Ethics statement

Ethical approval was granted by Institutional Review Board in the Academy of Military Medical Sciences, Peoples Liberation Army P. R. China (AMMS-IRW-202000036). All experiments were performed in accordance with relevant guidelines and regulations.

Bandavirus dabieense: Severe fever with thrombocytopenia syndrome virus (SFTSV) Xinyang strain: (Accession no. JQ341188,JQ341189 and JQ341190 for N, S, L, directly submission): a virus strain isolated from a patient from Shangcheng County, Xinyang City, Henan Province, China.

Ticks: SFTSV-free Haemaphysalis longicornis: H. longicornis ticks originally collected from Changping District, Beijing, China, were maintained at the insectary of AMMS at 26°C with 85% humidity under a 12/12h light/dark photoperiod. All the tick individuals were harvested from SFTSV-free maternal adults and the subsequent offspring were sampled to verify the negative infection of SFTSV by the specific RT-PCR screening. Unfed nymphs were used in the experiments at the age between 10–20 days post-molt.

Ifnar^-/- mouse

Homozygous Ifnar^-/- mice, female, aged 6–8 weeks, were purchased from Cyagen Biosciences Inc (Guangzhou, China). All experiments were performed using institutional guidelines for animal care and experimentation under the official approvals obtained prior to the in vivo assays.

Surveillance of tick burden on Apodemus agrarius in SFTSV endemic area

Apodemus agrarius mice were trapped in Ningbo City, Zhejiang Province from 12 June, 2018–14 June, 2020 using perforated Sherman traps (LFAHD folding trap, 7.62 × 8.89 × 22.86 cm, Sherman Traps, Inc, US) baited with peanut butter, sunflower seeds and apples. Traps were placed in the late afternoon and processed the following morning. The number of trap squares used (4 traps per square, with an inter-trap distance of 20m) depended on the sample size. We calculated the rodent abundance index per 100 traps as the sum of rodents over the trapping effort to control for the differences in trapping effort between sampling sites. Power analysis showed that a capture success of 10% to 25% at each sampling site was sufficient for evaluation. The captured rodents were delivered to the laboratory, and anesthetized in a closed container with ether or chloroform for about 10min to prevent the escape and bite of various parasites on the rodent body surface, then A. agrarius mice were morphologically identified according to the identification pictures of common medical vectors [33], as they were the most common rodent species and potential host reservoir of SFTSV in our sampling sites. After anesthesia with isoflurane, the mice were thoroughly checked for all ticks with fine forceps. Ticks harvested were counted and identified to species and life stage by morphological characteristics following the taxonomic key developed by Teng and Jiang [34]. The mice were also dissected to obtain liver, spleen and lung tissues which were preserved at liquid nitrogen until further processing.

Experimental infected Ifnar^-/- Mice with SFTSV

Female Ifnar^-/- mice (n = 6/group) were intraperitoneally injected with SFTSV (Xinyang strain) at a tissue culture infectious dose of 50 (TCID₅₀) per mouse, as these doses ensured mouse survival after challenge and would benefit subsequent sample collection and observations, an equivalent volume of PBS was used as a control [35]. Body weight changes, anal temperatures, and SFTSV RNA loads were monitored daily for 12 days immediately after SFTSV infection. SFTSV RNA loads in tail vein blood samples were determined by quantitative RT-PCR (qRT-PCR) as described below.

Fecal and urine samples collected from SFTSV-infected Ifnar^-/- mice

The fecal and urine samples were collected from Ifnar^-/- mice infected and noninfected with SFTSV. After SFTSV RNA loads > the initial infection dose in the challenged mouse, the mouse was individually placed into a sterilized Tecniplast E-Chiller cabinet (Tecniplast, China) to collect the fresh fecal and urine samples at 4°C. About 0.2g of fecal samples were taken in 1mL falcon tubes, homogenized and centrifuged (50,000 × g at 10°C for 2h), the supernatant is collected. The urine samples of 0.2mL in falcon tubes were centrifuged to remove any debris (50,000 × g at 10°C for 15min). At the time of solid phase microextraction (SPME) analysis, each vial of fecal water or urine samples was placed in 60°C for 5min, then exposed to 120μm DVB/CWR/PDMS fiber (Agilent) for 15min at 60°C.

GC-MS analysis for differential volatile metabolites (DVMs) in infected Ifnar^-/- mice

