Detection of Mycobacterium kansasii using a combination of loop-mediated isothermal amplification (LAMP) and lateral flow biosensors

Mycobacterium kansasii is an opportunistic pathogen that causes both intrapulmonary and extrapulmonary infections. The symptoms of the pulmonary diseases caused by M. kansasii closely resemble Mycobacterium tuberculosis. Rapid and accurate differentiation of M. kansasii from M. tuberculosis, as well as other mycobacteria, is crucial for developing effective therapeutics and disease treatment. In this study, we combined loop-mediated isothermal amplification (LAMP) with lateral flow biosensors (LFB) to detect M. kansasii, by targeting the species-specific sequence of rpoB, a gene which encodes the β subunit of bacterial RNA polymerase. The assay was validated to ensure that it was highly selective by testing M. kansasii, M. tuberculosis, other species of respiratory bacteria, and other nontuberculous mycobacteria. The detection limit of the assay was 1 fg/μL of DNA and 50 CFU of bacilli in sputum. The M. kansasii-LAMP-LFB assay is a fast, cheap, and accurate method for detecting M. kansasii by constant temperature amplification and simple interpretation.


Introduction
Nontuberculous mycobacteria (NTM) comprise most mycobacteria except the Mycobacterium tuberculosis complex (MTBC) and Mycobacterium leprae. With an increase in the number of patients suffering from immune injury (HIV infection, organ transplant patients, etc.) (Griffith et al. 2007), the number of NTM diseases has increased significantly (Wassilew et al. 2016). NTM are widely distributed in the environment, mainly in soil and water, and are classified into fast-growing and slow-growing mycobacteria. Mycobacterium kansasii, an opportunistic pathogen, is a slow-growing NTM. Patients with chronic lung disease, mycobacterial disease, malignancy, and alcoholism have a higher risk of developing diseases caused by M. kansasii (Kim et al. 2019;Hirashima et al. 2014;Theodore and Charles 2002;Won-Jung et al. 2002). In China, M. kansasii is the third most common slow-growing NTM to cause clinical disease (Fang et al. 2019;Tan et al. 2018).
As one of the most important pathogenic nontuberculous mycobacteria, the clinical symptoms of M. kansasii infection are similar to those seen in patients with tuberculosis and other nontuberculous mycobacterial infections, but the treatment is different. It is therefore very important to avoid misdiagnosis of M. kansasii and to differentiate it from other mycobacteria. The traditional method of identifying M. kansasii is through biochemical identification after culturing, which is time intensive and yields in accurate results. Chromatography, including gas chromatography, high-pressure liquid chromatography, and time-of-flight mass spectrometry that have been used to detect M. kansasii, require expensive equipment and complex operation (George et al. 2018;Lau et al. 2015;Olivier and Loots 2012). Similarly, the molecular methods, such as GeneChip, reverse hybridization, and sequencing technology that have been used to identify M. kansasii, are still relatively expensive because of the costly instruments required, time consumption, and complicated procedures (Quezel-Guerraz et al. 2010;Suhail and Eiman 2019;Seagar et al. 2008).
The loop-mediated isothermal amplification (LAMP) method combined with use of a lateral flow biosensor (LFB) is a method of multiple cross-constant temperature amplification, which uses the LFB to observe the results using the naked eye (Wang et al. 2017;Wang et al. 2014). This method can be directly applied for the detection of clinical samples. According to previous studies, M. kansasii can be distinguished from other mycobacteria by targeting speciesspecific sequences, such as 16SrRNA, heat shock protein 65 (hsp65), RNA polymerase beta subunit (rpoB), and 16S-23S rRNA gene internal transcribed spacer (ITS) region (Chikamatsu et al. 2019;Rina et al. 2014;Zoe et al. 2011). In the current study, we developed a LAMP-based method to detect M. kansasii by targeting rpoB.

Materials and methods
Visual detection of LAMP products using the lateral flow biosensor A total of 0.5 μL of LAMP products and 100 μL of running buffer (10 mM PBS, pH 7.4 with 1% Tween-20) were deposited, separately, in the designated region of the lateral flow biosensor (Hai Tai Zheng Yuan Technology Co., Ltd., Beijing, China). Once the biosensor had absorbed the entire quantity of running buffer (~2 min), the subsequent detection of the LAMP products was computed in the form of red lines on the nitrocellulose filter membrane (NC membrane) of the lateral flow biosensor (LFB). The TL was test line, and the CL was control line. The principle of LAMP product detected in LFB was reported (Wang et al. 2016); on the NC membrane of LFB, two zones as the test zone (conjugated with anti-FAM) and control zone (conjugated with biotin-BSA) would be combined with FAM and biotin in the amplified product to visualize the result in the form of red lines on the NC membrane.

