Introduction

Through a combination of vector control, reactive screening, passive diagnosis, new therapeutic drugs and the support of national control programmes, donors and stakeholders, interrupting the transmission of gambiense human African trypanosomiasis (HAT) at pan-African level is in reach (Barrett et al., 2024). Reported annual cases have been maintained below 1,000 since 2018, making the elimination (zero transmission) of gambiense HAT (gHAT) by 2030, one of the major objectives of the World Health Organisation (WHO), achievable (Franco et al., 2024). Nevertheless, animal African trypanosomosis (AAT), also known as nagana, remains a significant burden to sustainable agricultural development, and represents a threat to livestock and small holder economies (Boulangé et al., 2022) with a total estimated loss of 4.75 billion USD per year in terms of agricultural gross domestic product in tsetse-infested lands (Budd, 2001). Unlike HAT, that is caused by either Trypanosoma brucei (T. b.) gambiense in West Africa, or T. b. rhodesiense in East Africa, AAT encompasses a wider range of trypanosomatid parasites. In sub-Saharan Africa, animal infecting trypanosomes including T. vivax, T. congolense, T. brucei, T. theileri, T. simiae and T. suis are transmitted by blood-sucking flies, such as Glossinidae and Tabanidae, to a wide range of wild and domestic mammals. Socio-environmental factors such as land cover, humidity, temperature and livestock migration impact trypanosome transmission (Karshima et al., 2016). This is particularly evident in rural areas of sub-Saharan Africa, where high incidences of T. brucei, T. congolense and T. vivax are observed, due to conditions favourable to the proliferation of flies (Okello et al., 2022).

Symptoms of AAT are host and pathogen-dependent, with an infection potentially leading to spontaneous abortions, reduction in milk and meat production, delays in sexual maturity, low calving rates, loss of draught power, and ultimately to high mortality, all of which compound poverty in many Sub-Saharan countries (Kristjanson et al., 1999). Indirect control strategies including the combination of vector control, breeding of trypanotolerant animals, and avoiding grazing in infested areas have proved to be efficient in reduce the number of infections (Meyer et al., 2016). Chemotherapy and chemoprophylaxis are the main means for controlling AAT, although inadequate usage, presence of fake drugs and emergent drug resistance are hampering trypanocide efficacy (Giordani et al., 2016; Richards et al., 2021). Reliable and adapted diagnostic tests are crucial for parasite detection to better target infected animals for treatment, as well as for parasite identification to understand the distribution of trypanosomes in endemic areas. Microscopic, serological and molecular methods are generally used to diagnose AAT. Microscopy based methods are the most commonly used diagnostic methods, but lack sensitivity with a limit of detection (LoD) ranging from 10⁴ to 10⁵ parasites/mL when using fresh blood, to 250 to 5,000 parasites/mL for methods combining centrifugation (Desquesnes et al., 2022a). Serological methods, such as the card agglutination test (CATT) and ELISA, can be used to identify anti-Trypanosoma antibodies, but their reliability is often compromised by the poor antigen characterization in animal infecting trypanosomes or by the detection of antibodies from previous infections. In contrast, molecular diagnostic techniques, including PCR and Loop-mediated isothermal amplification (LAMP), offer increased sensitivity (73-100 %) and LoD (1-1000 parasites/mL) by targeting specific regions of the parasites DNA. For the main AAT-related Trypanosoma species, pan-species primers (targeting ITS1 for instance) and genus/species/type-specific primers are available and currently used to detect the DNA of parasites from the subgenus Nannomonas (T. congolense), Duttonella (T. vivax) and Trypanozoon (T.b.brucei, T. evansi and T. equiperdum) (Álvarez-Rodríguez et al., 2022; Desquesnes et al., 2022a).

AAT-related trypanosome species circulate co-endemically with T. b. gambiense in West and Central Africa, where domestic animals such as pigs, dogs, cattle, sheep and goats are suspected reservoirs of T. b. gambiense (Magang et al., 2023; Traoré et al., 2021). In these regions, Suids (warthogs, domestic and bush pigs) and bovids (bushbucks and buffalos) are the preferred hosts for most tsetse fly species (Serem et al., 2024). In a historical HAT focus of Côte d’Ivoire, free-range pigs sharing the same environment as humans were identified as reservoirs for T. b. gambiense (Traoré et al., 2021). The diagnostic techniques employed in this study, involving single round (Radwanska et al., 2002) or nested PCR (Morrison et al., 2008), however tended to overestimate the prevalence of the human infective parasites in pigs (Traoré et al., 2021). Therefore, an active and accurate surveillance of the infection status of domestic animals using a more reliable diagnostic tool will be key to reach and sustain gHAT elimination (Simo and Rayaisse, 2015; Vourchakbé et al., 2020). In addition, a comprehensive map of the animal infective trypanosomes circulating in tsetse-infested areas would enable an improved control strategy and prevent or limit further economic losses.

In this context, developing new diagnostic tools that are species-specific and highly sensitive is essential to accurately diagnose infected individuals and define the circulation of trypanosomes in domestic animals. For this purpose, we have adapted the Specific High-sensitivity Enzymatic Reporter unLOCKing (SHERLOCK) detection that offers unprecedented levels of diagnostic specificity and sensitivity (Gootenberg et al., 2017), and has been adapted for the detection of HAT (Sima et al., 2022). SHERLOCK combines an isothermal amplification by recombinase polymerase (RPA) with Cas13a detection of the target RNA (Gootenberg et al., 2017; Kellner et al., 2019; Piepenburg et al., 2006). The RPA step, in combination with a reverse transcriptase, allows the use of total nucleic acid (TNA) as an input, increasing the flexibility of the assay. Here, we have developed a broad panel of SHERLOCK assays for the detection of the six most prevalent AAT-related species. We have applied the SHERLOCK4AAT panel to two populations of domestic pigs from two eco-epidemiological regions of West Africa to determine their importance as reservoirs of both animal and human-infective trypanosomes. A total of 424 blood samples were collected (1) from free-ranging pigs in the last hypo-endemic foci of gHAT in Côte d’Ivoire, and (2) from farm pigs bred in the historical gHAT focus of forested Guinea where unconfirmed seropositive cases of gHAT are still sporadically detected (Courtin et al., 2019). These samples were screened using the complete SHERLOCK4AAT panel to provide an in-depth profile of the trypanosome species circulating among these domestic animals.