After sampling, desorption of the VOCs from the fiber coating was carried out in the injection port of the GC apparatus (Model 8890; Agilent) at 250°C for 5 min in the splitless mode. The identification and quantification of VOCs was carried out using an Agilent Model 8890 GC and a 7000D mass spectrometer (Agilent), equipped with a 30m × 0.25mm × 0.25μm DB-5MS (5% phenyl-polymethylsiloxane) capillary column. Helium was used as the carrier gas at a linear velocity of 1.2mL/min. Temperatures for inlets and MS source were taken as 250°C and 230°C, respectively. The oven temperature was programmed from 40°C (3.5min), increasing at 10°C/min to 100°C, at 7°C/min to 180°C, at 25°C/min to 280°C, hold for 5min. Mass spectra were recorded in electron impact (EI) ionization mode at 70eV. The quadrupole mass detector, ion source and transfer line temperatures were set at 150, 230 and 280°C respectively. The ion monitoring (SIM) mode of the MS was selected for the identification and quantification of the analytes. Briefly, the total ion chromatogram was obtained and then the mass spectra were identified. The detected metabolite peaks were identified using the NIST08 (National Institute of Standards and Technology) mass spectral library, and the extracted compounds were aligned and normalized to the internal standard ribityl [36]. (1) Principal component analysis (PCA). Unsupervised PCA was performed using the prcomp statistical function within R (www.r-project.org). The data were unit variance scaled prior to unsupervised PCA. (2) Hierarchical cluster analysis (HCA) and Pearson correlation coefficients (PCC). The HCA results of samples and metabolites were presented as heatmaps with dendrograms, while the PCC between samples was calculated by the corfunction in R. Both HCA and PCC were carried out by R package Complex-Heatmap. For HCA, normalized signal intensities of metabolites (unit variance scaling) are visualized as a color spectrum. (3) Selected differential metabolites. For the paired group analysis, differential volatile metabolites (DVMs) were determined by variable importance in the projections (VIP) (VIP > 1) and absolute Log₂Fold Changes (|Log₂FC|) ≥ 1.0. VIP values were extracted from the partial least squares discriminant analysis (PLS-DA) results, including score plots and permutation plots, generated using the R package MetaboAnalystR [37]. The data were logistically transformed (log2) and mean centroided prior to PLS-DA. A permutation test (200 permutations) was performed to avoid overfitting. (4) KEGG Annotation and Enrichment Analysis. Identified metabolites were annotated using the KEGG compound database (http://www.kegg.jp/kegg/compound/), the annotated metabolites were subsequently mapped to the KEGG pathway database (http://www.kegg.jp/kegg/pathway.html). Pathways with significantly regulated metabolites mapped were then fed into MSEA (metabolite sets enrichment analysis), their significance was determined by hypergeometric test P-values.

SFTSV screening

SFTSV was tested in samples of live ticks and mouse tissues (liver, spleen and lung). Samples were homogenized and centrifuged. Total nucleic acids (including RNA and DNA) were extracted from the homogenates using a Viral RNA MiniKit (52904) (QIAGEN, China) according to the manufacturer’s instructions. Real-time fluorescence quantitative PCR assays for the target genes were then performed using the Takara one-step RT-PCR kit (Takara, Japan) [38]. The total volume of the reaction system of SFTSV was 25 μL, including 2 × reaction mix 12.5 μL, probe (10 μM) 0.3 μL, upstream and downstream primers (S2 Table) (10 μM) 0.5 μL, RNA template 5 μL, enzyme mix 1.0 μL, and deionized water supplement. The cycling parameters were 50°C for 30min (one cycle), 95°C for 10min (one cycle), 95°C for 15s and 60°C for 45s (40 cycles). Cycle threshold values of SFTSV was set as ≤35. Data were analyzed using the software supplied by the manufacturer.

Choice tests

H. longicornis nymphs were allowed to choose between Y-tube olfactometer [1] arms by random assignment, with one arm containing SFTSV-positive Ifnar^-/- mice, healthy mice or DVM candidates (indole or 3-methylindole), and the other arm as an air control. The flow of odor-laden air was maintained at a rate of 0.2L/min for 2 min before the tests began. Each tick that proceeded 8.0 cm or more into either a treatment or control side arm as its first choice within 12min was considered a responder. All other ticks were considered non-responders.

Potent electrophysiological responses of H. longicornis to volatile indole compounds

H. longicornis nymphs were subjected to test the olfactory response with EAD following Josek’s protocol [39]. Briefly, the tick individual was attached ventrally to a circular metal plate (Ø 1 cm) with double-sided adhesive tape. Both the indifferent and the recording electrodes were filled with 10⁻²M KCl and 1% solution of polyvinylpyrolidone K90 (Fluka, Switzerland). The indifferent electrode, connected to the ground via a chloridized wire, was inserted in the region posterior to the scutum, after piercing it with a fine forceps. One of the forelegs was orientated to expose the anterior sensilla of Haller’s organ and immobilized with an adhesive tape. Pedal nerves of the forelegs were destroyed by pinching the coxa with fine forceps to prevent muscle activity during electrophysiological recording. In order to improve contact, the tips of the distal knoll sensilla were cut with metal knives fitted on micromanipulators. Preparation of the ticks and subsequent imaging were performed under visual control (Leica MZ12 stereomicroscope, 350 × magnification). Prior to imaging, the tip of the sensillum was cut with a piece of a razor blade in a holder mounted in a micromanipulator (NMN25, Narishige) under visual control. Recordings from the olfactory receptors were accomplished with glass electrodes connected to a high-input impedance preamplifier (10×) (Syntech INR-5, Hilversum) and brought into contact with the cut tip of the sensillum with the aid of micromanipulators. The recordings were sampled (13,714.3samples/s) and filtered (10–3000Hz, with 50/60Hz suppression) via USB-IDAC connection to a computer (Syntech, Hilversum). At least 10 ticks were tested with both indole and 3-methylindole at concentrations of 0.1, 1, 10 and 100μg/mL. The action potentials were extracted as digital spikes according to top–top amplitudes, using the Autospike software (version 3.9, 16 June, 2009, Syntech NL). The recording duration was 10s, and the stimulus was applied 500ms after the start of the recording. Responses were evaluated according to the difference in the number of spikes between the 500ms stimulation period and the 500ms period before the start of stimulation (dsf = difference in spike frequencies). One-way analysis of variance with repeated measurements (ANOVAR) followed by Bonferroni’s post hoc correction test was used to analyze the mean dsf discharged at each concentration of each treatment (substance tested), using the SPSS 17.0 software package (SPSS Inc., Chicago, IL, USA). Statistical significance was achieved when P < 0.05. Odorants that elicited a significantly higher dsf than the blank stimulus were considered active. Two-way ANOVAR with Bonferroni’s post hoc correction test was used to compare the effects of pairs of stimulating odorants. In cases where the sphericity assumption was violated, Greenhouse-Geisser (G-G) or Huynh-Feldt (H-F) corrections were applied when ε’ > 0.75 or ε’ < 0.75, respectively.

