Study area

Snails were collected during malacological surveys carried out in September 2020, mainly in the Senegal River Delta (SRD) in the North-West (NW) of Senegal. Specimens were collected in September and November in the Central Senegal (CS), in the Diourbel region, and in the Souht-East (SE) of the country, in the Kedougou region, respectively (Fig. 1a) Both the SRD and the SE of the country are endemic areas for schistosomiasis [11]. We also included in our study a Biomphalaria strain originating from Kedougou, which had been raised for two years in a laboratory.

Fig. 1

a The map of Senegal showing the snail sampling sites, produced using the Geographic Information System software QGIS v3.18.3-Zürich: http://www.qgis.org. b The different snail species collected were (1) Bi. pfeifferi SRD (NW), (2) Bu. forskalii SRD (NW), (3) Bu. forskalii Diourbel (CS), (4) Bi. pfeifferi Kedougou (SE) laboratory strain, and (5) Bi. pfeifferi SE wild strain. c An explanatory flowchart of the MALDI–TOF MS protocol. ACN: Acetonitrile, CHCA: α-cyano-4-hydroxycinnamic acid (created using BioRender.com)

The SRD area is the terminal part of the Senegal River and is located in the North-West of the country between latitudes 16° and 14°40′ North, and longitudes 15°30′ and 16°30′ West [11]. The SRD is a coastal area [32] extending from the Dagana region to Saint-Louis, and covering an area of ~ 6000 km2 (Fig. 1a). It is characterised by a semi-desert climate [11] with average minimum temperature of 25.1 ℃ and a maximum of 33.2 ℃. In September 2020, the average rainfall was 188 mm (mean relative humidity, 80.9%) [33]. The area became endemic for schistosomiasis after the construction of the Diama and Manantali dams in the 1980s, and is characterised by stable transmission [34]. In the Delta zone, we surveyed seven sites (Savoigne: 16°10′N/16°18′W, Keur Samba: 16°11′N/16°16′W, Kabane: 16°30′N/16°24′W, Ndiawdoune: 16°40′N/16°23′W, Minguene: 16°01′N/16°21′W, Mbakhana: 16°05′N/16°22′W and Ndiol Maure: 16°09′N/16°18′W), where we collected snails (n = 16, n = 60, n = 39, n = 14, n = 12, n = 6, and n = 15, respectively).

In South-Eastern Senegal, snails (n = 16) were collected in the Kedougou region at a single site in the village of Ngari (12°32′N, 12°11′W) (Fig. 1a). The Kedougou region is crossed by the Gambia River, close to its source in the Fouta Djallon and tributaries such as the Niokolo Koba [35]. It is characterised by a Sudano-Guinean climate with a single rainy season from May to November [36]. Average minimum temperatures of 23.1 ℃ and 20.7 ℃ and maximum temperatures of 33.1 ℃ and 37.5 ℃, with average monthly rainfall of 226.57 mm and 0.0 mm (mean relative humidity of 79.4% and 42.2%) in September and November 2020, respectively, were recorded [33]. The area features natural water sources, which create favourable biotopes for snails, the intermediate hosts of schistosomiasis [37]. The continued transmission of schistosomiasis in this area is due to the stability of the breeding sites, provided in certain places by dense vegetation that limits the action of solar radiation during the hot periods (40 ℃ between May and June). These permanent watering places concentrate all the villages’ water-based activities [37].

In contrast, Central Senegal is characterised by temporary pools of water, associated with a seasonal transmission of urinary schistosomiasis, depending on the rainy season [7, 17]. Our collection site here was located in the Diourbel region, 150 km east of Dakar, between latitudes 14°30′ and 15° North and longitudes 15°40′ and 16°40′ West [38] (Fig. 1a). It is a semi-urban area, currently not described as endemic for schistosomiasis, but featuring temporary pools. The climate in this area is Sudano-Sahelian, characterised by relatively high temperatures, a long dry season (November to June), and a four-month rainy season (July to October). In September 2020, the average minimum temperature was 24.8 ℃, the maximum temperature was 33.6 ℃ and there was an average monthly rainfall of 419.36 mm (mean relative humidity 82%) [33]. One site in the village of Touba-Ndiareme (14°37′N/16°12′W) was surveyed and four snails were collected.

