Introduction
Delphinium luteum was historically treated as a variety of D. nudicaule or a hybrid between D. nudicaule and D. decorum due to morphological similarities. However, it is now recognized as a distinct and endangered species based on its unique yellow flower colour, specific habitat and genetic distinctness (Koontz et al. Reference Koontz, Soltis and Brunsfeld2001, Reference Koontz, Soltis and Soltis2004). In Delphinium breeding, D. grandiflorum and D. elatum are primarily used, resulting in a rich variety of blue flowers but a limited range of other colours. Consequently, various interspecific crosses have been attempted to introduce new flower colours (Honda and Tsutsui Reference Honda and Tsutsui1997; Saito et al. Reference Saito, Toki, Suga and Honda1998; Honda et al. Reference Honda, Tsutsui and Hosokawa1999, Reference Honda, Watanabe and Tsutsui2003; Cheng et al. Reference Cheng, Guan and Li2020). Historically, the yellow-flowered D. zalil has been used for such hybridization (Legro Reference Legro1961). For instance, crosses between D. zalil and the red-flowered D. cardinale produced F1 hybrids with purple flowers, accumulating cyanidin-based anthocyanins, which represented a novel flower colour (Sakaguchi et al. Reference Sakaguchi, Isobe, Fujita, Ozeki and Miyahara2019). However, D. zalil and D. cardinale are genetically distant, leading to sterility in hybrid progenies, which renders this combination unsuitable for breeding. In contrast, D. nudicaule and D. luteum, the focus of this study, are genetically closely related (Koontz et al. Reference Koontz, Soltis and Brunsfeld2001), and making them a promising combination for interspecific hybridization. Although D. luteum is a rare endangered species, cultivated populations maintain unique genetic diversity that is valuable for conservation and breeding (Koontz et al. Reference Koontz, Soltis and Soltis2004). In this study, we performed interspecific hybridization between D. nudicaule and D. luteum to utilize this genetic resource for creating novel flower colours and to analyze the inheritance of flower pigments.
Experimental
Plant materials and breeding of D. nudicaule and D. luteum
Seedlings of D. nudicaule and D. luteum were obtained from Miyoshi & Co. and cultivated in a greenhouse at their Research and Development Center. In 2002, interspecific crosses were performed by hand-pollination of emasculated D. nudicaule flowers with D. luteum pollen. The hybrid lines were produced via self-pollination for several generations, until 2005, when seeds of 03N-XX (03N-12 and 03N-15) and 05N-XX (05N-05, 05N-21, 05N-22, 05N-34, 05N-42, 05N-56 and 05N-61) were obtained as genetically fixed lines. F1 lines were produced following the procedure used for D. nudicaule and D. luteum hybrid plants. It should be noted that D. luteum is presently an endangered species, making it difficult to obtain fresh plant material (Koontz et al. Reference Koontz, Soltis and Brunsfeld2001).
Flower colour assessment
The colour code, colour group and colour name of the darkest parts of the sepals of D. nudicaule and the hybrid lines of D. nudicaule and D. luteum were classified according to the Royal Horticultural Society Colour Chart Sixth Edition (2015).
Anthocyanin extraction and LC-MS/MS analysis
For D. nudicaule, 03N-15, 05N-05 and 05N-34, anthocyanins were extracted from sepals using 80% methanol containing 1% formic acid. The extracts were partially purified using an InertSep Slim-J C18 cartridge (GL Sciences, Tokyo, Japan), evaporated to dryness and re-dissolved in 80% methanol containing 1% formic acid. These samples were analyzed at the Center for Analytical Instrumentation, Chiba University. For N20-1 and N20-7, mass spectrometric analysis of freeze-dried sepals was performed at Kanazawa University. Putative anthocyanin structures were identified by comparing the obtained molecular weights with the KNApSAcK Core System database (Afendi et al. Reference Afendi, Okada, Yamazaki, Hirai-Morita, Nakamura, Nakamura, Ikeda, Takahashi, Altaf-Ul-Amin, Darusman, Saito and Kanaya2012) and previous reports (Hashimoto et al. Reference Hashimoto, Tanaka, Maeda, Shimizu and Sakata2000; Tanikawa et al. Reference Tanikawa, Seto, Suzuki, Omori and Tatsuzawa2024).
