Identification of a Tsal152–75 salivary synthetic peptide to monitor cattle exposure to tsetse flies

The saliva of tsetse flies contains a cocktail of bioactive molecules inducing specific antibody responses in hosts exposed to bites. We have previously shown that an indirect-ELISA test using whole salivary extracts from Glossina morsitans submorsitans was able to discriminate between (i) cattle from tsetse infested and tsetse free areas and (ii) animals experimentally exposed to low or high numbers of tsetse flies. In the present study, our aim was to identify specific salivary synthetic peptides that could be used to develop simple immunoassays to measure cattle exposure to tsetse flies. In a first step, 2D-electrophoresis immunoblotting, using sera from animals exposed to a variety of bloodsucking arthropods, was performed to identify specific salivary proteins recognised in cattle exposed to tsetse bites. Linear epitope prediction software and Blast analysis were then used to design synthetic peptides within the identified salivary proteins. Finally, candidate peptides were tested by indirect-ELISA on serum samples from tsetse infested and tsetse free areas, and from exposure experiments. The combined immunoblotting and bioinformatics analyses led to the identification of five peptides carrying putative linear epitopes within two salivary proteins: the tsetse salivary gland protein 1 (Tsal1) and the Salivary Secreted Adenosine (SSA). Of these, two were synthesised and tested further based on the absence of sequence homology with other arthropods or pathogen species. IgG responses to the Tsal152–75 synthetic peptide were shown to be specific of tsetse exposure in both naturally and experimentally exposed hosts. Nevertheless, anti-Tsal152–75 IgG responses were absent in animals exposed to high tsetse biting rates. These results suggest that Tsal152–75 specific antibodies represent a biomarker of low cattle exposure to tsetse fly. These results are discussed in the light of the other available tsetse saliva based-immunoassays and in the perspective of developing a simple serological tool for tsetse eradication campaigns to assess the tsetse free status or to detect tsetse reemergence in previously cleared areas.


Background
African trypanosomosis, a parasitic vector-borne disease that constitutes a major constraint to development in sub-Saharan Africa, exists under two forms: Human African Trypanosomosis (HAT) known also as sleeping sickness and African Animal Trypanosomosis (AAT) or Nagana. Tsetse flies (Glossina spp.) are the cyclical vectors of the trypanosome species causing these diseases. Some 46 million cattle distributed over 10 million km 2 in 38 sub-Saharan African countries [1,2] are estimated to be at risk of contracting AAT and hamper significantly the socioeconomic development of these African regions [3,4]. Among the 38 tsetse-infested countries, 34 are amongst the poorest in the world [5] and have included tsetse and trypanosomosis as a constraint in their poverty reduction strategy papers under the heavily indebted poor countries initiative [6].
The main strategies used to control or eradicate AAT remain (i) chemoprophylaxis and chemotherapy with trypanocidal drugs, (ii) promoting trypanotolerant cattle, and (iii) tsetse control or eradication programmes [1]. The Pan African Tsetse and Trypanosomiasis Eradication Campaign (PATTEC) initiative promoting integrated control of AAT and large tsetse eradication campaigns are underway in Uganda, Kenya and Ethiopia in East Africa and in Ghana, Burkina Faso and Mali in West Africa, with the aim of improving breeding and agriculture by creating new tsetse free areas [7].
Entomological evaluation tools represent an important component of any vector based control programme to appropriately target implementation areas and to evaluate their efficacy in time [8]. To date, the conventional method used within tsetse eradication campaigns is to estimate tsetse fly densities with traps deployed at fixed or temporary sites [9]. Important constraints are nevertheless associated with this method. The deployment and monitoring of traps, most of the time in very large areas (20 km/day/person in walking) with poor accessibility, is costly and demanding in terms of human resources and logistics. It is also known that traps have a poor efficiency (below 1 % of the flies present in a 1 km 2 around the trap are captured daily) and are becoming even less efficient at low tsetse densities [8,10,11]. Furthermore, traps are generally set up in sentinel fixed sites and thus only provide an indirect estimate of cattle exposure to tsetse bites especially in agro-pastoral areas where herds are very mobile. Alternative methods based on the evaluation of the antibody (Ab) responses directed against bloodsucking arthropod salivary antigens have been developed in the last few years. During the blood meal, hematophagous arthropods inject a mixture of anti-haemostatic, anti-inflammatory and immunomodulatory molecules into the skin of their hosts. These molecules play a crucial role in achieving an effective blood meal [12], but also in the establishment or not of pathogens into the vertebrate host [13]. The antigenic properties of these molecules have also been used to develop a range of immunoassays to detect associated specific Abs and to assess host exposure to a range of arthropod vectors of human pathogens [14].
