Optimization of a Loop-Mediated Isothermal Ampli cation (LAMP) Assay for Schistosoma Mansoni (Trematoda: Digenea) Detection in Biomphalaria spp. from Endemic Areas for Schistosomiasis in Brazil

Silvia Gonçalves Mesquita (  silviagmesquita@gmail.com ) Fundacao Oswaldo Cruz Instituto Rene Rachou https://orcid.org/0000-0003-2698-8574 Floria Gabriela dos Santos Neves Fundacao Oswaldo Cruz Instituto Rene Rachou Ronaldo Guilherme Carvalho Scholte Fundacao Oswaldo Cruz Instituto Rene Rachou Omar dos Santos Carvalho Fundacao Oswaldo Cruz Instituto Rene Rachou Cristina Toscano Fonseca Fundacao Oswaldo Cruz Instituto Rene Rachou Roberta Lima Caldeira Fundacao Oswaldo Cruz Instituto Rene Rachou


Background
Schistosomiasis is a parasitic disease that affect nearly 240 million people in the world. The disease is highly associated with low sanitation conditions and poverty, which lead people to use contaminated water for work and leisure (1). It is estimated that over 25 million people live in areas with a high-risk of schistosomiasis in the Americas. In Latin America, approximately 7.1 million people are infected with the etiological agent Schistosoma mansoni Sambon, 1907, with 95% of them living in Brazil (2), where the Northeast and Southeast regions are the most affected (3). Biomphalaria (Preston, 1910) snails are essential for maintenance of the parasite life cycle. Eleven Biomphalaria species and one subspecies have been reported in Brazil, and three of them -Biomphalaria glabrata (Say, 1818), Biomphalaria tenagophila (d'Orbigny, 1835) and Biomphalaria straminea (Dunker, 1848) -have been found naturally infected with S. mansoni (4). The presence of the susceptible snail hosts in water bodies is crucial for parasite development, and determines the distribution of the disease (5). Knowledge regarding the geographic distribution of Biomphalaria snails in Brazil is being progressively updated and demonstrates that intermediate host species are spreading to new locations (3).
Among the control measures available to eliminate schistosomiasis, the surveillance of potential and active transmission foci, together with snail control measures, are highly recommended (6). Traditionally, the detection of infected snail hosts is performed by either inducing cercarial shedding through arti cial light exposure (7) or using the shell-crushing method, both followed by stereomicroscope examination to detect either cercariae or sporocysts found in snail tissue (8,9). However, many factors can limit the effectiveness of these parasitological methods. Inducing cercarial shedding does not detect the early stages of snail infection, and can result in misidenti cation, as the larvae of other trematode species are morphologically similar to each other, requiring an experienced person for their accurate identi cation. The shell-crushing method does not allow the speci c identi cation of sporocysts and can damage the cercarial tissue hindering the observation of differential morphological characters. In addition, neither methods can be performed using dead snails (10)(11)(12)(13).
In order to overcome these limitations, several alternative methods for the xenomonitoring of human schistosomes have been described. Molecular approaches, such as conventional polymerase chain reaction (PCR) (14)(15)(16)(17), low-stringency polymerase chain reaction (LS-PCR) (18), restriction fragment length polymorphism (PCR-RFLP) (19), nested PCR (17), multiplex PCR (20)(21)(22), real-time quantitative PCR (qPCR) (23,24), DNA sequencing (25), and loop-mediated isothermal ampli cation (LAMP) (26)(27)(28)(29)(30) have all proved to be more accurate and sensitive alternatives than the traditional microscope-based methods. Despite the high sensitivity and speci city of molecular methods, their cost and requirement of laboratory infrastructure have limited their usage in surveillance. In this context, LAMP assay stands out as a promising method for the direct detection of S. mansoni infection in the eld, as it does not require laboratory equipment, such as a PCR machine or electrophoresis apparatus (27).
The estimated annual cost of schistosomiasis in Brazil is over US$41 million, with more than 90% of this economic burden being related to indirect costs (e.g., loss of productivity and wages due to sick-leave, hospitalization, and premature death) (31). This high economic burden, and the persistence of schistosomiasis transmission in many areas in Brazil, highlight the need for additional tools to control and eliminate the disease. By the optimization of the LAMP assay performed in this study, we provided a rapid, accurate, speci c, and sensitive isothermal method as an alternative approach for mapping and monitoring S. mansoni infection in Biomphalaria snail hosts. The study was conducted in the municipalities of Franciscópolis (-17.9579, -42.0079) and Malacacheta (-17.84379959, -42. One-thousand-and-one snails were collected between July and August 2019 by team members of the Helminthology and Medical Malacology Research Group (René Rachou Institute / Fiocruz-Minas). All collection sites were georeferenced using global positioning system (GPS) technology (Additional File 2). The snails were transported to Fiocruz Minas, and after all the analyses reported here, part of the collected snails were deposited in the Medical Malacology Collection (Fiocruz-CMM).

