Rapid detection of a novel B1-β-lactamase gene, blaAFM-1 using a loop-mediated isothermal amplification (LAMP) assay

BlaAFM-1 (GenBank Accession No. 143105.1) is a new B1 subclass metallo-β-lactamase gene discovered by our group, and isolated from an Alcaligenes faecalis plasmid that renders carbapenem antibiotics ineffective. In this study, we generated a fast and reliable assay for blaAFM-1 detection. We designed optimum loop-mediated isothermal amplification (LAMP) primers and constructed a recombinant plasmid AFM-1 to specifically detect blaAFM-1. Optimal LAMP primers were used to assess sensitivity of the recombinant plasmid AFM-1 and blaAFM-1-supplemented samples (simulated sputum and simulated feces). Fifty two samples, without blaAFM-1, were used to assess LAMP real-time assay specificity; these samples were verified by conventional PCR and sequencing for the absence of blaAFM-1. Three hundred clinical Gram-negative carbapenem-resistant strains were tested by LAMP assay for strains carrying blaAFM-1, which were confirmed by conventional PCR and Sanger sequencing. We calculated the sensitivity and its 95% confidence interval (95% CI), specificity and its 95% CI, and predictive values of the LAMP assay and conventional PCR/sequencing by investigating positive and negative clinical strains. The lowest limit of detection for the recombinant plasmid AFM-1 and blaAFM-1-supplemented samples (in both simulated sputum and simulated feces) was 101 copies/reaction. All amplification curves of the 52 blaAFM-1-free bacteria strains were negative, suggesting the LAMP assay had excellent specificity for detecting blaAFM-1. Among the 300 clinical strains, eight were positive for blaAFM-1 using LAMP. These LAMP results were consistent with conventional PCR and Sanger sequencing data. As with conventional PCR/sequencing, the LAMP method exhibits 100% sensitivity (95% CI 59.8–100%) and 100% specificity (95% CI 98.4–100%) for blaAFM-1 detection. The LAMP assay is also time-efficient (1 h) for blaAFM-1 detection. We established a new LAMP assay with high sensitivity and specificity to detect the novel B1-β-lactamase gene, blaAFM-1.

In this study, a novel B1-β-lactamase gene blaAFM-1 (GenBank Accession No. 143105.1) was identified by our group on a plasmid from Alcaligenes faecalis. When compared with nucleotide and amino acid homologies of other B1-β-lactamase genes, the highest comparative identities were 85% and 87%, respectively. Conjugation assays indicated that blaAFM-1 is transferred from the donor strain (A. faecalis) to the recipient strain (Escherichia coli). Moreover, the Carba NP test also showed that carbapenemase AFM-1 could hydrolyze carbapenems. More interestingly, several strains, i.e., Comamonas testosteroni, Stenotrophomonas maltophilia, Bordetella trematum, and Comamonas aquatica were shown to harbor blaAFM-1 in clinical carbapenem-resistant Gram-negative strains.
Loop-mediated isothermal amplification (LAMP) uses six specific primer sets for six different regions of a target sequence. The Bst polymerase, with strand displacement activity and nucleic acid amplification functions, is used to perform nucleic acid amplifications under constant temperature conditions [9][10][11]. In this study, we established and evaluated a fast and reliable assay for detecting blaAFM-1 using the LAMP assay.

Primer design
The blaAFM-1 sequence was downloaded from GenBank (GenBank Accession No. 143105.1). To determine the optimal primer set, several sets of primers were designed by Primer Explorer V4 software (http:// prime rexpl orer. jp/e/). Each set consisted of six primer sequences, one pair of inner primers (BIP and FIP), one outer pair (F3 and B3), and one pair of loop primers (LF and LB). We selected the optimal LAMP primer set, with the highest amplification efficiency. Primers were designed by Guangzhou Bangce Biotechnology Co., Ltd (Guangzhou, China), and synthesized by Sangon Biotech Co., Ltd (Shanghai, China) ( Table 1).

