Evaluation of a new rapid immunochromatographic assay for the detection of GES-producing Gram-negative bacteria

Abstract

Background

As carbapenemase-producing Enterobacterales are increasingly reported worldwide, their rapid detection is crucial to reduce their spread and prevent infections and outbreaks. Lateral flow immunoassays (LFIAs) have become major tools for the detection of carbapenemases. However, as for most commercially available assays, only the five main carbapenemases are targeted.

Objectives

Here, we have developed and evaluated an LFIA prototype for the rapid and reliable detection of the increasingly identified GES-type β-lactamases.

Methods

The GES LFIA was validated on 103 well-characterized Gram-negative isolates expressing various β-lactamases grown on Mueller–Hinton (MH) agar, chromogenic, and chromogenic/selective media.

Results

The limit of detection of the assay was 10⁶ cfu per test with bacteria grown on MH agar plates. GES LFIA accurately detected GES-type β-lactamases irrespective of the culture media and the bacterial host. The GES LFIA was not able to distinguish between GES-ESBLs and GES-carbapenemases. Because GES enzymes are still rare, their detection as an ESBL or a carbapenemase remains important, especially because extensive use of carbapenems to treat ESBL infections may select for GES variants capable of hydrolysing carbapenems.

Conclusions

The GES LFIA is efficient, rapid and easy to implement in the routine workflow of a clinical microbiology laboratory for the confirmation of GES-type β-lactamases. Combining it with immunochromatographic assays targeting the five main carbapenemases (KPC, NDM, VIM, IMP and OXA-48) would improve the overall sensitivity for the most frequently encountered carbapenemases and ESBLs, especially in non-fermenters.

Introduction

The dissemination of carbapenemase-producing (CP) Gram-negatives (GN) is a matter of great clinical concern given (i) the major role of these pathogens as causes of nosocomial infections (and, for Escherichia coli, also of community-acquired infections), and (ii) the major role of carbapenems in the treatment of those infections.¹ The activity of carbapenems is compromised by the spread of the five main carbapenemases (KPC, OXA-48, NDM, VIM, IMP), but a variety of minor carbapenemases have also been reported,^1–3 and are increasingly isolated and often responsible for outbreaks.³

Among these minor carbapenemases, GES-type enzymes are increasingly identified in clinical GN rods, including in Pseudomonas aeruginosa, Acinetobacter baumannii, E. coli and Klebsiella pneumoniae.³ GES-1 (for ‘Guiana extended spectrum’) was initially described in 2000 in a K. pneumoniae isolate from French Guiana.⁴ The rapidly increasing number of GES variants differing by 1 to 16 amino acid substitutions (n = 53; http://www.bldb.eu)^4,5 and their geographical spread in the different compartments (human, animals and environmental) signals a worrying global evolution and spread of this family of enzymes.^2,3 Among the 53 known GES variants, 32 (60.4%) with either Gly170Asn (n = 2; 3.8%) or Gly170Ser (n = 30; 56.6%) mutations are carbapenemases, whereas the others are categorized as minor ESBLs without any carbapenem hydrolytic activity.^2,3,5

In France, GES enzymes are still rare in Enterobacterales and A. baumannii, because only one GES-5-producing K. pneumoniae (0.03% of total CPEs) and four GES-producing A. baumannii isolates (0.7% of total CP-A. baumannii) were received by the French National Reference Center (F-NRC) for Antibiotic Resistance in 2020.^6,7 However, in P. aeruginosa a constant increase in the proportion of GES enzymes has been evidenced since 2016. GES-ESBLs represented the first (n = 35; 30% of all) ESBLs in P. aeruginosa isolates from France in 2020 before SHV-ESBLs (n = 22: 19%) and PER-1 (n = 19; 17%).⁷ Similarly, the proportion of GES-CP P. aeruginosa increased between 2018 and 2020 (11.8% to 19.5%), and in 2020 represented the second most prevalent carbapenemase in P. aeruginosa, after VIM (53%) and before IMP (14%) carbapenemases.⁷

