with the highest value observed in strain SCPM-O-B-14105 (30). Thus, strains SCPM-O-B-14109 and SCPM-O-B-14105 stand out from the rest in terms of a number of virulence characteristics. The SCPM-O-B-14109 strain is characterized by the highest total number of genes encoding virulence factors and their high variability in the “Immune Evasion” and “Iron Acquisition” classes. The SCPM-O-B-14105 strain is also characterized by a wide range in the “Iron Acquisition” and “Unknown Class” classes. The similarity of genetic virulence determinants in most strains may indicate their common origin or horizontal gene transfer. The significant differences observed in the classes “Immunity evasion,” “Iron acquisition,” “Iron uptake,” and “Unknown class” may reflect the adaptation of strains to different ecological niches. At the same time, the presence of a large number of “Unknown class” genes in some strains indicates the necessity for further research to identify and characterize their function.

A comparative analysis of the distribution of genetic determinants of virulence and determinants of resistance to antimicrobials revealed statistically significant differences (p < 0.001) (Table 3).

 

Table 3. Comparison of genetic determinants of virulence by groups and resistance to antimicrobials

Factor group	Number of genetic determinants		p
	Me	Q1; Q3
1. Adhesion	23	23; 23	p11–12 = 0.04; p11–1 < 0.001 p11–5 < 0.001; p11–8 < 0.001 p11–13 < 0.001; p10–12 = 0.02 p10–1 = 0.01; p10–5 = 0.01 p10–8 < 0.01; p10–13 < 0.01 p3–1 = 0.005; p3–5 = 0.004 p3–8 < 0.001; p3–13 < 0.001 p4–8 = 0.014; p4–13 = 0.013 p2–8 = 0.049; p2–13 = 0.046
2. Anti-phagocytosis	4	4; 6
3. System efflux	2	2; 2
4. Immune evasion	3	2; 3
5. Iron acquisition	23	18; 25
6. Iron uptake	9	9; 9
7. Regulators	5	5; 5
8. Secretion systems	33	32; 33
9. Resistance serum	8	6; 8
10. Trophic factors	0	0; 0
11. Toxins	0	0; 0
12. Unknown class	17	16; 19
13. Antimicrobial resistance	34	21; 37

Original Study Article DOI: https://doi.org/10.36233/0372-9311-717

 

A correlation analysis was performed to assess the relationship between the number of genetic determinants of antimicrobial resistance and the number of genetic determinants of virulence factors. The results of the analysis demonstrated a statistically significant inverse correlation (p = 0.034) between these two indicators. The correlation coefficient (r = –0.705), interpreted according to the Chaddock scale, indicates a high degree of closeness of the identified inverse relationship. Thus, the data obtained indicate a pronounced tendency toward a decrease in the number of genetic determinants of virulence factors with an increase in the number of genetic determinants of resistance to antimicrobials.

Discussion

As part of the analysis of genetic determinants of antimicrobial resistance in the studied K. pneumoniae strains, the group of antimicrobial inactivators demonstrated the greatest diversity and prevalence of genes. This group is largely represented by genes encoding BLRS, which play a key role in resistance to β-lactam antimicrobials, including penicillins, cephalosporins, monobactams and other antimicrobial groups. This study revealed the predominance of the antimicrobial inactivation mechanism, mainly due to BLRS and carbapenemases, which is consistent with global trends reflected in large-scale studies. In particular, a study by J. Yang et al. demonstrated the key role of enzymatic inactivation, especially the blaCTX-M-15 gene, in clinical isolates of K. pneumoniae [14].

The group of genes encoding efflux pumps is a significant determinant of antimicrobial resistance in K. pneumoniae. The strains we studied showed the presence of the following genes: baeR — a gene encoding the BaeR regulatory protein, which controls the expression of the BaeABC efflux pump involved in resistance to a wide range of antimicrobials and stress factors; kpnE, kpnF, kpnG, kpnH genes encoding components of the KpnEFGH efflux system, which is believed to be involved in resistance to some antimicrobials and other toxic compounds; the marA gene encoding the MarA regulatory protein, which controls the expression of several genes, including efflux pump genes such as acrAB, and is involved in resistance to a variety of antimicrobials; the lptD gene, which encodes the LptD protein, a component of the lipopolysaccharide transport system in the outer membrane of Gram-negative bacteria, and the h-NS gene, which encodes the histone-like protein H-NS, a global regulator of gene expression in bacteria. H-NS can influence the expression of efflux pump genes and, consequently, resistance to antimicrobials [9]; the cRP gene encoding the CRP protein, which is a cAMP-activated transcriptional regulator. The cRP gene is involved in the regulation of the expression of various genes, including efflux pump genes; the emrR gene encoding the regulatory protein EmrR, which controls the expression of the EmrAB efflux pump involved in resistance to various antimicrobials [10]; the oqxA and oqxB genes encoding components of the OqxAB efflux system, which provides resistance to fluoroquinolones, chloramphenicol and other antimicrobials [11]. Our study identified a significant proportion of determinants associated with efflux pumps, which correlates with the results of R. Oliveira et al. [15]. This study emphasizes the importance of active efflux systems, such as AcrAB-TolC and OqxAB, in the development of multidrug resistance in K. pneumoniae.

