Varroa destructor is an effective vector of Israeli acute paralysis virus in the honeybee, Apis mellifera

The Israeli acute paralysis virus (IAPV) is a significant marker of honeybee colony collapse disorder (CCD). In the present work, we provide the first evidence that Varroa destructor is IAPV replication-competent and capable of vectoring IAPV in honeybees. The honeybees became infected with IAPV after exposure to Varroa mites that carried the virus. The copy number of IAPV in bees was positively correlated with the density of Varroa mites and time period of exposure to Varroa mites. Further, we showed that the mite–virus association could possibly reduce host immunity and therefore promote elevated levels of virus replication. This study defines an active role of Varroa mites in IAPV transmission and sheds light on the epidemiology of IAPV infection in honeybees.

Varroatosis is one of the most serious honeybee diseases caused by a parasitic mite, Varroa destructor. Varroa mites use their piercing mouthparts to suck out haemolymph from honeybees. The repeated feeding results in a decline of colony vigour, shortened life span of the honeybees and eventual perishing of colonies (Rosenkranz et al., 2010). Further, the feeding of Varroa mites gives mites the potential to act as vectors of bee diseases, presenting additional threats to bee health. The role of Varroa mite as a vector in the transmission of bee viruses has been well documented (Bowen-Walker et al., 1999;Chen et al., 2004;Shen et al., 2005). However, it has not yet been demonstrated if this mite could act as a vector of Israeli acute paralysis virus (IAPV), a virus that was tightly correlated with honeybee colony collapse disorder (CCD), a malady that has decimated honeybee colonies across the USA (Cox-Foster et al., 2007;vanEngelsdorp et al., 2007) and around the world (Neumann & Carreck, 2010). The present study was undertaken to determine the possible role of Varroa mites in the transmission of IAPV and in the promotion of its replication in honeybees. Moreover, the expression of immune-related gene transcripts, apidaecin and eater that are involved in honeybee humoral and cellular immunity (Ertürk-Hasdemir & Silverman, 2005;Evans et al., 2006;Kocks et al., 2005;Simone et al., 2009) was measured in Varroa-challenged bees. Apidaecin is a proline-rich peptide and one of the most prominent components of the honeybee humoral defence against microbial invasion (Evans et al. 2006). Eater is a member of epidermal growth factor-family protein involved cellular immunity, as it plays an important role in phagocytosis (Ertürk-Hasdemir & Silverman, 2005;Kocks et al., 2005).
Two strong bee colonies (i.e. with at least seven frames filled with capped brood and food, and covered with adult bees) and two weak colonies (i.e. with a low-adult bee population and fewer than four frames with a small patch of capped brood) that were determined to be free of IAPV infection by RT-PCR assay were selected for transmission studies. Further, the expression levels of vitellogenin, an indicator of the general health of the colonies (Amdam et al., 2005;Simone et al., 2009), were determined for the two selected groups. The relative transcript levels of vitellogenin in bees from strong colonies were at least threefold higher than in bees from weak colonies with a peak of 4.46±0.31 (t53.084, P50.0059), which confirmed the health conditions visually determined above. Varroa mites that were to be used in transmission assays were collected from a colony that was heavily infested with mites; both honeybees and mites were shown to have IAPV infections by RT-PCR assay. The ubiquitous presence of deformed wing virus (Lanzi et al., 2006) in bees and of Varroa destructor virus-1 (VDV-1) (Ongus et al., 2004) in Varroa mites was considered as a background infection.
The transmission studies were set up following the design illustrated in Fig. 1. Briefly, newly emerged bees from both strong and weak colonies were divided into four subgroups (n540) individually, based on the level of Varroa mites (0, 10, 20 and 30 %) they would receive in a bee rearing cup ). In addition, for both strong and weak colonies, a subgroup of ten newly emerged bees was collected prior to transmission assay and used as negative control (defined as day 0). The first half of bees, along with half of experimental Varroa mites, were collected from each cup at day 3 after mite infestation, and the second half of bees and Varroa mites, were collected at day 7. The quantification of IAPV was performed by TaqMan realtime quantitative PCR (qRT-PCR). A PCR product of 586 bp was generated from the following pair of IAPV primers (sense: 59-GCGGAGAATATAAGGCTCAG-39; antisense: 59-CTTGCAAGATAAGAAAGGGGG-39). A 0.2 mM TaqMan probe (59-FAM-CGCCTGCACTGTCGTCATTA-GTTA-TAMRA-39), and 500 ng total RNA were incorporated in the reaction mixture of Access RT-PCR system (Promega). The absolute quantification method was used for analysing the qRT-PCR data of IAPV. The copy numbers of the IAPV were determined by relating the C t value of each sample to an established standard curve. The standard curve was established by plotting the logarithm of initial quantities of 10-fold serial diluted IAPV-recombinant plasmid, ranging from 10 3 to 10 14 copies, against corresponding threshold value (C t ). A linear relationship was observed between C t values and input recombinant plasmid DNA ( Supplementary  Fig. S1, available in JGV Online). The PCR amplification efficiency was calculated as 157 % based on the slope of the standard curve (slope522.43), by using the formula: E510 (21/slope) 21 (Application Note: http://www.stratagene. com/lit_items/appnotes10:pdf), confirming the validity of the assay used in this study.