Materials and Methods

Ethics statement

The only reliable method available to maintain and amplify field-isolated animal trypanosomes is to experimentally infect laboratory animals, here mice. For this study, experimental infections were carried out at UMR INTERTRYP and were approved by the regional ethics committee for animal experimentation as part of the project APAFIS #34149-2020091118136527v5 and authorized by the French Ministry for Higher Education and Research. In Guinea, ethical approval was obtained from the Comité National d’Ethique pour la Recherche en Santé (CNERS) under agreement number 102/CNERS/19. In Côte d’Ivoire, sample collection was conducted within the framework of epidemiological surveillance activities supervised by the HAT National Elimination Program (HAT-NEP). No ethical statement is required by local authorities for domestic animal sampling. Any veterinarian may carry out blood sampling on domestic animals, with the authorization of the owner, as it is performed during prophylaxis or diagnostic campaign. Breeders gave their consent for animal sampling after being informed of the objectives of the study. For pig care, venous sampling was performed by a veterinary of the Laboratoire National d’Appui au Développement Rural (Ministry of Agriculture).

Mice infection and parasite purification

24 hours prior to infection with trypanosome parasites, 8-week-old female BALB/c mice (Charles River Laboratory) were immunosuppressed by intraperitoneal injection of 0.1 mL Endoxan 500 mg (Baxter). T. congolense Savannah subgroup IL1180 strain (Tanzania, Serengeti 1971, STIB212), T. vivax Magnan strain (Burkina Faso, Magnan 2011) and T. theileri (Nederland, Utrecht 2003, UNUT) parasites (INTERTRYP unit’s cryobank) were inoculated by intraperitoneal injection of 0.15 mL of the cryopreserved blood containing parasites. Two mice were infected per strain and parasitaemia was monitored daily by tail blood sampling. Parasite yields were estimated by counting in a fast read 102 counting cells (Kova). Whole blood was collected by cardiac puncture under deep isoflurane anaesthesia when parasitaemia reached 1 x 10⁸ cells per mL. Infected blood was passed over a diethylaminoethyl (DEAE)-cellulose column and the column washed using phosphate buffer saline (PBS) supplemented with 1 % glucose (PBSG). Parasites were collected in PBSG and concentrated by centrifugation (2000 rpm, 10 min, RT).

Cell culture

T. b. brucei Lister 427 bloodstream form and T. b. gambiense ELIANE cells were cultured in HMI-11 medium (ThermoFisher scientific) supplemented with 10 % fetal bovine serum (Sigma) at 37°C with 5 % CO2. Human embryonic kidney (HEK) 293T cells were cultured in complemented dMEM (Merck) at 37°C with 5 % CO2.

Epidemiological surveys

Pig blood samples were collected during two independent surveys in Côte d’Ivoire and Guinea. In Côte d’Ivoire, 224 free-ranging pigs were swampled in May, June and July 2021 in the Bonon and Sinfra hypoendemic HAT foci, the Bouaflé and Vavoua historical HAT foci and the Brobo HAT-free area {Kaba, 2023 #8655}. In Guinea, 200 farm pigs were sampled in March 2023 in the N’zérékoré prefecture, an ancient gHAT focus. Four drops of blood were dried on 4^th grade Whatman filter paper (Cytivia Life Sciences) and conserved in sealed plastic bags with silica gel until total nucleic acid (TNA) extraction.

Nucleic acid extraction

RNA was extracted from bloodstream parasites (T. b. brucei Lister 427, T. b. gambiense ELIANE, T. congolense Savannah subgroup IL1180, T. vivax Magnan and T. theileri), HEK 293T cells and whole blood from French cattle, pig, sheep and goat (Bergerie de la Combe aux Loups). Bloodstream form parasites from infected mice, as well as cultured HEK 293T cells, were harvested in TRIzol (Life Technologies) and Chloroform:Isoamyl alcohol (24:1, Sigma-Aldrich) was added, followed by centrifugation at 13,000 rpm at 4°C for 15 min. The upper aqueous phase was transferred to a fresh tube containing 0.5 mL of 75 % ethanol. After incubation for 10 min at room temperature, RNA was extracted using a NucleoSpin RNA column (Macherey Nagel) or a RNAeasy kit (Qiagen) according to the manufacturer’s instructions. 500 µL of blood sampled from French cattle, pig, sheep and goats was treated with 500 µL lysis buffer supplemented by 50 µL Proteinase K and incubated for 30 min at 56°C. The resulting mixture was transferred into a QIAshredder column (Qiagen) and centrifuged at 13,000 rpm for 2 min. The flow-through was then used for RNA extraction using a RNeasy Mini kit (Qiagen) according to the manufacturer recommendations. After extraction, qualities and quantities of RNA were evaluated using the NanoDrop™ One (Ozyme).

For extraction of TNA from dry blood spots (DBS), three 6 mm-punches from one spot of each animal DBS were resuspended in 300 µL of lysis buffer supplemented by 30 µL of Proteinase K and incubated for 30 min at 56°C. The punches and liquid were transferred into a QIAshredder column (Qiagen) and centrifuged at 13,000 rpm for 2 min. TNA was extracted from the flow-through using the Maxwell RSC Buccal Swab Kit (AS1640) with the Maxwell RSC Instrument (AS4500) according to the manufacturer’s instructions. For each sample, all SHERLOCK assays were performed from a single TNA extraction.

Target selection, RPA primers and crRNAs design

Alignments of the 18S rRNA, IFX and GAPDH genes were performed on sequences from different species (Supplementary Table 1) using Clustal Omega (Sievers and Higgins, 2018). The distance matrix representing the percentage identity between each sequence was represented using GraphPad Prism software (V.10).

BLAST analysis (McGinnis and Madden, 2004) of the 18S rRNA and GAPDH genes selected target sequences was performed in the NCBI standard database of using the following selection of organisms: Microsporidia (taxid:6029), Ciliophora (taxid:5878), Haemosporida (taxid:5819), Diplomonadida (taxid:5738), Symbiontida (taxid:797087), Diplonemea (taxid:191814), Euglenida (taxid:3035), Kinetoplastea (taxid:5653), Metakinetosplastina (taxid:2704647), Trypanosomatida (taxid:2704949), Trypanosomatidae (taxid:5654) and Euglenozoa (taxid:33682). The program selection was optimized for somewhat similar sequences (blastn) and in algorithm parameters, the maximum number of aligned sequences to display was set as 5000.