Genetic determinants that enable bacteria to damage the antimicrobial target are represented by genes:

• armA, encoding the 16S rRNA methyltransferase ArmA. Enzymes encoded by genes similar to armA catalyze the methylation of 16S rRNA, which leads to a decrease in the affinity of aminoglycosides for the ribosome and, as a result, to the development of a high level of resistance to a wide range of aminoglycosides;
• gyrA, encoding subunit A of DNA gyrase (topoisomerase II). Mutations in the gyrA gene can alter the structure of DNA gyrase, reducing its affinity for fluoroquinolones and leading to the development of resistance;
• parC, encoding subunit A of topoisomerase IV. The mechanism of resistance to fluoroquinolones involves mutations in the parC gene, which encodes topoisomerase IV [12]. The most common genes in this class are those encoding 16S rRNA methyltransferase.

The group of genetic determinants responsible for target protection and reduced permeability to them is represented by the following genes: ompK37 encodes porin OmpK37, which is a K. pneumoniae outer membrane protein that forms water channels that facilitate the transport of nutrients and other molecules, including some antimicrobials, across the outer membrane. The marA gene encodes the MarA protein, a member of the AraC/XylS family of transcription regulators. The mdtQ gene encodes a component of the MdtABC RND efflux system. Although the main mechanism of action of MdtABC is to pump antimicrobials out of the cell, this mechanism can also be considered a form of “target protection” because it reduces the concentration of antimicrobials near its target. Furthermore, MdtQ is involved in maintaining cell wall integrity, and its expression can affect cell permeability. The QnrB1 protein is one of the most common PMQR proteins that protect the bacterial enzymes DNA gyrase and topoisomerase IV from inhibition by fluoroquinolones.

Analysis of K. pneumoniae isolates revealed significant genetic diversity at the K and O loci, as well as considerable variability in virulence profiles even among strains with closely related capsular types. A discrepancy was noted between genetic type (wzi/K locus) and pathogenic potential: some isolates with the same K type differed in their set of virulence genes, including the presence or absence of key genes characteristic of hypervirulent strains (hvKP). At the same time, signs of high virulence were detected in representatives of different genetic lines. The data obtained emphasize the limitations of capsule or serological typing alone and justify the need for a comprehensive approach combining K/O typing and virulence gene analysis for an adequate assessment of the pathogenic potential of clinical isolates of K. pneumoniae.

It should also be noted that the study conducted a direct analysis of the statistically significant inverse correlation (r = –0.705) between the number of resistance and virulence genes, which reflects the current scientific paradigm. Our empirical data are consistent with the concept of fitness trade-off [4]. At the same time, the study discusses exceptions to this rule — strains SCPM-O-B-14105 and SCPM-O-B-14109, which combine high levels of both resistance and virulence, which corresponds to the global trend of the formation of hybrid MDR-hvKp strains, described earlier in the literature [4].

For the nine strains of K. pneumoniae that were sequenced, the following genes were common: arnT, baeR, cRP, emrR, eptB, uhpT, h-NS, kpnE, kpnF, kpnG, kpnH, ompK37, lptD, marA, msbA, ompA and oqxA were common, and they belonged to three classes of resistance mechanisms: damage to the antimicrobial target, reduced permeability to the antibiotic, and efflux pumps. The presence of identical resistance genes in the K. pneumoniae strains presented may indicate several key mechanisms. The first mechanism is associated with horizontal gene transfer. Plasmids and mobile genetic elements (e.g., transposons) actively spread resistance genes between strains and even species [13]. The second mechanism causes clonal expansion of successful lineages. Certain clonal groups (e.g., ST11 in Asia, ST258 in Europe and America) carry stable plasmids with carbapenemase genes (e.g., KPC-2). These lineages dominate in clinical isolates due to a combination of resistance and virulence [14]. The third mechanism is related to selective pressure from antimicrobials. The widespread use of β-lactams and carbapenems in healthcare facilities contributes to the preservation of BLRS and carbapenemase genes [15]. Thus, identical resistance genes in the strains presented may reflect a combination of evolutionarily successful strategies: horizontal transfer between species, clonal expansion and selection under the influence of antimicrobial therapy.

