Imprudently using antibiotics in medicine and farming has resulted in increasingly severe bacterial resistance problems around the world (Fu et al., 2022). According to the World Health Organization (WHO), antibiotic resistance remarkably threatens food security, and global health and development (World Health Organization, 2022). Multidrug resistance (MDR) occurs when bacteria obtain resistance to three or more categories of antibiotics, causing them to pose more urgent threats (Cheng et al., 2019). Among the currently available antibiotics, carbapenems, colistin as well as tigecycline are last-resort antibiotics for managing MDR bacterial infections, especially those caused by Enterobacteriaceae (Tang et al., 2021). Unfortunately, the worrying emergence of the predominantly plasmid-borne colistin resistance gene mcr-1, carbapenem resistance gene bla_NDM-1 as well as tigecycline resistance gene tet(X4) in Enterobacteriaceae pathogens from different sources, such as animals, food and humans, has increasingly been reported in different continents (Singh et al., 2018; Zhong et al., 2019; Anyanwu et al., 2022; Lu et al., 2022; Zhang et al., 2022). Additionally, tet(X) genes have regularly been reported to coexist with bla_NDM-1 and/or mcr-1 genes (Chen et al., 2019; Sun C. et al., 2019), and such plasmid-borne antibiotic resistance genes (ARGs) are capable of transferring between epidemic strains of Enterobacteriaceae (Leshaba et al., 2022). Thus, a fast, efficient and accurate detection method for simultaneously screening and monitoring the above-mentioned ARGs in Enterobacteriaceae pathogens has become urgently needed for managing the dissemination of resistance in food animal and human populations throughout the food chain and relevant environments.

The most widely used molecular-based method for MDR gene detection is polymerase chain reaction (PCR), including conventional PCR (Adekunle et al., 2021; Cooper et al., 2021), real-time quantitative PCR (Park et al., 2021; Tolosi et al., 2021) and digital PCR (Xu et al., 2021; Merino et al., 2022). However, most farms in China are equipped with laboratories capable of isolating bacteria from environmental samples, but are generally not equipped with complex large-scale instruments such as qPCR (Ma et al., 2020). A method which is simpler to operate and less dependent on instruments is urgently needed. To date, researchers have developed novel simple, rapid and efficient and cost-effective molecular biology techniques, such as isothermal nucleic acid amplification, as alternative protocols for the rapid on-site evaluation of MDR genes (Wong et al., 2018; Li et al., 2019; Tran et al., 2022). Recombinase polymerase amplification (RPA) acts as a common nucleic acid isothermal amplification technologies first developed by Piepenburg et al. (2006). It adopts a phage recombinase for forming complexes with oligonucleotide primers, thereby facilitating oligonucleotide primers to bind to homologous sequences of the double-stranded DNA molecules (Lobato and O'Sullivan, 2018). It is allowed to accomplish amplification within 30 min at a constant low temperature around 37°C by combining with a single-stranded DNA binding protein and a strand-displacing polymerase (Lobato and O'Sullivan, 2018). In addition, only a few copies of target DNA are initially needed, and highly specific DNA amplification at detectable levels can be achieved within a short time (Kersting et al., 2014). Furthermore, multiplexing RPA amplifications in the same solution is feasible, despite its strong dependence on the amplicon size, target sequences, as well as primer design (Kersting et al., 2014). Detection of RPA products is mainly completed via probe-based florescence (Kim and Lee, 2016), gel electrophoresis (Cao et al., 2018), or visualization by nucleic acid lateral flow dipstick (LFD) immunoassays (Xu et al., 2018). Among these, LFD immunoassays are simple, fast (~5 min), cost-effective and accurate for detection of amplified products and are more suitable for point-of-care testing (Lin et al., 2022). Additionally, multiplex LFD immunoassays that analyze multiple targets simultaneously have emerged in several research fields. The most common solution is to prepare several test lines on one test strip (Di Nardo et al., 2021). For instance, Cavalera et al. reported a multi-target LFD immunoassay with two test lines that bound to several classes of immunoglobulins, which realized detection of total antibodies against SARS-CoV-2 (Cavalera et al., 2021). Such designs further reduce cost and increase throughput of detection, which is vital to the study of MDR genes. Recently, to meet increased accuracy requirements for quantitative detection, test strip readers (Li et al., 2019) and image software (Wang et al., 2021) have been employed to scan and evaluate the color signals of LFD.

The study focused on developing an RPA-LFD method for simultaneously identifying the highly prevalent genes mcr-1, bla_NDM-1 and tet(X4) in certain Enterobacteriaceae pathogens isolated from different food and animal fecal samples. In addition, a TSR-200 test strip reader (Allsheng Instruments Co., Ltd., Hangzhou, China) assisted in evaluating color signals to use LFD for accurate quantitative analysis. This allows for immediate monitoring of MDR genes in pathogenic bacteria in the field and will inform risk assessment and remediation strategies in a timely manner.