The presence of single nucleotide polymorphism (SNPs) in the target regions was assessed by aligning sequence variants of the parasite species using Clustal Omega (Sievers and Higgins, 2018). Sequence variants used are listed in Supplementary Table 1. SNPs identified in the TriTrypDB (https://tritrypdb.org) database were also taken into account in the design: NGS_SNP.Tb927_03: v5.1.921431, v5.1.921465, v5.1.921571, v5.1.921612, v5.1.921937, v5.1.922234, v5.1.922298, v5.1.922511, v5.1.922637, v5.1.922803, v5.1.923180, v5.1.959274; NGS_SNP.Tb927_07: v5.1.1951525, v5.1.1952283, v5.1.1951291; NGS_SNP.Tb927_02: v5.1.371291, v5.1.371584 and v5.1.370058.

Primers for Recombinase Polymerase Amplification (RPA) of targeted regions were designed in Primer-BLAST (Ye et al., 2012) using the customised parameters described by (Kellner et al., 2019). Oligonucleotides prone to generate self-dimers or cross primer dimers detected by Multiple Primer Analyzer (ThermoFisher) were discarded. Primers selected for each SHERLOCK assays are listed and highlighted in Supplementary Table 2. The T7 promoter sequence (5’ GAAATTAATACGACTCACTATAGGG) was added at each 5’ end of the selected forward primers to enable the in vitro transcription necessary for Cas13a detection (Kellner et al., 2019).

Production of crRNAs and positive controls

crRNA production was carried out as previously described (Kellner et al., 2019; Sima et al., 2022). Briefly, each crRNA DNA template, containing the spacer (specific targeted region) followed by the direct repeat (5’ GATTTAGACTACCCCAAAAACGAAGGGGACTAAAAC) and the T7 promoter sequence, were in vitro transcribed using the HiScribe™ T7 Quick High Yield RNA Synthesis Kit (NEB) and purified using magnetic beads (Agencourt RNAClean XP). DNA template sequences used for the synthesis of each crRNA are detailed in Supplementary Table 2.

For the production of the synthetic positive controls, in vitro transcription of the targets was performed as previously described (Sima et al., 2022) using RNA from the target species for each specific assay. The amount of template used to produce the positive control for each assay were as follows: 200 ng of RNA from T. b. brucei Lister 427 for 18S pan-trypanosomatid and 18S pan-Trypanozoon assays, 120 ng of RNA from T. b. gambiense ELIANE strain for TgSGP assay, 120 ng of RNA from T. vivax Magnan strain for IFX assay, 230 ng of RNA from T. congolense Savannah subgroup IL1180 strain for 18S T. congolense-specific assay and 5 ng of RNA from T. theileri for 18S T. theileri-specific assay. Due to a lack of T. simiae and T. suis genetic material, the positive controls were synthesized at Eurogentec using the in-silico predicted sequence of the target region. Primers and target regions used to produce positive controls are listed in the Supplementary Table 3.

SHERLOCK assays

SHERLOCK assays were performed as previously described (Kellner et al., 2019; Sima et al., 2022). RPA was performed using the TwistAmp Basic kit (TwistDx) following manufacturer’s instructions with slight modifications. To enable reverse transcription of the RPA (RT-RPA), 2.2 U of transcriptor reverse transcriptase (Roche) were added to the reaction. Specific RPA primer pairs were used for each target amplification at a concentration of 480 nM per primer, and the amplification reaction was carried out in a heating block at 42°C for 45 minutes (Sima et al., 2022). Leptotrichia wadeii Cas13a (LwCas13a) enzyme was produced as previously described (Kellner et al., 2019; Sima et al., 2022). Briefly, pC013-Twinstrep-SUMO-huLwCas13a plasmid (Addgene plasmid #90097) was inserted into Escherichia coli RosettaTM 2(DE3) pLysS competent cells for LwCas13a expression and purification (Gootenberg et al. 2017). Purified LwCas13a was conserved in Storage buffer (SB: Tris-HCl 1M, NaCl 5M, DTT 1M and glycerol) (Kellner et al., 2019) in single-use aliquots at −80°C to avoid freeze-thaw cycles.

The Cas13a-crRNA mediated target detection reaction was performed simultaneously with the in vitro transcription of the target amplified by RT-RPA. In vitro transcription of the amplified target was performed using T7 RNA polymerase (Biosearch technology). Each Cas13a detection reaction was run in triplicate on a TECAN plate reader (INFINITE F200 PRO M PLEX) and fluorescence readouts were collected every 10 minutes for 3 hours. All reactions were carried out with the inclusion of a negative control template (NCT) adding nuclease-free water as input, and a positive control template (PCT), here the in vitro produced positive control for each specific SHERLOCK. For screening samples from epidemiological surveys, the same protocol as described above was used, except a TECAN plate reader (INFINITE F200 PRO Option Infinite F Nano +) was used and the assay was performed without replicates. Optimizations were carried out either at Institut Pasteur in Paris (France) or at INTERTRYP unit in Montpellier (France), while analysis of samples from epidemiological surveys was carried out at Institut Pasteur of Guinea.

Data representation and statistical analyses

Fold-change over background fluorescence was used to represent fluorescence readouts for all SHERLOCK assays. Background fluorescence is given by the NCT evaluation, where water is used as input material for the RPA reaction. For each test, sample fluorescence obtained after 3 hours of reading was divided by NCT fluorescence at the same time.

Where FC represents Fold-Change over background fluorescence, Fsample t=3h represents sample fluorescence obtained after 3 hours and Fnct mean t=3h represents NCT fluorescence after 3 hours of reactions.

Analytical characteristics including threshold, sensitivity and specificity of each SHERLOCK assay were evaluated using Receiver Operating Characteristic (ROC) curves calculation comparing NCT versus PCT at controlled dilutions of parasites or hosts genetic materials. Results obtained from SHERLOCK specific assays on targeted parasites were gathered in a positive sample column, while results obtained from NCT, hosts and non-targeted parasites analysed with the assay, were placed in a negative sample column for ROC curve calculation and threshold evaluation. ROC curves were calculated using the Wilson/Brown method with a 95 % confidence interval and threshold values were selected based on the Youden’s index (sensitivity + specificity – 1), that defines an optimal threshold as the one that maximizes the index of classification accuracy (Brown et al., 2001; Schisterman et al., 2005). The threshold of positivity for each SHERLOCK assay was selected from ROC analyses as the value corresponding to the maximum specificity (100 %). The limit of detection (LoD) of each SHERLOCK reaction was assessed using specific tests on serial dilutions of target parasite RNA. Dilutions ranged from 5.10³ pg/µL (equivalent to 10⁷ parasites/mL) to 5.10^-5 pg/µL (equivalent to 0.1 parasites/mL). All plots and statistical analyses were performed with GraphPad Prism software (V.10). The Venn diagrams were made based on the Venny model by Oliveros J.C. (BioinfoGP, CNB-CSIC).