Monitoring the transcript level of candidate olfactory proteins with qRT-PCR

The transcript profiles for OBPL and NPC2 during the induction of indole and 3-methylindole at different concentrations were measured by quantitative RT-PCR using an ABI 7500 Real-Time PCR System (Applied Biosystems, Carlsbad, CA, USA). The housekeeping gene, β-actin, was used to normalize target gene expression and to correct for sample-to-sample variation. TaqMan primers (Table 1) for the amplification of β-actin, NPC2 and OBPL were reported by Cui et al. (2022) [40]. For the qRT-PCR reaction, cDNA was diluted to 200ng/mL. Each reaction was performed in a 20μL final volume containing 10.0μL TB Green Premix Ex Taq (TaKaRa), 0.8μL of each primer (10mM), 0.5μL probe (10 mM), 0.4μL Rox Reference Dye II, 2μL sample cDNA (200ng) and 6.0μL sterilized H₂O. The reaction cycling parameters were as follows: 95°C for 30 s, then 40 cycles at 95°C for 5s and 60°C for 34 s. And then, the reaction of 95°C for 15s, 60°C for 1min and 95°C for 15s was added to establish the melting curve. For data reproducibility, reactions were performed in triplicate, and three biological replicates were assessed. Negative controls were non-template reactions (sterilized H₂O instead of cDNA). According to the method described, the relative quantification of the target genes’ transcripts between different treatments was calculated using the comparative 2^(-△△Ct) method [39] with thresholds of |Log₂Fold Change| ≥ 0.2, and P-value≤0.05 set significant differences. Comparative analyses of the transcript level of target genes amongst different treatments were conducted using a one-way nested analysis of variance followed by Tukey’s honestly significant difference test. The relative mRNA transcript levels in nymphs’ forelegs between treatments were compared using Student’s t-test. All analyses were conducted using SPSS STATISTICS version 18.0 software (SPSS Inc., Chicago, IL, USA).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. The primer sequences used in the study.

https://doi.org/10.1371/journal.pntd.0013139.t001

Annotation of candidate olfactory proteins and their predicted 3D structures

Total foreleg RNA was isolated by trizol reagent (Invitrogen, Carlsbad, CA, USA) to construct cDNA library using the Creator SMART cDNA Library Construction Kit (Clontech, Mountain, CA, USA). The cDNA library was sequenced with Illumina sequencer. Candidate OBPL and NPC2 genes were identified by BlastX and MotifSearch program. OBPL and NPC2 protein sequences identified in H. longicornis and others reported in arthropod’s species were aligned using ClustalX 1.83 respectively. The putative N-terminal signal peptides were predicted by the SignalP V5.0 program (http://www.cbs.dtu.dk/services/SignalP/) [41]. A Swiss model method [42] was used to search structural templates for the two candidate olfactory proteins. Several identified OBPs structures were used as templates to construct 3D structures for the two candidates, the one with the highest score of Profiles-3D was retained. The Profiles-3D method and Ramachandran plot were used to evaluate the rationality of the established 3D model [43]. Molecular docking was performed by the on-line program SWISSDOCK using default parameters [44], the docking scores of AC and SwissParam were used to evaluate the fitness of ligand and target proteins [45].

Expression and purification of recombinant OBPL and NPC2 proteins

Two specific primer sets are designed to clone the coding region of OBPL/NPC2 in H. longicornis as followed: OBPL-F: 5‘- GTCATATGGCTGCCACGTACACGTCC-3’, OBPL-R: 5‘-TGAAGCTTTCAGTGGCTTC CGGGCAA-3’; NPC2-F: 5‘- AACATATGAAATACTACACGGATTG-3’, NPC2-R: 5‘- AAGAAT TCTTATTGTATCTTGGCGGC-3’ (Underlined showed Nde I and Hind III enzyme sites in the forward and reverse primer, respectively.) The PCR products were cloned into the bacterial expression vector pET28b (Promega, Madison, WI) between the Nde I and Hind III restriction sites, and verified by sequencing (S6 Fig). Plasmid containing the correct insert was extracted and transformed into Escherichia coli BL21(DE3) competent cells. A verified single colony was grown overnight in 50mL LB broth (including 100mg/mL Kanamycin). Five liters of LB medium was inoculated with the 50mL overnight culture at 37°C for 2–3 hours until the absorbance at OD₆₀₀ reached 0.6. The proteins were then induced with isopropyl-b-D-thiogalactopryranoside (IPTG) with a final concentration of 1 mM at 37°C for 6h. The bacterial cells were harvested by centrifugation (8000g, 10min), resuspended in the lysis buffer (80mM Tris-HCl, 200mM NaCl, 1mM EDTA, 4% glycerol, pH7.2, 0.5mM PMSF), lysed by sonication (10sec, 5 passes) and centrifuged again (12000g, 10min). The soluble fraction and the whole pellet were analyzed by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and found the target proteins mainly present in the inclusion bodies. Insoluble proteins were washed with 0.2% triton X-100 in 50mM Tris buffer (pH6.8) and then dissolved in 6M guanidinium hydrochloride, the protein refolding protocols performed using the redox methods [46]. Soluble and refolded target proteins were purified with two rounds of Ni-21 ion affinity chromatography (GE Healthcare), and the His-tag was removed using recombinant enterokinase (Novagen, Madison, WI, USA) respectively. The highly purified protein was desalted through extensive dialysis, and the size and purity of the recombinant protein were confirmed by 15% SDS-PAGE. The concentration of the purified target proteins was measured by the Bradford method using BSA as standard protein [47].