The detection of negative-strand RNA of IAPV was conducted to assess whether Varroa mites could allow replication of IAPV. Briefly, the cDNA complementary to the negative-strand RNA of IAPV was synthesized from RNA extracted from Varroa mites, with Tag-IAPV-sense primer (59-agcctgcgcacgtggGCGGAGAATATAAGGCTC-AG-39. The sequence of Tag is shown in lower case (Yue & Genersch, 2005). To prevent amplification of nonstrand-specific products, the synthesized cDNAs were column purified (Qiagen) to remove short fragments of oligonucleotides and residues of enzymic reagents (Boncristiani et al., 2009). The purified cDNAs were then PCR amplified with Tag and IAPV-antisense primer. The specificity of PCR products was confirmed by sequencing analysis.
The relative quantification of apidaecin and eater transcripts was performed by SYBR Green qRT-PCR. The interpretation of qRT-PCR data was conducted using the comparative C t method (DDC t method) as described previously (Chen et al., 2005). After the validation of approximately equal amplification efficiencies of immune transcripts and the reference gene ( Supplementary Fig. S2a, b, available in JGV Online), the expression of each immune transcript was determined by normalizing the C t value of the target to the reference gene b-actin (Chen et al., 2005) and relative to a calibrator, a sample representing the minimal level of the transcript. Negative controls from both healthy and weak groups (bees at day 0), showing the lowest level of gene expression and not showing significant difference between each other, were used as calibrator individually. The relative quantities of immune transcript expression in bees are expressed as an n-fold change relative to the calibrator following the equation 2 2DDCt .
Normality of data was checked by Shapiro-Wilk test, using SYSTAT software, and Chi-squared and Jarque-Bera tests, using PAST software (Palaeontological Statistic software package), while homoscedasticity was checked with Levene's test, using SYSTAT software. IAPV, apidaecin and eater data were translated as log 10 (original data+1) before analysis to obtain a normal distribution. A threeway factorial ANOVA was used to compare genome Fig. 1. Experimental design to assess the role of Varroa mites in the IAPV transmission. Bees were transferred into rearing cups and exposed to different levels of Varroa mites. Samples of bees and associated mites were collected at days 3 and 7 post-exposure. Bees collected prior to transmission assay (day 0) were used as negative control. equivalents of the virus and the relative levels of immune transcripts considering three factorial levels: the colony (strong versus weak), the Varroa mite inoculation (0, 10, 20 and 30 %) and the day post-treatment (3 versus 7). In order to measure the strength of the linear relationship, the coefficient of correlation, r 2 (Zar, 2009), was calculated between the viral copy number and the per cent of Varroa mite, as well as between the relative expression of immune genes and the per cent of Varroa-mite exposure. To test the effects of Varroa mite inoculation level on copies of IAPV among colonies (i.e. strong and weak) and among replicates within colony, two separate analyses of covariance (ANCOVA), at days 3 and 7, were carried out, with colony as main effect and replicates nested within colony.
