Application of Loop-Mediated Isothermal Amplification combined with colorimetric and lateral flow dipstick visualization as the potential point-of-care testing for diphtheria

Background Diphtheria outbreaks occurred in endemic areas and imported and indigenous cases are reported in UE/EEA. Because of the high infectiveness and severity of the disease, early and accurate diagnosis of each suspected case is essential for the treatment and management of the case and close contacts. The aim of the study was to establish simple and rapid testing methods based on Loop-Mediated Isothermal Amplification (LAMP) assay for the detection of Corynebacterium diphtheriae and differentiation between toxigenic and non-toxigenic strains.Methods Corynebacterium diphtheriae and Corynebacterium ulcerans isolates from the National Institute of Public Health-National Institute of Hygiene collection were used for the development of LAMP assay for the diagnosis of diphtheria and nontoxigenic C. diphtheriae infections. Various colorimetric methods for visualization of results were investigated. Sensitivity and specificity of the assay were examined using a collection of DNA samples from various gram-positive and gram-negative bacteria.Results The LAMP assay for tox and dtxR genes was developed. The sensitivity and specificity of the assay were calculated as 100%. The detection limit was estimated as 1.42 pg/µl concentration of DNA template when the reaction was conducted for 60 min. However, the detection limit was lowered 10 times for every 10 minutes of reduction in the time of incubation during the reaction. Positive results were successfully detected colorimetrically using hydroxynaphthol blue, calcein, QuantiFluor, and lateral flow Milenia HybriDetect dipsticks.Conclusion The assay developed in the study might be applied for point-of-care testing of diphtheria and other C. diphtheriae infections. It is highly sensitive, specific, inexpensive, easy to use, and suitable for low-resource settings.


Background
Diphtheria is an acute, potentially lethal infectious disease of humans caused by diphtheria toxinproducing strains of Corynebacterium diphtheriae, Corynebacterium ulcerans, and Corynebacterium pseudotuberculosis. It is a highly infectious disease. The infection can be transmitted through contact with infected persons and objects that are touched by them. Depending on the anatomic site that is affected by the disease, it could be respiratory or cutaneous diphtheria. Rarely other sites can also be affected such as eye, ear, and vulva. Diphtheria toxin absorbed from the mucosal or cutaneous lesions causes toxic damage to organs such as the myocardium, kidneys and nervous system. In respiratory diphtheria cases, formed pseudomembranes can cause obstruction in the airways [1].
After the introduction of vaccination against diphtheria in the 1940s, the infections caused by toxigenic corynebacteria seemed to be well controlled in developed countries. However, infections recorded during the last several years point at C. ulcerans and C. diphtheriae as reemerging human pathogens. According to ECDC data, the number of confirmed diphtheria cases in EU/EEA increased over three times from 2011 to 2015 [2]. Domestic pets and other animals have been described as novel reservoirs and sources of diphtheria infection [3,4,5] Because of the high infectiveness and severity of the disease, early and accurate diagnosis of each suspected case is essential for the treatment and management of the case and close contacts. Rapid microbiological tests are of high value because clinical diagnosis is not easy and might be confused with other causes, such as streptococcal sore throat or tonsillitis [10]. Misdiagnosis is the high risk particularly in countries where the diphtheria is uncommon. Point-of-care diphtheria testing is especially important in refugee camps and developing countries, where access to medical laboratories is extremely limited as well as in the investigation of an infection source.

Bacterial strains and DNA extraction
Nine toxigenic and 31 nontoxigenic Corynebacterium strains were used in the study. The toxigenic strains included 5 C. diphtheriae clinical isolates, one C. ulcerans clinical isolate, and reference C. diphtheriae strains such as PW8, NCTC 10648, and NCTC 3984. The nontoxigenic strains included 30 Milenia HybriDetect dipsticks (Milenia Biotec, Germany) were used for the detection of the amplified products labeled with biotin and FITC. Ten microliters of the reaction mixture were pipetted directly on the sample application area on the dipstick. Then, the dipstick was placed with the same application area into 100 µl of HybriDetect assay buffer and incubated for 5-15 min in an upright position. The results were regarded as positive when two bands were visible (a control band and a test band) or as negative when only a control band was visible.

Colorimetric detection of amplified products
For the colorimetric detection of amplified products, 5 indicators were used comparatively: neutral red, phenol red, hydroxynaphthol blue (HNB), calcein and QuantiFluor. Neutral red and phenol red are pH indicators. They are added to the pre-reaction solution. The progress of LAMP reaction is related to lowering of the solution pH, which can be observed directly as color change of faint orange to pink (neutral red) or red to yellow (phenol red) [11]. Hydroxynaphthol blue and calcein are metal ion indicators. They are also added to the pre-reaction solution. When Mg 2+ ion concentration decreases in the progress of LAMP reaction, the color change of the indicators can be observed directly [12]. The color shift is violet to blue for HNB and orange to fluorescent green for calcein. QuantiFluor is a DNA intercalating dye. It is added to the solution after the reaction is completed. When the LAMP reaction is positive, a color change of orange to fluorescent yellow is observed under ambient light condition.
Neutral red (Sigma-Aldrich, USA) and phenol red (Sigma-Aldrich, USA) were dissolved in deionized water or 1 M NaOH, respectively, at 50 mM to prepare a stock solution and diluted to 2.5 mM. in deionized water. To select an optimal concentration, the following volumes of the stock solution were added to the reaction solution: 0.25, 0.5, 1, 1.5, and 2 µl. The amount of QuantiFluor (Promega, Germany) in the post-reaction solution was optimized by the addition of the following volumes of the dye: 2, 1, and 0.5 µl.

