Development and evaluation of a rapid visual loop-mediated isothermal amplification assay for the tcdA gene in Clostridioides difficile detection

Introduction

Clostridioides difficile (C. difficile) is the major cause of antibiotic-associated diarrhea worldwide (Rodríguez et al., 2020). Although the estimated national burden of C. difficile infection (CDI) and associated hospitalizations decreased from 2011 through 2017 in 10 states in the US, the number of C. difficile cases in the US remained as high as 15,512 in 2017, with an estimated total national burden of 462,100 cases and an estimated incidence of 144 cases per 100,000 population (Guh et al., 2020). In the Shandong and Zhejiang provinces, China, tertiary hospitals reported a consistent 14% incidence of CDI (72/504) among hospitalized patients with suspected CDI in Shandong and 14% (115/804) among acute gastroenteritis outpatients in a Zhejiang pediatric hospital (Luo et al., 2018; Shuai et al., 2020). These findings are consistent with other studies from mainland China, which report a 14% crude incidence of toxigenic C. difficile in diarrheal patients (Tang et al., 2016).

The primary virulence factors of C. difficile are two structurally similar toxins-toxin A and toxin B, which are encoded by the tcdA and tcdB genes respectively (Kuehne, Cartman & Minton, 2011). Most cases of CDI are attributed to strains expressing both toxins A and B (A+B+) (Drudy, Fanning & Kyne, 2007). Although the toxin A−B+ strain is uncommon, it can also cause disease and has been relevant to previous outbreaks of CDI (Alfa et al., 2000; Kuijper et al., 2001; Drudy, Fanning & Kyne, 2007). Earlier reports indicated that the toxin A+B− strain only rarely caused human diseases (Rupnik, 2008). As a result, numerous studies have emphasized the importance of toxin B in the pathogenesis of CDI, while downplaying the importance of toxin A. However, a study recently reported the discovery of clinical pathogenic C. difficile strains that produce high levels of toxin A but minimal or no toxin B, indicating that toxin A alone can cause CDI (Lin et al., 2020). This pattern of toxin production, observed in more than 5% of isolates, is consistently found both in vitro and in vivo in humans and mice (Lin et al., 2020). Furthermore, the production of either toxin A or toxin B by these isolates is sufficient to induce the full spectrum of CDI symptoms (Drudy et al., 2007; Freeman et al., 2010). Additionally, both toxins A and B can independently cause disease in animal models (Kuehne, Cartman & Minton, 2011; Kuehne et al., 2010).

Current laboratory tests for the diagnosis of toxin A in C. difficile strains include the C.difficile cytotoxin neutralization assay (CCNA), toxigenic culture (TC), toxin A enzyme immunoassay (EIA), glutamate dehydrogenase (GDH) assay, and nucleic acid amplification test (NAAT).

Although CCNA and TC remain the current gold standards, their use for routine clinical detection is challenging due to their requirements for harsh culture conditions, involvement of highly technical and complex operations, and time-consuming nature (Shah et al., 2020; Liu et al., 2021). Enzyme immunoassays are specific and rapid, but not sensitive (Nicholson & Donskey, 2023). The glutamate dehydrogenase assay is sensitive and rapid, but it has some disadvantages such as cross-reactivity, poor specificity, and a high false positive rate (Bartlett, 2010; Crobach et al., 2016). Assays for the tcdA gene by NAAT include polymerase chain reaction (PCR; Kim et al., 2022b), multiplex-PCR (Moosavian et al., 2022), quantitative real-time PCR (Brennhofer et al., 2022), and multiplex real-time PCR (Novakova et al., 2021). Despite the specificity and sensitivity of these diagnostic methods, their suitability for rapid detection in primary hospitals and on-site detection is limited due to their time-consuming and complex nature, as well as the requirement for expensive equipment. Thus, a rapid, simple, and cost-effective assay is needed to complement current PCR methods for detecting the tcdA gene.

