Evaluation of the molluscicidal activity of bacillus spp. isolates to control aquatic intermediate host snails of liver fluke (fasciola spp.).

Abstract

Aquatic snails are involved in harmful disease cycles of Fasciola (liver fluke) that affect both humans and livestock in agriculture. The focus of this study was to isolate and identify candidate bacterial isolates antagonistic to aquatic snails, with the ultimate goal of controlling the host snails responsible for the transmission of liver flukes in South Africa. A bacterial antagonist would offer a novel means of snail population control and reduce the dependence on chemicals, and the growing risk of resistance. Due to their molluscicidal capabilities reported in the literature, and the benefits of being endospore-formers, strains within the family Bacillaceae were targeted as candidate biocontrol agents. A population of the freshwater snail Physella acuta (Draparnaud, 1805) was established and used for screening bacterial candidates as an easily-reared, proxy species for the Fasciola spp. intermediate host snails. Aerobic endospore-forming bacteria were isolated from aquatic soil collected primarily in the KwaZulu-Natal province, South Africa, utilising several approaches, including a general endospore heat-shock isolation method and two Bacillus thuringiensis (Bt) (Berliner, 1915) specific isolation methods. Initial screening for molluscicidal activity of 1180 isolates did not yield any strong performers; however, a subset of 124 isolates demonstrated some evidence of potential activity in preliminary screening, that in the absence of any strong molluscicidal isolates, warranted a more in-depth investigation. Subsequently, when the bioassays were repeated none of the 124 isolates showed strong molluscicidal activity; however, the eight isolates causing the highest mortality rates (16.7 – 50%) were selected for further analysis to determine whether these isolates exhibited weak molluscicidal activity. When spent culture supernatant was evaluated in molluscicidal bioassays 12 of the 124 isolates demonstrated strong molluscicidal activity. Selected isolates underwent identification and characterisation using DNA-based methods for the purposes of species identification and de-duplication of isolates on the basis of sequencing data and end point PCR. Isolates were identified using the benchmark technique of 16S rDNA gene fragment sequencing. For discrimination of closely related isolates, sequence analysis of two additional housekeeping genes was performed using rpoB and dnaJ. In addition, a number of end point PCRs were used to support sequence data by the presence or absence of PCR products for the lipopeptide marker genes for surfactin, fengycin, bacillomycin and iturin, as well as a Bacillus velezensis (Ruiz-García et al., 2005) specific primer. All the isolates were confirmed to be members of the Bacillaceae. Isolates selected for their production of molluscicidal supernatants were found to be closely related and identified as strains of B. velezensis, Bacillus amyloliquefaciens (Priest et al., 1987) and Bacillus subtilis (Ehrenberg, 1835). Isolates with potential molluscicidal activity were found to be more diverse, with representatives of B. velezensis, Bacillus cereus (Frankland and Frankland, 1887), Bacillus mycoides (Flügge, 1886), Paenibacillus sp., Priestia sp., and Gottfriedia sp. being identified. Isolates were deduplicated by removal of replicates with identical results for the various gene sequence data and end point PCRs. From the original 12 isolates producing molluscicidal supernatant, 6 were retained and all 8 isolates with potential molluscicidal activity were carried forward for further investigation. Molluscicidal supernatant producers underwent lipopeptide extraction procedures and bioassays to determine whether this class of compounds may be responsible for their observed molluscicidal activity. The 8 potential molluscicidal isolates, along with the 6 molluscicidal supernatant producers, were assayed for molluscicidal activity against P. acuta snails in a longterm (i.e. 2 weeks feeding endospore-impregnated food pellets followed by 4 weeks observation) endospore exposure assay. Molluscicidal active fractions were successfully extracted and concentrated from broth culture supernatants via a lipopeptide acid precipitation extraction protocol. This suggested that lipopeptides, or compounds extractable by acid precipitation, were responsible for the observed molluscicidal activity. Lipopeptide extracts showed molluscicidal activity at between 25 and 200 μg.mL-1 concentrations. In contrast, extended endospore exposure assays yielded no significant molluscicidal results. For the 8 isolates brought forward based on their potential molluscicidal activity, this result confirmed the lack of direct molluscicidal activity identified in the initial screening. Additionally, for isolates that produced molluscicidal supernatant, exposure of snails to these isolates’ endospores did not result in any observable molluscicidal activity, suggesting that only the excreted metabolites were responsible for their molluscicidal activity and that this did not occur at sufficient concentrations in the snail-tank system to demonstrate an effect. Subsequent investigations focused on the extracted molluscicidal compounds, their characterisation and identification, as well as the measurement of their lethal concentrations. Crude lipopeptide extracts from Landy and TSB growth media with molluscicidal activity against P. acuta underwent UPLC-ESI-MS to identify the lipopeptide components. In addition,

