A Loop-Mediated Isothermal Amplification Assay Targeting Neisseria gonorrhoeae penA-60.001

INTRODUCTION

Gonorrhea is one of the major sexually transmitted infectious diseases, affecting about 78 million people worldwide (1). Treatment of gonorrhea is severely compromised by the rapid emergence and spread of resistance to the available antimicrobials (1). Although ceftriaxone (CRO) is a key first-line antimicrobial treatment for uncomplicated anogenital and pharyngeal gonorrhea, used at a high dose alone (1 g, intravenously) or in combination with azithromycin (AZM), several resistant Neisseria gonorrhoeae strains have been identified since 2009 (2, 3).

N. gonorrhoeae FC428, which has a CRO resistance penA-60.001 allele, was identified in 2015 (4) and has now expanded worldwide (5–8). The 3′-terminal half of the penA-60.001 nucleotide sequence is almost identical to that of the penA genes in CRO-resistant Neisseria cinerea strains, while the 5′-terminal half of penA-60.001 has sequence similarity to CRO-sensitive N. gonorrhoeae penA (9).

The optimal intervention to contain the spread of an antibiotic-resistant pathogen requires rapid identification of antibiotic susceptibility and appropriate treatment. It is important to identify CRO-resistant N. gonorrhoeae strains carrying penA-60.001 because these strains have been spreading worldwide (5–8). Double-resistant N. gonorrhoeae strains, carrying penA-60.001 and high-level AZM resistance, have been reported (6, 7). Real-time PCR assays have been developed to detect penA-60.001 (10, 11). Although real-time PCR assays are accurate and sensitive, they require expensive equipment and are usually performed in laboratories. Thus, a loop-mediated isothermal amplification (LAMP) method that does not require expensive equipment might be suitable for point-of-care tests in clinics. The LAMP method has emerged as a useful alternative for detection of pathogenic microorganism genes in a variety of clinical specimens (12, 13). In addition, single-nucleotide polymorphism (SNP) genotype analysis can be carried out using an amplification-refractory mutation system (ARMS) (14, 15). Here, we report the development of a LAMP assay combined with an ARMS strategy to detect CRO-resistant gonococcal genomic DNA.

We have reported that penA-60.001 has significant sequence similarity to some commensal N. cinerea penA genes (9, 11). Therefore, we developed two independent LAMP assays (described in the supplemental material). One assay amplified a unique region of the N. gonorrhoeae penA gene (NG-penA-LAMP1), and the other assay specifically amplified the penA-60.001 allele (penA-LAMP2) (Table 1). Positive results from the two assays indicated the presence of N. gonorrhoeae penA-60.001.

We designed a primer set for NG-penA-LAMP1 assays that amplified only the N. gonorrhoeae penA gene. The 5′-terminal half of the N. gonorrhoeae penA sequence is similar to that of penA-60.001 but is significantly different from that of penA in other Neisseria spp. (9, 11). Therefore, from a BLAST search of the GenBank database, we selected a N. gonorrhoeae sequence in the penA 5′-terminal half, i.e., nucleotides 65 to 266 from the start codon. Consequently, we created a primer set for the NG-penA-LAMP1 assays to differentiate N. gonorrhoeae penA from other Neisseria penA genes (see Fig. S1A in the supplemental material).

The penA-60.001 allele contains mutations, including A311V (GCC to GTC) in penicillin-binding protein 2 (PBP2), which is one of the critical mutations involved in the CRO resistance of N. gonorrhoeae (16). This SNP (i.e., thymidine at nucleotide 932) is not located within penA sequences in CRO-sensitive N. gonorrhoeae strains (4, 17). Therefore, to specifically amplify penA-60.001 by utilizing this SNP, a primer set (penA-LAMP2) was selected from the region containing this thymidine (Fig. S1B).

We initially evaluated the NG-penA-LAMP1 and penA-LAMP2 primer sets using 10-ng genomic DNA samples from 14 World Health Organization (WHO) gonococcal reference strains (i.e., F, G, K, L, M, N, O, P, U, V, W, X, Y, and Z) and the N. gonorrhoeae strains GU140106 (18) and FC428 (4). The WHO reference strains have different CRO MIC values and penA alleles (17). Positive results were observed for all of the strains described above with the NG-penA-LAMP1 primer set but were observed for only strain FC428 with the penA-LAMP2 primer set.

We next studied the sensitivity and specificity of our assay system using genomic DNA (10 ng/reaction) from 204 N. gonorrhoeae strains and 95 strains of other Neisseria spp. (Table 1), prepared as described in the supplemental material. These strains were isolated during our previous studies (11, 19). DNA amplification using the NG-penA-LAMP1 primer set was observed only with N. gonorrhoeae strains. However, a positive reaction using the penA-LAMP2 primer set was detected only with genomic DNA of strains carrying the mosaic penA-60.001 allele, i.e., strains FC428, FC460, and FC498 (4, 8, 19). In addition, mosaic penA alleles other than penA-60.001 were carried by 102 of 204 N. gonorrhoeae strains examined in this study; 82 strains carried penA-10.001 and its derivatives, and 20 strains carried penA-34.001 and its derivatives (19). N. gonorrhoeae strains carrying these penA types often exhibit reduced susceptibility to cephalosporins (20). Specific detection of A311V using the penA-LAMP2 primer set was dependent on the B1 region (Fig. S1B) of the penA-60.001-BIP primer in penA-LAMP2 assays (Table S1). Among Neisseria spp. other than N. gonorrhoeae in this study, 9 strains, whose penA sequences had been determined in a previous study (9, 11), had penA-60.001-like alleles encoding valine, corresponding to A311V as in N. cinerea strains AM1601 and SH43-3 (9, 11); the B1 region of all 9 alleles was identical to that of penA-60.001. However, negative penA-LAMP2 amplification results were obtained for all Neisseria spp., including the 9 isolates with penA-60.001-like alleles. Other primers in the penA-LAMP2 set might contribute to the specific amplification of penA-60.001 with the penA-LAMP2 set. Consequently, combined NG-penA-LAMP1 and penA-LAMP2 assays have high sensitivity (100%) and specificity (100%) for identifying N. gonorrhoeae genomic DNA carrying penA-60.001.

