Prevalence of the CRISPR-cas system and its association with antibiotic resistance in clinical Klebsiella pneumoniae isolates

Background and objective(s) CRISPR-Cas is a prokaryotic adaptive immune system that protects bacteria and archaea against mobile genetic elements (MGEs) such as bacteriophages plasmids, and transposons. In this study, we aimed to assess the prevalence of the CRISPR-Cas systems and their association with antibiotic resistance in one of the most challenging bacterial pathogens, Klebsiella pneumoniae. Materials and methods A total of 105 K. pneumoniae isolates were collected from various clinical infections. Extended-spectrum β-lactamases (ESBLs) phenotypically were detected and the presence of ESBL, aminoglycoside-modifying enzymes (AME), and CRISPR-Cas system subtype genes were identified using PCR. Moreover, the diversity of the isolates was determined by enterobacterial repetitive intergenic consensus (ERIC)-PCR. Results Phenotypically, 41.9% (44/105) of the isolates were found to be ESBL producers. A significant inverse correlation existed between the subtype I-E CRISPR-Cas system’s presence and ESBL production in K. pneumoniae isolates. Additionally, the frequency of the ESBL genes blaCTX−M1 (3%), blaCTX−M9 (12.1%), blaSHV (51.5%), and blaTEM (33.3%), as well as some AME genes such as aac(3)-Iva (21.2%) and ant(2’’)-Ia (3%) was significantly lower in the isolates with the subtype I-E CRISPR-Cas system in comparison to CRISPR-negative isolates. There was a significant inverse correlation between the presence of ESBL and some AME genes with subtype I-E CRISPR-Cas system. Conclusion The presence of the subtype I-E CRISPR-Cas system was correlated with the antibiotic-resistant gene (ARGs). The isolates with subtype I-E CRISPR-Cas system had a lower frequency of ESBL genes and some AME genes than CRISPR-negative isolates.


Introduction
Acquisition of DNA elements such as fitness, antibiotic resistance, and virulence genes through horizontal gene transfer (HGT) is a crucial step in bacterial adaptation to various hosts and environments [1,2].Furthermore, many species of bacteria have developed an adaptive immune system known as clustered regularly interspaced short palindromic repeats and their associated Cas proteins (CRISPR-Cas), which helps to restrict the acquisition of external genetic elements and protect against invasive plasmids and bacteriophages [3,4].These defense systems are composed of a leader sequence, cas genes, and a CRISPR array.A CRISPR array typically consists of highly conserved short direct repeats (DR), separated by unique sequences (spacers) acquired from mobile genetic elements (MGEs) [5][6][7].The number of cas genes in the CRISPR-Cas locus is variable and are often located next to CRISPR repeat-spacer arrays [8].Based on the cas operon architecture, cas gene content, and Cas protein sequences, CRISPR-Cas systems are classified into two main classes, comprising 6 major types and 33 distinct subtypes [9].The Cas proteins possess a variety of enzymatic domains with helicase, polymerase, or nuclease activity and are essential for the functioning of the CRISPR-Cas system [10,11].
Acquiring new DNA, often encoded on MGEs like transposons and plasmids, is the mechanism by which many bacteria develop resistance to antibiotics.Investigations conducted on various bacteria have demonstrated the significance of CRISPR-Cas systems in the exchange of genetic material and their potential to impact the rate of evolution [12].Palmer et al. [13] demonstrated a significant negative correlation between antibiotic resistance acquisition and the CRISPR-Cas system's presence in Enterococcus faecalis isolates.Moreover, a few studies suggested that the CRISPR-Cas system could regulate the pathogenicity of bacteria.The CRISPR-Cas system in Pseudomonas aeruginosa allows for the modulation of biofilm formation, which is a crucial factor in the pathogenicity of a variety of microorganisms [14].CRISPR-Cas system modulates the prophage's contents in Streptococcus pyogenes and, consequently, its virulence [15].However, Touchon et al. [16] demonstrated that the CRISPR-Cas system is not an effective barrier against antibiotic resistance and plasmid spreading in Escherichia coli.For these reasons, there is much potential for research into how this system affects various bacterial pathogens' virulence and antibiotic resistance.
Owing to the extensive dissemination and high rate of antibiotic resistance, Klebsiella pneumoniae has become a predominant opportunistic pathogen in hospital environments [17].High molecular weight plasmids are associated with hypervirulent phenotype and multidrug resistance in K. pneumoniae [18,19].So far, two types of CRISPR-Cas systems including type I (types I-F, I-E, and I-E*) and IV (primarily type IV-A) systems have been identified in Klebsiella spp.The type I CRISPR system is located mainly in chromosome, while the type IV system is exclusively found in plasmids [8,20,21].Type I-E is the canonical type I-E CRISPR-Cas system (located in the cysH-iap region), containing a cas operon, consistent direct repeats (29 bp), and a CRISPR array (designated as CRISPR1) which is situated downstream of the cas genes.Type I-E* is variable in comparison to type I-E, which is located in the ABC transport system-glyoxalase area and occasionally has a transposase-encoding gene integrated into the cas operon.Besides, this type contains two CRISPR arrays (designated as CRISPR2 and CRISPR3, respectively) that bracket the cas genes [22][23][24].The marker gene of the type I CRISPR-Cas system is cas3, and cas1 is a universal cas gene in all CRISPR-Cas types [25].Whether the CRISPR-Cas system in K. pneumoniae facilitates HGT or functions as an immune system is still a question.Therefore, this study aimed to determine the correlation between the presence of the CRISPR-Cas system and ESBL and aminoglycoside genes in clinical K. pneumoniae isolates.