the dose response curves were measured for 24 h and 72 h contact times. The LC50 for crude lipopeptide extracts with a 24 h exposure ranged between 16.37 μg.mL-1 and 36.16 μg.mL-1, and the LC95 was between 20.05 and 48.68 μg.mL-1. The LC50 for the crude extracts with a 72 h exposure was between 11.89 μg.mL-1 and 27.21 μg.mL-1, and the LC95 was between 15.09 and 30.93 μg.mL-1. Analysis of the UPLC-ESI-MS spectra for each molluscicidal crude extract indicated the presence of various lipopeptide isoforms, including surfactin, iturin, fengycin and bacillomycin-D and -L. Surfactin was common to all molluscicidal crude lipopeptide extracts examined, which suggests that surfactin was a major contributor to the observed molluscicidal activity. The molluscicidal properties of pure surfactin was evaluated against two aquatic snail species, P. acuta, which was used in initial screening, and secondly against Pseudosuccinea columella (Say, 1817), a host snail of Fasciola liver flukes. The molluscicidal efficacy of the crude lipopeptide extracts and pure surfactin were assessed against a test fish species, Danio rerio (F. Hamilton, 1822), as a first step in evaluating ecotoxicity. Commercially-available pure surfactin was found to be molluscicidal against P. acuta and P. columella, with LC50 values of 10.04 μg.mL-1 and 16.58 μg.mL-1, respectively, and LC90 of 12.29 μg.mL-1 and 19.15 μg.mL- 1, respectively, with a 24 h contact time. Crude lipopeptide extracts demonstrated lethal effects on 24 hpf D. rerio embryos with an LC50 of between 10.19 and 44.93 μg.mL-1 with a 24 h contact time. Danio rerio exposure to pure surfactin was found to be lethal, with LC50 and LC90 values of 7.96 and 11.45 μg.mL-1, respectively, with a 24 h exposure period. In comparison to 24 hpf embryos, 24 hpf eggs are still encased in the chorion. The 96 hpf embryos showed lower sensitivity to surfactin with LC50 values of 13.96 and 12.26 μg.mL-1, and LC90 values of 17.61 and 16.87 μg.mL-1, respectively, with 24 h contact times. This is the first report of surfactin having molluscicidal characteristics but it does raise some ecotoxicological concerns for its potential use as a molluscicide. This research is the first broad-scale screening of aerobic endospore-forming bacteria for use as molluscicidal biocontrol agents. An initial screening of 1180 isolates did not reveal any strong molluscicidal isolates with direct activity on aquatic snails. However, an investigation into secreted metabolites determined that lipopeptide extracts exhibited molluscicidal effects, of which surfactin was common to all active fractions. Pure surfactin assays proved that this lipopeptide has molluscicidal activity at low concentrations, with LC50 values of

10.04 μg.mL- 1 and 16.58 μg.mL-1 for P. acuta and P. columella, respectively, with a 24 h exposure time. This is a novel finding, increasing the range of activity of this biotechnologically useful compound. While no isolates from this research have demonstrated utility as biocontrol agents, the finding of molluscicidal lipopeptides opens a new avenue for the biocontrol of aquatic snail pest species, which may impact the neglected tropical diseases fascioliasis and schistosomiasis, with further potential for the control of other aquatic snail pest species.