Recent studies have shown that the human IgG response against whole salivary extracts (WSE) of several Glossina (G.) species (G. morsitans (m.) morsitans, G. fuscipes (f.) fuscipes and G. palpalis (p.) gambiensis) was correlated with human exposure to tsetse flies in different HAT endemic areas [15][16][17][18]. Similar results were observed in outbred cattle from a tsetse infested area in South-West Burkina Faso that were shown to harbor higher IgG responses than cattle in the North where tsetse flies are absent. These results were further confirmed in cows experimentally exposed to the bite of a range of Glossina species and other bloodsucking arthropods [19]. However, the use of WSE in immunoassays is likely impaired by the existence of potential cross-reactions with Abs directed against common saliva antigens that are shared by different arthropod species [20]. Other important drawbacks are the difficulty in achieving mass production of WSE in a standardized manner and the storage of these antigens over long periods. WSE are thus poorly suited for large scale studies [21,22]. Resorting to recombinant salivary proteins has enabled to overcome some of these limitations and a number of immunoassays based on specific recombinant salivary proteins have been developed to detect exposure to a range of arthropods [23][24][25][26][27]. An alternative strategy has been to identify specific linear epitopes by in silico approaches in order to design peptides to be used as the immunoassay antigens. Production of short synthetic peptides at high purity can easily be entrusted to a private company; shipment and storage are facilitated as they can be lyophilised. In the last few years, such approaches were successfully applied to develop salivary biomarkers of human exposure to Aedes aegypti [28] and Anopheles (An.). gambiae [21,22]. Recently, the IgG response to a synthetic peptide designed from the Tsetse Saliva Growth Factor-1 (Tsgf1) sequence, was shown to be specific of human exposure to tsetse flies [29] and was successfully used to assess the evolution of human tsetse contacts during a vector control intervention in Guinea [30].
In the present study, our aim was to design peptides that could be used to assess exposure of cattle to tsetse flies. An immuno-proteomic approach, using sera from animals experimentally exposed to several arthropod species and WSE from G. m. submorsitans, was first used to identify the most specific immunogenic salivary proteins to detect Glossina exposure. Several epitope prediction and protein conformation software were then used in combination with Blast analysis to design linear peptides within the identified proteins. Synthetic peptides were finally produced and evaluated in an indirect-ELISA test with serum samples from (i) cattle from tsetse free or tsetse infested areas and (ii) cows experimentally exposed to low or high numbers of tsetse flies.

Serum samples
All bovine serum samples used in this study were collected from cattle outbred in different environmental settings from Burkina Faso or were obtained by experimental exposure of cows to several arthropod species and are described in detail elsewhere [19]. Samples from outbred cattle included 17 samples collected in a tsetse free area in Northern Burkina Faso and 43 samples collected in a tsetse infested area in the South-West part of the country. Concerning experimentally exposed sera, we included six samples collected from six cows individually exposed weekly to the bite of G. m. submorsitans, G. p. gambiensis, Amblyomma (A.) variegatum, An. gambiae, Tabanidae spp. or Stomoxys spp. Samples used in the present study were those collected at the end of the exposure experiments, after 12 weeks for An. gambiae and 23 weeks for all other species. In addition we also used 96 serum samples collected weekly from two groups of four animals that were exposed to 50 G. m. submorsitans flies twice a week during 11 weeks (high exposure group) or 10 flies weekly during the same period (low exposure group).