Fluorescence competitive binding assay

Both indole and 3-methylindole were selected for fluorescence competitive binding assays according to the DVMs with significant olfactory responses in SFTSV infected mice. This experiment was performed on an F-380 fluorescence spectrophotometer (Gangdong Sci & Tech Development. Co., Ltd, Tianjin, China) at room temperature (25°C) with a 1 cm light path quartz cuvette and 10-nm slits for excitation and emission. The excitation wavelength was 337nm, and the emission spectrum was recorded between 390 and 460nm. The dissociation constants of OBPL or NPC2 with the fluorescent probe 1-NPN was measured, and a final concentration of 2mM protein solution in 50mM Tris-HCl (pH7.4) was titrated with aliquots of 1mM 1-NPN dissolved in methanol to final concentrations ranging from 1 to 16mM. The binding affinities of the chemicals were tested through competitive binding assays using 1-NPN as the fluorescent reporter at a concentration of 2mM, and the concentration of each competitor was varied from 0 to 32mM. The fluorescence intensities at the maximum fluorescence emission between 390 and 460nm were plotted against the free ligand concentration to determine the binding constants. The bound chemical was evaluated based on fluorescence intensity with the assumption that the protein was 100% active, with a saturation stoichiometry of 1:1 (protein: ligand). The binding curves were linearized using a Scatchard plot, and the dissociation constants of the competitors were calculated from the corresponding IC₅₀ values based on the following equation: K_D= [IC₅₀]/(1+ [1-NPN]/K_1-NPN), where [1-NPN] is the free concentration of 1-NPN and K_1-NPN is the dissociation constant of the complex protein/1-NPN [48].

Statistical analysis

Data from qPCR and EAD tests were analyzed using SPSS 17.0 (SPSS Inc., Chicago, IL, USA). ANOVA and Tukey’s Honestly Significant Difference (HSD, P = 0.05) were used to determine differences in OBPL or NPC2 mRNA levels or whether EAD recordings were significant among different treatment groups. The t-test was employed to evaluate significant differences in OBPL or NPC2 mRNA levels in the forelegs of H. longicornis between different treatments.

High H. longicornis burden on SFTSV-positive Apodemus agrarius in field trials

In our field trials, a total of 262 Apodemus agrarius (178 males and 84 females) were trapped and 19.84% (35 males and 17 females) were infested with immature H. longicornis ticks. Of the infested mice, 44 individuals (30 males and 14 females) were tested SFTSV positive and 8 (5 males and 3 females) were free of SFTSV. There were no significant difference of the tick infestation ratio and SFTSV infection rate between the sexes of A. agrarius. However, A. agrarius suffering from SFTSV infections were likely to carry more H. longicornis ticks than negative ones as both the overall density and intensity of infested immature H. longicornis on SFTSV-positive A. agrarius appear much higher than those on SFTSV-free mice (Pearson Chi-Square test, χ² = 37.2, P < 0.001). (S1 Fig). The heavy burden of H. longicornis on the SFTSV-positive A. agrarius suggests the infestation predilection of H. longicornis for the infected hosts induced by SFTSV infection. And in the following, we tested our hypothesis that H. longicornis ticks are attracted to the characteristic host clues of SFTSV-positive mice.

Differential volatile metabolites in fecal and urine samples of SFTSV infected Ifnar^-/- mice

We established an untargeted GC-MS metabolomics method for fecal and urine samples of Ifnar^-/- mice infected with SFTSV. We substantially extracted 672 volatile metabolites including amino acids, phenolics, indoles, dicarboxylic acids and other metabolites of microbial origin from fecal samples (S1 Data). The same method was applied to urine samples and it enabled the detection of 449 volatile metabolites as listed (S2 Data). Two technical replicates were run for each sample and only the metabolites detected in both the cases were listed, demonstrating the reproducibility of the GC-MS method. Under the criteria that VIP value >1 and |Log₂FC| ≥ 1.0 [49], the presence of 161 volatile compounds differentially expressed in the fecal samples of the SFTSV-positive were detected (Fig 1A). Of these metabolites, 11 compounds including indoles, phenols and esters were significantly up-regulated, while another 14 compounds including bourbonene, benzaldehyde, cyclohexane and eucalyptol were down-regulated (Fig 1B). The pathways analysis performed also revealed that the differential metabolites were demonstrated close associations with their corresponding pathways although most of the metabolites were involved in more than one pathway. For example, the metabolites of tryptophan metabolism (serotonin, kynurenines, tryptamine and indolic compounds) were the most abundant. In addition, there were 55 differential metabolites in urine samples from SFTSV-positive Ifnar^-/- mice, of which 26 were upregulated and 29 were downregulated (S2 Data). The group of SFTSV-infected mice differentially expressed the inflammatory mediator-regulating pathway of the phenyl-TRP channel and phenylalanine metabolism. Among these, the significantly upregulated pathway involving phenylalanine acid, the downregulated pathways involved in eucalyptol (S2A and S2B Fig). Since indoles are recognized as quorum-sensing molecules signaling interspecies even interkingdom that meet the criteria for a perceptible host VOC cue [2–4], the present study selected two of these DVM candidates, indole or 3-methylindole, to test the potential olfactory response of H. longicornis and bridge the ecological gap between vector ticks and infected host.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Differential volatile metabolites in the fecal sample of SFTSV infected Ifnar^-/- mice.