Construction of the recombinant plasmid AFM-1
BlaAFM-1 was chemically synthesized and ligated into the pET28-a(+) vector (Beijing Liuhe Huada Gene Technology Co., Ltd, Beijing, China) to construct the recombinant plasmid AFM-1 (Beijing Liuhe Huada Gene Technology Co., Ltd). Next, the plasmid was transformed into competent E. coli Top10 cells (Beijing Liuhe Huada Gene Technology Co., Ltd), and plated onto Luria Broth agar containing 4 μg/mL meropenem for overnight incubation at 37 °C. The next day, plasmid DNA was extracted from colonies using the Qiagen bacterial plasmid extraction kit (TianKangxin (Beijing) Technology Co., Ltd. Beijing, China), and the primers pET-F/R (Sangon Biotech (Shanghai) Co., Ltd) ( Table 2) were used to PCR amplify the recombinant plasmid AFM-1. The plasmid was sequenced to verify the correct blaAFM-1 insert.
The turbidity of A. faecalis AN70 strain bacterial solution expressing blaAFM-1 was adjusted to 0.5 McFarland units, and diluted with 0.9% NaCl solution (Sangon Biotech). This solution was mixed with the AN70 strain suspension. This solution was then mixed with blaAFM-1-free sputum and blaAFM-1-free feces samples, respectively, to generate blaAFM-1-supplemented sputum simulated sample and blaAFM-1-supplemented fecal simulated sample, respectively. Then genomic DNA was extracted with MagBeads sputum DNA Extraction Kit (Beijing Newcomb Technology Development Co., Ltd. Beijing, China) and MagBeads Feces DNA Extraction Kit (Beijing Newcomb Technology Development Co., Ltd.), respectively. BlaAFM-1-free sputum and blaAFM-1free feces acted as negative controls. The final simulated sample concentrations were serially diluted 10 5 , 10 4 , 10 3 , 10 2 , 10 1 , and 10 0 copies/reaction, from high to low. The recombinant plasmid AFM-1 and simulated samples were then used in the LAMP assay to assess sensitivity. We then compared and evaluated the sensitivity of the recombinant plasmid AFM-1 and simulated samples.

LAMP assay specificity
LAMP assay specificity was determined using 52 blaAFM-1-free strains (mainly collected from Nanfang hospital which is a tertiary comprehensive hospital), including six standard strains [obtained from National Center for Clinical Laboratories (Beijing, China)], 16 templates carrying the β-lactamase gene or carbapenemase gene, nine common Enterobacterales strains, six common non-fermenting strains, ten common Grampositive cocci, four common fungi, and a human leukocyte sample. Recombinant plasmid AFM-1 was used as a positive control (Additional file 1: Table S1). Strain DNA  was extracted using the Ezup column bacterial genomic DNA extraction kit (Sangon Biotech), and strains containing blaAFM-1 were confirmed by conventional PCR and sequencing. All strains were subjected to LAMP assay, under the same conditions, with specificity identified from amplification curves.

LAMP assay validation with clinical strains
Three hundred clinical Gram-negative carbapenemresistant strains were collected to validate the LAMP assay. All carbapenem-resistant Gram-negative strains were identified and antimicrobial susceptibility tests were performed using the BD phoenix 100 Automated Microbiology System (BD Diagnostics, Franklin Lakes, NJ, USA). DNA from strains was extracted using the Ezup column bacterial genomic DNA extraction kit (Sangon Biotech). The blaAFM-1 gene was detected by LAMP and conventional PCR/sequencing. LAMP and PCR reactions were processed and sequenced as previously described.
Recombinant plasmid AFM-1 and sterilized distilled water were used as positive and negative controls, respectively. We then compared differences between results.

Statistical analysis
Positive and negative numbers of the blaAFM-1 gene detected by the LAMP assay and conventional PCR/ sequencing were counted, respectively. A 2 × 2 diagnostic test characteristics of the LAMP assay against PCR/ sequencing table was drawn. Sensitivity, specificity and their 95% CIs, predictive values of clinical strains were calculated by these two methods.

Optimal LAMP primer set
Using the same reaction conditions, we screened our optimal primer sets for the highest amplification efficiency (Fig. 1a).

LAMP assay sensitivity
LAMP amplification curves indicated that the minimum limit of detection for blaAFM-1 from the recombinant plasmid AFM-1, two simulated (sputum simulated and feces simulated) samples were both 10 1 copies/reaction (Fig. 1b-d).