Loop-mediated isothermal amplification (LAMP) is a powerful molecular technique for nucleic acid amplification. LAMP leverages the strand displacement activity of Bst DNA polymerase, which facilitates DNA amplification under isothermal conditions (Notomi et al., 2000; Ushikubo, 2004). The high amplification efficiency of this technique, capable of generating up to 10⁹ copies of target DNA within an hour, underscores its potential in rapid diagnostic applications. The excellent specificity of LAMP is attributed to its use of four (or six) primers, which can identify six (or eight) distinct regions on the target DNA or RNA (Notomi et al., 2000; Parida et al., 2008). Additionally, the detection limit of LAMP surpasses that of PCR, and the results can be visually interpreted without the need for sophisticated equipment. LAMP has been effectively used to detect various pathogens, including bacteria (Hong-Min et al., 2023), viruses (Nawab et al., 2024), parasites (Chen et al., 2023), and fungi (Badparva et al., 2022), as well as different toxin types (Norén et al., 2011; Pancholi et al., 2012), binary toxin genes (Yu et al., 2017), and resistant genes (Lin et al., 2015, 2022) of C.difficile.

This study designed three novel sets of LAMP primers and optimized LAMP for tcdA detection. To ascertain the specificity of the tcdA primer within the LAMP assay, 26 distinct pathogenic bacterial strains devoid of the tcdA gene were analyzed as negative controls. Primer sensitivity was assessed by conducting serial dilutions of C. difficile VPI10463 DNA. Finally, the study compared the consistency of LAMP and PCR methods in detecting the tcdA gene of C. difficile in 164 stool specimens.

Materials and Methods

Results

Optimal primers for rapid detection of tcdA

All three sets of primers designed herein (Table 2) produced turbidity after 26 min (Fig. 1). Primer set tcdA12 showed the fastest amplification (Fig. 1), so tcdA12 primers were selected for optimization. The reaction time of tcdA12 primers with additional loop primers (LB and LF) was less than one-half that of the tcdA8 primers without loop primers.

Appropriate temperature for tcdA LAMP reaction

The tcdA12 primer set was evaluated across a temperature range of 58 °C to 69 °C at intervals of 1 °C. The ideal temperature range for the tcdA12 primer set was determined to be 60 °C to 67 °C, with peak amplification efficiency identified at 66 °C (Fig. 2).

Specificity of tcdA LAMP reaction

Primer set tcdA12 only amplified when C. difficile VPI10463 was used as the template; all reactions with other bacterial species and no template controls were negative (Fig. 3A). Results of the chromogenic reaction confirmed from the results of the real-time monitoring of turbidity (Fig 3B).

Sensitivity of tcdA LAMP reaction

As demonstrated in Figs. 4A and 4B, the detection limit of real-time turbidity was 20.700 pg/µl, which was identical to that of visual detection. The NK1 and NK2 primers with the same concentration of C.difficile VPI10463 DNA were also evaluated using PCR. The detection limit for PCR was 207 pg/µl (Fig. 4C), which was 10-fold lower than that of the LAMP reaction.

Evaluation of the assay with stool specimens

A total of 164 stool specimens were considered eligible and suitable for this study. One patient was diagnosed with pseudomembranous enteritis which is a common symptom of CDI (Figs. 5A, 5B). Electronic colonoscopy revealed numerous scattered pale yellow pseudomembranes and areas of congested, brittle mucosa (Fig. 5A). Histopathological imaging (Fig. 5B) revealed an infiltration of inflammatory cells into the mucosal lamina propria.

In the stool specimens, the detection rate of the tcdA gene using the LAMP method was 17% (28/164), significantly higher than the 10% (16/164) detection rate of the PCR method (X² = 47, df = 1, P < 0.01) (Table 3). The consistency between the LAMP and PCR methods was moderate (Kappa = 0.533, p < 0.01). Notably, 15 of the stool specimens that were negative for the tcdA gene in PCR but positive in LAMP were subsequently confirmed to be positive through Sanger sequencing performed at Sangon Biotech (Shanghai, China). Of the three specimens that were negative in LAMP but positive in PCR, only one was found to be positive using tcdA gene sequencing. Using sequencing as the reference standard, the LAMP assay outperformed the PCR assay with a sensitivity of 97% compared to 48%, and a specificity of 100%, nearly identical to PCR’s 99%. The LAMP assay achieved a PPV of 100% and an NPV of 99%, whereas the PCR assay had a PPV of 88% and an NPV of 90% (Tables 4 and 5).