Determination of specificity, sensitivity, detection limit, and minimal reaction time
Specificity and sensitivity of the LAMP were investigated using abovementioned bacterial species that can be present in respiratory tracts. The sensitivity was calculated as follows: A/(A + C) × 100%, and the specificity was calculated as follows: D/(B + D) × 100%, where A is the number of true positive results, B is the number of false-positive results, C is the number of false-negative results, and D is the number of true negative results.
The limit of detection was investigated using 10-fold serial dilutions of the total genomic DNA.
To determine required minimal LAMP reaction time, we examined the results of the reactions for tox and dtxR markers after 10, 20, 30, 40, 50, and 60 min of incubation using 10-fold serial dilutions of the total genomic DNA as a reaction template.

Results
The species-specific dtxR gene present in all C. diphtheriae strains and the tox gene present only in potentially toxigenic C. diphtheriae, C. ulcerans, and C. pseudotuberculosis strains were selected as target genes for designing the LAMP primers. Initially, three sets of primers for each of the genetic markers investigated were designed but only the sets presented in Table 1 did not yield false-positive results and therefore were selected for the study. The concentration of each of the primer as well as other reagents in the reaction mixture was optimized. Labeling of the primers FIB and BIP with biotin and FITC did not influence the amplification reaction, as it was assessed based on agarose gel electrophoresis results. The efficiency of the LAMP reaction was comparable in the temperature ranging from 62°C to 70°C (Figure 1). For the study, we selected 65°C as recommended by the manufacturer of the Bst 2.0 DNA polymerase.
We could detect positive LAMP reactions with the naked eye using HNB, calcein, QuantiFluor, and Milenia HybriDetect dipsticks. The positive reaction was clearly visible when the used HNB concentration was 0.125, 0.16, and 0.25 mM. For further studies, we selected the concentration 0.16 mM of HNB. The optimal amounts of calcein and QuatiFluor per reaction were 0.5 and 2 µl, respectively. By using Milenia HybriDetect dipsticks, we observed atypical results for samples with a high concentration of DNA. According to manufacturer's instructions, two color bands should be visible on the dipstick for positive results: test band and control band, whereas only control band should be visible for negative results. However, we observed that when the concentration of amplicons was high, the control band was not visible (Figure 2). This issue was overcome by the dilution of the amplified product. We could not detect positive LAMP reaction when the neutral red and phenol red were used. It was probably because the pH changes during the reaction were very subtle.
The sensitivity and specificity of the LAMP reaction for tox and dtxR markers were comparable using HNB, calcein, and QuantiFluor, as well as Milenia HybriDetect dipsticks, and both were calculated as 100% ( Table 2). The detection limit was also comparable for both genetic markers and all product detection methods and estimated as 1.42 pg/µl concentration of DNA template, which means 2.84 pg of DNA in 25 µl of the reaction mixture, when the reaction was conducted for 60 min. However, the detection limit lowered 10 times for every 10 minutes of reduction in the time of incubation during the reaction (Table 3, Figure 3).

Discussion
Diphtheria is a vaccine-preventable disease. Currently, the diphtheria vaccination coverage varied from 42% in some developing countries to 99% in some developed countries [13]. Diphtheria outbreaks occur in endemic countries, and diphtheria cases are reported every year in UE/EEA. According to ECDC data, 216 diphtheria cases were reported in UE/EEA in the period 2012-2016. Most of them were imported from endemic geographical areas and some were indigenous cases [14].
Epidemiological data on diphtheria and a growing problem of nontoxigenic C. diphtheriae invasive infections [15,16,17] have revealed the need for point-of-care testing (POCT) technology for the detection of C. diphtheriae in carriers, suspected cases, and contacted persons. Such POCT technology would be of great value especially in endemic regions of the disease, where access to health care is limited, and in refugee camps, to timely start appropriate treatment and further prevent the spread of the C. diphtheriae and the outbreak. Our study aimed to establish simple and rapid testing methods based on LAMP assay for the detection of C. diphtheriae and differentiation between toxigenic and non-toxigenic strains. Additionally, we compared various methods for visualization of amplified products. The developed method was highly sensitive and specific and showed a very low detection limit. It was reported by other researchers that LAMP detection limit is lower than polymerase chain reaction (PCR) [18]. However, we found that the incubation time necessary to obtain positive results depends on the amount of the target DNA in the sample. The methods of visualization the results, including the use of HNB, QuantiFluor, calcein, and Milenia HybriDetect dipsticks, did not influence the detection limit, but colorimetric detection using HNB and calcein, which were added to the reaction mixture before incubation, are superior to QuantiFluor and Milenia HybriDetect dipsticks because they enable faster detection of positive reaction in the realtime mode, and no additional handling after reaction is needed. The opening tubes after the reaction is associated with an increased risk of contamination of other subsequent LAMP reaction solutions.
The LAMP reaction temperature ranging from 62°C to 70°C, as revealed in our study, shows that the One of the disadvantages of most molecular methods is the requirement for storage conditions of reagents, such as polymerases, which usually have to be kept in freezing. However, a lyophilized mastermix containing all reagents required for LAMP assay was developed for the detection of some other pathogens [20,21]. Furthermore, LAMP reagents are commercially available in a dry format currently, which can be stored at room temperature.
At the stage of development of the potential point-of-care test for diphtheria, we used DNA samples.