LAMP assay specificity
Due to LAMP specificity, except for recombinant plasmid AFM-1 which contained blaAFM-1 and generated S-shaped curves within 40 cycles, the other 52 samples without blaAFM-1 were not amplified. These data agreed with PCR amplification/sequencing data and indicated that the LAMP real-time assay exhibited high specificity for blaAFM-1.

Clinical strain validation
We used our LAMP assay to investigate 300 Gram-negative carbapenem-resistant strains; eight strains were identified as carrying blaAFM-1 (Fig. 2, Additional file 1: Table S2 for the information on positive strains), in agreement with conventional PCR results. Importantly, no false positive results were identified using the LAMP assay (Table 3). For clinical validation, when compared with conventional PCR/sequencing, the sensitivity, specificity, and predictive values of the LAMP method for detecting blaAFM-1 were 100% ( Table 3). The 95% CI for sensitivity was 59.8-100%, and 98.4-100% for specificity.

Discussion
MBLs hydrolyze almost all types of β-lactams (including meropenem), but not aztreonam [5,12]. Equally, lactamase inhibitors are also ineffective against MBLs [5,12]. Due to disease risks and limited therapies for carbapenemase-producing organisms, especially MBL-producing organisms, the rapid and early detection of carbapenemases is an urgent requirement. BlaAFM-1, as a new member of the subclass B1 carbapenemase gene family, was first discovered on a broad host conjugative IncW plasmid, which easily conferred blaAFM-1 to other bacteria, therefore, AFM-1 enzyme is likely a certain risk.
To the best of our knowledge, LAMP technologies have been extensively used for bacteria, virus, and parasites [11]. LAMP internal and outer primers can increase reaction specificity, and loop primers can shorten reaction times, and increase sensitivity. It has been reported that when a recombinant plasmid containing the target gene, LAMP sensitivity was generally 100-10,000 times that of the PCR method [13]. Also, LAMP amplification does not require temperature cycling, and crude unpurified DNA can be used as a template to detect blaAFM-1 by LAMP assay [11,14], which greatly shortens detection times. The LAMP assay takes 1 h to complete, is userfriendly, and simply requires the mixing of reaction solution, primers, Bst polymerase and templates in a constant temperature instrument. Due to strategic LAMP primer design, these primers not only detected blaAFM-1, but also distinguished other MBL genes, such as blaNDM-1; this gene displayed the highest identity (85%) with blaAFM-1, but was not amplified by the LAMP assay.
Notably, given the high sensitivity of the LAMP assay, it may also produce false positive results during reaction times due to high amplification efficiencies [11,15,16], To reduce false positives, aerosol pollution must be prevented during procedures, therefore paraffin oil addition will prevent this [11].
This study had several limitations. Firstly, the number of clinically verified strains was small and errors may have occurred during sensitivity and specificity assay  1 a Amplification curve of the optimal primer sets. b-d The sensitivity of recombinant plasmid AFM-1, a sputum-simulated sample, and a feces-simulated sample, respectively. Amplification curves were obtained by diluting recombinant plasmid AFM-1 DNA, sputum-simulated sample, and feces-simulated sample from 10 5 copies/reaction to 10 0 copies/reaction, respectively. Sterile distilled water was used as a negative control Fig. 2 Evaluation of the LAMP method with 300 clinical strains, eight strains (out of 300) generated S-shaped type amplification curves using LAMP. calculations, therefore larger sample sizes are required for comprehensive clinical verification. Secondly, when compared with conventional culture assays, LAMP is rapid and time-efficient. However, its application to clinical specimens remains to be thoroughly explored and verified. Lastly, the β-lactamase and carbapenemase genes, used for LAMP-specificity evaluation, were single drug resistance genes. The coexistence of two or more drug resistance genes were not evaluated for LAMP specificity, thus further investigations are needed in the future.

Conclusions
The LAMP method was used to detect blaAFM-1 for the first time. When compared with conventional PCR/ sequencing, our method was rapid, sensitive, and specific, and is suitable for laboratory applications.