Although rare, GES carbapenemases have now been identified worldwide, most frequently associated with single occurrences but several outbreaks of GES-producing Enterobacterales, P. aeruginosa and A. baumannii have also been described.^2,3,8–14 GES-5 CP-GN have increasingly been described in multispecies outbreaks on different continents and in several environmental studies.^2,3,12 In most of these outbreak cases, the index patient was missed because most commercially available diagnostic assays detect/confirm only the presence of the five main carbapenemases and epidemiological follow-up was difficult, because it relied on homemade PCR or on WGS, which is costly and time-consuming for efficient outbreak management.^8,9,15

Phenotypic approaches such as antimicrobial susceptibility are often difficult to interpret because GES producers are often susceptible to carbapenems, GES enzymes are not well inhibited by clavulanic acid, and thus do not yield a clear synergy image, and the carbapenem hydrolysis is often weak in certain genetic backgrounds, which yield negative test results with assays based on carbapenem hydrolysis (such as the Carba NP, β-CARBA^™ tests, Star BL, Bruker and rCIM).^3,15 Molecular detection of GES-type genes is often done using homemade PCR or by WGS.^8,15–17 Only two commercially available PCR assays, Amplidiag CarbaR+ MCR Kit (Hologic, Villepinte, France) and the EasyScreen^TM ESBL/CPO Detection Kit (Genetic Signatures, Newtown, Australia), include GES enzymes in their targets.¹⁸ But these assays require specific equipment and trained personnel and are costly.^18,19 Rapid lateral flow immunoassay (LFIA)-based diagnostic tests have been successfully used to detect CTX-M-like (NG-Test^® CTX-M MULTI; NG-Biotech, Guipry, France) and the five main carbapenemase enzymes (NG-Test^® CARBA-5; NG-Biotech).²⁰

The aim of this study was to develop and evaluate the analytical performances of a prototype LFIA capable of detecting GES-type β-lactamases on a collection of GN isolates with well-characterized β-lactam resistance mechanisms.

Materials and methods

Ethics statement

All experiments were performed in compliance with French and European regulations on the care of laboratory animals [European Community (EC) Directive 86/609, French Law 2001-486, 6 June 2001] and with the agreements of the ethics committee of the Commissariat à l’Energie Atomique (CEtEA ‘Comité d’Ethique en Expérimentation Animale’ n° 44) no. 12-026 and 15-055 delivered to S. Simon by the French Veterinary Services and CEA agreement D-91-272-106 from the Veterinary Inspection Department of Essonne (France).

β-Lactamase purification

The bla_GES-5 gene fragment corresponding to the mature β-lactamase was cloned into the expression vector pET41b(+) (Novagen, VWR International, Fontenay-sous-Bois, France).²¹ GES-5 was expressed and purified with the hexa-histidine tag on its C-terminus in one-step pseudo-affinity chromatography using an NTA-Ni column (GE Healthcare, Les Ulis, France) as previously described.²¹

mAb and assay development

Biozzi mice (10 weeks old) were immunized with the recombinant GES-5 β-lactamase.²¹ Twenty mAbs were tested and the pairs of antibodies showing the best limit of detection for GES-5-expressing bacteria were selected, produced on a large scale and provided to NG-Biotech (Guipry, France) for the development of the GES LFIA as previously described.²⁰ An isolated colony was suspended in five drops (∼150 µL) of the extraction buffer (lysis step) and vortexed for a few seconds. Subsequently, 100 µL of this extract was dispensed on the cassette and the migration was allowed for 15 min. The results were eye read by monitoring the appearance of a red band on the test line, along with a band corresponding to the internal control (Figure 1).