Preparation of strains and genomic template
The 85 mycobacteria and 3 respiratory pathogenic bacteria used in this study (Table 1) were obtained from the National Tuberculosis Reference Laboratory (NTRL) of China. All mycobacteria were cultured at 37°C on modified Loewenstein-Jensen medium, purchased from Celnovte-Bio Co. Ltd. (Zhenzhou, China). Genomic DNA from these strains was extracted using the CTAB-phenol-chloroform extraction method. The CTAB was purchased from Amresco (USA). The genomic DNA was quantified using a Nanodrop ND-100 instrument (Calibre, Beijing, China) and diluted serially to 1 ng/μL, 10 pg/μL, 1 pg/μL, 100 fg/μL, 10 fg/μL, 1 fg/μL, and 0.1 fg/μL to analyze the detection limit of the M. kansasii-LAMP-LFB assay.

Design of M. kansasii-LAMP-LFB assay primers
The set of primers used for the M. kansasii-LAMP-LFB assay were designed by Primer ExplorerV4 (Eiken Chemical, Japan) and the primer software PRIMER PREMIER 5.0, based on the rpoB reference sequence (GenBank accession no. NC_022663.1). The primer sequences shown in Table 2 were further analyzed by BLAST and confirmed to be specific for M. kansasii. The locations of the primers are shown in Figure S1.
When using colorimetric indicator method, we evaluated the results and expected the color of the amplified products to change from violet to azure in the positive tube, while negative controls and blank controls remained violet. When using LAMP-LFB assay, the positive responses should indicate two visible red lines (test line, TL; control line, CL), while negative and blank controls showed only the control lines.
In these experiments, M. kansasii (ATCC12478) was used as the positive strain, M. tuberculosis H37Rv (ATCC 27294) was the negative control, and double distilled water was the blank control. To verify the optimal amplification temperature, at which there was the fastest reaction rate, a total of six temperatures (65-70°C, with 1°C interval) were compared under standard LAMP protocol for 60 min, then hold the reaction at 85°C for 5 min. The reaction products were analyzed by LFB detection.

Specificity and detection limit of the M. kansasii-LAMP-LFB assay
The specificity of the M. kansasii-LAMP-LFB was analyzed using the DNA template from 25 M. kansasii strains, 60 mycobacterial strains, and 3 strains from other respiratory bacteria (Table 1). These tests were repeated three times. The detection limit was determined using serial dilutions (1 ng, 10 pg, 1 pg, 100 fg, 10 fg, 1 fg, and 0.1 fg per microliter) of the genomic DNA template. The results of the M. kansasii-LAMP-LFB were compared with values collected using the colorimetric indicator and real-time turbidity methods. Three independent replicates were tested. Optimal amplification time for the M. kansasii-LAMP-LFB assay The optimal amplification time for the M. kansasii-LAMP-LFB was confirmed by increasing the reaction time from 30 to 40 min, at 10-min intervals, and then comparing those results with that obtained from the experiment with a 60-min reaction time. The reaction products were then analyzed by LFB detection. Three independent replicates were analyzed.

M. kansasii-LAMP-LFB assay testing artificial sputum samples
The above experiments were used to determine optimum amplification temperature and time. Based on these optimal reaction conditions of the M. kansasii-LAMP-LFB assay, artificial sputum samples of M. kansasii (ATCC 12478) and M. tuberculosis H37Rv (ATCC 27294) were used to evaluate its accuracy. The artificial sputum used in this study was obtained from the National Tuberculosis Reference Laboratory of China (NTRL). Its main components were normal saline, sterile purified water, lecithin, salmon sperm DNA, bovine serum albumin, and mucin (precise composition was not listed because of the patents). Mycobacterium kansasii solution (5 μL), at a concentration of 10 5 CFU/mL, was added into the tube with 995 μL artificial sputum medium, and at last, the concentration of M. kansasii artificial sputum sample was 500 CFU/mL. The M. kansasii DNA was extracted from M. kansasii artificial sputum sample by the boiling method using 100 μL ddH 2 O. And then, 5 μL of DNA template solution from the 100 μL boiling method product was used in the 25 μL amplification system. There were only 25 CFU M. kansasii DNA in the amplification system at the end. According to above method, 10 μL 10 5 CFU/mL of M. kansasii added to 990 μL artificial sputum medium to obtain 1000 CFU/mL M. kansasii artificial sputum sample, and 5 μL DNA template solution in the final 25 μL amplification system comprised just 50 CFU bacteria DNA. Then using the above method, 5 μL template with 5000 CFU and 500 CFU M. kansasii strain DNA was obtained.

Results
Optimization of the temperature for M. kansasii-LAMP-LFB assay Mycobacterium kansasii ATCC12478 DNA (1 ng per reaction) was used as the positive control, to verify the optimal amplification temperature. When the assay was carried out at temperatures ranging from 65°C to 70°C, the reaction was monitored by the real-time turbidity. This experiment demonstrated that the most suitable temperature was 67°C, at which the shortest reaction time was observed ( Figure S2).