Panel A: volcanic diagram; Panel B: identified DVMs and their fold changes.

https://doi.org/10.1371/journal.pntd.0013139.g001

3. Indolic compounds trigger the forage behavior of H. longicornis

3.1 Choice test of ticks for indolic compounds.

H. longicornis nymphs were allowed to choose between Y-tube olfactometer arms, with one arm containing SFTSV-positive Ifnar^-/- mice, healthy mice or DVM candidates (indole or 3-methylindole), and the other arm as an air control. The binomial tests showed that the ticks significantly chose the arms containing mice or candidate DVMs compared with those with either control air or solvent treatment. H. longicornis ticks chose healthy mice with less frequency than those to infected mice and the candidate DVMs (N = 100, P = 0.03). And the binomial proportion of the ticks’ choices to the two candidate DVMs appeared almost identical to the choices to the SFTSV-infected mice (N = 80, P > 0.5). No significant difference was observed between the proportion of H. longicornis choosing indole or 3-methylindole (N = 40, P > 0.5) (Fig 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Nymph H. longicorinis choices in Y-tubes olfactometers.

Indole (40 mg indole in a solution of 98% water and 2% ethanol); 3-methylindole (40 mg 3-methylindole in a solution of 98% water and 2% ethanol); solvent (2% ethanol, 98% water). Healthy mouse (healthy Ifnar^-/- mouse); SFTSV-mouse (SFTSV infected Ifnar^-/- mouse).

https://doi.org/10.1371/journal.pntd.0013139.g002

The predicted 3D structural models and the ligand binding properties of NPC2 and OBPL in H. longicornis

To investigate the mechanisms of the specific attraction behavior, the olfactory genes had been retrieved from the transcriptomics data of H. longicornis achieved previously. As results, we failed to obtain any odorant receptor (OR) genes and odorant binding proteins (OBP) genes, which were popularly in insects of many kinds. Fortunately, a total of 8 transcripts obtained in the transcriptomics data of H. longicornis, which were categorized into 3 groups, NPC2, OBPL and Microplusin-like (ML), based on their sequences’ identity, motif similarity (S4 Fig) and following 2D and 3D structures. With the on-line program SWISS-MODEL, the NPC2 protein structures were predicted based on the crystal structure of Bovine NPC2 (PDB ID code 1NEP) [50] due to their highest identity (95.0%) and the largest coverage of amino acid sequences (Fig 3A). The most reliable 3D model indicated that NPC2 of H. longicornis possesses an β-sandwich fold consisting of two orthogonally arranged β-sheets (ca. 30° rotation) in the helical region (S3 Table). The first β-sheet has three β-strands (β2, β4, and β6) and the second sheet has four β-strands (residues β5, β7, β8, and β9). The fourth and fifth β-strands are connected by a short half-turn α-helix (η1). There are three disulfide bonds connecting residues Cys64 in coil1 and Cys177 in β9 (S1); Cys79in α1 and Cys83 in β3 (S2); Cys130 in the η1 and Cys136 (S3). The innermost cavity in the helical region of NPC2 is formed by the following residues: Val74, Leu94, Glu95, Leu150, Val163, Trp165, and Phe178 and thought to control access of ligands to the polar-faced, water-lined internal cavity [50] (S5A and S5B Fig) although several small cavities present in the ‘bottom’ part of the hydrophobic core of NPC2 protein. Another cavity, formed by residues Phe107, Leu151, Val152, Leu153, Phe67, and Val180, is noticed connected to a pocket on the surface of the protein, formed by Ser116, Leu120, Phe122, Glu155, Phe156, and Pro101 residues, through a small opening topped by a ‘gate’ formed by Phe-156 and Phe-66. We speculate that, for larger ligand to bind, the gate would have to open up and the two β-sheets. Similarly, the 3D structure of H. longicornis OBPL protein were also predicted using the on-line program SWISS-MODEL based on the crystal structure of a classic OBP (PDB ID code 7NYJ) [51] from Varroa destructor as their paired identity and coverage of amino acids sequences (Fig 3B). The scores of global model quality estimation (GMQE) and QMEAN were 0.52 and -3.65, respectively. Therefore, the predicted model of OBPL seems reasonable and reliable (S3 Table). As shown in the secondary structure, H. longicornis OBPL is composited of signal peptide (1–23aa), 5 α helices, 2 β folds and 2 TTs. The five α-helices locate at residues Lys48-Lys64 (α1), Glu75-Ser92 (α2), Phe104-Asn113 (α3), Glu121-Lys 138(α4), Phe143-Cys162 (α5), and three disulfide bonds connect Cys47 in coil1 and Cys167 in α5 (S1); Cys67 in α1 and Cys94 in α2 (S2); Cys137 in the α4 and Cys158 in the α5 (S3). The hydrophobic residues Phe15, Leu58, Phe59, Ala62, Val64, Leu73, Leu76, Ala79, Leu80, Ala88, Leu89, Gly92, Leu 96, Phe123, Leu124 and Ile125 from helices 1, 3, 4, 5, and loops between helices 3 and 4, and 5 form the binding cavity (S5C and S5D Fig). Interestingly, the C-terminus is pulled to the core of the protein to form part of the binding pocket wall, which can function as a “lid” for the release of ligands. The overall fold of five helices knitted together by three disulfide bridges and containing a hydrophobic binding cavity has been observed in OBPL of H. longicornis.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. The putative secondary structure of NPC2 and OBPL derived from H. longicornis.