Table 3:

Comparison of the results of LAMP and PCR for detecting the tcdA gene in stool specimens.

PCR	LAMP		Total
	Positive	Negative
Positive	13	3	16
Negative	15	133	148
Total	28	136	164

DOI: 10.7717/peerj.17776/table-3

Table 4:

Comparison of the results of LAMP and sequencing for detecting the tcdA gene in stool specimens.

Sequencing	LAMP		Total
	Positive	Negative
Positive	28	1	29
Negative	0	135	135
Total	28	136	164

DOI: 10.7717/peerj.17776/table-4

Table 5:

Comparison of the results of PCR and sequencing for detecting the tcdA gene in stool specimens.

Sequencing	PCR		Total
	Positive	Negative
Positive	14	15	29
Negative	2	133	135
Total	16	148	164

DOI: 10.7717/peerj.17776/table-5

Discussion

This study marks a significant advancement in the rapid detection of C. difficile. The optimized LAMP assay reduced the time to detect the tcdA gene to under 60 min. Incorporating loop primers (LB and LF) reduced LAMP reaction times by over 50%. Including loop primers also enhanced the LAMP reaction’s efficiency and sensitivity by multiplying the initiation sites for amplification. Loop primers are designed to be activated singularly during the synthesis of the artificial template (Nagamine, Hase & Notomi, 2002). LAMP-based detection yielded negative results for 26 pathogenic bacterial strains from different genera than C. difficile, demonstrating the primers’ exceptional specificity. The 10-fold higher sensitivity of the LAMP method compared to conventional PCR corroborates findings from prior studies (Kim et al., 2022a; Carvajal-Gamez et al., 2023) and supports the integration of this LAMP assay into routine diagnostic workflows.

The LAMP method yielded a detection rate of 17% for the tcdA gene of C. difficile in stool specimens, which is similar to rates observed in diarrheal stool samples of diverse populations across China. For instance, a multicenter study in Shanghai, China, reported a detection rate of 18% (93/531) for the tcdA gene (Mi et al., 2020), while an independent cross-sectional study in Southwest China noted a detection rate of 14% (125/978; Liao et al., 2018). These findings suggest that the LAMP method’s efficacy in detecting the tcdA gene is comparable to the methods used in previous studies. Moreover, this study also revealed a significant discrepancy in the detection rates of the tcdA gene in stool specimens between LAMP and PCR. LAMP showed a superior detection rate. This was further evidenced by the fact that 15 of the stool specimens that were negative in PCR, but positive in LAMP, were subsequently confirmed to be true positives through sequencing of the tcdA gene. In our previous study of 300 cultured C. difficile strains from Southern Medical University Nanfang Hospital, a perfect concordance (kappa = 1) was observed between LAMP and PCR detections of the tcdA gene (Lin et al., 2022), supporting the notion that LAMP is equally reliable when bacterial DNA is present in higher quantities. The difference in consistency between LAMP and PCR in detecting C. difficile strains in stool specimens in this study may be attributed to the variable concentration of the C. difficile DNA present. In stool specimens where C. difficile DNA may be at relatively low concentrations, LAMP’s lower detection limit allows for a higher detection rate. Conversely, the increased DNA content in cultured C. difficile strains following enrichment culture may mask the differences in detection limits between the two methods, leading to a high degree of consistency in detecting the tcdA gene in these strains. These findings collectively advocate for the integration of LAMP in clinical diagnostics of CDI, particularly for cases where PCR may not provide sufficient sensitivity.