Maps were generated with the QGIS software version 3.42.0 with backgrounds retrieved from data.gouv.fr (OSM2IGEO - Ivory Coast - SHP - EPSG:4326 for Côte d’Ivoire, and OSM2IGEO - Guinea - SHP - EPSG:4326 for Guinea).

In the absence of reliable infection status in the field data from pigs, the posterior distributions of performance estimates, such as sensitivity (Se), were jointly estimated with a Bayesian latent class model (Joseph et al., 1995). The model takes boolean outcomes (either positive or negative) of several tests as input. This Monte Carlo method operates based on a Gibbs sampling process. The model iteratively updates both the test performance parameters (sensitivity, specificity, and frequency of positive individuals in sample collection) and the latent infection status. Flat prior was used. After visually controlling that stationarity of the chain was quickly reached, we discarded the first 150 burn-in iterations and calculated posterior densities based on > 10,000 iterations. 95% confidence intervals were calculated for each proportion to quantify the uncertainty of the estimates, providing a scientific range within which the population proportions are likely to be accurate (Brown et al., 2001).

Results

Selection of targets for broad and species-specific SHERLOCK assays targeting AAT species (SHERLOCK4AAT)

Target gene selection was based on two requirements: to develop (a) a broad SHERLOCK assay for the detection of all trypanosomatid parasites and (b) for species discrimination. To adapt SHERLOCK for the detection of AAT-related species, we followed the criteria as described in (Sima et al., 2022). Briefly, candidate genes must (1) be expressed in the bloodstream form of the parasites, (2) be conserved between strains, (3) have few or no single nucleotide polymorphisms (SNPs), or (4) show sequence divergence that would allow for species discrimination. All RPA primers and crRNAs designed for each SHERLOCK assays are listed in Supplementary Table 2.

Pan-trypanosomatid assay

SHERLOCK has the capacity to discriminate between two sequences differing by a single nucleotide (Gootenberg et al., 2017). This study therefore focused on regions of the 18S rRNA gene that show limited sequence homology between species. In addition to being a multicopy gene, the 18S rRNA gene shares over 70 % identity between Kinetoplastea members (Supplementary Figure 1). Using BLAST analysis, the region between 1535-1672 bp (XR_002989632.1) was found to be conserved between all trypanosomatid species and was selected as the pan-trypanosomatid SHERLOCK target. RPA primers were designed to amplify this 138 bp region using Primer-Blast (Ye et al., 2012) (Supplementary Figure 2A and Supplementary Table 2). The level of identity of the target region was assessed between the different sequences described in the Trypanosomatidae family, as well as in the order, subclass and class that contain this family (Supplementary Figure 3). Examination of the Euglenozoa phylum, excluding the Kinetoplastea class, showed 98.4 % identity within the Trypanosomatidae family and 97.3 % identity within the Kinetoplastea class (Supplementary Figure 3). This was reduced to 82.9 % for the Phylum Euglenozoa, excluding the Kinetoplastea sequences (Supplementary Figure 3). Three distinct crRNAs were designed to this region, which can be used individually or multiplexed to increase the sensitivity (Supplementary Figure 4A and Supplementary Table 2). The sequences described in each group were 100 % conserved to the crRNA. RPA primers and crRNA targets of the pan-trypanosomatid SHERLOCK assay target regions (Kinetoplastea) were then compared to the homologous sequences from other classes within the Phylum Euglenozoa (Euglenida, Diplonemea, Symbiontida and Kinetoplastea) (Supplementary Figure 2A and 3). In addition, sequences from related protozoan parasite classes (Diplomonadida, Haemosporida and Ciliophora) and from the genetically close but phylogenetically distinct group Microsporidia were also included in the analysis (Supplementary Figure 2A).

Pan-Trypanozoon assay

A pan-Trypanozoon SHERLOCK assay was designed for increased selectivity to especially detect T. brucei s.l.. The regions within the 18S rRNA that were highly conserved among the species of the Trypanozoon subgenus, while showing significant divergence from other trypanosome species not belonging to the Trypanozoon subgenus, were analysed (Supplementary Figure 2B). A region ranging from 1318 to 1433 bp (XR_002989632.1) (Figure 1A and Supplementary Table 2) was selected as it showed high conservation among the Trypanozoon subgenus (99.12 - 100 % percentage of identity, query coverage >90 %), except for three T. evansi sequences (91.96 - 96.49 %, OL869590.1, KT844944.1, MF142286.1). The crRNA target region selected was also conserved in the subgenus Trypanozoon and divergent in other parasites. Specific alignments with trypanosome species that could share the same hosts (T. congolense, T. vivax, Leishmania spp., etc) were performed to assess target amplicon specificity (Supplementary Figure 2B). The pan-Trypanozoon SHERLOCK assay can discriminate between the species belonging to the Trypanozoon subgenus (T. b. spp., T. evansi types A and B, and T. equiperdum) from other non-Trypanozoon trypanosomes or other co-endemic parasites.

Figure 1.

Species-specific assays

For species discrimination within a single sample, several SHERLOCK assays were developed to detect the most prevalent species causing AAT. These include T. congolense, T. vivax, T. theileri, T. simiae and T. suis.

T. congolense. Regions conserved in the 18S rRNA gene for all described T. congolense sequences, including the Savannah, Forest, Kilifi and Dzanga-Sangha groups, were analysed. The target region selected from position 116 to 246 bp of the 18S rRNA gene (AJ009146.1) shows 100 % identity between the described T. congolense sequences. BLAST analysis of the T. congolense 18S rRNA gene sequences, excluding the target species (T. congolense, (taxid: 5692)) was also performed to determine the percentage of identity of closely related species. T. simiae (86.72 to 90.62 %), T. sapaensis (88.28 %), T. godfreyi (88.28 %), and T. otospermophili (87.02 %) had the highest percentage identity observed (Supplementary Figure 5). Species with common hosts and of veterinary relevance such as the two analysed T. vivax strains showed 82.81% and 87.5% identity. In addition, none of these sequences were fully conserved for the selected crRNA (Supplementary Figure 5), ensuring the specificity of the test and the absence of cross-detection with other species.

T. vivax. The invariant flagellum antigen (IFX) gene has no apparent homologue in any other Trypanosoma species or other related parasites, and was therefore selected (Autheman et al., 2021; Romero-Ramirez et al., 2022). BLAST analysis, using T. vivax Y486 (HE573024.1) as a reference genome, resulted in no matches (Figure 1A and Supplementary Table 2).

