Evaluation of the EasyScreen™ ESBL/CPO Detection Kit for the Detection of ß-Lactam Resistance Genes

Early detection of multidrug resistant bacteria is of paramount importance for implementing appropriate infection control strategies and proper antibacterial therapies. We have evaluated a novel real-time PCR assay using fluorescent probes and 3base® technology, the EasyScreenTM ESBL/CPO Detection Kit (Genetic Signatures, Newtown, Australia), for the detection of 15 β-lactamase genes (blaVIM, blaNDM, blaIMP, blaOXA-48, blaKPC, blaOXA-23, blaOXA-51, blaSME, blaIMI, blaGES, blaTEM, blaSHV, blaCTX-M, blaCMY, blaDHA) and colistin resistance mcr-1 gene from 341 bacterial isolates (219 Enterobacterales, 66 P. aeruginosa and 56 A. baumannii) that were grown on Mueller–Hinton (MH) agar plates. One colony was suspended in provided extraction buffer, which lyses and converts the nucleic acids into a 3base®-DNA form (cytosines are converted into uracil, and subsequently thymine during PCR). The converted bacterial DNA is then added to the 6 PCR mixes, with primers for three targets plus one internal control. The EasyScreenTM ESBL/CPO Detection Kit was able to detect the 5-major (NDM, VIM, IMP, KPC, OXA-48) and 2-minor (IMI, Sme) carbapenemases and their variants irrespective of the species expressing them with nearly 100% sensitivity and specificity. With cephalosporinases CMY (82% of sensitivity) and DHA (87% of sensitivity) detection of chromosomally encoded variants was less efficient. Similarly, the chromosomally encoded OXA-51 variants were not consistently detected in A. baumannii. Despite being capable of efficiently detecting blaCTX-M-, blaTEM-, blaSHV- and blaGES-like genes, the EasyScreen™ ESBL/CPO Detection Kit was not able to distinguish between penicillinases and ESBL-variants of TEM and SHV and between GES-ESBLs and GES-carbapenemases. As GES enzymes are still rare, their detection as an ESBL or a carbapenemase remains important. Detection of mcr-1 was efficient, but none of the other mcr-alleles were detected in the 341 bacterial isolates tested. The EasyScreenTM ESBL/CPO Detection Kit is adapted for the detection of the most prevalent carbapenemases encountered in Gram-negatives isolated worldwide.


Introduction
Antimicrobial agents played a major role in improving the well-being and health of humans all over the world. These molecules are a mainstay of public health [1]. However, while antibiotics have been successful in reducing the burden infectious diseases, their use has increased exponentially, leading to the emergence and spread of antibiotic resistant bacteria. Gram-negative bacteria (GNB), and especially Enterobacterales, Pseudomonas aeruginosa and Acinetobacter baumannii have emerged as major players in resistance [2]. In these species, resistance may affect all major classes of anti-gram-negative agents (e.g., β-lactams, fluoroquinolones and aminoglycosides), while multidrug resistance (MDR) is relatively common and the rate of infections caused by MDR-GNB is the qualitative detection of nine carbapenemases, four ESBLs and two cephalosporinases widespread in clinically relevant Gram-negatives (NDM, VIM, IMP, KPC, IMI, SME, OXA-48, OXA-23, OXA-51, CMY, DHA, GES, SHV, TEM and CTX-M) and the colistin resistance gene mcr-1 directly from culture plates.

EasyScreen™ Sample Processing Kit
The EasyScreen™ Sample Processing Kit is designed to rapidly isolate nucleic acids (DNA and RNA) directly from bacterial culture agar plate or broth. The nucleic acids are converted into 3base ® sequences prior to purification/or direct use with the EasyScreen TM ESBL/CPO Detection Kit. The lysis reagent was reconstituted, as recommended by the manufacturer by mixing reagent 1 and 2, and subsequent heating at 80 • C until complete dissolution ( Figure 1). Twenty microliters of the combined reagents was dispensed into the provided screw cap tubes (Reaction tube) and one colony, picked using a sterile pipette from culture plates, was resuspended by scraping the tip against the tube and by vortexing. The samples were incubated at 95 • C for 15 min, vortexed and centrifuged before addition of 980 µL of dilution buffer mixed by inversion and centrifugation. For each run, a Negative Process Control (consisting of conversion solution into which a sterile toothpick was dipped) is included.