Bacterial strains

To evaluate the GES LFIA, 103 GN bacterial isolates with PCR-characterized β-lactamase content were used. This collection included 54 Enterobacterales, 30 P. aeruginosa and 19 A. baumannii isolates, harbouring single or multiple β-lactamase genes: carbapenemases, ESBLs, and plasmid- or chromosome-encoded AmpCs grown on Mueller–Hinton (MH) agar plates. Among these isolates 13 produced GES-ESBLs and 15 GES-carbapenemases. GES-1, -2, -5, -6, -7, -9, -11, -12 and -14 alleles were evaluated (Table 1). The limit of detection was determined as previously described.²¹ MICs were determined using frozen 96-well broth microdilution panels with preloaded antibiotic growth medium supplied by International Health Management Associates, Inc. (IHMA; Schaumburg, IL, USA). Cefiderocol was tested in iron-depleted cation-adjusted MH broth as recently approved by the CLSI (https://clsi.org/standards/products/microbiology/documents/m100/), whereas comparators were tested in CAMHB. MICs were interpreted as previously indicated.²²

Culture media tested

Four representative isolates (K. pneumoniae GES-1 D6R4, P. aeruginosa GES-5 B2O58, A. baumannii GES-14 E10O42 and K. pneumoniae CTX-M-15 16.29) were grown on six commonly used culture media: MH agar plate, URISelect^TM 4 (Uri4; Biorad, Marne-la-Coquette, France), Columbia agar plus 5% horse blood (COH), ChromID^TM ESBL agar, ChromID^TM CarbaSmart and Drigalski agar (bioMérieux, Marcy l’Etoile, France).

Results

Evaluation of NG-Test^® GES on reference isolates grown on MH agar plates

The GES LFIA was able to detect all the tested GES variants (GES-1, -2, -3, -5, -6, -7, -9, -11. -12 and -14) with a 100% sensitivity and specificity, on pure isolates grown on agar plates (Table 1). However, it was not able to distinguish between GES-ESBLs and GES-carbapenemases. Non-targeted oxacillinases, such as OXA-143-like, OXA-48-like or OXA-1/2/9/10/35-like enzymes, or other resistance mechanisms such as cephalosporinases or ESBLs, gave negative results in every species tested. Finally, WT isolates (Enterobacterales, P. aeruginosa and A. baumannii) also gave negative results using the GES LFIA. The limit of detection of the assay was 10⁶ cfu per test with bacteria grown on MH agar plates (data not shown).

Evaluation of the GES LFIA on different culture media

The four representative isolates were grown on six of the most used media for bacterial growth (Table S1, available as Supplementary data at JAC Online). Some media currently used for the identification and/or selection of ESBL- or carbapenemase-expressing strains generate colonies with genus-specific colours (blue, green, pink or dark purple on Uri-4 plates, for example). These coloured colonies, once suspended in the extraction buffer, stained the latter in a similar manner. This staining did not interfere with the results. The three GES-expressing strains gave positive results whereas the CTX-M-15 producer gave a negative result. Colony staining did not change the appearance of the nitrocellulose membrane and still yielded easily interpretable results (data not shown).

Discussion

The production of carbapenemases and ESBLs among Enterobacterales, Pseudomonas spp. and Acinetobacter spp. has become a signature for MDR. The rapid and effective detection of antibiotic-resistant bacteria is a critical step for antibiotic stewardship and infection control. Despite technological improvements, the identification of pathogenic bacteria, as well as the detection of antibiotic resistance, remains complex and time-consuming, with time to results often >24 h.^3,15,16 LFIAs have proven to be useful as easy, rapid and reliable confirmatory tests for detection of the five main carbapenemases in GNs.^19,20 However, most commercially available assays detect only the five main carbapenemases. Yet minor carbapenemases are increasingly isolated and often responsible for outbreaks, and are not detected by commercially available assays.^8,9