Parasitological examination and morphological identi cation of trematode larvae
In the Lobato Paraense Mollusk Room (LPMR) at the René Rachou Institute / Fiocruz-Minas, snails were separated into pools according to their collection site and then submitted to the shell-crushing method to detect natural infection with S. mansoni or other trematodes. The squeezed material was examined under a stereomicroscope to detect the presence of cercariae and/or sporocysts. The detected cercariae were isolated and then observed under an optical microscope using non-permanent preparations for morphological identi cation. The morphological identi cation step was carried out according to the identi cation keys and descriptive works of different authors (32)(33)(34)(35). Cercariae were preserved in ethanol for future analysis.

Pepsin digestion and DNA extraction
The squeezed snail pools were transferred to 50 mL centrifuge tubes labelled with a collection site code, and submitted to pepsin digestion following the protocol of Wallace & Rosen (36), and sedimentation by the Baermann-Moraes method. The sediment was centrifuged for 20 min at 5,000 g, the supernatant was removed, and the remaining pellet was cryopreserved at -80°C until DNA extraction.
DNA extraction from the digested pool of snails was performed using the Wizard® Genomic DNA Puri cation Kit (Promega, Madison, USA), according to the manufacturer's instructions. The DNA from the cercariae isolated from snails was extracted using the DNeasy Blood and Tissue Kit (Qiagen, MD, USA), again following the manufacturer's protocol.

PCR-RFLP for species-speci c molecular identi cation of snails
Genomic DNA (gDNA) obtained from snail samples from all collection sites were used as template for a species-speci c identi cation using a PCR-RFLP assay. The species-speci c pro les generated after the digestion of the ampli ed ITS fragment by the Dde I restriction enzyme (Promega, Madison, USA) were used to identify the snails present in each pool, using as a reference the pro les previously described by Caldeira et al (19). The results were visualized on silver-stained 6% polyacrylamide gels.

Multiplex PCR for family-speci c molecular identi cation of trematodes.
In order to investigate the presence of trematode infection, the gDNA obtained from snail samples from all collection sites were used as template for a trematode-family-speci c multiplex PCR according to Mesquita et al (37). The gDNA from isolated cercariae were also used to con rm morphological identi cation. In order to compare the size of the ampli ed DNA fragments obtained from the eld material, various positive controls were included using gDNA from cercariae belonging to the following trematode families: the Clinostomidae Lühe, 1901, the Echinostomatidae Lühe, 1901, the Schistosomatidae Stiles & Hassall, 1898, and the Strigeidae Railliet, 1919. These samples were provided by the Laboratory of Trematode Biology, Department of Parasitology, Federal University of Minas Gerais, Brazil. Negative controls with no DNA were included in each reaction. The resulting PCR products were visualized on silver-stained 6% polyacrylamide gels.

LS-PCR for molecular detection of the presence of Schistosoma mansoni infection in snails.
The gDNA obtained from snail samples from all collection sites were used as template for the LS-PCR to detect S. mansoni infection in snails using the primers for the minisatellite region-mtDNA and protocol described by Jannotti-Passos et al (18). A sample from S. mansoni (10 ng/µl) was included as a positive control, and as a negative control no DNA was used. The ampli cation pro le of the positive control was used as a standard for the ampli cation pro le obtained from the unknown DNA samples.

Conventional PCR for molecular detection of the presence of Schistosoma mansoni in snails.
The outer primers F3 and B3 designed by Fernández-Soto et al (38) were used in conventional PCR to amplify a 203 bp mitochondrial fragment from the S. mansoni samples from the Mucuri and Jequitinhonha Valleys. Positive (10 ng/µl of S. mansoni gDNA) and negative controls (no DNA) were included. The reaction was carried out in a nal volume of 25 µl containing: 1X PCR Buffer (Invitrogen, USA), 1.5mM MgCl 2 (Invitrogen, USA), 0.25 mM each dNTP (Invitrogen, USA), 2 pmol/µl of each primer (F3 and B3), 1.5 U of Platinum™ Taq DNA Polymerase (Invitrogen, USA) and 2 µl of the DNA. The reaction was set up as follows: (i) initial denaturation at 94°C for 1 min, followed by (ii) 30 cycles of 20 s at 94°C, 20 s at 60°C and 30 s at 72°C, and then (iii) a nal extension at 72°C for 10 min. The PCR products were run and visualized on silver-stained 6% polyacrylamide gels.