Snail collection and morphological identification

With regards to collecting snails from ponds, flexible forceps were used to collect snails from the vegetation surrounding the waterholes or from any other material, including branches and dead leaves. In the river, we scraped the vegetation with a long-handled landing net. The vegetation was then shaken into a container, where the snails fell and were collected. Snails from the same water point were identified morphologically and kept in the same pre-labelled container, noting the collection site, in order to facilitate their transportation. The identification of the snails was based on shell morphology [39]. Some specimens, which were difficult to identify with the naked eye, were observed using a binocular magnifying glass and the Zeiss Axio Zoom V16 microscope (Zeiss, Marly-le-Roi, France) (Fig. 1b). The snails were classified according to species and geographical area, and were then stored at –80 ℃.

Sample preparation for MALDI–TOF MS analysis

Each snail specimen was carefully extracted from its shell and dissected with a new sterile blade to collect the foot. Each foot was successively rinsed with 70% ethanol and distilled water for two minutes and dried on sterile filter paper for MALDI–TOF MS analysis, while the remainder of the body was stored at −20 ℃ for further genomic analysis. The snail foot was selected because it provides a better source of protein than other tissues for MALDI–TOF MS identification [31, 40]. The feet of the Bi. pfeifferi and Bu. forskalii snails were placed individually in 1.5 ml Eppendorf tubes with glass beads, ≤ 106 μm (Sigma, Lyon, France), and mix containing 70% (v/v) formic acid (Sigma-Aldrich, Lyon, France), 50% (v/v) acetonitrile (Fluka, Buchs, Switzerland) and high-quality liquid chromatography (HPLC) water. All samples were ground with 30 μl of the mixture using a TissueLyser II (Qiagen, Hilden, Germany) over three one-minute cycles at 30 Hz.

Sample loading on the target plate and MALDI–TOF MS settings

The samples were then centrifuged at 2000×g for 30 s and 1 μl of the supernatant of each homogenate was deposited on a MALDI–TOF MS target plate (Bruker Daltonics, Wissembourg, France) in ten copies. Each deposit was covered with one microlitre of a CHCA matrix suspension, composed of saturated α-cyano-4-hydroxycynnamic acid (Sigma, Lyon. France), 50% acetonitrile (v/v), 2.5% trifluoroacetic acid (v/v) (Aldrich, Dorset, United Kingdom), and high-performance liquid chromatography (HPLC) water to allow for co-crystallisation. After drying for several minutes at room temperature, the target was introduced into the Microflex LT instrument (Bruker Daltonics, Bremen, Germany) for analysis (Fig. 1c).

MALDI–TOF spectral analysis

Protein mass profiles were obtained using a Microflex LT instrument (Bruker Daltonics, Germany), using Flex Control version 2.4 software (Bruker Daltonics). This performs positive ion measurements in linear mode at a laser frequency of 50 Hz, in a mass range of 2 kDa to 20 kDa. Mass spectra were analysed over an m/z range of 2000 to 20,000. Each spectrum corresponds to the ions obtained from 240 laser shots fired in six regions of the same deposit on the ground plate and acquired automatically using the AutoXecute function of the FlexControl v.2.4 software (Bruker Daltonics GmbH & Co. KG, Bremen, Germany). Spectral profiles obtained from Bi. pfeifferi and Bu. forskalii snail feet were displayed with FlexAnalysis v.3.3 software and exported to ClinProTools v.2.2 (Bruker Daltonics) and MALDI Biotyper v.3.0 (Bruker Daltonics, Germany) for data processing.

Intraspecific reproducibility and interspecific specificity were assessed by comparing the spectral profiles obtained from the ten spots of each snail specimen. Spectral quality was confirmed using FlexAnalysis software making it possible to assess the intensity, peak regularity, baseline flatness, and inter- and intra-group reproducibility of the snails. The original spectra of each snail species were imported into ClinProTools for principal component analysis (PCA). Poor quality spectra (low peak intensity < 3000 arbitrary units (au), and/or no reproducibility) were excluded from the analysis. A dendrogram was created using MALDI Biotyper software to visualise the heterogeneity level of MS spectra from snail groups.