Production of whole salivary extracts WSE were obtained from 10-12 day-old G. m. submorsitans uninfected males and females from the IRD/CIRAD colony in Montpellier (France). The tsetse saliva was collected as described previously [17] by a salivation technique that does not require the dissection of salivary glands to avoid the presence of non-salivary antigens in WSE. Briefly, tsetse flies (4-6 flies) were enclosed in 50 ml Falcon tubes closed by a mosquito net and placed above a drop of salivation buffer on warm slides. Buffer drops were collected after 10 min of salivation and were stored at −80°C before use. Prior to electrophoresis, WSE were desalted and concentrated using the 2-D Clean-Up Kit (GE Healthcare, Germany) according the manufacturer instructions. Protein concentrations of WSE were assessed by the Bradford method.
After the second dimension, a preparative 2DE-gel was fixed for 20 min in 50 % ethanol/5 % acetic acid solution, and then 10 min in 50 % ethanol solution, washed three times in ultrapure water. Finally, this gel was stained with a colloidal blue solution (Fermentas, Saint-Remy les Chevreuse, France) overnight and washed four times in ultrapure water.

Identification of G. m. submorsitans salivary proteins by mass spectrometry
Visible spots after colloidal blue staining were manually excised under a laminar flow hood and enzymatic in-gel digestion was performed automatically (Tecan freedom evo® proteomics) according to the Shevchenko modified protocol [31].
Briefly, protein spots were digested using 150 ng of trypsin, peptide extraction was performed using 5 sonication cycles of 2 min each and peptides were concentrated 1 h at 50°C in a heat block. Peptide samples were automatically spotted (Tecan freedom evo® proteomics). For this step, 0.5 μl of peptide sample and 0.5 μl of αcyano-4-hydroxy-trans-cinnamic acid (a saturated solution prepared in acetonitrile/trifluoroacetic acid, 50: 0.1 %, vortexed, sonicated 30 s and microcentrifuged 30 s with a 1/3 dilution of the supernatant used as the matrix) were deposited on a 384-well MALDI anchorship target using the dry-droplet procedure [32] and air dried at room temperature. Peptide samples were then desalted using a 10 mM phosphate buffer and dried again at room temperature. MALDI-TOF MS analysis was performed using UltraFlex MALDI TOF-TOF mass spectrometer (Brucker Daltonics, Bremen, Germany) in the reflecton mode with a 26 kV accelerating voltage and a 50 ns delayed extraction. The AutoXecute™ module of Flexcontrol™ v3.0 (Bruker Daltonics) (laser power ranged from 40 to 50 %, 600 shots) was used to acquire mass spectra. Spectra were analysed using FlexAnalysis™ software v3.0 (Bruker Daltonics) and calibrated internally with the autoproteolysis peptides of trypsin (m/z: 842.51; 1045.556; 2211.10). Peptides were selected in the mass range of 900-3000 Da.
Peptide Mass Fingerprint identification of proteins was performed by searching against the Glossina entries of either the Swiss-Prot or TrEMBL databases (http:// www.expasy.ch) and by using the MASCOT v2.3 algorithm (http://www.matrixscience.com) with trypsin enzyme specificity and one trypsin missed cleavage allowed [33]. Carbamidomethyl was set as fixed cystein modification and oxidation was set as variable methionine modification for searches. A mass tolerance of 50 ppm was allowed for identification. Matching peptides with one missed cleavage were considered as pertinent when there were two consecutive basic residues or when arginine and lysine residues were in an acidic context. MASCOT scores higher than 47 were considered as significant (p < 0.05) for Swiss-Prot and TrEMBL (v 2011_04) database interrogations.