Optimization of a LAMP assay for speci c detection of Schistosoma mansoni infection in snails.
In order to detect S. mansoni infection, gDNA obtained from pooled snails were used as a template for the LAMP assay using the primers described by , and 2 µl of the DNA. The reaction tubes were incubated at 65°C for 50 min, and then heated at 80°C for 5 min to stop the reaction. The ampli cation product was detected using silverstained 6% polyacrylamide gels. The result was also visualized by color change after the addition of 2 µl of a SYBR Green I 1,000 X (Life Technologies, California, USA) by the naked eye (positive: yellow-green, negative: orange) and by UV light exposure (positive: uorescent, negative: non-uorescent).
In order to access the speci city of the optimized assay, gDNA from trematodes commonly found parasitizing Biomphalaria snails in the Neotropical region were used as templates for the reaction described above. Cercariae samples of the following trematode families were used for this purpose: the Clinostomidae, the Echinostomatidae, the Strigeidae, the Spirorchiidae Stunkard, 1921, the Diplostomidae Poirier, 1886 and the Notocotylidae Lühe, 1909. These samples were provided by the Laboratory of Trematode Biology, Department of Parasitology, Federal University of Minas Gerais, Brazil. Biomphalaria glabrata positive and negative for S. mansoni infection, as well as adult worms of S. mansoni, were also used (samples provided by the Fiocruz-CMM).
The analytical detection limit of the LAMP assay was determined by serial dilutions of S. mansoni from 10 ng/µl to 0.1 fg/µl.

2.9
Validation of optimized LAMP assay using laboratory and eld samples.
The validation step using laboratory samples evaluated the capacity of the assay to detect different stages of infection, and in pools having a different proportion of negative and positive snails. For that, B. glabrata snails were obtained from the LPMR and the following squeezed samples prepared: single individual snails obtained at either 1, 7, 14 or 28 days post-infection (dpi); a single pool containing 20 negative snails and 1 snail in the pre-patent period of infection; and another single pool containing 20 negative snails and 1 snail shedding cercariae. The extracted gDNA were used as templates for the optimized LAMP assay.
For validation with eld samples, the gDNA from snails collected at the Mucuri and Jequitinhonha Valley regions were used as a template for the optimized LAMP assay.

Statistical analysis
To analyze the agreement between diagnostic tests, the Kappa index and its 95% con dence interval were calculated using the GraphPad online tool (www.graphpad.com/quickcalcs/kappa1/). The Landis and Koch (39) scale of agreement was used to analyze the data The sensitivity and speci city of the LAMP assay were calculated using a combination of the results from the LS-PCR and conventional PCR as references and the following formulas: sensitivity = (number of LAMP-positive results/ number of infected snails) x 100; speci city = (number of LAMP-negative results/number of non-infected snails) x 100. Cercariae isolated from the snails collected at the MV 41 site in the Mucuri Valley, and JV 04 in the Jequitinhonha Valley, had their identi cation con rmed as belonging to the Schistosomatidae family. No ampli cation was observed in snails collected at JV 03 ( Fig. 2A and B). Using the gDNA extracted from

Optimized LAMP assay for the high-precision detection of S. mansoni infection in Biomphalaria host snails
We found that with some modi cations to the Fernández-Soto et al (38) protocol, the optimized LAMP assay was effective in detecting S. mansoni infection in snails, with no cross-reactivity to other trematode species that also parasitize Biomphalaria spp. (e.g. the families Clinostomidae, Diplostomidae, Echinostomatidae, Notocotylidae, Spirorchiidae and Strigeidae) (Fig. 4A). The assay had a detection limit of 0.1 ng of the parasite DNA (Fig. 4B). The assay was also able to detect infection by S. mansoni in laboratory samples when polyacrylamide gels or SYBR Green I were used to visualize the ampli cation products. When the visual inspection by SYBR Green I was used, the infection could be detected as early as 7 days after the exposure of the snail to the parasite (Fig. 4C).