Blind tests

To confirm the morphological identification of the snails, the MALDI–TOF MS spectra obtained from the foot of each specimen were queried using MALDI Biotyper against our in-house reference spectra database including 64 spectra from eight snail species collected in two areas of Senegal, namely Richard-Toll in the north, and Niakhar in the centre. The level of similarity is estimated by the log score value (LSV) that correspond to the degree of homology between the query and the reference spectra in the database. The blind test report score can range from 0 to 3 depending on the degree of matching between the spectral signal intensities. A sample was considered to be correctly identified when the spectrum analysed provided an LSV value ≥ 1.7. For each sample, four spectra with the highest LSV of the ten were selected for further analysis. The in-lab database contains reference spectra of a number of freshwater gastropod species and is available online at https://doi.org/https://doi.org/10.35088/f605-3922: raw-data-frozen-and-ethanol-stored-snails.

MS data analysis and interpretation

The ClinProTools and MALDI Biotyper software packages (Bruker Daltonics GmbH & Co. KG, Bremen, Germany) were used for data analysis and interpretation. The Flex Analysis software was used to visualise the raw spectra obtained using the MALDI–TOF Microflex LT mass spectrometer. The raw spectral data were processed by PCA to visualise the inter-species MS profile dissimilarities produced by ClinProTools. The same software was used to identify the discriminating peaks between specimens in each of the two species Bu. forskalii and Bi. pfeifferi, based on their geographical origin. A dendrogram was also drawn with MALDI Biotyper v.3.0.

The spectra of Bi. pfeifferi snails collected in the populations located in North-Western Senegal (in the SRD), South-Eastern Senegal (in Kedougou), including both field (WS) and laboratory strains (LS), as well as those of Bu. forskalii collected in North-Western and Central Senegal (Diourbel), were analysed using ClinProTools. In our analyses, we selected four spectra for each specimen in each geographical zone. For the PCA, all Bi. pfeifferi specimens collected from the field in Kedougou (WS) (n = 6) (MP5Bi1, MP5Bi2, MP5Bi4, MP5Bi5, MP5Bi6 and MP5Bi7) and from the laboratory (LS) (n = 10) (KgBi1, KgBi12, KgBi13, KgBi14, KgBi15, KgBi16, KgBi17, KgBi18, KgBi19 and KgBi20) were included. To balance the number of specimens of each geographical origin, we randomly selected ten snails collected in the SRD area for each species (Bi. pfeifferi: KABiI2, KABiI3, KABiI4, KABiI6, KABiI7, KABi1, KABi5, KABi7, KABi8 and KABi9, and Bu. forskalii: KABfI1, KABfI2, KABfI3, KABfI4, KABfI5, KABf1, KABf2, KABf3, KABf5 and KABf7). In the centre, in the Diourbel area, we used three specimens of Bu. forskalii (KhBf1, KhBf2 and KhBf3).

MALDI–TOF MS biomarker mass set

To determine differential peaks, spectral mass profiles were loaded into ClinProTools (Bruker Daltonics) based on a subset of three Bu. forskalii samples and twelve Bi. pfeifferi samples. The parameter settings in ClinProTools for spectral data preparation consisted in baseline subtraction (top-hat; minimum baseline width of 10%), recalibration (maximum peak shift of 1000 ppm and 30% match to calibrant peaks, and exclusion of impossible-to-recalibrate spectra), calculation of the average spectrum (resolution 800), calculation of the average peak list (signal-to-noise threshold 2.00), calculation of peaks in individual spectra, and normalisation of peak lists.

Groups were created for the different snail populations, including group A (population 1 = Bu. forskalii NW samples and population 2 = Bu. forskalii Central Senegal samples) and group B (population 3 = Bi. pfeifferi NW samples and population 4 = Bi. pfeifferi SE lab strain and wild strain samples).

