Varroa destructor mites cause honey bee colony mortality, vector other bee pathogens, and impose economic costs on the apiculture industry and other agricultural enterprises that benefit from honey bee pollination. This search for molecular markers that associate with Varroa resistance will enable more efficient and effective selection, breeding and propagation of honey bees resistant to Varroa destructor. Effective selection, breeding and propagation of honey bees resistant to Varroa mites will reduce apicultural dependence upon chemical mite controls that may adversely affect honey bee development, reproduction and survival. OBJECTIVES: The principal aim of the proposed Phase 1 Honey bee SNP Association research project is to perform a whole-genome association analysis of 1536 Single Nucleotide Polymorphisms (SNPs) to find molecular markers that are associated with relevant phenotypic variation in Varroa destructor mite resistance. This Phase I research will draw extensively from the technical expertise and groundbreaking insights we have achieved in our prior honey bee SNP genotyping study of representative African, European and New World hybrid populations published in Science. Whitfield, C., S. K. Behura, S. H. Berlocher, A. G. Clark, J. S. Johnston, W. S. Sheppard, D. R. Smith, A. V. Suarez, D. B. Weaver, N. D. Tsutsui. 2006. Thrice out of Africa: Ancient and modern expansions of the honey bee, Apis mellifera. Science 314: 642-645; and Pinto, M. A., W. L. Rubink, J. C. Patton, R. N. Coulson, J. S. Johnston. 2005. Africanization in the United States: Replacement of Feral European Honey Bees (Apis mellifera L.) by an African Hybrid Swarm. Genetics 170: 1653-1665. We first will collect and genotype feral bees from along a transect in Texas, designed to represent various local feral populations which now consist of survivors that persisted after the Varroa-precipated population collapse. Sampling from various Ecoregions across Texas should reduce the risk of spurious SNP associations attributable to local adaptation or subpopulation structure masking the signals attributable to Varroa resistance. Additional samples, including drones whenever possible, will be collected from several lines of Varroa resistant stocks developed by the USDA-ARS Baton Rouge Bee Lab - ARS Russian, and ARS SMR/VSH; two Varroa resistant populations selected and developed by Daniel Weaver and Bee Weaver Apiaries, Inc., known as BeeWeaver AllStars and BeeWeaver Buckfast. Varroa susceptible colonies will be identified from field trials using a variety of other commercially available lines that demonstrate low levels of Varroa tolerance and samples collected. In addition we will include a number of previously identified and genotyped Varroa susceptible bees which originated from feral colonies which survived less than 2 years after the arrival of Varroa in the Texas feral population. These Varroa susceptible samples will be the controls for purposes of defining a null set of SNP allele frequencies and haplotypes that do not associate with Varroa resistance. The project director, Daniel Weaver, through affiliated companies, owns or manages more than 10,000 colonies of honey bees throughout the U.S., sell queen honey bees throughout the US and the World, and imports bees and queens into the US from Australia. The project director thus has ready access to a wide variety of honey bee stocks and germplasm, and all necessary equipment and facilities for apicultural research and bee sample collection, DNA extraction, and genotyping analysis. Most notably, Bee Weaver Apiaries, Inc., has been selecting and propagating Varroa destructor and Acarapis woodii resistant honey bees for many years. APPROACH: The advantages of an association study over more traditional linkage mapping and/or candidate gene approaches are several. First, in genome wide association studies no assumptions need be made about the location of causal genetic variants, and this is appropriate for a study searching for markers of mite resistance, because there are no obvious candidate genes to inform the more traditional candidate-gene or linkage analysis study. Since an association approach is a comprehensive method that can be attempted in the absence of any information about the genomic location or function of the causal variation, association approaches are suited to the technical objectives in this instance. Second, linkage mapping is theoretically most effective in identifying genomic regions that have a high probability of containing a single, relatively rare, highly penetrant gene giving rise to the Mendelian trait effecting the phenotype in question. But there is no a priori reason to suspect a simple monogenic Mendelian trait is responsible for mite resistance. In fact, there are reasons to suspect that the complex behavioral phenotypes that are likely to affect mite resistance may arise from the epistatic interactions of multiple genomic actors making linkage approaches even more problematic. And importantly, linkage analysis is less powerful than association analysis in identifying common genetic variants that have a modest effect on phenotype. In fact, in those few linkage studies to date that have successfully identified genes contributing to disease or phenotype causation, the relevant genes have failed to contribute to a substantial fraction of the overall heritability of the disease. For example, there may be positive correlation between a particular gene mutation identified in linkage studies and breast cancer, but unfortunately the identified genetic variant is responsible for only a small fraction of total breast cancer cases. By contrast, association studies are theoretically much more powerful at elucidating multiple genetic factors, where any individual variant contributes only a small fraction of the total relative genetic risk for a particular phenotype. Carlson, C.S., Eberle, M.A., Kruglyak, L., & Nickerson, D.A., Mapping Complex Disease Loci in Whole-Genome Association Studies. Nature 429: 448, 451 (May 24, 2004). Whole genome association studies are possible only when there are sufficiently large numbers of available markers suitably distributed across the entire genome. Because of the data generated in the Honey Bee Genome Project, we now have at our disposal a set of ~1.1 million SNPs generated by comparison of high quality genomic Africanized honey bee traces against the assembled European honey bee genome and the two underlying European haplotypes represented in the trace reads. Moreover, we intend to use the Illumina GoldenGate genotyping platform that allows parallel calling of up to 1536 SNPs per array, allowing us to choose a favorably unbiased distribution of SNPs approximately evenly spaced across the genome