T. theileri. A conserved region between the three T. theileri lineages (TthI, TthII and TthIII) (Brotánková et al., 2022) was selected. The target region, ranging from 1014 to 1158 bp of the 18S rRNA gene (AJ009163.1), has 100 % identity between the described sequences of the T. theileri lineages, and the most closely related sequences were from T. cruzi, T. rangeli and T. dionisii, with a 97.89 % identity (query coverage =65 %). The segment 1052-1081 bp corresponding to the crRNA sequence, is highly conserved between T. theileri variants (Figure 1A and Supplementary Table 2), however it contains a SNP (1060 C/T) in a number of the sequences analysed. Hence, one RPA primer pair and two multiplexed crRNAs were designed to target both version of the SNP for the T. theileri SHERLOCK assay (Supplementary Figure 4E).

T. simiae. The region from 1246-1346 bp of the 18S rRNA gene (AJ404608.1) was selected as the targeted region for RPA amplification of the T. simiae-specific SHERLOCK assay. The segment 1273-1299 bp that corresponds to the crRNA sequence, has 100 % identity between the described sequences of T. simiae (Figure 1A and Supplementary Table 2).

T. suis. The multicopy GAPDH gene was used, as it can be phylogenetically separated into T. suis and T. suis-like strains from other African trypanosomes (Rodrigues et al., 2020). The region from 389 to 528 bp of the GAPDH gene (MK962702.1) has been selected as the targeted region for RPA amplification of the T. suis-specific SHERLOCK assay (Figure 1A and Supplementary Table 2). This region is 100 % conserved between the T. suis species described, and outside this clade, the most closely related sequences were found in T. congolense, with a 90 % of identity (query coverage >90 %). The segment 410-437 bp of the gGAPDH gene (MK962702.1), that corresponds to the crRNA target region, is highly conserved between T. suis species, with 100 % identity, and the BLAST analysis showed that this target fragment can only align to T. suis sequences but not to that of other Kinetoplastea species (taxid:5653).

SHERLOCK can discriminate between closely related AAT-causing trypanosomatid species

All RPA primers and crRNAs for each targeted species were screened to select the optimal combinations, with the highest specificity and selectivity. For the pan-trypanosomatid assay, RPA primer pair A was used to screen three crRNAs (A:crRNA1, A:crRNA2 and A:crRNA3; Supplementary Figure 4A). From this, crRNA 3 was first selected for the simple pan-trypanosomatid assay. For the pan-Trypanozoon assay, three RPA primer pairs (A, B and C) were screened with two crRNA candidates each (A:crRNA1, A:crRNA2, B:crRNA1, B:crRNA2, C:crRNA1, C:crRNA2, Supplementary Figure 4B). The A:crRNA1 combination was selected for the pan-Trypanozoon assay, being the one that obtained a higher fold change for the positive control and a lower fold change for the negative control. For the IFX T. vivax species-specific SHERLOCK assay, two RPA primer pairs were assessed (A and B) in combination with six crRNA candidates (A:crRNA1, A:crRNA2, A:crRNA3, B:crRNA4, B:crRNA5, B:crRNA6). RPA primer pair B and crRNA 6 showed the lowest variability between replicates and without any cross-reaction with other trypanosome species (Supplementary Figure 4C). For the T. congolense-specific assay, one pair of RPA primer (A) was combined with ten crRNA candidates (A:crRNA1, A:crRNA2, A:crRNA3, A:crRNA4, A:crRNA5, A:crRNA6, A:crRNA7, A:crRNA8, A:crRNA9, A:crRNA10). The crRNA1 that exhibited the lowest inter-replicate deviation and no cross-reactivity when exposed to the genetic material of non-target trypanosomatid parasites was selected (Supplementary Figure 4D). For the T. theileri-specific SHERLOCK assay, the combination of the RPA primer pair A with the two designed crRNAs (1 and 2) was used to cover the specific detection of all the T. theileri described sequences (Supplementary Figure 4E). For T. simiae and T. suis, single combinations of RPA primer and crRNA were tested. The specificity of each of these combinations was also verified by comparing the specific recognition of each RNA target to RNA from human embryonic kidney cell (HEK) (for pan-trypanosomatid and pan-Trypanozoon tests) or RNA from T. brucei (for T. vivax, T. congolense and T. theileri tests) (Supplementary Figure 4). In total, a RPA primer-crRNA combination was selected for each assay, based on the highest fold change over background fluorescence value, the highest specificity, and the minimum standard deviation (Figure 1A).

The specificity of the RPA primer-crRNA combinations selected for each assay was then cross-assessed against closely related parasites (Figure 1B-C). The single pan-trypanosomatid assay detected RNA from all trypanosomatid species tested (T. b. brucei, T. b. gambiense, T. b. rhodesiense, T. evansi (type A and B), T. equiperdum,

T. congolense, T. vivax, T. theileri, T. cruzi and Leishmania major) but not from HEK cells, synthetic controls for T. simiae and T. suis or co-endemic parasites such as Plasmodium falciparum (Figure 1B). The pan-Trypanozoon assay detected only RNA from members of the Trypanozoon subgenus (T. b. brucei, T. b. gambiense, T. b. rhodesiense, T. evansi (type A and B) and T. equiperdum) but not from HEK cells, synthetic controls for T. simiae and T. suis or co-endemic parasites such as T. congolense, T. vivax, T. theileri and Plasmodium falciparum (Figure 1B). The species-specific T. congolense, T. vivax, T. theileri, T. simiae and T. suis assays only detected their respective targeted parasite RNA or synthetic controls with no cross reactivity with any other RNAs (Figure 1C). Additionally, no unspecific cross-reaction was observed with the host-derived genetic material, as assessed by testing RNA from pig blood (SHERLOCK single pan-trypanosomatid and pan-Trypanozoon assays in Supplementary Figure 7).

The analytical sensitivity of each assay was quantified using serial dilutions of positive control RNAs from known numbers of parasites. For the single pan-trypanosomatid assay, the LoD was between 10 to 10³ parasites per mL. The pan-Trypanozoon and T. theileri assays tested 10 parasites per mL. For the T. congolense assay, the LoD was found between 10 and 10³ parasites per mL, and between 10⁴ and 10⁵ parasites per mL for the T. vivax test (Figure 2A). The sensitivity of T. simiae and T. suis tests were evaluated on synthetic positive controls, as no genetic materials were available, and both showed a LoD between 10 and 100 fM (Figure 2B).

Figure 2.