Panel A: NPC2; Panel B: OBPL.

https://doi.org/10.1371/journal.pntd.0013139.g003

We also utilized molecular docking (SwissDock) to analyze the interactions between two putative ligands, indole and 3-methylindole with NPC2 and OBPL respectively. The results showed that all two ligands were located in the binding cavity of OBPL but interacted with different amino acids from the α-helical and motifs of loop regions in H. longicornis OBPL. The van der Waals interactions from hydrophobic residues and the hydrogen bonds formed by oxygen-containing functional groups and hydrogen donor residues are important linkages between the two ligands and OBPL (Fig 4A and 4B). However, the binding cavity of the two ligands to NPC2 were distinct. Both indole and 3-methylindole bind NPC2 at one surface binding cavity, instead of the buried cavity in the innermost core of OBPL. The van der Waals interactions were mainly formed between carbon-carbon double bonds and the aromatic residue Phe107. Only the indole base is predicted to overlap the center of the NPC2-binding cavity of H. longicornis (Fig 4C and 4D). Thus, the peculiar 3D structure of OBPL in H. longicornis might confer Haller’s organ of H. longicornis with a compact pocket for odorant accommodation from indole or 3-methylindole (S5C and S5D Fig). Since the AC score in SwissDock consists of the CHARMM force field energy plus the FACTS solvation energy terms and provides an estimate of the binding free energy as a weighted sum of the polar and non-polar terms, the affinity of NPC2 (AC score -1.791514 ~ -2.308619) and OBPL (AC score -1.696810 ~ -2.351299) to the indoles were ranked in the same category. Due to its smaller size and weight, indole has a much greater affinity to the same target protein than 3-methylinole as shown in their SwissParam scores.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. The protein-ligand interactions of NPC2 and OBPL of H. longicornis to indole and 3-methylindole predicted with Swiss-docking.

Pane A: OBPL- Indole; Pane B: OBPL-3-methylindole; Panel C: NPC2-Indole; Panel D: NPC2-3-methylindole; (viewed and generated by PyMol 2.6.2). blue block: Protein; yellow block: Ligand; red pod: Metal Ion; dash line: Hydrophobic Interaction; blue line: Hydrogen Bond.

https://doi.org/10.1371/journal.pntd.0013139.g004

Transcript levels of the NPC2 and OBPL genes in the foreleg of H. longicornis responding to indole and 3-methylindole stimulation

To test the responses of H. longicornis to the stimulations of indole or 3-methylindole, newly molted nymphs of H. longicornis was allowed to exposed to indole or 3-methylindole at different concentrations and then the relative transcript level of the NPC2 and OBPL genes were measured. Results indicated that the relative transcript levels of NPC2 and OBPL genes changed with different patterns. The stimulations of both indole and 3-methylindole resulted in 1.2-1.4 folds change in the transcript level of OBPL gene under the stimulations of lower concentration (1µg/mL and 10µg/mL) (Fig 5). There was no significant difference of the relative transcript level of OBPL observed under the same stimulations. In comparison of OBPL, the NPC2 gene reduced its transcription level to 0.6-0.8 folds change under the stimulations of higher concentration (1µg/mL and 10µg/mL) of indole or 3-methylindole (Fig 5). The up-regulated transcription of OBPL and down-regulated transcription of NPC2 indicates the corresponding associations of indolergic odorants with OBPL in H. longicornis.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. The relative transcript level of NPC2 and OBPL in the forelegs of H. longicornis under indoles stimulations.

https://doi.org/10.1371/journal.pntd.0013139.g005

Specific binding of OBPL with indole or 3-methylindole in vitro

The recombinant expression vector plasmid pET-28b-OBPL was successfully constructed (S6 Fig) and then transfected into host E. coli BL21 (DE3) and optimized host E. coli T7E. When induced by 1 mM IPTG under 37°C, 4h, the OBPL protein was expressed as an inclusion body with a molecular weight ca. 20.28kDa (Fig 6). The similar procedure was also performed for NPC2 protein and achieved its molecular weight ca. 15.34kDa. The purified OBPL or NPC2 (S7 Fig) proteins were submitted to the fluorescence competition binding assays of indole and 3-methylindole to validate the binding of target protein to indolergic odorants. The ligand binding characteristics of recombinant target proteins were demonstrated with competitive fluorescence binding assays with 1-NPN as competitor. The dissociation constant of the OBPL/1-NPN complex was 2.56 μM (Fig 7A) and that of the NPC2/1-NPN complex was 3.18 μM (Fig 7B). Both recombinant OBPL and NPC2 were able to bind the 2 volatile indolergic odorants, the former achieving higher affinities with Ki = 2.256 (2.043-3.135) μM for indole and Ki = 4.191 (3.357-5.277) μM for 3-methylindole (Fig 7C), compared to the moderate affinity obtained by NPC2 with Ki = 10.32 (9.540 to 13.49) μM for indole and Ki = 15.36 (11.153 to 25.85) μM for 3-methylindole (Fig 7D). The Haller’s organ specific expression of the OBPL protein, and its high affinity binding to biologically active volatiles supports a possible functional role of the chemosensory proteins OBPL in the perception of general host derived odorants in H. longicornis. Thus, the Haller’s organ-biased OBPL may mediate host recognition in the H. longicornis and represent new interesting targets for population control in the prevention and control strategies for tick borne diseases.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 6. The expression and purification of OBPL protein derived from H. longicornis in vitro.