Isolation and identification of K. pneumoniae
In this prospective study, one hundred and five nonduplicate and non-consecutive clinical K. pneumoniae isolates were gathered from various clinical specimens including urine, sputum, blood, and wound from the patients admitted to Imam Reza Teaching and Treatment Hospital in Tabriz, Iran.These isolates were initially identified by conventional bacteriology tests such as gram staining, colony morphology, the reaction in triple sugar iron agar (TSA), lysine iron agar (LIA), citrate utilization, indole production, and motility.Subsequently, molecular identification was carried out using polymerase chain reaction (PCR) as described elsewhere [26].Ultimately, the identified isolates were stored in Tryptic Soy Broth (TSB, Merk) containing 10% (v/v) glycerol and kept at -70 °C until used.

Phenotypic detection of extended-spectrum β-lactamases (ESBLs) production
ESBL phenotypic detection was accomplished by the combination disk diffusion test (CDDT) in accordance with the Clinical and Laboratory Standards Institute (CLSI) guidelines [27].Ceftazidime (30 µg), cefotaxime (30 µg), ceftazidime-clavulanic acid (30/10 µg), and ceftazidime-clavulanic acid (30/10 µg) disks (Mast Group Ltd., Merseyside, UK) were used in the CDDT.The discs were placed onto a Mueller Hinton agar (MHA) plate that had been inoculated with the test strain.When the diameter of the inhibition zone surrounding the combination disks was greater than that of the cefotaxime (30 µg) or ceftazidime (30 µg) disks alone by at least 5 mm, the isolates of K. pneumoniae were considered as ESBL-producers.K. pneumoniae ATCC 700,603 and Escherichia coli ATCC 25,922 were used as the positive and negative controls for the production of ESBLs, respectively.

Genotypic detection of CRISPR and antibiotic resistance genes
The tissue buffer boiling method (0.05 M NaOH and 0.25% sodium dodecyl sulfate (SDS) was utilized to extract the total DNA of the isolates.The CRISPR1, CRISPR2, and CRISPR3 genes were detected via PCR in order to confirm the existence of the CRISPR-Cas system.The cas1 and cas3 genes were also identified to detect the type of CRISPR-Cas system.Moreover, the presence of aminoglycoside-modifying enzymes (AME) and ESBL genes were identified using PCR amplification.Table 1 displays the primer sequences along with the amplified size.A 25 µl reaction mixture was used for the PCR amplification, which included 12.5 µL of Taq DNA Polymerase 2X Master Mix RED (Amplicon Co., Denmark), 1 µL of extracted DNA as the template, 1 µL of each primer (10 pmol), and 9.5 µL of DW in the BIO-RAD C1000 thermal cycler (Applied Biosystems, USA).A standard UV transilluminator was used to view the stained gels after the PCR products were electrophoresed on a 1.5% agarose gel in 1X TBE buffer.