Specificity of M. kansasii-LAMP-LFB assay
A total of 1 ng of all strains in Table 1 was used in the M. kansasii-LAMP-LFB assays at 67°C for 40 min. The DNA from M. kansasii ATCC 12478 and 24 clinical isolated strains had positive results. The other mycobacteria species, as well as the three respiratory pathogenic bacteria (Klebsiella FIP*, 5′-labeled with 5-carboxyfluorescein (5-FAM) used in LAMP-LFB assay; LF*, 5′-labeled with biotin used in LAMP-LFB assay pneumoniae, Neisseria meningitidis, Streptococcus pneumoniae), and double distilled water were negative (Fig.  2). On the lateral flow biosensor, two red lines (TL and CL) appeared positive, and only one red line (CL) appeared negative. The above results suggest the assay could specifically detect and focus on M. kansasii relative to other mycobacteria and other common respiratory pathogens.
Sensitivity of M. kansasii-LAMP-LFB assay of artificial sputum samples DNA template (5 μL) was added in the M. kansasii-LAMP-LFB assay, and 50 CFU of the specific M. kansasii strain could be detected in the sputum by the lateral flow biosensor (Fig. 3).

Discussion
Mycobacterium kansasii infection was first detected in Kansas City, USA in 1953. Mycobacterium kansasii is a photochromogenic bacterium and is one of the most important nontuberculous mycobacteria found in clinical settings (Zofia et al. 2018). It is pathogenic to humans, and primarily causes human pulmonary infections and extrapulmonary disseminated infections (James et al. 2017;Kaur et al. 2011;Moon et al. 2015). The isolation rate of M. kansasii is very high in Europe, America, and Japan, especially in patients with acute immunodeficiency syndrome (Basille et al. 2018;Furuuchi et al. 2019;Gary 2017;Santin et al. 2018). In a report, in distribution of nontuberculous mycobacteria species isolated from pulmonary NTM patients in South China, the separation rate of M. kansasii was ranked forth, following Mycobacterium abscessus, Mycobacterium intracellulare, and Mycobacterium avium (Tan et al. 2018). It is therefore necessary to detect M. kansasii using rapid and accurate detection methods. The method that we developed in this study is a simple, rapid, specific, and inexpensive constant temperature amplification method. Constant temperature gene amplification can save time during the amplification and detection steps (Gandelman et al. 2011;Tomita et al. 2008).
In this study, we chose a specific gene, rpoB of M. kansasii, to design the LAMP primers. Using BLAST analysis, we found that the rpoB sequence we selected is highly species specific for detecting M. kansasii. The M. kansasii-LAMP-LFB assay showed positive results with only M. kansasii, suggesting that this assay is highly specific. In the assay, the amplification step was conducted at 67°C for 40 min and Fig. 1 The detection limit of the M. kansasii-LAMP assay. Serial dilutions (1 ng/μL, 10 pg/μL, 1 pg/μL, 100 fg/μL, 10 fg/μL, 1 fg/ μL, 0.1 fg/μL) of the M. kansasii DNA template were used, and amplification products were monitored by turbidimeters (a), colorimetric indicator (b), and LFB detection (c) stopped while held at 85°C for 5 min. The amplified products were added to the LFB and the results could be seen in almost 2 min. The whole detection process from isothermal amplification to result observation only needs about 50 min, which is significantly faster than other molecular methods like GeneChip, reverse hybridization, and sequencing technology.
Moreover, it showed excellent sensitivity and a detection limit of 1 fg, allowing even 50 CFU of M. kansasii strains in sputum to be tested. For the entirety of this experimental process, only a constant temperature machine was needed. The requirements for the equipment were extremely simple, and it was very convenient to observe the results without any additional apparatus required. Notably, the whole process cost including primers, isothermal amplification reagents, and lateral flow biosensor is only about 10 dollars. Therefore, the M. kansasii-LAMP-LFB assay could be a very simple, fast, specific, and inexpensive way to detect clinical disease caused by M. kansasii infection. And also, this is the first time LAMP-LFB used for testing M. kansasii.
Acknowledgments We greatly appreciate Dr. Yanlin Zhao, the head of the National Tuberculosis Reference Laboratory (NTRL) of China for the support of this study.
Funding This work was supported by the Sichuan Medical Association (Grant No. S16066).
Data availability All data used or analyzed during this study are included within this article.

Compliance with ethical standards
Shu Zhang has applied for a patent which covers the primers set included in this study (application number CN 201910700776.7) from the State Intellectual Property Office of China.

Conflict of interest
The authors declare that they have no conflict of interest.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.