Location:2013 Annual Report
1a. Objectives (from AD-416):
Use genomic selection to enhance genetic resistance to Marek's disease.
1b. Approach (from AD-416):
With high-density chicken rearing, control of infectious diseases is critical for economic viability and maintaining public confidence in poultry products. Among poultry diseases, Marek’s disease (MD), a lymphoproliferative disease caused by the highly oncogenic a-herpesvirus Marek's disease virus (MDV), continues to be a major concern. The fear of MD is further enhanced by unpredictable vaccine breaks that result in devastating losses. The field of genomics offers one of the more exciting avenues for enhancing control of MD. By identifying genes that confer genetic resistance, it should become possible to select for birds with superior disease resistance. Genetic resistance to MD is a complex trait controlled by many genes. Most genome-wide efforts for complex traits rely on linkage between the causative gene and a genetic marker, which results in limited detection power and resolution of gene location. An alternate method is to screen for allele-specific expression (ASE), a simple yet powerful approach, where the expression of each gene allele is compared within an RNA sample. When the expression of the alleles is not equal, then one can unequivocally declare ASE and the presence of a polymorphic cis-acting (genetic) element for that gene as linkage disequilibrium (LD) is confined to the transcriptional unit. The only requirements for ASE to work are: (1) the assumption that variation in expression between alleles of a gene are responsible for part of the phenotypic variation, and (2) the existence of a cSNP to monitor the alleles. Response to MDV infection in chickens was evaluated in F1 progeny from ADOL inbred lines 6 and 7 using next generation sequencing on a limited number of samples to query for ASE followed by Illumina GoldenGate assays to validate and expand the number of samples. In brief, our ability to identify and validate almost all of the 5,360 cSNPs in 3,773 genes clearly demonstrates that ASE is an efficient method to identify potentially most or all of the genes responding to viral infection for this complex trait. In this proposal, we now address whether the expression differences are linked to measurable phenotypic changes (e.g., disease incidence) and, if so, can the cSNPs be used in a genomic selection program? Specifically, we address the following objectives: 1. Using lines 6 and 7 advanced intercross lines, determine the relatively influence of each cSNP on MD incidence and which allele confers MD resistance or susceptibility. 2. Validate predictions by selecting birds that are predicted to be most MD resistant or susceptible, intermate within predicted susceptibility class, and conduct MDV challenge studies from their progeny. 3. Repeat the same process using commercial layer lines, which will translate the results into the public domain.
3. Progress Report:
This project is directly linked to project 3635-31320-009-10R titled "Enhancing Genetic Resistance to Marek’s Disease in Chicken via Allele-Specific Expression Screens and Genome-Wide Selection." Marek's disease (MD) is one of the most serious chronic disease threats to the U.S. poultry industry. Selecting for increased genetic resistance to MD is a control strategy that can augment current control measures. Allele-specific expression (ASE) is a simple yet powerful approach to identify high confidence candidate genes for complex traits where the expression of each gene allele is compared within an RNA sample. When allelic imbalance is observed, then one can unequivocally declare ASE and the presence of a polymorphic cis-acting (genetic) element for the gene. We’ve incorporated a new genome-wide ASE screen for non-MHC genes that respond to Marek’s disease virus (MDV) infection. Lines 6 and 7 were intermated to produce F1 progeny. Half of the progeny were challenged with MDV at 2 weeks of age. At 1, 4, 7, 11, 13, and 15 dpi, 12 birds for each treatment group were euthanized, and RNA from the spleen was isolated. To get a genome-wide and unbiased survey of all the expressed genes and an indication of ASE, 7 individual RNAs from both uninfected and MDV-infected birds were sequenced using an Illumina platform. This generated up to 95 million 100 base reads, or 9.5 Gb of sequence per sample. To analyze the data, SNPs were first identified, and then the ratio of the two alleles in uninfected and MDV-infected animals was determined. If the expression ratio changes in response to viral infection, it indicates that there was a cis-acting regulatory element affecting the expression of the gene or genetic element in the gene containing the SNP. Our study identified 4,528 SNPs in 3,718 genes exhibiting ASE in response to infection. Using the SNP exhibiting ASE in response to MDV infection, we are addressing whether the expression differences account for MD incidence and, if so, use the SNPs in a genomic selection program. Specifically, we developed approximately 1,100 6x7 F6 MD resource population. DNAs from these birds were genotyped with a custom Affymetrix 15K SNP array that includes 2,256 ASE SNPs, 4,497 SNPs associated with selective sweeps for MD genetic resistance, and 8,097 previously validated SNPs to fill gaps. A genome-wide association analysis has been performed and revealed that the most significant associations are with ASE SNPs. Furthermore, on a per SNP basis, SNPs associated with ASE explain twice the genetic variance compared to the other SNP classes. In addition, all the SNPs account for 0.44 of the total variance or 80+% of the genetic variance, therefore, we have little to no missing heritability. Thus, we strongly believe that variation in transcriptional regulation is the dominant driving force in the genetic architecture in variation of MD incidence. To validate these results, genomic estimated breeding values (GEBVs) have been determined for 200+ 6x7 F7 sires. We are currently planning to mate the top 30 and bottom 30 males to 6 random 6x7 F7 females for a progeny test; parallel studies are being performed on commercial layers. Our results strongly suggest that the genetic architecture for MD resistance is controlled by polymorphisms that influence transcriptional variation. If so, then ASE screens should prove to be a powerful method to identify SNPs that account for genetic variation of other complex traits, especially those that involve response to pathogens. Furthermore, this allows the selection of SNPs that can be incorporated in genomic selection programs like commercial layers, and are unlikely to need “retraining” as the SNPs will be in very high LD to the causative mutation.