Analysis of genotype by enterobacterial repetitive intergenic consensus (ERIC)-PCR
The ERIC-PCR method was used to assess the isolates' genetic relatedness to one another.The single primer ERIC1 with a sequence of 5'-A T G T A A G C T C C T G G G G A T T C A C-3' was used for genotype of all isolates.ERIC-PCR was performed in a volume of 25 µl containing 12.5 µL of TEMPase DNA Polymerase Hot Start 2x Master Mix BLUE PCR (Amplicon Co., Denmark), 2 µL of template DNA, 1 µL of each primer (10 pmol), and 9.5 µL of nuclease-free water.The thermal cycling conditions were as follows: initial denaturation at 94ºC for 5 min, followed by 35 repeated cycles of DNA denaturation at 94ºC for 30 s, annealing at 48ºC for 1 min, extension of primer at 72 ºC for 2 min, and a final extension at 72 °C for 5 min.Following electrophoresis on 1.5% (w/v) agarose gel, the amplified fragments were stained with DNA-safe stain (Sinaclon Co., Tehran, Iran), visualized with UV light, and captured with an ultraviolet gel documentation device (Uvitec, UK).New England Biolabs' 100 bp DNA ladders were used as molecular size markers to estimate product size.The similarity between strains was found based on the analysis of the banding by GelJ software, and dendrograms were generated using the Dice similarity method and the Unweighted Pair Group Method with Arithmetic Averages (UPGMA) technique.

Statistical analyses
Descriptive statistics were analyzed by SPSS software (version 27.0 SPSS Inc., Chicago, IL, USA).Pearson chisquare or One-tailed Fisher's exact tests (when one or more of the cell counts is less than 5) were used to compare the occurence of the CRISPR-Cas system and its subtypes and the presence of ESBL and aminoglycoside genes among K. pneumoniaee isolates.In addition, the correlation between the presence of CRISPR-Cas systems and different ARGs was calculated by Spearman's rank correlation coefficient among the isolates.The discriminatory power was measured by http://insilico.ehu.eus/.The p-value < 0.05 was considered statistically significant.

S rRNA F: A T T T G A A G A G G T T G C
aac (6)´-Ib

Bacterial isolates
From

Correlation between CRISPR-Cas system and ESBL production
Of the 105 K. pneumoniae isolates that were not susceptible to cefotaxime, 44 (41.9%) were found to be ESBL producers using the combination disk diffusion test (CDDT).
The majority of isolates containing the CRISPR system were unable to produce ESBLs.Of the isolates containing subtype I-E CRISPR-Cas system, only 5 (15.2%) were ESBL producers, whereas the remaining 28 (84.8%)isolates did not.Statistical analysis revealed a significant inverse correlation between the subtype I-E CRISPR-Cas system's presence and ESBL production in K. pneumoniae isolates (P-value > 0.001, correlation coefficient = -0.367).

CRISPR-Cas systems and AME genes
PCR results also revealed that 91 (88.6%) out of 105 isolates contained AME genes.The aac(6´)-Ib gene (83.8%,88/105) was the most common AME gene, followed by  3)-IVa, and 1 harbored ant(2'')-Ia.Moreover, there was a significant correlation between the presence of subtype I-E CRISPR-Cas system and some AME genes (Table 3).The results revealed that the frequency of the AME genes aac(3)-Iva (7/33, 21.2%) and ant(2'')-Ia (1/33, 3%) was significantly lower in the isolates with the subtype I-E CRISPR-Cas system in comparison to CRISPR-negative isolates (P = 0.05 and P = 0.01, respectively).The association between the presence of CRISPR systems and ESBL production and ESBLs and AME genes among the K. pneumoniae isolates is shown in Fig. 1.