The detection limit of this assay was determined using different amounts of strain FC428 genomic DNA (1.0 to 1.0 × 10⁶ genome copies) as the template. Both NG-penA-LAMP1 and penA-LAMP2 could amplify DNA using more than 1.0 × 10⁴ genome copies per reaction.

The detection limit was also determined using human urine from healthy volunteers spiked with various concentrations of cultured strain FC428. After being boiled at 95°C for 5 min, human urine samples were centrifuged at 200 × g, and the supernatant was used for the spiking assay (21). A bacterial culture that had been boiled at 95°C for 5 min was added to the supernatant, followed by an assay using NG-penA-LAMP1 and penA-LAMP2. CFU were estimated from data demonstrating that an optical density at 600 nm of 0.4 corresponded to a bacterial density of approximately 1.0 × 10⁸ CFU/ml. Minimum levels of 1.0 × 10⁴ and 1.0 × 10⁵ CFU per reaction were detectable using NG-penA-LAMP1 and penA-LAMP2, respectively.

Furthermore, direct detection from clinical specimens was examined. Seven N. gonorrhoeae culture-positive specimens and 5 negative specimens corresponding to suspected gonococcal infections were used. Each urethritis swab was suspended in 200 μl of Tris-EDTA buffer. After being boiled for 5 min, the samples were centrifuged at 9,000 × g, and the supernatant was kept on ice. For the LAMP reaction, 2 μl of supernatant was used. The supernatants from the 7 culture-positive samples were positive only in the NG-penA-LAMP1 assay, because none of the strains carried the penA-60.001 gene. The supernatants from the 5 culture-negative samples were negative in both assays. The culture-positive samples with N. gonorrhoeae carrying the penA-60.001 gene were expected to show a positive reaction in the penA-LAMP2 assay, considering that NG-penA-LAMP1 and penA-LAMP2 show the same sensitivities. This study was approved by the Institutional Review Board of the National Institute of Infectious Diseases (approval number 993).

The method reported here used two independent LAMP assays to enable the detection of N. gonorrhoeae strains carrying penA-60.001. In validation experiments using a number of Neisseria species isolates, false-positive reactions were not found using the combination of the independent NG-penA-LAMP1 and penA-LAMP2 assays. It is noteworthy that the penA-LAMP2 primer set, which was designed to detect a specific penA-60.001 sequence from N. cinerea (9), could specifically produce amplicons from penA-60.001 of CRO-resistant N. gonorrhoeae FC428-like strains. However, we could not completely exclude the possibility of a cross-reaction with an unidentified penA gene in Neisseria spp., particularly when clinical specimens from the pharynx were being tested, because Neisseria spp. in the oropharyngeal flora can be highly diverse (22), although our combined assay system reported here was highly specific.

The bacterial loads in gonococcal urethritis specimens from patients with symptomatic and asymptomatic gonococcal urethritis have been reported to be 3.7 × 10⁶ copies per swab and 2.0 × 10⁵ copies per swab, respectively (23). The detection sensitivities of the NG-penA-LAMP1 and penA-LAMP2 systems in assays of FC428-spiked urine were 1.0 × 10⁴ and 1.0 × 10⁵ CFU per reaction, respectively. Assuming similar bacterial loads in specimens from actual clinical samples, our assay systems would be useful for analyzing clinical specimens, especially urethritis specimens. The detection limit for this LAMP method was higher than those for real-time PCR methods, which have limits of 3 to 10 genome copies (10, 11), but LAMP assays can be performed in one step using heat-resistant strand-displacement DNA polymerase, without DNA denaturation (12, 13). The results of the developed LAMP assay were obtained within approximately 60 min, whereas real-time PCR assays require approximately 100 min (11). Furthermore, the LAMP assay does not require expensive dedicated devices and fluorescent probes, which are usually required for real-time PCR assays. LAMP assays can be conducted in resource-limited laboratories in which a water bath is available (24). Many nucleic acid amplification test (NAAT) applications have been reported for point-of-care testing for gonococcal infections (25). To improve the applicability of the LAMP assays reported here, validation of this assay method using actual clinical specimens will be required.

In conclusion, in this study, a novel LAMP method was developed for the specific detection of gonococcal strains carrying penA-60.001. We designed two LAMP assays, one for the specific detection of conserved N. gonorrhoeae penA sequences and the other for unique penA-60.001 sequences. In combination these two highly specific detection assays decreased false-positive results, enabling the specific detection of CRO-resistant gonococcal strains. Therefore, we expect the assay method reported here to contribute to enhanced surveillance of N. gonorrhoeae antimicrobial resistance.