Then, to improve the sensitivity of the simple pan-trypanosomatid assay, crRNA1 was tested in combination with crRNA3 for target recognition, as both crRNAs fall within the same RPA amplicon A (Figure 3A). Comparative analyses between the two versions were performed and multiplexing was shown to increase the performance of the assay, especially for the lower dilutions evaluated (100, 500 and 10³ p/mL) by increasing the fold-change readouts of these positive controls by 1, 2.5 and 3-fold-change units, respectively (Supplementary Figure 6). Using results obtained on the blood samples from pigs in Côte d’Ivoire (see next section), we compared the two versions of the pan-trypanosomatid SHERLOCK assays to determine the posterior distribution of sensitivity estimates. We applied a Monte Carlo method (Gibbs sampling) proceeding in an iterative chain using the infection frequency, specificity, and sensitivity of the two tests to update the infection status. We determined the posterior sensitivity distribution of the simple versus multiplex pan-trypanosomatid SHERLOCK assays. Multiplexing crRNAs increased the median sensitivity from 66.1 % to 74.7 %, while keeping specificity above 80 % (Figure 3B). Furthermore, a higher concordance was observed between the results obtained on the same samples with the pan-Trypanozoon SHERLOCK assay and the multiplexed pan-trypanosomatid SHERLOCK assay (25.7-% of double positives), as compared to the simple version (7.8 % of double positives) (Figure 3C). Given its improved detection efficiency, the multiplex version of the 18S rRNA pan-trypanosomatid assay was subsequently used exclusively and referred to as the multiplex pan-trypanosomatid assay.

Figure 3.

To determine a positivity threshold value (fold-change cut-off) for each SHERLOCK assay, ROC curve analyses were performed and use to evaluate each following analytical sensitivity and specificity. The positivity threshold values were 6.050 for the pan-trypanosomatid test, 11.70 for the pan-Trypanozoon test, 1.923 for the T. congolense test, 1.831 for the T. vivax test, 1.805 for the T. theileri test, 1.546 for the T. simiae test, and 2.609 for the T. suis test (Supplementary Figure 8). Subsequently, for a given test, any sample with a fold change value above the threshold was considered positive for this test. Unsurprisingly, assays designed against multiple copy genes (18S rRNA and GAPDH) showed a greater sensitivity than those targeting single copy genes (IFX). Taken together, all the validated tests were able to specifically detect the RNA targets for which they were designed, making it possible to distinguish between closely related trypanosomatid parasite species. We then applied this novel SHERLOCK4AAT toolbox was applied to the analysis of field samples in a molecular eco-epidemiological survey.

Application of the SHERLOCK4AAT toolbox to eco-epidemiological surveys in pigs

In transmission foci of western Africa, domestic pigs can host multiple trypanosome species. To investigate this, the SHERLOCK4AAT toolbox was used to survey trypanosomes circulating in domestic pigs from hypo-endemic and historical foci in Côte d’Ivoire and Guinea. In Côte d’Ivoire, blood samples were collected from 224 free-ranging pigs across the rural districts of Bonon, Bouaflé, Sinfra, Vavoua, and Brobo (Figure 4A). In the N’zérékoré prefecture of Forested Guinea, 200 farm pigs were sampled, with a focus on organized farming communities (Figure 5A). In addition to the SHERLOCK4AAT toolbox assays, all samples were also screened for T. b. gambiense using the previously described TgsGP SHERLOCK assay (Sima et al., 2022).

Figure 4.

Figure 5.

In total, 74.6 % [68.9, 80.3] (167/224) of the samples tested positive in at least one SHERLOCK assay in Côte d’Ivoire (Figure 4B), and 49.5 % [42.6, 56.9] (99/200) in Guinea (Figure 5B) representing a high apparent prevalence of trypanosome infections in these two populations of pigs. In parallel, 47.3 % [40.8, 53.8] (106/224) of the pigs from Côte d’Ivoire tested positive using at least one parasitological method (buffy coat examination, mAECT and /or isolation in mice; results detailed in a separate study, manuscript in preparation). No systematic parasitological data was available for Guinean samples. The difference in positivity rates between SHERLOCK detection and parasitological approaches on Ivoirian samples was likely due to the higher sensitivity of SHERLOCK as compared to classical methods. Indeed, we compared three SHERLOCK assays (multiplex pan-trypanosomatid, pan-Trypanozoon and T. congolense-specific SHERLOCK assays) to the three parasitological tests (buffy coat examination, mAECT and /or isolation in mice) to determine the posterior distribution of sensitivity estimates. A Monte Carlo method (Gibbs sampling) was used by proceeding in an iterative chain using the infection frequency, specificity, and sensitivity of the tests to update the infection status. This analysis suggested that, in terms of sensitivity, the combination of SHERLOCK4AAT assays may compete with combined parasitological methods (i.e., median sensitivity 91.4 % vs 84.3 %) while controlling median specificity above 75 % (Supplementary Figure 9). Importantly, in contrast to parasitological tests, SHERLOCK4AAT allowed for the identification of the trypanosome species involved in more than half of these infections.

Using the multiplex pan-trypanosomatid assay, 41.1 % [34.6, 47.6] of the samples from Côte d’Ivoire (90/219) and 29.5 % [23.2, 35.8] from Guinea (59/200) were positive (Figure 4C-5C). The pan-Trypanozoon assay identified 38.0 % [31.6, 44.4] (85/224) and 39.0 % [32.2, 45.8] (78/200) of samples in each country respectively (Figure 4C–5C). As there is no specific test for T. b. brucei, samples that were positive in the pan-Trypanozoon assay but negative in the TgsGP assay were presumed to be T. b. brucei infections. Only one sample tested positive in the T. b. gambiense-specific assay in each country (Figure 4B-5B). The TgsGP-positive sample from the district of Bonon, Côte d’Ivoire was negative for all other SHERLOCK assays, as well as in all parasitological tests. In contrast, the TgsGP-positive sample from Guinea was positive in the multiplex pan-trypanosomatid SHERLOCK assay, as well as in the CATT and mAECT tests, but it negative in the pan-Trypanozoon SHERLOCK assay. To further explore trypanosome diversity, species-specific SHERLOCK assays for AAT were applied to the samples. In total, 23.2 % [17.7, 28.7] (52/224) of the samples in Côte d’Ivoire and 7.5 % [3.8, 11.2] (15/200) in Guinea were positive for T. congolense. Additionally, a single sample from Côte d’Ivoire tested positive in the T. suis-specific SHERLOCK assay. Furthermore, no RNA from T. vivax or T. theileri was detected with the IFX SHERLOCK assay.

