Genotyping-by-Sequencing on Pooled Samples and its Use in Measuring Segregation Bias during the Course of Androgenesis in Barley.

Estimation of allelic frequencies is often required in breeding but genotyping many individuals at many loci can be expensive. We have developed a genotyping-by-sequencing (GBS) approach for estimating allelic frequencies on pooled samples (Pool-GBS) and used it to examine segregation distortion in doubled haploid (DH) populations of barley ( L.). In the first phase, we genotyped each line individually and exploited these data to explore a strategy to call single nucleotide polymorphisms (SNPs) on pooled reads. We measured both the number of SNPs called and the variance of the estimated allelic frequencies at various depths of coverage on a subset of reads containing 5 to 25 million reads. We show that allelic frequencies could be cost-effectively and accurately estimated at a depth of 50 reads per SNP using 15 million reads. This Pool-GBS approach yielded 1984 SNPs whose allelic frequency estimates were highly reproducible (CV = 10.4%) and correlated ( = 0.9167) with the "true" frequency derived from analysis of individual lines. In a second phase, we used Pool-GBS to investigate segregation bias throughout androgenesis from microspores to a population of regenerated plants. No strong bias was detected among the microspores resulting from the meiotic divisions, whereas significant biases could be shown to arise during embryo formation and plant regeneration. In summary, this methodology provides an approach to estimate allelic frequencies more efficiently and on materials that are unsuitable for individual analysis. In addition, it allowed us to shed light on the process of androgenesis in barley.

Extent and overlap of segregation distortion regions in 12 barley crosses determined via a Pool-GBS approach.

KEY MESSAGE: Extent and overlap of segregation distortion regions in 12 barley crosses determined via a Pool-GBS approach. Segregation distortion is undesirable as it alters the frequency of alleles and can reduce the chances of obtaining a particular combination of alleles. In this work, we have used a pooled genotyping-by-sequencing (Pool-GBS) approach to estimate allelic frequencies and used it to examine segregation distortion in 12 segregating populations of barley derived from androgenesis. Thanks to the extensive genome-wide SNP coverage achieved (between 674 and 1744 markers), we determined that the proportion of distorted markers averaged 28.9 % while 25.3 % of the genetic map fell within segregation distortion regions (SDRs). These SDRs were characterized and identified based on the position of the marker showing the largest distortion and the span of each SDR. Summed across all 12 crosses, 36 different SDR peaks could be distinguished from a total of 50 SDRs and a majority of these SDRs (27 of 36) were observed in only one population. While most shared SDRs were common to only two crosses, two SDRs (SDR3.1 and SDR4.2) were exceptionally recurrent (seen in five and four crosses, respectively). Because of the broad span of most SDRs, an average of 30 % of crosses showed segregation distortion in any given chromosomal segment. In reciprocal crosses, although some SDRs were clearly shared, others were unique to a single direction. In summary, segregation distortion is highly variable in its extent and the number of loci underpinning these distortions seems to be quite large even in a narrow germplasm such as six-row spring barley.

Improving the efficiency of isolated microspore culture in six-row spring barley: II-exploring novel growth regulators to maximize embryogenesis and reduce albinism.

Two alternative cytokinins, thidiazuron and meta-topoline, were tested in isolated microspore culture on recalcitrant barley genotypes (six-row, spring), and green plant regeneration was improved substantially. Doubled-haploid (DH) plants are coveted in plant breeding and in genetic studies, since they are rapidly obtained and perfectly homozygous. In barley, DHs are produced mainly via androgenesis, and isolated microspore culture (IMC) constitutes the method offering the greatest potential efficiency. However, IMC can often be challenging in some genotypes because of low yield of microspores, low regeneration and high incidence of albinism. Six-row spring-type barleys, the predominant type grown in Eastern Canada, are considered recalcitrant in this regard. Our general objective was to optimize an IMC protocol for DH production in six-row spring barley. In particular, we explored the use of alternative hormones in the induction medium (thidiazuron and dicamba), and in the regeneration medium (meta-topoline). This optimization was performed on two typical six-row spring (ACCA and Léger), a two-row spring (Gobernadora) and a two-row winter (Igri) barley cultivar. When 6-benzyl-aminopurine (BAP) was replaced by a combination of thidiazuron and dicamba in the induction medium, a 5.1-fold increase (P < 0.01) in the production of green plants resulted. This increase was mainly achieved by a reduction of albinism. Moreover, a 2.9-fold increase (P < 0.01) in embryo differentiation into green plants was obtained using meta-topoline instead of BAP in the regeneration medium. Together, these innovations allowed us to achieve a substantial improvement in the efficiency of IMC in this recalcitrant type of barley. These results were later successfully validated using sets of F1s from a six-row spring barley breeding program.

Androgenic response of barley accessions and F1s with Fusarium head blight resistance.

Most Fusarium head blight (FHB) resistant barley (Hordeum vulgare L.) accessions perform relatively poorly from an agronomic point of view. Due to the polygenic inheritance of FHB resistance, introgression of this complex trait into well-adapted elite germplasm will likely require multiple cycles of hybridization and selection to combine resistance and agronomic performance. The use of anther culture to produce doubled haploids would seem well justified to reduce the time required to achieve this goal. Unfortunately, little is known concerning the androgenic response of the small number of genotypes with known partial FHB resistance. To make the best use of such FHB resistance donors in a barley improvement program, we first characterized the FHB resistance of eight reported FHB resistance sources (Chevron, Gobernadora, Seijo II, Shyri, Svanhals, Zhedar I, F104-250-9 and C97-21-38-3) in our own FHB nursery in Quebec City (QC, Canada). In parallel, we assessed the androgenic response of these same eight lines with that of three cultivars (ACCA, Léger and Cadette) of known androgenic response. Finally, the androgenic response of F(1) hybrids involving some of these genotypes used as parents was measured and compared to that of the parental genotypes. Very large and significant differences were observed in the number of green plants produced by the different accessions and F(1)s. Although anther culture seemed very promising for some accessions, for others, the androgenic response was so low that a conventional approach would seem more appropriate.