4.2 Non-random association and interactions between DOC loci and chloroplastic RpoC1
Since its establishment, the Barley1K infrastructure has been an instrumental research tool to explore landscape genomics (Hubneret al. , 2009; Hübner et al. , 2013; Chang et al. , 2022) , plant biotic and abiotic interactions (Sade et al ., 2012; Dakhiya et al. 2017; Alegria Terrazas et al. 2020; Dakhiya & Green 2023), and more recently as reference for the whole barley Pan genome (Jayakodi et al. 2020). Here, we extend the use of the B1K by 1) developing a more dedicated genotyping platform for the wild barley to allow GWAS and 2) start exploring the chloroplast variation and its possible relationship with fitness traits. At first glance, the results obtained in this study could not directly be translated to crop improvement, mainly since, as we show here, adaptation strategies in the wild and agricultural set up are different (biomass vs reproductive output responses to heat, see above). However, the B1K resource bear genetic attributes that could assist in targeting specific genes and gene complexes mainly owing to the non-randomness and long years of recombination within and between nuclear and chloroplastic genomes and the way natural selection works on these intra and inter-loci combinations. For the nuclear genome, it is clear that the LD decay is enormously shorter than that one could find in interspecific populations (Morrell et al. 2005). DOC3.2 may be a good example for this “gene discovery” use of the B1K panel for zooming-in on a causal diversity that otherwise would be delimited to 7.7 Mbp in using the interspecific population (Prusty et al., 2021). Here, a closer look on at the allelic diversity indicates the QTL for amplitude under HT in long days, which was identified in the previous study, most probably corresponds to the signal identified at chr3H_67267835. This marker is flanked by the non-significant markers chr3H_66,838,086 and chr3H_68,838,244, therefore limiting the source of phenotypic variation to approximately 2 Mbp for this QTL.
For the chloroplast genome per se, and for the allelic combinations with the nuclear genome, this study shows clearly the rich diversity repertoire found in this organelle within the wild (Table S7 ) and its relevance to plant adaptation. Moreover, we could identify both LD and CNI between DOC and the rpoC1 alleles in the B1K panel, where in an extreme case of the DOC3.2 this locus showed segregation only among carriers of the RpoC1G1713T . Non-random association of alleles in the nucleus and cytoplasmic organelles, or cyto-nuclear LD, is both an important component of a number of evolutionary processes and a statistical indicator of others (Fieldset al., 2014). We also followed up on the GWAS to examine inheritance of these CNI intereaction, and, beyond validating the environment-specific effects of the DOCs , we could also get indication for preferred homogeneity of the CNI. For example, in the case of CNI for the DOC5.1 combination of plasmotype and nuclear alleles from same origin (B1K-05-07 or B1K-50-07, the two parental lines of the F2) was associated with higher amplitude. This correspond well with recent study which suggested that increased amplitude under heat is positively correlated with higher plant vitality (Dakhiya & Green, 2023). In case of cereal evolution, it was hypothesized and in silico tested that in ancient hybridization, such as the one between the A and D-genomes of modern wheat there has been biased maintenance of maternal A-genome ancestry in nuclear genes encoding cytonuclear enzyme complexes (CECs) (Li et al., 2019). It is tempting to hypothesize that we observe such scenario in the case of the DOC and rpoC1 loci in the B1K.