In this study we showed that the GES LFIA may be used for the detection of GES-type enzymes in Enterobacterales, P. aeruginosa and A. baumannii without any specific equipment and reduced hands-on time. Indeed, the fact that it can be stored and performed at room temperature opens up the opportunity of reaching laboratories with limited equipment such as those in low-incomes countries. Combining it with immunochromatographic assays targeting the five main carbapenemases (KPC, NDM, VIM, IMP and OXA-48) would allow detection of a larger panel of enzymes especially in areas where GES enzymes are more prevalent. Considering the French epidemiology of CP-P. aeruginosa, the NG-Test^® CARBA 5 could detect 76% of carbapenemases (VIM, NDM, IMP and KPC),^19,22 but by combining it with GES LFIA, nearly 96% of carbapenemase producers would be detected.⁷ The limit of detection (10⁶ cfu/test), which is comparable to those of other LFIAs detecting β-lactamases, is sufficient for efficient GES detection from bacterial cultures.^19,20

The GES LFIA is not able to discriminate between GES-ESBLs and GES-carbapenemases. Detection of both enzymes is equally important, because GES enzymes are still rare and their detection as an ESBL or a carbapenemase remains important. Indeed, it has been demonstrated that antibiotic pressure with not only imipenem but also cefotaxime or aztreonam was able to select resistance to imipenem with a low susceptibility to commercially available inhibitors (clavulanate, tazobactam), with GES-producing enzymes.²³ The emergence of true carbapenem-hydrolysing class A enzymes foreshadows a dark future on the overall picture of antibiotic resistance, especially with the high rates of ESBL-producing enterobacterial isolates worldwide, and the increasing use of carbapenems to treat infections with these bacteria. In addition, recent observations made by the F-NRC revealed that GES-producing P. aeruginosa often have MICs for ceftazidime/avibactam close to or above the susceptibility breakpoints (disc diameter 17 mm), thus urging the need for precise MIC determination with GES-producing P. aeruginosa before treatment may be given.⁷ In our study, all 10 GES-producing P. aeruginosa had MICs of ceftazidime/avibactam ranging from 8 mg/L to >64 mg/L with an MIC₅₀ of 32 mg/L (Table 1). Similarly, GES-producing Enterobacterales, P. aeruginosa and A. baumannii tended to have higher MICs of cefiderocol, with a MIC₅₀ of 2 mg/L, 0.5 mg/L and 1 mg/L, respectively (Table 1).

A limitation of our study is that we could only evaluate 10 of the 53 known GES variants. However, these 10 variants are the most prevalent ones, especially GES-5; further studies including more variants are necessary to determine the performance of the GES LFIA on minor GES variants. Finally, testing on biological matrices, such as blood cultures, urines or rectal swabs, is also necessary.

Conclusions

The GES LFIA is efficient, rapid (results in  ≤15 min) and user friendly. It can be used with bacteria grown on most standard agar plates used in clinical microbiology laboratories. The GES LFIA detected all the GES-expressing GN strains without ambiguous interpretation. This assay is an ideal companion assay for immunochromatographic assays targeting the five main carbapenemases (KPC, NDM, VIM, IMP and OXA-48), in areas where GES enzymes are more prevalent, and for non-fermenters, especially P. aeruginosa.

Acknowledgements

We acknowledge Elodie CRETON for technical assistance.

Funding

This work was partially supported by the Assistance Publique-Hôpitaux de Paris, the Université Paris-Saclay, the Institut National de la Santé et de la Recherche Médicale (INSERM), and grants from the French National Research Agency for the Laboratory of Excellence in Research on Medication and Innovative Therapeutics (LERMIT) (ANR-10-LABX-33) and Innovantibio project (ANR-17-ASTR-0018-03) and from EIT Health for the project AMR Detectool.

Transparency declarations

L.N., C.L., L.F., A.C. and A.V. are employees of NG-Biotech. They were involved in LFIA strip development, but not in the validation or in the interpretation of the test results. All other authors: none to declare.

Supplementary data

Table S1 is available as Supplementary data at JAC Online.

References