Fig 1.
Annotation of the predicted Varroa destructor secretome.
(A) Abundance of Gene Ontology terms Level 2 from the predicted secretome (orange) and the whole predicted protein set (gray) of V. destructor plotted using WEGO v. 2.0. Percentage and number of genes in Log10 scale are reported on left and right Y-axes, respectively. Of the total 1,704 proteins in the predicted secretome, only 1,030 matched in the Swiss-Prot database and led to the GO terms represented above (orange). WEGO analysis showed that “extracellular region”, “molecular function regulator”, “structural molecule activity”, “antioxidant activity” and “biological adhesion” are overrepresented GO terms in the predicted secretome, compared to the whole predicted proteome. (B) Venn diagram showing number and percentage of secretome sequences matching in three different databases. Swiss-Prot (blue), salivary proteins of Acarina (yellow) and venom proteins of Hymenoptera (green). Venom proteins of Hymenoptera species were downloaded from NCBI using keyword ‘venom AND Hymenoptera [organism]’), while “saliva-related” tick’s and mite’s proteins were obtained using keyword ‘saliva AND Acarina [organism]’. The predicted secretome was blasted using an E value cut-off of 10−5. Venn diagram of annotated proteins was plotted using the online tool Venny v. 2.1.
Fig 2.
Salivary gland expression of putative host regulation factors found in the predicted secretome of Varroa destructor.
(A) Relative expression data of 3 selected candidates are presented as mean fold changes of 3–4 independent biological replicates. Each replicate consisted of a pool of 5–10 mites and comprised two samples: salivary glands (SG), and rest of the whole body, deprived of salivary glands (Whole Body–SG). Values on Y-axis are reported in Log10 scale. Error bars represent standard deviation (SD). Statistically significant differences are denoted with an asterisk (P < 0.005). (B) In-situ hybridization of DIG-labeled RNA probe for the salivary chitinase (Vd-CHIsal). Salivary glands hybridized with the antisense probe showed positive blue signals, while neither background nor unspecific signals were observed when the sense probe was used (negative control). B: Brain; C: Chelicera; P: Pedipalps; Sg: Salivary glands. Scale bars: 100 μm.
Fig 3.
Maximum-likelihood phylogenetic tree of Vd-CHIsal protein.
Bootstrap support values ≥60% are indicated. The accession number of each sequence is followed by the taxon name. Vd-CHIsal is indicated by a red arrow. The branch highlighted in pink represents the monophyletic group which Vd-CHIsal forms with few chitinases of predator and parasitic mites. Other highlighted branches represent chitinases with a known role in parasitoid-host interactions (blue), chitinases of phytophagous mites (green) and chitinases with a known function in the molting process (yellow). A chitinase of Amblyomma americanum (GenBank: AIR95100) showed the longest branch in the unrooted tree and was used as the outgroup reference.
Fig 4.
Survival of Varroa destructor as affected by RNAi-mediated silencing of the gene encoding Vd-CHIsal.
(A) Relative expression of Vd-CHIsal after mite soaking in a dsRNA solution. qRT-PCR data are presented as mean fold changes of 7 independent biological replicates. Each replicate consisted in a pool of 2 mites. Each time point was separately analyzed and the 0.9% NaCl control sample was used as calibrator. Mean dCt values within each time point were compared by one-way ANOVA followed by Tukey’s post-hoc test. Mean values denoted with different letters are significantly different. Error bars represent standard deviation (SD). (B) Kaplan-Meier survival curves of mites soaked in a solution of dsRNA targeting Vd-CHIsal. Saline controls (0.9% NaCl), GFP dsRNA and Vd-CHIsal dsRNA soaked mites were individually maintained on the same host pupa throughout the whole duration of the assay. Subjects at risk were 18, 30 and 45 for 0.9% NaCl, GFP dsRNA and Vd-CHIsal dsRNA treatments, respectively. In a concurrent set of trials, host pupae were replaced every 24 h for both saline controls (0.9% NaCl) and Vd-CHIsal dsRNA soaked mites. Subjects at risk were 29 and 22 for 0.9% NaCl and Vd-CHIsal dsRNA, respectively. Statistical details are in the text.
Fig 5.
Differentially expressed genes in honey bee pupae artificially infested with mites delivering saliva with the full repertoire of proteins or lacking Vd-CHIsal.
(A) Differential expression of 13 honey bee genes, as affected by presence of Vd-CHIsal in the saliva (KD/WS). DESeq2 adjusted P was < 0.05 and FDR was set at 5%. Log transformed mean FPKM values are reported on Y axis. For each experimental condition 3 separate pupae were analyzed. Error bars represent SD. Summary data sheets of differential expression analysis are presented in S2 and S3 Tables. (B) Relative expression of immune genes in honey bee pupae as affected by Vd-CHIsal expression in Varroa destructor infesting mites. Each mean value is obtained on 7–10 pupae, individually analyzed. Results of qRT-PCR are presented as mean fold changes relative to non-infested pupae used as calibrator. Values on Y axis are reported in Log10 scale. Error bars represent standard deviation (SD). Mean values were compared by one-way ANOVA, followed by Tukey post-hoc test, and values statistically different are denoted with different letters (P<0.05). Details of statistical analyses are presented in S4 Table. NP: non-parasitized controls; WS: pupae infested with mites soaked in saline solution; KD: pupae infested with mites soaked in a solution of Vd-CHIsal dsRNA.