ERIC-PCR analysis
The determination of genomic diversity of 105 clinical K. pneumoniae isolates demonstrated that there were one hundred ERIC types, including 95 singletons and 5 common, using ERIC-PCR at an 80% similarity cut-off value with a discriminatory power of 0.9991 (Fig. 2).In addition, there were 10 clusters at 53% similarity cut-off value with a discriminatory power of 0.8703.Generally, the number of bands in the electronic analysis of the PCR products ranged from 4 to 13 with the sizes ranging from about 300 bp to more than 1500 bp.Out of 105 isolates, a total of 9 (8.6%) isolates belonged to cluster 1, 7 (6.7%)isolates to cluster 2, 8 (7.6%) to cluster 3, 11 (10.5%) to cluster 4, 24 (22.9%) to cluster 5, 6 (5.7%) to cluster 6, 20 (19%) to cluster 7, 3 (2.9%) to cluster 8, 4 (3.8%) to cluster 9 and 13 (12.4%) to cluster 10.The isolates were not significant among the ERIC clusters (P = 0.1285).There were no isolates with CRISPR-Cas systems in cluster 8 and cluster 5 had the most isolates containing the CRISPR-Cas system.In addition, the AME and ESBL genes distribution pattern demonstrated no significant association with ERIC clusters (P = 0.1312 and P = 0.738, respectively).

