Honey bees (Apis mellifera) are an important agricultural commodity providing honey, other bee products, and pollination services [1, 2]. Domesticated honey bees in the United States and elsewhere have been in decline in recent years, despite an increasing need for honey bee pollination services . This fact is often blamed on increasing challenges from pests and pathogens, as well as episodes of severe decline such as the enigmatic 'colony collapse disorder' (CCD) .
Among the most detrimental of honey bee pests is the ectoparasitic mite Varroa destructor . V. destructor and its closely related congener, V. jacobsoni, are native to Asia where they parasitize the Eastern honey bee, A. cerana.
V. destructor was only identified as a morphologically and genetically distinct species from V. jacobsoni relatively recently . V. destructor began to appear in Asian colonies of A. mellifera during the last century and is now widely distributed, inadvertently aided by trade in bees and bee products.
Mite-infested bee colonies suffer directly from parasitism of pupae and adults, and indirectly from viral and microbial pathogens that the mites vector [7, 8]. Feeding by mites induces an immunosupression in bees that leads to increased titres of pre-existing infections , further compounding their impact. The economic toll of V. destructor on apiculture is estimated to be millions of U.S. dollars per year, and chemical control agents are worrisome both for their collateral effects on bee health and the potential for honey contamination .
Varroa -honey bee interactions are mediated to a large extent via chemical cues, and bees have numerous mechanisms to control Varroa populations (reviewed in [5, 11]). Varroa mites reproduce on honey bee pupae, using chemical signals produced by the developing honey bee larvae to target appropriately aged hosts. The mature female offspring of reproductive Varroa emerge with the adult honey bee, and subsequently move to nurse bees (which are engaged in brood care), thereby allowing them to remain in close proximity to the brood [12, 13]. Honey bees resist 'Varroatosis', the infestation of colonies by Varroa mites, via grooming of adult infested bees, removal of infested pupae (hygienic behavior), and physiological resistance mechanisms . Recent successes in breeding Varroa -resistant bees, including the selection of 'Russian' bees with longstanding exposure to mites [14, 15], indicate that a better understanding of how bees and mites interact with each other can lead to novel management strategies.
Comparative studies of the fragility of the A. mellifera - V. destructor interaction, which has apparently prevented most Asian lineages of V. destructor as well as other Varroa species from colonizing A. mellifera [6, 16–18], supports the hypothesis that mite olfaction or other requirements for mite reproduction may be suitable control targets. A molecular-genetic approach to develop such innovative controls would clearly benefit from further insights into Varroa genomics, which could be exploited in conjunction with tools already extant for honey bee. Prior to this study, genes for only two non-mitochondrial V. destructor proteins had been deposited in GenBank, a sodium channel gene (AAN37408.1) and a glycoprotein (ACU30143.1). Genome sequencing will greatly expand this gene catalog, and may also uncover unforeseen targets for novel and specific acaricides, such as divergence in metabolic pathways between mites and bees or the discovery of important microbial interactions.
High-throughput, shotgun sequencing of whole genomes allows the rapid identification of thousands of genic sequences, greatly facilitating molecular and population-genetic studies that would otherwise proceed in piecemeal and laborious fashion. Here we report an initial sequence survey of the V. destructor genome in conjunction with a flow-cytometric estimate of genome size. Our annotations and analysis should aid investigators seeking molecular approaches to mite control. They will also provide a guide for a planned full genome project for this species , one of several genomics initiatives that are unfolding the molecular interactions between honey bees and a constellation of potentially interacting pathogens [4, 7, 20, 21].
Of the eight genetically distinct lineages of V. destructor that parasitize A. cerana in Asia, two have been identified on A. mellifera [6, 18, 22, 23]. Anderson  designated these lineages the Japan (J) and Korea (K) 'haplotypes' in reference to mitochondrial DNA makers, but they are concordantly distinct at nuclear markers as well . Genetic differentiation within lineages is low , likely reflecting the population-genetic impact of life-history traits  such as full-sib mating and male haploidy , as well as potential population bottlenecks tied to host-shift events and subsequent range expansion [18, 23]. In this study, we have analyzed the K haplotype of V. destructor from A. mellifera, the predominant haplotype presently found in North America . We have identified over 13,000 contigs with sequences homologous to other species; many of these have recognized domains and/or functional annotations transferred from other arthropods. Interestingly, V. destructor appears to have experienced a higher rate of protein evolution than Ixodes scapularis since their divergence from the most recent common ancestor over 300 million years ago. Sequences attributable to a range of microbes were identified, including a large number of sequences from one or more novel actinomycete bacteria, the presence of which was confirmed by PCR in individual mites but not in adult honey bees. We also identified a novel virus related to the Baculoviridae that was abundant in the genomic survey. Finally, we found a low level of nucleotide polymorphism in the sequenced sample of ~1,000 mites, consistent with expectation . This bodes well for future efforts to sequence and assemble a reference genome for this species and to identify genetic variation that correlates with host-interaction traits among Varroa strains and species.