While all bees from negative control as well as bees not exposed to Varroa mites tested negative for IAPV, bees that were exposed to Varroa mites became IAPV positive. The copy numbers of IAPV were significantly and positively correlated with the level of Varroa mites introduced in bees from both strong colonies (day 3: r 2 50.86, P,0.001; day 7: r 2 50.84, P,0.001) and weak colonies (day 3: r 2 50.98, P,0.001; day 7: r 2 50.95, P,0.001). As shown in Fig. 2(a,  b), the more mites introduced, the higher IAPV copy number was detected in infected bees, which also increased over time. The copy number of IAPV in bees at day 7 was significantly higher than that in bees at day 3 (F514.09; d.f. 1, 32; P,0.01). Bees originating from strong and weak colonies showed different levels of response to the challenge of Varroa mites. The IAPV concentrations in bees originating from strong colonies were significantly lower (Fig. 2a) than bees from weak colonies (Fig. 2b), at the different levels of Varroa-mite challenge (F559.17; d.f. 1, 32; P,0.001). While the copy number of IAPV in bees increased significantly as the load of Varroa mites increased (F5320.07; d.f. 3, 32; P,0.001), the magnitude of the IAPV titre change in bees from strong colonies was lower than in weak colonies. In fact, the effect of replicates was not significant, both at days 3 and 7, indicating that the relationship between copies of IAPV and levels of Varroa is the same for all the replicates, within each colony type. In contrast, the relationship between copies of IAPV and levels of Varroa were significantly different between strong and weak colonies, as indicated by significant differences of slopes (F518.1; d.f.51, 12; P,0.001 at day 7; F523.2; d.f.51, 12; P,0.001 at day 3) and intercepts (F514.7; d.f.51, 12; P,0.001 at day 7; F510.5; d.f.51, 12; P,0.001 at day 3).
While evaluation of Varroa mites prior to transmission assays showed that 70 % Varroa mites (14/20) collected from the field were positive for IAPV, 100 % of the Varroa mites (72/72) originating from the same colony and collected after the inoculating experiment tested positive for IAPV. Further, the strand-specific RT-PCR assay showed the presence of negative-strand IAPV RNA in all Varroa-mite samples examined (Fig. 2c). Like other positive-stranded RNA viruses, the replication of IAPV (Maori et al., 2007) proceeds via the production of negative-stranded intermediates. Therefore, the presence of negative-strand RNA of IAPV in Varroa mite appears to be evidence of virus replication and suggests that the

Varroa destructor is a vector of honeybee virus
Varroa mite could be a biological vector that supports the replication of the virus. Further studies to confirm this hypothesis will be necessary.
Our results also showed that Varroa mites not only served as a vector to transmit the virus but also impacted on the course of IAPV infection by possibly weakening the bees and reducing host immunity, and therefore leading to elevated levels of virus replication.  reported a suppression of the immune response by Varroa-mite feeding, which subsequently triggered viral replication in bees. The occurrence of immune suppression of apidaecin was observed in bees from weak colonies at the highest level of Varroa infestation. Moreover, the expression of immune transcripts was dependent upon host health backgrounds, and somehow affected by the cupinduced stress, which triggered gene transcript expression even in the absence of Varroa. While the correlation between the expression level of apidaecin and the load of Varroa mites in bees from strong colonies was not statistically significant (F52.53; d.f. 3, 16; P50.09) (Fig. 3a), the increases in level of Varroa exposure were accompanied by a decrease in relative levels of apidaecin in bees originating from weak colonies (F55.22; d.f. 3, 16; P50.011) (Fig. 3b). The expression of eater was evoked in bees by Varroa-IAPV association. While the correlation between the levels of eater and the levels of Varroa-mite exposure was not statistically significant (F52.24; d.f. 3, 32; P50.10) (Fig. 3c, d), bees originating from weak colonies had a significantly lower relative transcript abundance than those from strong colonies (F5121.18; d.f. 1, 32; P,0.001) (Fig. 3c, d). This difference may partially explain the results reported in the literature (Gregory et al., 2005;Navajas et al., 2008;, which could have been due to overlooked health conditions of the experimental populations. Host condition is an important factor that influences the bee response to any environmental challenge and the complex cross-talk between stress and immunity. An in-depth understanding at the molecular level of the stress impact on immunomodulation of latent pathogens is of key importance for shedding light on how different environmental factors can influence honeybee population stability. Collectively, our laboratory data show that: (i) Varroa mites are capable of transmitting IAPV in honeybees; (ii) there is a significant positive correlation between the rate of IAPV transmission and the density of Varroa mite; (iii) mite-to-mite indirect transmission is possible by cofeeding on the same bees of the virus-positive mites and virus-negative mites; (iv) Varroa mites are IAPV replication-competent; and (v) host-health conditions could alter outcomes of virus transmission and host-vector-virus interactions in a significant way. Based on this, we can conclude that Varroa mites provide a plausible route of IAPV transmission in the field, and may significantly contribute to the honeybee disease problems often associated with colony collapse. Fig. 3. Relative expression of apidaecin transcript in response to Varroa-IAPV association in bees originating from strong colonies (a) and from weak colonies (b). Relative expression of eater transcript in response to Varroa-IAPV association in bees originating from strong colonies (c) and from weak colonies (d).