Colistin, carbapenems and tigecycline are considered last-resort antibiotics for defense against MDR bacterial infections (Tang et al., 2021). However, these antibiotics become less effective because their respective predominant resistance genes, mcr, bla_NDM and tet(X), emerge and spread in animals, food and humans (Singh et al., 2018; Zhong et al., 2019; Anyanwu et al., 2022; Lu et al., 2022; Zhang et al., 2022). Among these last-resort antibiotics, colistin was heavily used in food-producing animal industries in China for a long period of time until it was banned as a growth promoter in 2017 (Walsh and Wu, 2016). However, colistin-resistant Gram-negative pathogens, especially Enterobacteriaceae pathogens, still have high presence in poultry/livestock farm environments (Liu et al., 2016; Tang et al., 2022), and the plasmid-borne resistance gene mcr-1 is the main factor contributing to widespread colistin resistance (Arcilla et al., 2016). In addition, although the use of carbapenems, broad-spectrum β-lactam class antibiotics, is authorized only in clinics (Ahmad et al., 2019; Farhat and Khan, 2020; Nasser et al., 2020), their main resistance gene, bla_NDM-1, is still frequently identified in Enterobacteriaceae pathogens in poultry/livestock farm environments. This is because of the extensive application of all β-lactam antibiotics (except carbapenems) in the prophylaxis and growth promotion in food-producing animal industries over the last decade in China and have provided long-term selective pressure for the bla_NDM-1 gene (Shen et al., 2016; Shi et al., 2021). Similarly, although the animal husbandry field has never adopted tigecycline, a third-generation tetracycline, excessively using of first- and second-generation tetracyclines may have provided a selective pressure for the tet(X4) gene in poultry/livestock farm environments (Tang et al., 2021). Of note, a growing number of reports indicate that tet(X4) is often detected together with bla_NDM-1 genes or mcr-1 genes (Bai et al., 2019; Sun C. et al., 2019). Additionally, tet(X4)-positive strains also harboring both mcr-1 and bla_NDM genes have been found in food animal samples (Sun J. et al., 2019).

MDR Enterobacteriaceae pathogens in poultry/livestock farm environments are considered the most common reservoirs of the mcr-1, bla_NDM-1 and tet(X4) genes (Xiong et al., 2018; Xia et al., 2019; Shi et al., 2021). Besides, such ARGs may be spread along the food chain in poultry/livestock farm environments and meat production facilities, and finally be applied in community and even hospital environments (Wang et al., 2017; Shi et al., 2021). As resistance genes to last-resort antibiotics potentially threaten the food safety as well as public health, predominant ARGs such as mcr-1, bla_NDM-1 and tet(X4) in their common Enterobacteriaceae pathogen hosts should be more closely monitored along the food chain worldwide over the long-term (Shi et al., 2021). Therefore, it is in urgent need to develop a rapid, and accurate detection method to simultaneously screen and monitor multiple resistance genes to last-resort antibiotics in the field more effectively.

Multiple nucleic acid detection methods, including multiplex PCR, multiplex loop-mediated isothermal amplification (LAMP) and multiplex RPA assays, can achieve rapid, accurate and simultaneous amplification of multiple resistance genes to last-resort antibiotics in a single reaction. However, multiplex PCR assays require thermal cycling steps and specialized instruments, giving them limited application in the field (Ma et al., 2020). Similarly, multiplex LAMP assays require more target-specific primers (normally four to six) for amplification under isothermal conditions between 60 and 65°C in 40 to 60 min (Fang et al., 2018). Compared with these two methods, multiplex RPA assays require fewer target-specific primers (two primers or two primers with one probe) for amplification, lower incubation temperatures (25–43°C) and shorter incubation times (less than 30 min) and have a high tolerance to sample impurities (Lillis et al., 2014; Ma et al., 2020). Thus, in our study, a fast, innovative boil-up method for crude DNA extraction (taking around 10 min for the entire extraction process) from Enterobacteriaceae pathogens was used to increase the time savings of the multiplex RPA procedure. Studies have confirmed the application of multiplex RPA assays for simultaneously detecting multiple foodborne pathogens (Ma et al., 2020; Jin et al., 2022). However, studies on its detection for ARG remain less. Our study focused on explaining certain and quantitative multiplex RPA-LFD method to simultaneously identify multiple resistance genes to last-resort antibiotics in certain Enterobacteriaceae pathogens when the temperature remained unchanged and no specialized instruments were used, which has great potential for applications in the field.