Discussion
A recent study examined CRISPR-Cas systems in the genomes of K. pneumoniae [28].According to bioinformatics studies, only 6 out of 52 K. pneumoniae strains with the available draft or complete genomes had a complete CRISPR-Cas system, indicating a limited distribution of the CRISPR-Cas systems in K. pneumoniae.
Out of 116 spacer sequences, 38 spacers were found to be extremely similar to the bacterial, plasmid, or phage genome sequences using blast search analysis [28].In addition, Liao et al. [29] revealed that only 14.9% (25/168) of clinical K. pneumoniae isolates had the CRISPR-Cas system.There could be multiple reasons for the restricted distribution of the CRISPR-Cas systems in K. pneumoniae.First, the loss of the CRISPR-Cas system, may be due to strong selective pressure to acquire antibiotic resistance or virulence genes [30].Secondly, the CRISPR-Cas system itself might be an MGE associated with various habitats, which could be transferred into K. pneumoniae strains [31].Thirdly, CRISPR-Cas systems distribution is non-random and Multilocus sequence typing (MLST)-dependent [29,32].As the CRISPR-Cas system appears not to be widely prevalent in K. pneumoniae species, it remains to be determined how the system's presence or absence could contribute to the evolution of K. pneumoniae strains [33].
Based on Cas1 and Cas3's amino sequences and their genomic position, Klebsiella pneumoniae's CRISPR-Cas system could be classified into type I-E and subtype I-E* Fig. 1 The correlations between CRISPR systems and phenotypic and genotypic ESBLs and AME genes among the K. pneumoniae isolates.The presence of most of the ESBL and aminoglycoside genes was correlated to the absence of the cas3 gene among the isolates (P-value < 0.05).In addition, the phenotypic ESBLs were directly correlated to genotypic ESBLs (P-value < 0.001) and inversely correlated to genotypic AMEs (P-value < 0.05).Furthermore, genotypic AMEs were directly correlated to aminoglycoside resistance genes (P-value < 0.001).The green stains are demonstrated the inverse correlations and the red stains are demonstrated direct correlations.The correlations were tested using Spearman's rank test.*: P-value < 0.05; **: P-value < 0.001 Fig. 2 The genomic diversity of the clinical K. pneumoniae isolates by using ERIC-PCR at 80% similarity cut-off value [8].Wang et al. [2] revealed that the presence of cas1 and cas3 in conjunction with CRISPR arrays are indicators that the isolates of K. pneumoniae carry the CRISPR-Cas system.It is interesting that all or some isolates with Cas3 did not have Cas1.Cas1 is typically involved in the adaptation phase of CRISPR immunity, where new spacers are integrated into the CRISPR array.The absence of Cas1 suggests potential limitations in the ability of these isolates to acquire new immunity against phages or plasmids.This phenomenon is shown in other studies such as Jwair et al. [34] and Li et al. [8] that were observed in K. pneumoniae or Gholizadeh et al. [35] Hullahalli [36] and Palmer [37] that were observed in E. faecalis.In E. faecalis, there are three types of CRISPR including CRISPR1cas, CRISPR2, and CRISPR3-cas.CRISPR2 is an orphan CRISPR (lacks cas genes) and uses the cas genes of other CRISPRs.In addition, the function or activity of cas genes of CRISPR-Cas types of K. pneumoniae remains unclear and needs to be determined.In our study, 34.2% of K. pneumoniae isolates were found to harbor the CRISPR-Cas system which is considered a low proportion.The relatively low prevalence of CRISPR systems in this study can be attributed to the fact that most of the strains contained investigated antibiotic-resistant genes; therefore, they were found to be negative for these systems.We also found the subtype I-E* CRISPR-Cas system to have a lower prevalence (2.8%) than the subtype I-E CRISPR-Cas system (31.4%).Similar to our results, Li et al. [8] reported the prevalence of CRISPR-Cas system in K. pneumoniae was 30.7% (54/176) and Alcompoz et al. [38] reported 25.4% (46/181), and most of them have also belonged to type I-E.However, several studies reported the prevalence of type I-E* CRISPR-Cas system in K. pneumoniae was higher than in type I-E [29,39,40].Furthermore, in contrast to our results, the findings of Hu et al. [39] demonstrated that the examined isolates showed co-existing type I-E and subtype I-E*.These can be because of the difference in the phylogenetic traits of bacterial isolates across different geographic areas as well as the diverse origins of the isolates used in the investigations.In this regard, Kannadasan et al. [41] reported that the type of CRISPR-Cas systems found in K. pneumoniae can vary greatly depending on the geographical location.Regular monitoring of the proportion of subtype I-E* and type I-E strains of K. pneumoniae could be crucial because it may have an impact on the global patterns of evolution and development of multidrug resistance in this bacterium [41].
Our results indicated there was a significant inverse correlation between the presence of ESBL and some AME genes in the isolates with subtype I-E CRISPR-Cas system in comparison to CRISPR-negative.The findings raised the possibility that the subtype I-E CRISPR-Cas system effectively restricts the acquisition of acquired ARGs and external DNA fragments.Similarly, Lin et al. 's [42], Jwair et al. [34] and Wang et al. [2] demonstrated that there was a highly significant inverse association between the prevalence of CRISPR-Cas system and drug resistance in carbapenem-resistant and ESBL-producing K. pneumoniae.Generally, the CRISPR-Cas systems found in K. pneumoniae are not always correlated to a dearth of ARGs; rather, an enormous number of ARGs and CRISPR-Cas systems have been found co-existing in the analyzed genomes [38,43].In this regard, Alkompoz et al. [38] revealed that the frequency of the genes including bla VIM , bla NDM , ereA2, armA, msrE, florR, mcr-3, and tet(B) was significantly higher in the presence of CRISPR-Cas systems.However, other genes such as bla TEM , bla KPC , bla LAP−2 , rmtB, fosA, and catA3 were significantly higher in the genomes of the CRISPR/Casnegative strains.Studies carried out on different bacterial species also have demonstrated the contradiction in the CRISPR-Cas system's effect on preventing the spread of ARGs and, as a result, antibiotic resistance.It was previously found that the CRISPR-Cas system is significantly correlated with the absence of ARGs and high drug sensitivity in Pseudomonas aeruginosa [44,45] and Enterococcus faecalis [35,46].In contrast, it was associated with increased antibiotic resistance in Campylobacter jejuni [47].There are several reasons why the CRISPR-Cas system's presence does not always prevent the spread of ARGs in bacteria.Strong selective pressure for ARGs acquisition could lead to CRISPR repression, and the presence of self-targeting spacers also may render many CRISPR-harboring strains immunologically inactive [22,30,32].Moreover, phages that express anti-CRISPR proteins (Acrs) have the ability to deactivate the bacterial CRISPR-Cas system, which could lead to the spread of ARGs, as has been observed in P. aeruginosa [48][49][50].In addition, as previously reported in Shigella species, insertion sequence-mediated mutations and point mutations in the cas1 and cas2 genes were associated with the spread of MDR strains [51].The presence of point mutations in the protospacer adjacent motif (PAM) sequence or mismatches between invader DNA and spacer curbs CRISPR interference and drastically decreases the affinity of the cascade-crRNA complex for target DNA.This prevents the cleavage of DNA even in the presence of the CRISPR system since matched protospacer sequences are necessary for CRISPR scanning [52,53].For CRISPR interference activity, spacer GC content and proximity to the leader sequence are crucial because the leader sequence functions as a promotor to regulate the transcription process and is a preferred site for the insertion of further spacers [38,53,54].Moreover, the restrictionmodification (R-M) systems may play an important role in preventing the spread of ARGs, in addition to the CRISPR-Cas system [55].H-NS proteins also bind to the cas operon's promoter in addition to DNA-binding proteins, leading to a reduction in cas3 expression and, as a result, a loss of CRISPR-Cas activity.Previous studies have shown that imipenem treatment induces H-NS expression, which results in a loss of CRISPR system activity [42,54,[56][57][58].Inhibition of cas3 expression in K. pneumoniae through the stimulation of the transcriptional repressor H-NS leads to loss of the immunity of the CRISPR-Cas system and eventually ARGs acquisition [42].Collectively, these reasons increase the likelihood that MGEs evade the immunity of the CRISPR-Cas system.
Within this investigation, 105 clinical K. pneumoniae isolates were differentiated into 100 genotypes using ERIC-PCR.This finding indicated that the great majority of isolates were not clonally related and that the spread of K. pneumoniae was not correlated with a clonal outbreak.The results of the ERIC-PCR technique demonstrated that isolates were highly heterogeneous and genetically diverse.This could be correlated to the genetic variation in our isolates [59,60].Our results were consistent with previous studies.In a study conducted in Iran, Kiaei et al. [61] differentiated the 37 K. pneumoniae strains into 29 genotypes using the ERIC-PCR method.In Brazil, Ferreira et al. [62] also used ERIC-PCR to differentiate 25 K. pneumoniae strains into 23 genotypes.In another study, Ghamari et al. [59] identified 55 and 60 different genotypes among 60 carbapenem-resistant K. pneumoniae isolates using RAPD and ERIC-PCR methods, respectively.We also found that the CRISPR-Cas systemcontaining K. pneumoniae isolates belonged to different clusters, and the pattern of the distribution of the ESBL and aminoglycoside genes demonstrated that there was no significant association with ERIC clusters.However, in the Wasfi et al. [60] and Kashefieh et al. [63] studies, both ERIC-PCR and RAPD-PCR genotypic analyses demonstrated an association with resistance patterns of K. pneumonia.Even though RAPD-PCR and ERIC-PCR are quick, easy, and affordable genotyping techniques, their reproducibility is limited and contingent upon the PCR conditions and bacterial DNA quality [63,64].Alternative typing techniques, like MLST, have been developed to achieve more dependable results.MLST method relies on the sequencing of conserved housekeeping genes and has demonstrated reproducibility and high reliability in comparison to other typing methods [59,64].