Immunoblotting and specific proteins identification
The proteins separated by 2DE were then electrotransferred onto a PVDF (polyvinylidene dilfluoride, Biorad) membrane for western blotting as previously described [34]. The PVDF membrane was washed in Tris Buffer Saline (TBS) and then incubated in blocking buffer (TBS tween 0.05 % and 5 % dry milk) for 1 h 30 min. After washing three times with TBS tween 0.1 %, the membrane was incubated overnight at 4°C with bovine serum diluted at 1/100 in TBS tween 0.05 % with 2.5 % dry milk. The membranes were then washed three times with TBS tween 0.1 % and three times with TBS, and then equilibrated with 2.5 % dry milk in TBS for 15 min. Mouse anti-bovine IgG conjugated to peroxidase (Sigma, St Louis, MO, USA) was added at a dilution of 1/15 000 in TBS tween 2.5 % with 2.5 % dry milk for 2 h 30 at room temperature. After addition of the secondary antibody, the membranes were washed four times with TBS tween 0.1 % and four times with TBS. The immunogenic proteins in membranes were revealed using the West Pico ECL (Pierce, Rockford, IL, USA) and exposed to XCLXposure films (Pierce) for 4 min.
Digital images of both western blotting and 2D gel were captured by scanning at 16 bits resolution under non saturating conditions, 300 dpi and stored in TIFF format. The proteins visualised on 2D gels were matched with immunogenic proteins detected by western blotting using SameSpots TM Software 3.3 (Nonlinear Dynamics) in order to identify immunogenic spots.

Peptide design
The sequences of proteins of interest were downloaded on the ExPASy Proteomics Tools server (http:// www.expasy.ch). The signal peptides were predicted by SignalP 4.0 [35] and cleaved from the immature protein sequences. Secondary structures were then predicted with NetSurfP [36] and I-Tasser [37] servers. The surface exposure of proteins of interest has been viewed in 3D by the Pymol software (http://www.pymol.org).
The identification of putative linear B-cell epitopes of identified proteins was carried out with Bcepred [38], Bcpred [39] and Antigenicity plot server [40]. All epitopes that were identified by at least two out of three algorithms were selected. Peptides having a length less than 30 amino acids and including the maximum of epitopes were selected for further analyses. The specificity of peptide sequences to the Glossina genus was checked by requests on the NCBI Blast T non redundant databases (http://www.blast.ncbi.nlm.nih.gov/Blast.cgi). Synthetic peptides were synthesized and purified (95 %) by Genepep SA (St-Jean de Vedas, Montpellier, France). All peptides were shipped in lyophilised form and were then resuspended in ultrapure water (1 mg/ml) and stored as frozen aliquots.

Evaluation of bovine IgG Ab level against synthetic peptides and WSE
The anti-peptide IgG responses were measured by indirect-ELISA, according to Poinsignon et al. [22]  IgG responses against WSE were also evaluated on the same samples as described previously [19]. Each test sample was analysed in duplicate in antigen wells and, in parallel, in a blank well containing no peptide solution or no WSE (ODn) to control non-specific reactions between the serum and the reagents. Individual results were expressed as ΔOD value calculated according to the formula ΔOD = ODx -ODn, where ODx represents the mean of individual OD in both antigen wells, as previously used [22].

Statistical analysis
All data were analysed with GraphPad Prism5 software® (San Diego, CA, USA). After verifying that ΔOD values did not assume Gaussian distribution using Shapiro-Wilk test, the non-parametric Mann-Whitney U-test was used for comparison of Ab levels between two independent groups. All differences were considered significant at P < 0.05.

Results
Identification of specific immunogenic salivary proteins of Glossina exposure The sialome of G. m. submorsitans was investigated by 2D-gel electrophoresis separation of the WSE followed by colloidal blue staining (Fig. 1) and subsequent MALDI-TOF/MS identification. A total of 53 spots were observed with colloidal blue staining and could be picked for mass spectrometry analysis. Forty-seven spots were successfully identified by MALDI-TOF MS analysis providing a catalogue of seven salivary proteins (Table 1). For the spots of low molecular weight (< 25kD) no identification could be obtained because the amount of material was too scanty. In contrast, low molecular weight spots are much more intense in 2-D electrophoresis gels obtained with G. m. morsitans saliva extracts obtained by centrifugation of salivary glands [41]. These differences may be related to differences in the saliva composition of these two closely related subspecies but also to the saliva collection technique. Several spots with similar molecular weight but different isoelectric points led to the same identification suggesting the existence of isoforms for these proteins. All spots were identified as related to G. m. morsitans salivary proteins [41]: the Tsetse salivary gland protein 1 (Tsal1); the Tsetse salivary gland protein 2 (Tsal2) and its two isoforms (Tsal2A and Tsal2B); the Tsetse Salivary Growth Factor 1 (TSGF-1); the Salivary Secreted Adenosine (SSA); the Adenosine Deaminase-related Growth Factor C (ADGF-C); the 5′Nucleotidase family salivary protein (5′-nuc) and the Tsetse Antigen 5 (TAg5).