For multiplex RPA assays for ARG detection, primer/probe concentration, incubation temperature and time and magnesium-acetate concentration are the most vital extrinsic factors contributing to the assay’s maximum efficiency (Hu et al., 2019; Ma et al., 2020). Our study first adjusted the primer/probe concentration to 0.30 μM/0.06 μM for the mcr-1 gene, 0.42 μM/0.12 μM for the bla_NDM-1 gene and 0.30 μM/0.09 μM for the tet(X4) gene with 14 mM magnesium-acetate for obtaining an equivalent amplification for three ARGs. Second, we set the optimal time at 20 min and the temperature conditions at 37°C. This reaction time is significantly faster than that of multiplex PCR (around 4 h) and multiplex LAMP (40–60 min) assays. Furthermore, it is allowed to complete the RPA reaction in a heating block or water bath, even at ambient temperature or body temperature (Lillis et al., 2014; Wang et al., 2017).

Experiments confirm LFD as a faster and simpler tool for detecting RPA product, featuring an antibody labeled with antigen-specific gold nanoparticles, thereby ensuring that it is not needed to clean up nucleotide. The resulting visualization can be observed within 5 min (Wu et al., 2017). To achieve triplex detection in our study, we labeled each of the LF probes with biotin, Cy5 and TAMRA at the 5′ end, with a C3 spacer at the 3′ end and modified inside the LF probe with THF. We labeled all the reverse primers with digoxin at the 5′ end. We finally obtained 3 detectable double-labeled amplicons. Besides, the preparation of LFDs with three test lines served for RPA amplicon visualization (Ma et al., 2020). In general, LFD is a qualitative or semi-quantitative detection method, but the current test strip reader, which combines facile colorimetric readouts with test strips, can realize quantitative detection (Wang et al., 2022). In our study, sensitivity analyses showed a close association of color signals with template concentrations of mcr-1, bla_NDM-1 and tet(X4) genes in Enterobacteriaceae pathogens (R² = 0.9881, R² = 0.9745, and R² = 0.9807, respectively). Use of the test strip reader also eliminates errors associated with assessment by eye.

Sensitivity and specificity analyses verified that RPA-LFD assays were practical for multiple resistance genes to last-resort antibiotics detection in Enterobacteriaceae pathogens. In our study, this assay simultaneously detected as few as 10¹ copies/μl of mcr-1, bla_NDM-1 and tet(X4) genes in Enterobacteriaceae pathogens, with a sensitivity 10-fold higher than that of PCR (10² copies/μl). Furthermore, such ARGs were specifically detected in different Enterobacteriaceae pathogens. These results showed that RPA-LFD is a suitable method for detecting multiple resistance genes to last-resort antibiotics in foodborne pathogens and has potential applications in the field.

In conclusion, the multiplex RPA-LFD method is fast, sensitive, specific and user-friendly for direct detection of the mcr-1, bla_NDM-1 as well as tet(X4) genes of Enterobacteriaceae pathogens and does not require expensive equipment. It can provide immediate monitoring results of multiple resistance genes to last-resort antibiotics in foodborne pathogens in the field in a timely manner. On this basis, we will further study the pretreatment technology of the complex sample, and establish the RPA-LFD method for direct analysis of ARGs in water, soil and feces and other environmental samples.

The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding authors.

CL took charge of original draft writing, data curation, and review and editing. JW was responsible for original draft writing, data curation, and software. XG, WL, DC, CZ, QY, CX, and BT together took charge of data curation and software. LP, YT, and PL acquired the funding. GH also participated in review and editing. JF and HJ participated in funding acquisition, methodology, supervision, and review and editing. All authors contributed to the article and approved the submitted version.

This research was completed with the support of the Public Welfare Technology Application Research Project of Zhejiang Province (No. LGN22C200013), the Public Welfare Technology Application Research Project of Ningbo (No. 2022S014), the Major Science and Technology Project of Xiaoshan District (No. 2021226), the Open Fund of the Key Laboratory of Biosafety detection for Zhejiang Market Regulation (No. 2022BS002), Science and Technology Innovation Activity Plan for College Students in Zhejiang Province (No. 2022R409A057 and 2022R409A050), the National Science Foundation of China (Nos. 31901792 and 31801655), and the National Program on Key Research Project of China (No. 2019YFE0103900).

LP was employed by Zhejiang Hongzheng Testing Co., Ltd, Ningbo, Zhejiang, China. XG was employed by Zhejiang Gongzheng Testing Center Co., Ltd, Hangzhou, Zhejiang, China. YT was employed by Hangzhou Tiannie Technology Co., Ltd, Hangzhou, Zhejiang, China.

The remaining 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.