Discussion
In this study, interspecific hybridization between D. nudicaule and D. luteum, followed by self-pollination, established genetically fixed hybrid lines with various flower colours: red-purple (03N-12, 03N-15), red (05N-05), orange-red (05N-21, 05N-22, 05N-34) and yellow (05N-42, 05N-56, 05N-61) (Fig. 1, Table 1). Because red-purple and yellow are rare traits, further crosses were performed between these fixed lines, resulting in progenies (N20 series) with purple, red-purple, red, orange-red and yellow flowers.
Flower phenotype of hybrid progenies of D. nudicaule and D. luteum.
Figure 1 Long description
The image displays a collection of flower phenotypes from hybrid progenies of D. nudicaule and D. luteum. The top row includes D. nudicaule and hybrids 03N-12, 03N-15, 05N-05 and 05N-21, showing red-purple and red flowers. The second row features hybrids 05N-22, 05N-34, 05N-42, 05N-56 and 05N-61, displaying orange-red and yellow flowers. The third row presents hybrids N20-1, N20-4, N20-5, N20-6 and N20-7, with purple and red-purple flowers. The bottom row includes hybrids N20-8, N20-10, N20-11, N20-12, N20-16 and N20-20, showing orange-red and yellow flowers. Each flower is positioned on a grid background, highlighting their distinct colors and forms.
Phenotypic characteristics and anthocyanin composition of D. nudicaule, D. luteum and their hybrid progenies
Table 1 Long description
The table compares phenotypic characteristics and anthocyanin composition of D. nudicaule, D. luteum, and their hybrid progenies. D. nudicaule exhibits vivid reddish orange color with specific anthocyanins, while D. luteum's data is not tested. Hybrid lines show a range of colors from red-purple to yellow, with varying anthocyanin presence, some not detected or not tested. Fixed hybrid lines like 05N-05 share anthocyanin profiles with D. nudicaule, while crosses between fixed lines show broader anthocyanin diversity. The data suggests significant color variation among hybrids, with some lines showing unique anthocyanin combinations. Interpretation is limited by instances of untested or undetected anthocyanins.
Notes: NT indicated ‘not tested’. a: Peak numbers correspond to those in Supplementary Figure S1. LC-MS/MS analysis was performed on selected lines representing distinct colour groups.
LC-MS/MS analysis revealed that the red-purple flowers accumulated complex anthocyanins derived from pelargonidin, cyanidin and delphinidin (Table 1). In contrast, D. nudicaule and the red/orange progenies accumulated only pelargonidin-based anthocyanins. Generally, anthocyanin colour is determined by the number of hydroxyl groups on the B-ring; flavonoid 3′-hydroxylase (F3′H) is required to synthesize purple cyanidin, whereas flavonoid 3′,5′-hydroxylase (F3′5′H) is responsible for synthesizing delphinidin, the basis for blue pigment (Tanaka and Brugliera Reference Tanaka and Brugliera2013). Since D. nudicaule accumulates only pelargonidin-based anthocyanins, it is inferred to lack functional F3′H and F3′5′H genes or enzyme activities. In contrast, although D. luteum does not accumulate anthocyanins, the production of cyanidin and delphinidin in the hybrid progenies suggests that D. luteum possesses functional F3′H and F3′5′H enzymes, likely for the biosynthesis of flavonols. Consequently, the hybridization between D. nudicaule and D. luteum appears to have achieved complex anthocyanin synthesis through the mutual complementation of their biosynthetic pathways. A similar phenomenon was previously reported in the interspecific hybridization of D. zalil and D. cardinale (Sakaguchi et al. Reference Sakaguchi, Isobe, Fujita, Ozeki and Miyahara2019).
The chemical diversity observed in the hybrid progenies strongly supports distinct genetic backgrounds for D. nudicaule and D. luteum. No anthocyanins were detected in the yellow flowers of the progenies, suggesting they possess pigments similar to D. luteum. In the yellow-flowered species D. zalil, the colour is reported to be derived from flavonols (Miyahara et al. Reference Miyahara, Hamada, Okamoto, Hirose, Sakaguchi, Hatano and Ozeki2016). These results demonstrating that novel flower colours can be created by introducing genetic resources from unutilized wild species via traditional breeding, without relying on genetic modification.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/S1479262126100598.
Acknowledgements
This research was funded by the Japan Society for the Promotion of Science (JSPS) KAKENHI (Grant No. 19K15829). We thank the Center for Analytical Instrumentation, Chiba University, and Mr. Masataka Nakano at Center for Experimental Modeling of Human Disease, Kanazawa University for the LC-MS/MS analysis. During the preparation of this manuscript, the authors used Gemini (Google) in order to improve the English language. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.