Applicability of the optimized LAMP assay to eldcollected Biomphalaria snail hosts
Snails collected in the Mucuri and Jequitinhonha Valleys were examined using the optimized LAMP assay. The LAMP ampli cation product was detected in snails from the collection sites MV 03, MV 41, MV 45, MV 52, JV 02, JV 04 and JV 05 when visualized using silver-stained 6% polyacrylamide gels. After the addition of 2 µl of SYBR Green I 1,000X, the color change reaction was detected in snails from the same collection sites, except MV 03 (Fig. 5) (Additional le 4).
Considering LS-PCR and conventional PCR as the reference tests, the optimized LAMP assay presented a sensitivity of 100% and speci city of 91.66%. Kappa statistics showed an "almost perfect" agreement of 0.88 between the LAMP assay and the other molecular methods evaluated in this study. Therefore, our results showed that either LS-PCR, conventional PCR, and/or the optimized LAMP assay could be used for xenomonitoring of transmission areas (Table 1), and that S. mansoni infection could be detected by all three methods in collection sites MV 41, MV 45, MV 52, JV 02, JV 04 and JV 05 (Fig. 1).

Discussion
Our work has identi ed active foci of S. mansoni transmission in six collection sites from the municipalities of Franciscópolis, Jequitinhonha, Joaíma, Malacacheta, and Ponto dos Volantes, con rming that these are endemic areas in the state of Minas Gerais. Molecular approaches enabled the detection of infected snails with higher accuracy than parasitological methods, reinforcing the need for additional tools to precisely map and monitor endemic areas, and, in the future, achieve schistosomiasis control and elimination.
The transmission of S. mansoni has been reported so far in 19 Brazilian states. The state of Minas Gerais includes around 70% of the endemic areas, being the subject of many studies over the years (3,(40)(41)(42)(43)(44) . Poverty contributes to increased contact of individuals with contaminated water, as populations from underprivileged areas usually seek natural watercourses for work and leisure activities, raising schistosomiasis transmission rates (41,43). In addition to knowledge regarding socio-economic aspects and habits, biological and ecological features are relevant for understanding the transmission process in each region (47). The temperature from the water body is a determining factor for the development of both the snail and the parasite, and prospective studies have predicted the future impact of climate change on the dynamics of transmission (48). Speci c vegetation and parasitism in snails can affect not only their populational density but also cercarial abundance (49).
Rain volume has a close relationship with the density of Biomphalaria snails and the positivity rate found among them for S. mansoni infection. According to Calasans et al (47), the abundance of Biomphalaria snails is negatively related to the rainfall, while Biomphalaria infection rises during wet periods of the year. Historically, the Mucuri and Jequitinhonha Valleys have low rainfall levels in all seasons (50). Between July and August 2019, when our surveys were undertaken, the accumulated rainfall ranged from 10-30 mm 3 according to the Meteorology National Institute (INMET) (51), being the lowest pluviometric measurements in that year. This indicates that the density of infected Biomphalaria snails found in our survey might have been underestimated due to the characteristics of rainfall while it was undertaken.
In this study, a total of 1,001 snails were collected from 18 sites The presence of trematode infection in snails was investigated using a multiplex PCR protocol that enables the differentiation of four important families commonly found parasitizing Biomphalaria snails (37,(60)(61)(62). Schistosomatidae species were detected in 44.4% (8/18) of the study sites, while Echinostomatidae and Strigeidae were each found in 5.5% (1/18). Snails from the collection sites MV 41 and JV 04 that were found shedding S. mansoni cercariae in the parasitological examination had their infection con rmed by multiplex PCR due to the ampli cation of 140 bp target corresponding to the Schistosomatidae family. In ve further sites from the Mucuri Valley, and one from the Jequitinhonha Valley, the presence of Schistosomatidae infection in snails was also detected. Even though the ampli cation of Schistosomatidae DNA does not necessarily mean the presence of S. mansoni itself, this result raises concern that these areas might be potential foci for schistosomiasis. As expected, no ampli cation was observed in the snails collected at the JV 03 site, since the set of primers used does not cover the Spirorchiidae family isolated from this location during the parasitological examination of snails. As the primers used by the multiplex PCR only amplify four trematode families, it is not possible to con rm that snails from the remaining collection sites are not infected by other trematode families.