In the analysis of stool samples, the LAMP assay showed high sensitivity (97%) and perfect specificity (100%), along with a PPV of 100% and an NPV of 99%, when using sequencing as the gold standard. This suggests that the LAMP assay offers a high degree of reliability in both detecting the presence of the target pathogen and verifying its absence. In contrast, the PCR assay, despite its high specificity (99%), demonstrated significantly reduced sensitivity (48%), potentially limiting its effectiveness as a standalone diagnostic tool. The PCR’s PPV of 88% and NPV of 90% suggest that while positive results are likely accurate, the risk of false negatives is increased, which could contribute to under-diagnosis. Consequently, the LAMP assay could represent a more robust approach for precise detection of CDI within the examined population.

The LAMP method exhibits considerable advantages in the detection of the tcdA gene for C. difficile. First, eliminating the DNA denaturation step simplifies the operational workflow, as reactions are carried out at an isothermal condition of 60–65 °C using the strand displacement activity of Bst DNA polymerase, thus obviating the need for the precise thermal cycling equipment required for PCR (Soroka, Wasowicz & Rymaszewska, 2021). This significantly lowers the technical requirements of detection, allowing molecular diagnostic techniques to be performed in resource-limited settings. Second, the LAMP reactions can be directly visualized through color changes in the calcium-magnesium complex when exposed to UV light, circumventing the need for the complex gel electrophoresis step that is requisite in PCR. This advancement not only expedites detection speed, but also minimizes the dependency on carcinogenic dyes, thereby enhancing the safety profile of the procedure. Third, the sensitivity of LAMP is 10–100 times higher than that of conventional PCR (Kim et al., 2022a; Ashmi et al., 2023; Carvajal-Gamez et al., 2023). This significant enhancement in sensitivity ensures accurate detection of pathogens even at low concentrations, which is vital for early intervention and for controlling the spread of infection. Finally, the specificity of LAMP is notably enhanced through the employment of 4–6 primers, which identify up to eight specific locations on the DNA template. This significantly mitigates the likelihood of false positives, a potential risk associated with the two-primer system in PCR (Soroka, Wasowicz & Rymaszewska, 2021).

While the LAMP method stands out for its advantages, it is not without its limitations. Cross-contamination arising from multiple pipetting steps is a notable concern, particularly with material present in the aerosol. This study mitigated this issue by employing low-melting-point paraffin to seal the reaction mixtures, thereby preventing the spread of amplification products. Rigorous sample handling protocols and enhanced ventilation should be conducted to minimize contamination risks. Another limitation is the suitability of LAMP products for downstream applications, such as sequencing or cloning (Sahoo et al., 2016). This drawback underscores the need for further methodological refinements to expand the utility of LAMP products beyond mere detection. Additionally, the sensitivity of LAMP to inhibitors remains a complex challenge (Dong et al., 2014). This highlights a potential area for the development of more sophisticated controls and detection metrics within the LAMP protocol. Finally, primer-dimer formation in LAMP could lead to false positives.

A limitation of this study is the exclusive use of the C. difficile VPI10463 strain as a positive control, which may not represent the diversity of C. difficile strains. To address this limitation and to understand its potential impact on the conclusions drawn from the results, a more comprehensive analysis was performed. The subsequent sequencing analysis of 15 stool specimens that tested positive with the LAMP assay but negative with PCR provided additional insights. Not only did this analysis confirm the presence of C. difficile VPI10463 in two cases, but it also identified other strains in the remaining samples: seven cases of GZ14, two cases of ZR48, and one case each of ZR80, ZR50, GZ5, and SH8. These findings suggest that the LAMP assay developed in this study may have the capacity to detect a broader spectrum of C. difficile strains than initially anticipated, which is a significant consideration for the validity of this method. This understanding of the assay’s limitations and their possible impact on the conclusions drawn in this study is critical for interpreting the study’s results and for guiding future research directions.

Conclusions

In this study, the LAMP method proved to be an effective technique for the rapid and visual detection of the tcdA gene of C. difficile, demonstrating potential advantages over PCR in terms of speed, simplicity, and sensitivity. The tcdA-LAMP assay is particularly suitable for medical diagnostic environments with limited resources and represents a promising diagnostic strategy for the screening and detection of CDI in populations at high risk.

Supplemental Information