Interestingly, a substantial number of co-infections were observed. In total, 24/219 individuals (11.0 ± 4.1 %) in Côte d’Ivoire and 13/200 in Guinea (6.5 ± 3.4 %) tested positive in the multiplex pan-trypanosomatid, the pan-Trypanozoon and the T. congolense-specific assays. These triple positive results indicate co-infections involving both a Trypanozoon species (likely T. b. brucei) and T. congolense (Figure 4D-5D). In addition, 9/219 samples (4.1 ± 2.6 %) from Côte d’Ivoire tested positive in the pan-Trypanozoon and the T. congolense assays, which also represented co-infections with T. brucei s.l. and T. congolense. Overall, 15.2 % [10.5, 19.9] of the samples from Côte d’Ivoire showed co-infections with a Trypanozoon species and T. congolense. Additionally, 15 samples from Côte d’Ivoire tested positive in both the multiplex pan-trypanosomatid assay and the T. congolense-specific assay or were positive only for T. congolense. In Guinea, two samples tested positive in both the T. congolense-specific and multiplex pan-trypanosomatid SHERLOCK assays, indicating T. congolense mono-infections (Figure 4D-5D).

Furthermore, 56 samples (25.6 ± 5.8 %) in Côte d’Ivoire, and 38 samples (19.0 ± 5.4 %) in Guinea tested positive in both the multiplex pan-trypanosomatid and the pan-Trypanozoon SHERLOCK assays. In contrast, 17 samples from Côte d’Ivoire and 40 from Guinea tested positive in the pan-Trypanozoon assay but negative in the multiplex pan-trypanosomatid assay, likely due to the lower sensitivity of the latter. Lastly, a 19 samples (8.7 ± 3.7 %) in Côte d’Ivoire and 19 samples (9.5 ± 4.1 %) in Guinea tested positive in the multiplex pan-trypanosomatid assay only (Figure 4D-5D). The absence of positivity in the T. congolense, T. simiae and T. suis-specific assays may be due to their lower sensitivity, or to infections with other non-targeted trypanosome species.

Discussion

The SHERLOCK4AAT toolbox, a new panel of highly sensitive and specific high-throughput assays for detecting animal trypanosomes

Species identification is essential for assessing both the distribution of prominent parasites impacting veterinary health and spotting reservoirs for human-infective trypanosomes. In a context where symptoms are often mild and unspecific, especially in hosts like pigs, accurate species-specific detection methods are essential. Ideally, diagnostic assays should be able to analyse both fresh and stored (ambient temperature) samples to determine parasites circulating within a specific area.

In this study, conserved and species-specific targets have been used to adapt SHERLOCK for the detection of most members of the Trypanosomatidae family (Kinetoplastea class) (Supplementary Figure 3) of veterinary importance (pan-trypanosomatid), and species-specific assays to discriminate between the six most common animal trypanosomes species in the study regions: Trypanozoon (T. b. brucei, T. evansi and T. equiperdum), T. congolense (Savanah, Forest Kilifi and Dzanga-sangha), T. vivax, T. theileri (TthI, TthII and TthIII), T. simiae and T. suis. The design of both broad and species-specific assays was based primarily on sequences of the 18S rRNA and GAPDH genes which, as well as being conserved across species (Supplementary Figure 1), are highly expressed and contain regions that allowed for species discrimination. However, for T. vivax, the single-copy IFX gene was selected for its high specificity (Romero-Ramirez et al., 2022). The first simple version of our pan-trypanosomatid assay employed a single crRNA guide and was previously used to evaluated blood samples from gHAT seropositive individuals with a specificity of 93 % (95.4-98.6) (N’Djetchi et al., 2024). To increase the performances of this assay, we added a second crRNA guide into the reaction mix to target the same RPA amplicon during the SHERLOCK reaction, resulting in an increased test sensitivity (Figure 3 and Supplementary Figure 6). For Trypanozoon species, no sequences were identified to distinguish T. b. brucei, T. evansi (A and B) and T. equiperdum as these parasites cannot be easily differentiated using the 18S rRNA gene. In the future, other genes such as RoTat1.2 VSG or ESAG6/7 could be explored to discriminate T. evansi from other Trypanozoon by SHERLOCK (Claes et al., 2004; Holland et al., 2001). Overall, in silico and in vitro analyses enabled us to assess the analytical specificities of these new SHERLOCK assays, confirming that the targets selected were sufficiently specific to be used as diagnostic for the species and groups of species concerned.

Using parasite RNAs, the LoD of the SHERLOCK assays were estimated to range between 10 to 10³ parasites/mL for the pan-trypanosomatid assay, 10 parasites/mL for the pan-Trypanozoon and T. theileri assays, from 10 to 10³ parasites/mL for T. congolense and 10⁴-10⁵ parasites/mL for T. vivax. Most PCR-derived molecular techniques currently used for detecting animal trypanosomes have a detection limit of 1 to 10 parasites/mL using genus- or species-specific primers, 50 parasites/mL using pan-species ITS1 primers (Desquesnes et al., 2022a) or 5 parasites/mL using 18S rRNA-targeting primers (Deborggraeve et al., 2006). The ROC analyses of the T. brucei, T. congolense and T. vivax SHERLOCK assays (Supplementary Figure 8C) also shown sensitivity and specificity comparable to those of the recently developed 7SL sRNA diagnostic test for Animal trypanosomosis (Contreras Garcia et al., 2022). Therefore, in total, the analytical sensitivities obtained by the pan-trypanosomatid, pan-Trypanozoon, T. theileri and T. congolense SHERLOCK assays were comparable with those of other existing molecular tools. Yet, a significant difference in favour of using SHERLOCK4AAT is that it detects RNAs, probing for active infections. The T. vivax SHERLOCK assay, targeting the single-copy IFX gene (Autheman et al., 2021), did not meet the sensitivity requirements for the molecular diagnosis of animal trypanosomes and would require the identification of another target. As no genetic material from parasites was available for developing T. simiae and T. suis SHERLOCK assays, LoDs were determined using synthetic RNA controls. Both assays were able to detect 10 to 100 fM of the synthetic controls, far from the detection capacity of SHERLOCK, which can detect synthetic targets at a concentration of 2 aM in optimal conditions (Gootenberg et al., 2017).