The resistance of bacteria to antimicrobials depends not only on the presence of resistance genes, but also on their regulation, stability, and interaction with the environment. Gene expression can be suppressed by epigenetic mechanisms (e.g., DNA methylation in Neisseria gonorrhoeae) or depend on mobile elements (mcr plasmids in Escherichia coli). Post-translational modifications (incorrect protein folding) or the absence of key mutations (e.g., in gyrA for fluoroquinolones) also inactivate resistance. Methylation of rRNA (cfr gene) is necessary for resistance to aminoglycosides, and environmental conditions (biofilm in Haemophilus influenzae) temporarily modulate susceptibility. Even in the presence of genetic determinants of resistance, bacteria can exhibit tolerance through metabolic slowing (Helicobacter pylori). Thus, resistance is a multifactorial process in which the dynamic interaction of genetic and environmental factors plays a key role [8, 15]

Conclusion

A study of clinical strains of K. pneumoniae isolated from patients with COVID-19 revealed a complex pattern of antimicrobial resistance, with a predominance of enzymatic inactivation of antimicrobials (42.4%) due to the high prevalence of β-lactamase genes (blaCTX-M-15, blaOXA-48, blaNDM-1, blaTEM-1) and aminoglycoside-modifying enzymes (aac(6')-Ib10, armA), as well as the activity of efflux systems (36.3%) associated with the marA, oqxAB, kpnEFGH genes.

Genotyping revealed the circulation of strains with high virulence potential belonging to clinically significant clones. Analysis of K and O loci showed a predominance of capsular types associated with hypervirulent lines (KL1, KL64, KL57), and lipopolysaccharide types O1, O2a, and O2afg. At the same time, strains with the highest VS (4–5), such as SCPM-O-B-14105 (KL1, O1) and SCPM-O-B-14109 (KL64, O2a), carried genes encoding siderophore systems (iuc, iro) and regulators of the hypermucosal phenotype (rmpA2), indicating their potential belonging to hypervirulent pathotypes.

The key result of the study is the establishment of a statistically significant inverse correlation between the total number of genetic determinants of virulence and determinants of resistance to antimicrobials. This indicates the existence of a fitness trade-off, whereby the accumulation of resistance genes may be accompanied by a decrease in the virulence potential of circulating strains. However, the presence of high levels of both virulence and resistance in individual isolates (e.g., SCPM-O-B-14109) highlights the high risk of successful clones emerging that combine both dangerous traits.

The phenotypic resistance profile confirmed the high prevalence of multidrug-resistant strains: 66.7% of isolates showed resistance to all tested antimicrobials. At the same time, the identified cases of phenotypic susceptibility in strains carrying resistance genes (for example, to ertapenem in isolate SCPM-O-B-14104) emphasize the importance of the stability of genetic elements and the influence of environmental factors in the realization of resistance.

The data obtained, including the catalog of resistance genes, virulence profiles, and molecular typing data, are important for local epidemiological monitoring, allowing the spread of highly virulent and resistant clones to be tracked, as well as for the development of more effective antimicrobial therapy strategies in patients with COVID-19 co-infection and K. pneumoniae-related infections.

About the authors

Guzel Sh. Isaeva

Kazan State Medical University

Author for correspondence.
Email: guisaeva@rambler.ru
ORCID iD: 0000-0002-1462-8734

Dr. Sci. (Med.), Head, Department of microbiology named after Academician V.M. Aristovsky

Russian Federation, Kazan

Nikita S. Chumarev

Kazan State Medical University

Email: nikitasergeevichsno@gmail.com
ORCID iD: 0000-0001-6247-6184

postgraduate student, Department of microbiology named after Academician V.M. Aristovsky

Russian Federation, Kazan

Marina N. Belova

Republican Clinical Infectious Diseases Hospital named after Professor A.F. Agafonov

Email: belova.marina@tatar.ru
ORCID iD: 0000-0001-9579-3370

Head, Bacteriological laboratory

Russian Federation, Kazan

Natalya D. Shaykhrazieva

City Clinical Hospital No. 7, Kazan

Email: epid-gkb7@mail.ru
ORCID iD: 0000-0002-2241-3100

Cand. Sci. (Med.), epidemiologist, Head, Epidemiological department

Russian Federation, Kazan

References

Supplementary files