In order to identify G. m. submorsitans immunogenic salivary antigens that are specific of cattle exposure to tsetse flies, we established the 2D immunoblot profiles of serum samples from cows experimentally exposed to tsetse flies (G. m. submorsitans, G. p. gambiensis) or to other bloodsucking arthropods that are common in the study area (A. variegatum, An. gambiae, Tabanidae spp. or Stomoxys spp.). A number of spots were common between the two tsetse species but differences could be observed (Fig. 2). Whereas ADFG-C appears to be immunogenic in the cow exposed to G. p. gambiensis, this salivary protein did not react with the G. m. submorsitans serum. On the contrary, the G. p. gambiensis serum did not recognize any spot associated with the Tsal2 isoforms (Tsal2A and Tsal2B), TSGF-1, 5′-nuc and TAg5 families of salivary proteins. Importantly, cross reactions (although weak) were also observed with the sera of cows experimentally exposed to Stomoxys spp. or An. gambiae which recognised Tsal2, 5′-nuc, TAg5 and Tsal2, Tsal2A, 5′-nuc, TAg5 respectively ( Table 2). Based  on these results we decided to focus the peptide design on Tsal1 and SSA sequences as these salivary antigens were recognised only by the sera of animals individually exposed to both tsetse species and did not react with the sera of animals bitten by the other arthropod species.

Peptide design
Four and one sequences of 22 to 28 amino acid residues containing putative linear epitopes were identified from the G. m. morsitans Tsal1 and SSA sequences respectively (Additional file 1: Table S1). In order to avoid potential cross-reactivity with Abs against proteins from other bloodsucking arthropod species as well as from host pathogens, these 5 peptides were submitted to the NCBI Blast T non redundant databases (Additional file 1: Table  S1). The SSA 65-92 peptide showed some degree of homology with Culex quinquefasciatus and was therefore dropped out in further analyses. Similarly, among the 4 peptides identified to carry putative Tsal1 linear epitopes, we selected Tsal1 52-65 and Tsal1 145-166 as for the other two peptides, closest matches were obtained for organisms to which cattle may be naturally exposed (Plasmodium yoelii yoelii and An. darlingi).
After checking for Ab accessibility of these two peptides at the surface of the Tsal1 protein using 3D models (Fig. 3), Tsal1 52-75 and Tsal1 145-166 were synthesised and tested against cattle serum samples.  Table 2 Glossina morsitans submorsitans salivary secreted proteins recognised by cows exposed to tsetse and other hematophagous arthropods Arthropod species used for experimental exposure Tsal1  Tsal2  Tsal2A  Tsal2B  TSGF-1  SSA  ADGF-C  5′-nuc  TAg5 G. m. submorsitans indicates Glossina morsitans submorsitans salivary proteins recognised on 2D gels by western blot with sera from animals exposed experimentally to several arthropod species Ability of synthetic peptides to detect exposure to tsetse flies The ability of the two synthetic peptide candidates to detect exposure of cattle to tsetse flies was first assessed on serum samples from cattle bred in different ecoclimatic zones in Burkina Faso. Serum samples from the cows experimentally exposed to the none-tsetse species were also included as negative controls (Fig. 4). Anti-Tsal1 52-75 IgG responses were significantly higher (P = 0.009) in cattle from tsetse infested area as compared to those from tsetse free area, and were the lowest in the control animals exposed to tabanids, stable flies, mosquitoes or ticks. They also provided a better discrimination between animals from tsetse infested or free areas as compared to the response directed against WSE (P = 0.116). In contrast no significant differences in the IgG response specific to Tsal1 145-166 (P = 0.574) were observed according to the exposure status, and this peptide was dropped out in further analyses. This may be due to the fact that as shown in Fig. 3 the Tsal1 145-166 sequence is predicted to be involved in the formation of Tsal1 secondary structures, which may have altered its immunogenic properties.