Biomarker peaks were identified using the “Peak Statistic” function of ClinProTools, followed by manual confirmation that the same peaks could be distinguished using FlexAnalysis. A genetic algorithm in ClinProTools provided the highest “recognition capability” (RC) and “cross-validation” (CV) values with the lowest number of peaks. These values indicate the capacity to split between different classes of spectra based on the selected discriminating peaks.

DNA extraction and nucleotide sequence analysis

Samples which were well identified, both morphologically and by MALDI–TOF MS (LSV ≥ 1.7), were also randomly selected for further DNA-based identification. For each snail, the remaining specimen was rinsed with distilled water and placed in a 1.5 ml Eppendorf tube for genomic DNA extraction. In short, each sample was incubated at 56 ℃ overnight with 180 μl of G2 lysis buffer (Qiagen Hilden, Germany) and 20 μl of proteinase K (Qiagen Hilden, Germany). The supernatant was recovered in another tube and then extracted using the EZ1 BioRobot extraction device (Qiagen Hilden, Germany) employing the EZ1 DNA Tissue Kit (Qiagen) according to the manufacturer’s instructions. Genomic DNA of each sample was eluted with 200 μl of Tris–EDTA buffer (Qiagen) and stored at – 20 ℃ until use.

For DNA-based identification, the PCR template was DNA extracted from a specimen previously identified by MALDI–TOF MS. PCR reaction targeting a 710 bp region of cytochrome c oxidase subunit I (COI) and a 550 bp region of the 16S rRNA was performed in a thermal cycler (Applied Biosystems, 2720, Foster City, USA) with AmpliTaq Gold 360 PCR master mix (Applied Biosystems, Waltham, USA). The COI region was amplified using Folmer’s universal primers LCO1490 (5′-GGTCAACAAATCATAAAGAT ATTGG-3′), HCO2198 (5′-TAAACTTCAGGGTGACCAAAAAATCA-3′) [41, 42] and 16S with forward 16Sar-L (5′-CGCCTGTTTATCAAAAACAT-3′) and reverse 16Sbr-H (5′-CCG GTCTGAACTCAGATCACGT-3′) [43]. The amplification protocol consisted in an initial denaturation at 95 ℃ for 15 min, followed by 40 cycles (35 cycles for 16S) at 95 ℃ for 30 s, at 40 ℃ for 30 s (at 55 ℃ for 50 s for 16S), 72 for one minute 30 s (1 min for 16S) and a final step at 72 ℃ for 7 min. PCR product migration for 25 min at 180 V in a 1.5% agarose gel with SYBR Safe dye was read using a Gel Doc System (Bio-Rad, Hercules, USA). The amplicons were purified using Macherey Nagel plates (NucleoFast 96 PCR, Düren, Germany) and were sequenced using the same primers. The BigDye Terminator v1.1, v3.15 × Sequencing Buffer (Applied Biosystems, Warrington, UK) was run using an ABI 3100 automated sequencer (Applied Biosystems). The sequences obtained were assembled and analysed using the Chromas Pro v.1.77 software (Technelysium Pty. Ltd, Tewantin, Australia) and further queried against the National Center for Biotechnology Information (NCBI) online nucleotide database with the Basic Local Alignment Search Tool (BLAST) (http://blast.ncbi.nlm.nih.gov). A maximum likelihood phylogenetic tree was constructed with MEGA version 7.0.26 software [44, 45]. Statistical support of the internal tree branches was assessed by 1000 bootstrapping replicates.

Data analysis

To determine differential peaks, spectral mass profiles with ClinProTools (Bruker Daltonics), representative peaks among the different groups were selected using several statistical tests, including the t-test, the analysis of variance test (ANOVA), Wilcoxon or Kruskal–Wallis (W/KW) tests, and the Anderson–Darling (AD) test. A P-value of 0.05 was set as the statistical significance cutoff [46]. A characteristic peak was considered when P < 0.05 in the AD test and P-value in the W/KW test was also < 0.05. When P = 0.05 in the AD test, a characteristic peak was selected if the corresponding P-value in the ANOVA was < 0.05 [47]. Informative peaks were those that were statistically significantly different between populations.