- The European hazelnut (Corylus avellana L.) is a diploid (2n = 2x = 22) tree crop important to the economy of Oregon’s Willamette Valley, where 99% of hazelnut production in the United States is located. Corylus avellana exhibits sporophytic self-incompatibility (SSI), controlled by a single S-locus with at least 33 unique alleles. The alleles exhibit codominance in the stigma and dominance or codominance in the pollen. SSI is present in many plant families and is understood best at the molecular level in Brassica. However, Brassica gene sequences have not proven useful for investigations in Corylus. With new genomic tools available for the study of hazelnut, including a new Pacific Biosciences (PacBio) reference genome for ‘Jefferson’, more progress can be made toward the goal of identifying the genetic determinants of SSI.
BAC end sequences associated with the minimal tiling path of the S-locus were used to identify 36 PacBio contigs associated with the hazelnut S-locus (located on linkage group 5[LG5]). Simple sequence repeat (SSR) markers were developed from a subset of dinucleotide repeats found in those contigs and were characterized and mapped. Markers mapping near the S-locus on LG5 were screened against a population of 192 seedlings selected for recombination between the flanking markers of the S-locus, G05-510 and AU02-1350, for which the S-alleles had been determined. This yielded a 500 kb region containing the S-locus which was used to develop more SSR markers and a set of high-resolution melting (HRM) markers. When the new markers were used in fine mapping, they fully exploited all recombination in the fine mapping population and reduced the size of the S-locus region to 193.5 kb, containing 18 genes. The sequence of the mapped region most likely contains the S1 allele and a locus adjacent to it was also identified, and is likely to contain the S3 allele. The reference genome represents these genomic regions side-by-side but evidence suggests they represent two alleles at a single locus.
To study expression of S-locus genes following self- and cross-pollination a set of 5 reference genes was evaluated for qPCR relative expression analysis. ‘Jefferson’ stigmas were self-pollinated (incompatible) and cross-pollinated with ‘Felix’ and OSU 1253.064 (compatible). The tissue was collected at 0, 1, 2, 3, 5, and 7 hours following pollination. RNA was extracted and amplified using primers designed for reference genes. All five showed promise as references for relative expression analysis in pollinated stigmas and three were used to normalize expression of the genes of interest. Primers were designed for the 18 S-locus genes and were used to amplify cDNA from the pollinated style time course samples. Five genes showed no expression, four showed constitutive expression, five decreased in expression, and four increased in expression. Five genes showed an interaction between time and pollen type and one gene showed a difference in expression between types of pollen applied but without the interaction effect.
Long-range PCR was used to generate amplicons 8-12 kb in length for targeted resequencing of the S-locus region. For primer pairs designed using the ‘Jefferson’ reference genome, 29 out of 43 gave bands on agarose gels when amplified from ‘Jefferson’ DNA. The 43 primer pairs were tested in the parents of ‘Jefferson’ (OSU 252.146 and OSU 414.062) and two unrelated genotypes, OSU 1477.047 DNA and OSU 1026.073. Six fragments amplified from OSU 252.146 DNA, three from OSU 414.062 DNA, eight from OSU 1477.047 DNA, and two from OSU 1026.073 DNA. The results, obtained from the first attempt at primer development, demonstrate that long-range PCR could be a viable option to produce amplicons for targeted resequencing in ‘Jefferson’ hazelnut.