Finally the performance of the Tsal1 52-75 peptide as a biomarker of exposure was further explored with experimental sera obtained from cows exposed to low or high regimen of tsetse bites (Fig. 5). In the low exposure group (10 flies/once a week), an early increase of the specific IgG response was observed after only three weeks of exposure and was maintained over the exposure period (11 weeks). Unexpectedly, very low IgG Ab levels were observed in the high exposure group (50 flies/twice a week), throughout the exposure period  [37]. N-ter is the first amino acid of the protein and C-ter, the last. Candidate biomarker peptides are colored in red for Tsal1 52-75 and green for Tsal1 145-166 Fig. 4 Cattle IgG responses against WSE and, Tsal1 52-75 and Tsal1 145-166 peptides . The IgG responses directed against whole saliva extracts (WSE) and the two candidate synthetic peptides were investigated in 43 animals from a tsetse infested area, 17 animals from a tsetse free area and four animals exposed experimentally to A. variegatum, An. gambiae, Tabanidae spp. or Stomoxys spp. (other arthropods). Individual ΔOD values are represented by empty circles. In the scatter plot, the horizontal bars indicate the median value for each group. Statistical significance between the different groups is indicated (non-parametric Mann-Whitney U-test) despite the fact that these animals mounted high IgG responses to WSE.

Discussion
In this study, immuno-proteomics and bioinformatics tools were combined to design specific peptides as potential biomarker candidates for evaluating cattle exposure to tsetse flies. Indirect-ELISA tests using the identified peptides as antigens were then performed on sera from cattle exposed naturally or experimentally to tsetse bites. The Tsal1 52-75 peptide appears as a promising candidate as anti-Tsal1 52-75 IgG responses were detected in both naturally exposed animals and in cows submitted to low tsetse exposure levels that where close to the tsetse challenge observed in the study area [42].
A number of tests have been developed to assess host exposure to tsetse bites. Early studies focused on the detection of host Abs raised against WSE from several tsetse species: G. p. gambiensis to assess human exposure in West Africa [18]; G. m. submorsitans to assess cattle exposure in West Africa [19]; G. f. fuscipes to assess human exposure in Central Africa [16,17]; and G. m. morsitans to assess human exposure in East Africa [15]. These studies showed that proteins from the Tsal family are major constituents of tsetse saliva, and induce strong Ab responses in tsetse exposed hosts. These proteins were thus considered as interesting candidates to develop biomarkers of tsetse exposure. Consecutively, it was shown that a G. m. morsitans Tsal1 recombinant proteins could be used instead of WSE in mice and pigs experimentally exposed to tsetse flies [15,26]. Because the production of Tsal recombinant proteins in large quantities is difficult, possibly due to the DNA binding/ endonuclease activity of Tsal proteins [43], these authors also developed a nanobody-based competitive immunoassay to detect anti-Tsal Abs [44]. The advantage of this method is that the same test can be applied to a wide range of hosts; nevertheless the test still requires the use of WSE, the production/storage of which can be a limitation in the context of laboratories from developing countries. In the present study, we evaluated the immunogenic properties of G. m. submorsitans salivary antigens in cows exposed to two tsetse species and to other bloodsucking arthropods. Whereas G. m. submorsitans Tsal proteins were shown to be highly immunogenic, only anti-Tsal1 IgG Abs were specific to tsetse fly exposure as immune cross-reactions with Tsal2 proteins were observed in animals exposed to stable flies or An. gambiae. The results presented in the present work point out Tsal1 as the best salivary antigen candidate to develop a highly specific biomarker of cattle exposure. Nevertheless, a study carried out on humans led to different results [18] as both Tsal proteins were recognised by sera from unexposed individuals. Instead this work led to the identification of a specific epitope within the TSGF-1 salivary protein [29], which has now been validated to monitor human exposure to G. p. gambiensis during a vector control campaign in Guinea [30]. It is noteworthy that in the present study, Ab response was detected against TSGF-1 only in the cow exposed to G. m. submorsitans. This illustrates that results obtained in a given animal model