EasyScreen™ ESBL/CPO Detection Kit
The converted DNA is subsequently used for PCR amplification using the EasyScreen™ ESBL/CPO Detection Kit, a rapid in vitro nucleic acid amplification assay for the qualitative detection of ESBL and Carbapenemase Producing Organism (CPOs) in nucleic acid from rectal swabs and cultured bacteria. The targets are indicated in Table 1. The EasyScreen™ ESBL/CPO Detection Kit includes all reagents required for the detection of the different targets used in the real-time PCR amplification of the nucleic acid, primers and probes labeled with fluorophores as detected by the real-time PCR instrument. In addition, all reaction mixes are manufactured to include an Extraction Control (EC) to determine the reliability of the extracted nucleic acids and indicate the presence of any inhibitors after extraction from primary samples. The 16-plex real-time PCR was performed as recommended by the manufacturer on a Bio-Rad CFX384™ Thermal cycler (Bio-Rad), except for the number of cycles, which was reduced to 35 cycles, using the following protocol: denaturation step of 95 • C for 15 min, 35 cycles (95 • C/2 s; 55 • C/15 s; 60 • C/15 s), and final extension of 65 • C/15 s. The results were interpreted by GS-Call Software provided by the manufacturer. Complete sample processing (lysis of any microorganisms present in one colony and conversion of all cytosine bases into uracil (detected as thymine after PCR amplification) with combined Reagent 1 and 2. (B) Two microliters of DNA extract together with sixteen microliters of PCR mixes was added to each well of the 384-well microplate. (C) Real-time PCR assay (EasyScreen™) on CFX384™. PCR amplification and result interpretation using the manufacturer's proprietary software (GS-Call Software). Total duration of the assay for one sample: c.a. 1h50 up to 3h15 for 24 samples.

EasyScreen™ ESBL/CPO Detection Kit
The converted DNA is subsequently used for PCR amplification using the EasyScreen™ ESBL/CPO Detection Kit, a rapid in vitro nucleic acid amplification assay for the qualitative detection of ESBL and Carbapenemase Producing Organism (CPOs) in nucleic acid from rectal swabs and cultured bacteria. The targets are indicated in Table 1.

Discrepant Results Analysis
Discrepant results between supposed genotype and EasyScreen™ ESBL/CPO PCR results were confirmed using in house PCR, as previously described [20,27,28].

Bacterial Isolates
Bacterial isolates were from the French National Reference Center (F-NRC) for carbapenemresistant Enterobacterales strain collection located in the Bacteriology-Hygiene laboratory of the Bicêtre hospital, France. These are representatives of the bacterial isolates circulating in France during the last 10 years and characterized according to the workflow of F-NRC for antimicrobial resistances. A total of 341 clinical isolates comprising 219 Enterobacterales, 66 Pseudomonas spp. and 56 Acinetobacter spp. harboring single or multiple β-lactamase genes targeted by the assay and/or mcr-1 genes, and other resistance markers not targeted by the assay for specificity control (Tables 2-4 Table 4). Non-CP-GNB (n = 142) were 109 Enterobacterales, 20 P. aeruginosa and 13 A. baumannii.

Statistical Analysis
Performance parameters (sensitivity and specificity) presented here were calculated following resolution of discrepant results, taking the total number of targets as denominator. The sensitivity and specificity values of the EasyScreen TM ESBL/CPO Detection Kit were calculated with their respective confidence interval 95% (95% CI) using the free online software VassarStats: Website for statistical Computation on http://vassarstats.net/ (last accessed on 7 September 2022).