Conclusion
Our findings revealed that the presence of the subtype I-E CRISPR-Cas system is associated with the ARGs.Significantly, the isolates with subtype I-E CRISPR-Cas system had a lower frequency of the ESBL genes and some AME genes compared to CRISPR-negative isolates.Analysis of the correlation between the CRISPR-Cas system and antibiotic resistance will help to identify and better understand the mechanism of bacterial resistance and provide new instructions for the prevention and treatment of bacterial antibiotic resistance.Therefore, the CRISPR-Cas system along with other genetic markers could be used for infection control by resistant pathogens, to give insights into their genetic contents and phenotypic characteristics, and also to differentiate low-risk strains of pathogens from high-risk strains.
In addition, there were 11 different patterns associated with the distribution of ESBL genes (Table2).At least two or more ESBL genes were present in most isolates containing ESBL genes.The combination of bla CTX−M9 , bla TEM , and bla SHV was the most common (26.9%, 25/93), followed by bla TEM and bla SHV (21.5%, 20/93).29 (31.2%)out of 93 isolates also had a single ESBL gene; 15 of them harbored bla SHV , 12 harbored bla TEM , and 2 harbored bla CTX−M1 .Furthermore, there was a significant correlation between the presence of subtype I-E CRISPR-Cas system and ESBL genes (Table3).

Table 3
The association of drug-resistance genes with CRISPR-Cas Systems in K. pneumoniae isolates