Panel A: in vitro expression of OBPL in the recombined plasmid vector; NPE: supernatant DPE: inclusion protein; MW: Marker; Ø: host bacteria (without induction); strain No.1: BL21(DE3) strain; strain No.2: T7E strain.

https://doi.org/10.1371/journal.pntd.0013139.g006

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 7. The competitive fluorescence binding assays on OBPL/NPC2 to indoles.

Panel A: The dissolution constant of OBPL to 1-NPN; Panel B: the dissolution constant of NPC2 to 1-NPN; Panel C: The binding activity of OBPL to indole and 3-methylindole; Panel D: The binding activity of NPC2 to indole and 3-methylindole.

https://doi.org/10.1371/journal.pntd.0013139.g007

Upregulated indolic metabolites ameliorate pathogen-induced host damage and promote the foraging behavior of ticks for host

Research into the importance of indoles has increased dramatically, as the crosstalk between hosts and their microbiota is driven by tryptophan metabolites [52]. Various indoles have been shown to reduce bacterial biofilm formation and damage the cell surface by increasing endogenous oxidative stress, thereby inhibiting the growth of various intestinal bacteria including Salmonella, Eberthella, Shigella, Escherichia and others [53]. No wonder Bunnell et al. (2011) reported the significant increases in indole levels in sick hedgehogs suffering from any of 12 different ailments including lungworm (Crenosoma striatum), tapeworm, ringworm, breathing difficulties, hypothermia, and injury [12]. Furthermore, indolic metabolites have also been demonstrated to exert a remarkable effect in combating pathogenic viruses (Dengue, Zika and Chikungunya viruses) through drastic cell immobility and disintegration [54,55]. Viral infection triggers the stimulation of Toll-like receptors (TLRs), which induce the production of kynurenine by degrading tryptophan and activating indoleamine 2,3-dioxygenase (IDO1). IDO1 is a multifunctional enzyme and an endogenous immune checkpoint, which impairs effector T cell function, increases regulatory T cell (Treg) population and induces immune tolerance [56]. Alterations in tryptophan metabolism have been shown associated with levels of various anti-inflammatory cytokines, interleukin 6 (IL-6), tumor necrosis factor-alpha (TNFα) and other interferons (IFNs), which is produced in a rapid and transient manner in response to infection or tissue injury, and contributes to host defense by stimulating acute phase responses, haematopoiesis and immune responses [57]. Accordingly, indolic metabolism benefit the health of the host by antagonizing the proliferation of pathogens, ameliorating pathological damage and modulating host immune system. Through the complex interactions of microbiome-indoles-immune system, indoles achieve their microbiostatic functions and indicate pathological states of hosts by fostering premorbid risk factors. Therefore, the production and release of indoles in the host is an essential outcome of pathogen-host defense interactions, which could serve as a reasonable indicator of the health status of the host [58]. In the present study, the up-regulation of the tryptophan pathway and the elevated indolic metabolites in the SFTSV-infected mice reconfirmed that the production of indoles can be induced by viral infections and that the increased indoles potentially benefit the host with a remarkable amelioration of virological damage. More importantly, our study provides strong evidence that ectoparasitic ticks acutely sense the metabolic changes resulting from the virus-host interaction and then simultaneously trigger or promote the tick’s foraging behavior for the particular host suffering from viral infection. Among these, the perception of elevated indolic metabolites may be a crucial process to charge the tick’s foraging behavior, although the detailed signaling and response processes remain to be elucidated.