Experimental Setup and Evaluation of Culture Media
The EasyScreen™ Sample Processing Kits (Reference SP001) lyses any microorganisms present in the sample to be tested and converts the cytosine bases to uracil (detected as thymine after PCR amplification) to create 3base ® DNA and RNA. The 3base ® nucleic acids lead to increased homology as compared to the native four bases, thus reducing the complexity of genomes. The latter being more similar to each other enables the design of primers and probes with fewer mismatches and that result in better amplification and in less cross-reactivity.
Five CPOs (1 OXA-48-producing E. coli, 1 KPC-producing K. pneumoniae, 1 IMPproducing K. pneumoniae; 1 VIM-producing P. aeruginosa; and 1 NDM-OXA-23 producing A. baumannii) were grown on six different culture media (MH, TSA, Uri4, COH, ChromID ESBL and CARBA Smart) and were tested to see whether they are compatible with the extraction procedure. There was no difference between these media in terms of amplification (data not shown). Thus, during this retrospective evaluation, the assay was used as recommended by the manufacturer, on fresh overnight bacterial colonies grown on MH plates.
The procedure was easy to perform, taking approximately 16 min for one isolate and up to 45 min for 24 isolates for sample preparation, 2 min for one isolate and up to 60 min for 24 isolates for plate preparation, and a run time of approximately 90 min for 35 cycles using a the CFX 384 instrument (Bio-Rad). The results were interpreted by GS-Call Software (Genetic Signatures) and presented as positive or negative for a given gene with Ct values.
As the PCR were conducted on colonies, the number of cycles was reduced to 35, as preliminary results showed late Cts (>40), which likely corresponded to contamination due to high DNA concentrations. With 35 cycles, this problem was solved as no false or unexplained positive PCRs were recorded.
Finally, the EasyScreen™ ESBL/CPO Detection Kit is also able to detect mcr-1 colistin resistance genes with a specificity of 100%. However, it was not able to detect other MCR variants such as MCR-2, -3, -4 and -5.
As for Enterobacterales, all CTX-M, SHV and TEM-producers were detected. For SHV, both isolates tested expressed an SHV ESBL, while for TEM only one TEM-ESBL was present among the five tested positive TEM-producers. Thus, considering ESBL detection in P. aeruginosa isolates, the EasyScreen TM ESBL/CPO Detection Kit revealed a sensitivity and a specificity of 100.00% (CI95: 63.06% to 100.00%) and 92.98% (CI95: 83.00% to 98.05%), respectively.
The other resistant determinants were not present in the panel of tested isolates.
Similarly, 23 bla TEM genes were detected, while none corresponded to a TEM-ESBL. Only two SHV-5 ESBLs and two CTX-M-producers were detected.

Overall Assay Performance for Carbapenemase Detection Extrapolated to the French Epidemiology of CP-GNB
When extrapolating these results to the global French epidemiology, as represented by the carbapenemase genes identified by the French National Reference Center (F-NRC) for Carbapenem-resistant Enterobacterales, P. aeruginosa and A. baumannii, excellent specificity and sensitivity might be expected for the detection of CPEs (Table 5)   Taken together, our results obtained with EasyScreen TM ESBL/CPO Detection Kit on our collection of isolates, and when extrapolated to the global French epidemiology of CPOs, we would expect (in the best way) 99.97% sensitivity for CPE detection, 98.22% for CP Pseudomonas spp. and 88.25% for CP-Acinetobacter spp. These results are better than those of BD MAX Check-Points CPO Assay [30], which only detects the five main carbapenemases, with sensitivities of 99.26% for CPE detection, 93% for CP Pseudomonas spp. and only 12.5% for CP-Acinetobacter spp.