LS-PCR and conventional PCR were both able to detect the presence of S. mansoni in 33.3% (6/18) of the surveyed sites. The optimized LAMP assay developed in this work revealed the presence of in snails from 38.8% (7/18) of the collection sites, when ampli cation was visualized using polyacrylamide gels, having an "almost perfect" Kappa agreement with LS-PCR and conventional PCR, with 100% sensitivity, and 91.66% speci city. When the chosen method to check the result was visual inspection of reaction tubes after the addition of an intercalating dye, ampli cation was detected in six collection sites, presenting the same result obtained when LS-PCR and conventional PCR were used. A very weak amplicon corresponding to the Schistosomatidae family was detected in snails from the collection site MV 03 but no ampli cation was detected with LS-PCR and conventional PCR using this sample as a template. The apparent LAMP product from this sample, when visualized using a polyacrylamide gel, raises the hypothesis that the LAMP assay was more sensitive than LS-PCR and conventional PCR in detecting S. mansoni infection in snails. However, the limit of detection of each method indicates that this is not the case, since LS-PCR can detect up to 1 pg of S. mansoni DNA (18), conventional PCR up to 0.01 pg (Additional le 3) and optimized LAMP assay presented a limit of detection of 0.1 ng. Therefore, LAMP is less sensitive than the other evaluated methods, suggesting that snails from the site MV 03 are not infected with S. mansoni. Although several trematode samples have been used to test the speci city of the optimized LAMP assay, samples from other species that belong to the Schistosomatidae family have not been used. The analysis of the results from the LAMP assay combined with multiplex PCR suggests cross-reactivity between members of the same family. In the Brazilian context, other schistosomes do not have much relevance to human health, as only S. mansoni causes schistosomiasis in this country. Avian schistosomes have been reported causing cercarial dermatitis in countries from the Northern Hemisphere, but the occurrence of this condition has not been reported so far in Brazil (63-65). As an alternative to overcome this false-positive result, we suggest that the visual inspection of reaction tubes by the naked eye should be prioritized, instead of running the LAMP products in gels. This visualization strategy not only reduces the possibility of false results, but also makes the assay more applicable directly in the eld in low infrastructure conditions.
The LAMP assay was rst described in 2010 (66), and since then this technique has been used to detect many pathogens, including S. mansoni, but mostly in human samples (38,(67)(68)(69)(70)(71). The applicability of LAMP for screening snails to characterize transmission areas is very promising (49), and has been tested by several authors (26-30). Molecular techniques can detect the presence of S. mansoni even when snails are not shedding cercariae, which would provide valuable information for surveillance services, as in many endemic areas collected snails rarely shed cercariae even though schistosomiasis transmission remains present. This failure to nd cercarial shedding can be misleading, often giving the false impression of low or even absent transmission (49). Although molecular methods such as conventional PCR, DNA sequencing, and qPCR can properly ful ll this gap, these techniques are inappropriate for laboratories with limited resources, as they require expensive machinery and technical expertise, raising the associated costs of each reaction. Among all advantages of the using isothermal assays, the possibility of performing the test directly in the eld in laboratories with limited infrastructure is undeniable. Our group optimized the LAMP assay using the primers described by Fernández-Soto et al (38). When following the exact conditions described by Gandasegui et al (29), non-speci c ampli cation was detected in trematode samples belonging to the families Diplostomidae and Spirorchiidae (Additional le 5). Even though the optimization resulted in a reduction in the analytical limit of detection (from 1 fg to 0.1 ng), our ndings con rmed that the amount of S. mansoni DNA that can be detected by the assay is su cient to detect, by visual inspection alone, the presence of the parasite 7 days after exposure of the snails to 8 miracidia, and in pooled samples.
By the use of the optimized LAMP assay, we detected three times more the infection by S. mansoni in snails when compared to parasitological examination using shell-crushing method, revealing six active transmission areas for schistosomiasis in both Mucuri and Jequitinhonha Valleys. Molecular methods also allowed the mapping of potential transmission foci through the identi cation of B. glabrata in much of the surveyed area, as demonstrated in the maps generated by this study.

Conclusion
Parasitological methods based on the detection of S. mansoni larval forms in Biomphalaria snails are limited and are affected by the variation in disease prevalence in different regions, such that false negative results may often be obtained when these methods are applied. The optimization of a LAMP assay provides a sensitive, speci c, rapid, and precise diagnostic alternative, with a performance as good as other molecular approaches evaluated in this study. However, as an isothermal method, LAMP is easier to perform directly in the eld or in low-infrastructure laboratories. Considering the challenges to controlling schistosomiasis, or even to interrupt its transmission in endemic areas, mapping and monitoring transmission foci with higher accuracy will improve decision-making processes enabling more appropriate allocation of public funding and resources aimed at the elimination of schistosomiasis as a public health problem. mansoni infection was detected in snails after the use of molecular methods. Note: The designations employed and the presentation of the material on this map do not imply the expression of any opinion whatsoever on the part of Research Square concerning the legal status of any country, territory, city or area or of its authorities, or concerning the delimitation of its frontiers or boundaries. This map has been provided by the authors.