Identification and QTL Analysis of Flavonoids and Carotenoids in Tetraploid Roses Based on an Ultra-High-Density Genetic Map

Abstract

Roses are highly valuable within the flower industry. The metabolites of anthocyanins, flavonols, and carotenoids in rose petals are not only responsible for the various visible petal colors but also important bioactive compounds that are important for human health. In this study, we performed a QTL analysis on pigment contents to locate major loci that determine the flower color traits. An F1 population of tetraploid roses segregating for flower color was used to construct an ultra-high-density genetic linkage map using whole-genome resequencing technology to detect genome-wide SNPs. Previously developed SSR and SNP markers were also utilized to increase the marker density. Thus, a total of 9,259 markers were mapped onto seven linkage groups (LGs). The final length of the integrated map was 1285.11 cM, with an average distance of 0.14 cM between adjacent markers. The contents of anthocyanins, flavonols and carotenoids of the population were assayed to enable QTL analysis. Across the 33 components, 46 QTLs were detected, explaining 11.85–47.72% of the phenotypic variation. The mapped QTLs were physically clustered and primarily distributed on four linkage groups, namely LG2, LG4, LG6, and LG7. These results improve the basis for flower color marker-assisted breeding of tetraploid roses and guide the development of rose products.

Rose Rosette Disease Resistance Loci Detected in Two Interconnected Tetraploid Garden Rose Populations

“…The linkage maps developed with our two mapping populations presented in this study are on par with the other published high-density SNP-based linkage maps (Bourke et al, 2017b;Zurn et al, 2018Zurn et al, , 2020Cheng et al, 2021;Yu et al, 2021), bringing the total number of tetraploid rose linkage maps reported to seven. The total lengths of our maps were similar to the K5 cut rose mapping population (568.67 cM) (Bourke et al, 2017b), as we also used the default Haldane mapping function and the polymapR implementation of MDSmap (Preedy and Hackett, 2016).…”

“…Our maps were roughly 100 cM longer than the two garden rose mapping populations (421.92 and 405.42 cM) ( Zurn et al, 2018 , 2020 ) using the Kosambi mapping function. The linkage maps constructed with the Chinese tetraploid YS mapping population using whole-genome sequencing ( Cheng et al, 2021 ) (1285.11 cM) and specific locus amplified fragment sequencing (SLAF-seq) ( Yu et al, 2021 ) (1158.90 cM) were more than double the length of our maps. All the maps described above were constructed using R-package polymapR.…”

“…This software has been used extensively to study autohexaploid sweet potatoes for root-knot nematode resistance ( Oloka et al, 2021 ), carotenoid, and starch biosynthesis ( Gemenet et al, 2020 ) and yield component traits ( da Silva Pereira et al, 2020 ). Several autotetraploid rose maps have been constructed using polymapR ( Bourke et al, 2017b ; Zurn et al, 2018 , 2020 ; Cheng et al, 2021 ; Yu et al, 2021 ) and QTL for stem prickles and flavonoid and carotenoid levels using polyqtlR ( Bourke et al, 2021 ; Cheng et al, 2021 ). Interconnected population QTL analysis has been performed in autotetraploid potatoes using diaQTL analyzing tuber shape ( Amadeu et al, 2021 ).…”

“…All the maps described above were constructed using R-package polymapR. All the linkage maps were similar in size and quality with the exception of the YS maps ( Cheng et al, 2021 ; Yu et al, 2021 ). The difference in genotyping platforms may be one of the factors contributing to the discrepancies between the lengths of these maps as the maps that are similar in length were genotyped using the WagRhSNP 68K array while the YS mapping population was genotyped using whole-genome resequencing and SLAF-seq.…”

“…Thus, the higher genotyping error rate may have contributed to the data containing artificial double recombinations resulting in inflated maps ( Hackett and Broadfoot, 2003 ). Like the other five tetraploid rose mapping populations ( Bourke et al, 2017b ; Zurn et al, 2018 , 2020 ; Cheng et al, 2021 ; Yu et al, 2021 ), which have been maintained and are used to study other traits, our two mapping populations present a good genetic resource to help dissect other traits of interest via future QTL studies.…”