Discussion
The EasyScreen™ ESBL/CPO Detection Kit is an efficient and rapid (<3 h from colony to result) detection system for the most frequently encountered penicillinases (including ES-BLs), carbapenemases, cephalosporinases and MCR-1 in Enterobacterales, Pseudomonas spp. and A. baumannii. Concerning the evaluation of Pseudomonas spp. and Acinetobacter spp. strains, we observed more false-positive and negative results and consequently the percentage of specificity and sensitivity is lower especially with CMY, IMP, GES and OXA-51 variants.
The EasyScreen™ ESBL/CPO Detection Kit was performed on cultured bacteria using a boiling extraction protocol, which requires hands on time. However, the EasyScreen™ ESBL/CPO Detection Kit is compatible with most existing automated nucleic acid extraction and real-time PCR instruments and can thus be fully automated.
Irrespective of the host bacteria, the EasyScreen™ ESBL/CPO Detection Kit showed excellent biological performances (sensitivity and specificity) for the five most common carbapenemases, including IMP variants that constitute a very heterogeneous family of enzymes and that are not well detected by most molecular assays [27]. The results of our study revealed that the EasyScreen™ ESBL/CPO Detection Kit is well adapted to the French epidemiology of CPE and CP-Pa, which reflects that of many countries in Europe and around the world. The short turnaround time and its simplicity make it suitable for routine use in clinical microbiology laboratories. It can provide results from colonies that grew on most commonly used culture plates, including MH and complex chromogenic and selective screening media.
The kit also detects the main carbapenemases encountered in Enterobacteriaceae (KPC, GES, NDM, VIM, IMP, OXA-48-like) in P. aeruginosa (VIM, IMP, NDM and KPC) and in A. baumannii (OXA-23, VIM, IMP, NDM), except minor OXA-carbapenemases usually identified in A. baumannii (OXA-24/-40-like and OXA-58-like) that are not targeted by this assay. When extrapolated to the global French epidemiology of CPOs, we would expect (in the best way) 99.97% sensitivity for CPE detection, 98.22% for CP Pseudomonas sp. and 88.25% for CP-Acinetobacter spp. Thus, the Younden index for the French situation would be 99.27, 93.3, 88% for CP-E, CP-Pa and CP-Ab, respectively.
Minor carbapenemases are increasingly isolated and often responsible for outbreaks, as no commercially available assay targets these resistance determinants [37][38][39][40][41][42][43][44]. At the French national reference center in 2020, 22 strains of Enterobacter cloacae complex producing class A carbapenemases IMI/NMC-A, 11 strains of Proteus spp. producing OXA-23 and one strain of K. pneumoniae producing GES-5 have been identified, representing 1.1% of all carbapenemases received during that year [37]. Recently, in a multicentric study performed in 12 French hospitals, 26.9% (14/52) of the amoxicillin-clavulanate-resistant Proteus mirabilis isolates produced an OXA-23 carbapenemase, suggesting that this resistant determinant might be underestimated in France [38]. As early as 2013, Bush et al. have shown the silent spread of Sme enzymes in the United States and suggested these carbapenemase to be included in carbapenemase detection systems. Recently, Public Health England has revealed increasing identification of S. marcescens-producing Sme carbapenemases in the United Kingdom [39,40]. Similarly, GES-5 carbapenemase-producing GNB have increasingly been described in multispecies outbreaks on different continents [41][42][43][44]. In most of these cases, the index patient was missed as no commercially available assay targeted this carbapenemase, and epidemiological follow up was difficult, as it relied on home-made PCR, or on WGS which is costly and time consuming [41].
Similarly, it is the only commercially available assay that allows detection of ESC resistance by targeting the main ESBL, CTX-M-like, but also some minor ESBLs (VEB prevalent in South-East Asia, PER in Turkey, and GES in Greece and in the Arabic Peninsula), which turns this assay into a globally adapted assay [45]. Finally, it includes the two main plasmid-encoded AmpCs, currently spreading in most countries [36]. Regarding TEM and SHV, as it is not able to distinguish between penicillinase and ESBL variants, these results should be considered only when all the other ESBLs genes are negative, and yet the bacteria display a classical ESBL phenotype. In a recent study conducted in France, out of 100 consecutively isolated ESBLs, 98 were CTX-M variants and 2 E. cloacae were SHV-2a [20]. These latter using the EasyScreen™ ESBL/CPO Detection Kit were positive for SHV and displayed an ESBL phenotype on disc diffusion antibiogram. Its simplicity and short turnaround time (less than 2 h for one sample) makes it suitable for use in the routine microbiology laboratory. It can provide results from colonies that grew on MH but also from selective screening media.
The 3base ® nucleic acids reduces the complexity of genomes allowing the design of primers and probes with fewer mismatches, which results in better amplification and in less cross-reactivity.
Mutation and/or polymorphisms in the primer/probe binding region of the targeted gene may lead to false negative PCR results and thus to non-detection of some variants. The use of the 3base ® nucleic acids is less prone to mutations/polymorphisms, as it reduces the complexity of genomes allowing the design of primers and probes with fewer mismatches, which results in better amplification and increased detection of variants. This is best exemplified with IMP variants, a very heterogeneous family of carbapenemases. Nevertheless, as for every molecular assay, they must be evaluated on isolates corresponding to the local epidemiology in-order-to know which are the alleles that might be missed, especially in a constantly evolving field of carbapenemases. The penultimate goal is now the evaluation of the EasyScreen™ ESBL/CPO Detection Kit directly on clinical samples, especially on rectal swabs to identify carriers of carbapenemase/ESC resistance genes, and the use of Genetic Signatures' full automation solutions would be 'good to have' in order to speed up extraction and plate preparation.