cannot always be extrapolated to another. These differences are likely due (i) to the sequence diversity of salivary proteins between the different tsetse species; (ii) to host species specificities in immune recognition; but also (iii) to the range/level of biting insects or pathogens to which tsetse hosts are submitted to and that can vary greatly between mammals or eco-climatic contexts. Hence, available serological tools, especially those relying on the recognition of a limited number of epitopes (such as it is the case for recombinant proteins and synthetic peptides), should be carefully evaluated prior to implementation as specificity and sensitivity of a given test may vary greatly according to the different contexts. a b Fig. 5 Monitoring anti-Tsal1 52-75 and anti-tsetse saliva antibody responses in cows experimentally exposed to low and high levels of tsetse bites. a Low exposure group (10 flies weekly) and (b) high exposure group (50 flies twice a week). Vertical bars above or below the curves are the standard errors of the group mean Tsal1 52-75 -based immunoassays, appear as promising tools to assess cattle exposure in West Africa where G. morsitans and G. palpalis subspecies represent the main tsetse species. Because this peptide was designed from the G. m. morsitans Tsal1 sequence, it might also be applied more widely as suggested by the results obtained experimentally on mice and pigs with the G. m. morsitans r-Tsal1 protein [26]. A surprising result was the fact that in our experimental conditions, no Ab response to Tsal1 52-75 was observed in the group of cows submitted to intensive tsetse fly bites despite the fact these cows exhibited strong Ab responses against WSE. This suggests that single epitopes behave differently in terms of immunogenicity according to the exposure conditions. Noteworthy, r-Tsal1 indirect-ELISA tests [26] or Tsal specific monoclonal nanobodies [44] were also less efficient than WSE to discriminate between mice or pigs exposed to different biting regimens. According to our results, the anti-Tsal1 52-75 Ab response represents a biomarker of low exposure levels but is likely less useful to measure the intensity of cattle exposure. The mechanisms underlying this intriguing result are not yet understood but could be related to antigen specific B cell exhaustion or anergy induced by high antigenic stimulation levels. This is however, an interesting feature for a biomarker candidate as it suggests that the development of a qualitative Tsal1 52-75 synthetic peptide-based immunochromatographic rapid test to detect low tsetse exposure levels is a reachable goal.
Declaring tsetse free areas or detecting the possible reemergence or reintroduction of tsetse flies after interventions is an important aspect of tsetse eradication campaigns [45]. Using tsetse traps only is challenging because this entomological method is not sensitive, even less when tsetse densities are low [8]. Hence in such context, it underestimates the true tsetse density or incorrectly concludes to the absence of flies. Serological tests able to detect low exposure levels could thus represent important alternative and complementary tools. Such tests could be used in the field on cattle herds or sentinel animals that are mobile baits naturally attractive for tsetse flies. Such sentinel animals are already commonly used in the frame of tsetse vector control campaigns to monitor trypanosome infections, an indirect marker of tsetse exposure. In our experimental conditions, the bite by less than 30 flies over a period of three weeks was sufficient to induce anti-Tsal1 52-75 Ab responses. Further studies evaluating different biting regimens (number of flies, biting frequencies) as well as rechallenge experiments in previously exposed animals are required to determine more precisely the sensitivity of Tsal1 52-75 -based immunoassays. Indeed anti-saliva Abs were shown to be boosted by very low numbers of tsetse bites in re-challenged mice and pigs [26]. It will also be useful to determinate the persistency of Tsal1 52-75 Ab after an exposure to tsetse bites.

Conclusions
In conclusion, we identified a Tsal1 peptide whose Ab response is specific of cattle tsetse fly exposure. The IgG response directed to the Tsal1 52-75 synthetic peptide could be a biomarker of low cattle exposure. These are promising results in the framework of developing simple Tsal1 52-75 based immunoassays (such as rapid tests) to monitor the tsetse flies presence at low fly densities or to detect early reemergence in previously cleared areas.