SHERLOCK4AAT profiling in Guinea and Côte d’Ivoire

Trypanosome species identification can be used to finely map the risk of AAT parasites transmission in these areas. After cattle, pigs are the main domestic hosts infected by animal trypanosomes, but they can also potentially be reservoirs of T. b. gambiense (Büscher et al., 2018). This is all the more important than pigs are often raised in close proximity to humans, increasing the risk of animal to human trypanosome transmission. In blood samples from Ivorian pigs, analyses with the SHERLOCK4AAT toolbox revealed 37.9 % positive samples for Trypanozoon RNAs and 23.2 % positive samples for T. congolense RNAs, with 15.2 % samples positive for both tests. In addition, no sample was positive for T. vivax, T. theileri or T. simiae, but one sample was positive in the T. suis assay and another positive for T. b. gambiense. These findings are consistent with the most recent known distribution of animal trypanosomes in free-ranging pigs raised in Côte d’Ivoire’s historical gHAT foci. The use of conventional molecular diagnostics on 97 free-range pigs resulted in an overall positivity rate of 57 % for Trypanozoon DNA (PCR-TBR) and 24 % for T. congolense forest DNA (PCR-TCF), with 14 % of pigs positive in both PCRs and none positive for T. vivax (PCR-TVW) (N’Djetchi et al., 2017; Traoré et al., 2021). The epidemiological status of AAT in Guinea is unclear, as no large-scale molecular studies have been published so far. The blood samples from farmed pigs analysed here were collected in the prefecture of N’zérékoré, a former endemic focus of gHAT. Results showed a positivity rate of 39.0 % for the detection of Trypanozoon RNAs, like that observed in Côte d’Ivoire, but a lower rate of positivity for T. congolense RNAs (7.5 %), with 6.5 % of samples positive for both tests. In addition, one sample tested positive for T. b. gambiense, but no T. vivax, T. simiae, T. theileri nor T. suis was detected. The lack of any samples positive for T. vivax or T. theileri in any study areas confirms that pigs are not frequent carriers of these species (Fetene et al., 2021). However, the blood samples positive for T. b. gambiense in each study site provide further evidence of the ability of pigs to carry human trypanosomes and to act as potential reservoir. Similarly, the presence of T. b. gambiense genetic material in pig samples has already been suspected in Côte d’Ivoire (Traoré et al., 2021).

In both sites, a high prevalence of trypanosomatid parasites was observed using the multiplex pan-trypanosomatid and/or pan-Trypanozoon assays (42.0 % in Côte d’Ivoire and 48.5 % in Guinea). Having partially deconvolved the species present in these animals, there is still a small proportion (8.7 % in Côte d’Ivoire and 9.5 % in Guinea) of trypanosomatid parasites that remained uncharacterised by our tests on samples found only positive in the multiplex pan-trypanosomatid SHERLOCK assay. Firstly, these assays were designed by in silico determination of the most specific and conserved region to target the entire Trypanosomatidae family, and even Kinetoplastea class, or to discriminate between species, yet the limited number of sequences available for some species in the literature and in the online databases may under-represent the real genetic diversity of these parasites in natura. In line with this hypothesis, there may also exist uncharacterised trypanosome species circulating among these reservoirs, as described in small mammals in central and East Africa (Votypka et al., 2022), and that were only detected by the broad multiplex pan-trypanosomatid assay. Altogether, these results are consistent with those reported in the few available previous studies (N’Djetchi et al., 2017; Traoré et al., 2021) and confirm the effectiveness of the SHERLOCK4AAT toolbox in analysing the distribution of trypanosomes in hypo-endemic gHAT foci. During natural infections, trypanosome nucleic acids can be detected in the skin, adipose tissue and various organs even if the blood remains negative (Amisigo et al., 2024). Hence, applying SHERLOCK detection on skin samples in the future would be useful for monitoring both blood and skin-dwelling parasites. The future development of multiplexed assays combined with adapted lateral flow tests would also be highly desirable for reaching field-applicable point-of-care molecular diagnostic versions of the tests useable in the field. The absence of point of care diagnostic test for AAT is currently a major hurdle for its surveillance and control (Desquesnes et al., 2022b).

T. b. gambiense is suspected to circulate among domestic animals raised in close contact to humans, such as pigs. Here, T. b. gambiense has been detected in one animal in each study site. Considering the resulting low prevalence of T. b. gambiense infections in these two populations of domestic pigs (0.45 ± 0.9 % in Côte d’Ivoire to 0.50 ± 1.0 % in Guinea) and the short life expectancy of domestic pigs reared for commercial purposes in these two rural areas (6 to 8 months), these animals do likely not represent a significant risk to the elimination of gHAT. This is apparently confirmed by the absence of new human cases as well as by the low abundance of tsetse flies in the two sites (article in preparation). However, because of their proximity to humans and their easy access for frequent sampling, pigs could be used as sentinels to monitor T. b. gambiense circulation, possibly with the SHERLOCK4AAT toolbox.

Acknowledgements

The authors thank Annette MacLeod (University of Glasgow, Glasgow, UK) for providing RNA from T. b. rhodesiense, Nick Van Reet and Philippe Büscher (Institute of Tropical Medicine [ITM], Antwerp, Belgium) for providing RNA from T. equiperdum, T. evansi types A and B, Gerald Spaeth and Artur Scherf (Institut Pasteur, Paris, France) for providing RNA from Leishmania major and Plasmodium falciparum and John Kelly (The London School of Hygiene & Tropical Medicine, London, UK) for providing RNA from T. cruzi. We warmly thank all the colleagues at the University Jean Lorougnon Guédé (Daloa, Côte d’Ivoire) and at the Institut Pierre Richet (Bouaké, Côte d’Ivoire) as well as all the health workers in the districts of Sinfra and Bonon (Côte d’Ivoire) and N’zérékoré (Guinée) for their strong support in the field.

Additional information

Funding

This project was funded by the Agence Nationale pour la Recherche (ANR-21-CE17-0022-01 SherPa), the Programme Investissement d’Avenir of the French Government through the two Laboratoires d’Excellence, ANR-10-LABX-62-IBEID and ANR-11-LABX-0024-PARAFRAP, and the Pasteur Network (Bourse Calmette-Yersin to AC).

Author contributions

RE, EPA, AC, LG and BR conceived and designed the experiments. RE, EPA, AC, SB, ADH, MND, MT, YL, BD, ADC, MK, VJ and JMB performed the experiments. RE, EPA, AC, PC, LG and BR analysed the data. MK, DK, MC, VJ, ST, JMB, BR and LG contributed reagents, materials and analysis tools. RE, EPA, AC, LG and BR wrote the manuscript. MND, MK, VJ, ST, and JMB reviewed the manuscript.

Additional files

Supplementary Table 1. Accession numbers of organisms used to evaluate the presence of single nucleotide polymorphism (SNPs) in the different target regions.

Supplementary Table 2. Primers and crRNA guides used for the SHERLOCK4AAT toolbox.

Supplementary Table 3. RPA primers used to produce the in vitro transcribed targets (positive control templates).