Smell of indoles guide tick foraging behavior for infected host

Physiologically, ticks and other blood-feeding arthropods have overlapping receptive fields for many aromatics and heterocyclics [59]. Among these, indole and 3-methylindole, the derivatives from bacterial degradation of tryptophan, were also known to mediate long-range attraction in several culicine and anopheline species [60,61]. The former aids the location of blood host by host recognition, while the latter acts as an oviposition attractant. Indeed, indoles elicit strong physiological responses in adult antennal trichoid sensilla of An. gambiae [62,63], and olfactory receptor neurons (ORNs) activation in Ae. aegypti [64], C. quinquefasciatus [60], and C. tarsalis [65]. In addition, 4-methyl-phenol (4-MP or p-cresol), one aromatic compound found in human sweat and hay infusion [65], plays an important role as an attractant for tick’s blood feeding and Aedes, Culex, and Anopheles mosquitoes’ oviposition [60,61,64,65]. However, the function of the indoles to tick’s host foraging, toxin and predator avoidance, egg deposition, and mate selection has yet to be determined. In the present studies, the increased level of indolergic metabolites in the infected host resulted in high levels of H. longicornis infestation, suggesting the possible attractiveness of indoles to host-seeking ticks, which may partially explain the interesting phenomenon that higher tick burdens occur on the sick hedgehogs reported by Bunnell et al. (2011). Although in their later experiment, the addition of indole to the feces of healthy hedgehogs did not attract the ticks tested as expected [12], the announced proposal that ticks may choose their host based on indole linked to the host’s health status, as the attractiveness of these indolergic metabolites to ticks may be masked or inhibited by some unidentified substances and their potential synergistic effects. Most interestingly, indolergic metabolites were also found to attract or repellent to various vector insects with several odorant receptors (ORs) and odorant binding proteins (OBPs) involved, including CquiOBP1 (Culex quinquefasciatus), CquiOR2, CquiOR10, AalbOR10 (Aedes albopictus), AgamOR10 (An. gambiae), and AsinOR10 (An. sinensis) [66]. These include olfactory indolergic receptors, commonly found in mosquitoes and other highly chemo-sensitive insects, which allow this large and diverse taxonomic group to exploit a dazzling array of ecological niches. However, the known olfactory receptors are only present in insects, but not in Arachnida. Thus, the identification of OBP analogous and their appropriated ligands are considered crucial for the chemosensory behavior of the Arachnida species, since they can bind chemical cues and transport them to the appropriate receptors via the sensillum lymph [65]. To the best of our knowledge, efforts afford to identify olfactory associated proteins failed to yield any OBP in Arachnida species instead of NPC2 protein, Microplusin-like (ML) protein and OBPL protein, despite that large numbers of OBPs have been identified in hemipteran insect species [66]. Nowadays, as small soluble proteins belonging to the myeloid differentiation factor 2 (MD-2) related lipid-recognition protein family, NPC2 proteins are functionally similar to odorant binding proteins (OBPs) and involved in chemical communication in arthropods [67,68]. Moreover, the typical β-barrel or sandwich structure of the NPC2 protein, consisting of eight β-helices and a short α-helical segment, can bind the signaling chemicals extensively [69], suggesting a potential function as an olfactory recognition protein in the tick species. Till date, NPC2 has been found in Ixodes ricinus [70], Amblyomma americanum [71], I. scapularis [72], H. longicornis [40], Varroa destructor [73], Pardosa pseudoannulata [74], Camponotus japonicus [69], Helicoverpa armigera [67] and others arthropods [75]. However, the exact odor binding activities and their corresponding upstream and downstream signal transduction molecules of the NPC2 proteins in ticks are still unclear. Similar to NPC2, ML proteins also have been annotated from a foreleg transcriptome analysis [72] of I. scapularis and a proteome analysis of A. americanum [71], as well as large-scale, tick genomic studies [76]. Although they were exclusive and biased expressed in Haller’s organ, a single, comparable ligand-binding site might be insufficient to elucidate the olfactory chemo-sensation role of ML in ticks, although relevant functional experiments have yet to be performed. Till to now, only two foreleg-biased OBPL proteins, OBPL-1 (19 aa; 118 aa) and OBPL-2 (24 aa; 150 aa), that resemble insect odorant-binding proteins were discovered in a chemosensory proteome of A. americanum [71]. Concerning OBPL, a classic OBPL from H. longicornis was fortunately identified and its potential binding properties with indolic ligands were analyzed in our study. Our current functional analysis based on amino acid sequences and three-dimensional folding suggests that the classic OBPL comply with the binding protein criteria for odorant ligands proposed by Pelosi et al. [68]. (1) The classic OBPL presents a signal peptide revealing their secretory nature. (2) the classic OBPLs are made of five α-helical domains connected by short unstructured loops and knitted together in a compact and stable structure by three disulfide bridges [77]. (3) A comparable, central hydrophobic binding cavity predicted enable the classic OBPL with binding activity with appropriated odorant ligands. Together with the relative high expression level of the OBPL in forelegs of H. longicornis, we come to conclude that the classic OBPLs might function as one odorant binding protein to recognize and transport volatile odorant compounds. Indeed, our competitive fluorescent binding assays also confirmed a specific binding preference of the classic OBPLs for indoles, indole and 3-methylindole, with high affinities, which play a notable role in H. longicornis olfactory system.

Control strategies inspired by tick olfactory ecology

Ticks display robust olfactory-driven behaviors as its olfactory receptor neurons (ORNs) that innervate the sensilla in the Haller’s organ are able to detect various behavior-modifying semiochemicals extensively [78]. Most of these semiochemicals are distinct but limited range of volatiles from host animals, which can elicit specific behaviors such as host foraging or natural enemy avoidance [79]. For example, meth-cresol shows the strongest electrophysiological response and clearly attracts ticks to feed, while ortho-methylphenol enables ticks to efficiently find a mating partner [79]. Using the delicate and intricate olfactory mechanism, ticks bridge the gap between different pathogens and their hosts, and efficiently spread diverse tick-borne pathogens to humans. Till date, much researches have focused on the determine of tick’s odorant receptors (ORs), OBPs, and their appropriate volatiles, which will lead to the development of odor baits to be used in tick management programs [79]. However, little work has been done to exploit olfactory links between vector ticks and infected hosts under conditions of pathological damage. Since characteristic host odors are usually produced and released during infection with specific microbiota or set of pathogens, a full understanding of the characteristics of tick-borne disease transmission pathways or cycles would help develop scientific control strategies based on characteristic pathogen-driven host odors. Inspired by the close links between vector arthropods and hosts infected with various pathogens as described above, we used olfactory ecology theory to test the semiochemical links between vector ticks and SFTSV-infected hosts. Our studies, for the first time in the sight of olfactory ecology, have successfully revealed the olfactory link between SFTSV infected host and host seeking H. longicornis. Although our results are limited and infantile, the efficient olfactory links between H. longicornis and SFTSV infected mice may not be limited within indolic volatiles. Other differential metabolites should not be excluded bridging the gaps between the vector tick and SFTSV infected hosts. For which, more efforts in detail should be determined in the future. Nevertheless, our preliminary results would pave a way to device a newly control strategy for tick borne disease based on olfactory ecological theory.