Carotene Hydroxylase Activity Determines the Levels of Both α-Carotene and Total Carotenoids in Orange Carrots

Abstract

The typically intense carotenoid accumulation in cultivated orange-rooted carrots (Daucus carota) is determined by a high protein abundance of the rate-limiting enzyme for carotenoid biosynthesis, phytoene synthase (PSY), as compared with white-rooted cultivars. However, in contrast to other carotenoid accumulating systems, orange carrots are characterized by unusually high levels of α-carotene in addition to β-carotene. We found similarly increased α-carotene levels in leaves of orange carrots compared with white-rooted cultivars. This has also been observed in the Arabidopsis thaliana lut5 mutant carrying a defective carotene hydroxylase CYP97A3 gene. In fact, overexpression of CYP97A3 in orange carrots restored leaf carotenoid patterns almost to those found in white-rooted cultivars and strongly reduced α-carotene levels in the roots. Unexpectedly, this was accompanied by a 30 to 50% reduction in total root carotenoids and correlated with reduced PSY protein levels while PSY expression was unchanged. This suggests a negative feedback emerging from carotenoid metabolites determining PSY protein levels and, thus, total carotenoid flux. Furthermore, we identified a deficient CYP97A3 allele containing a frame-shift insertion in orange carrots. Association mapping analysis using a large carrot population revealed a significant association of this polymorphism with both α-carotene content and the α-/β-carotene ratio and explained a large proportion of the observed variation in carrots.

Figure 1.

The Carotenoid Biosynthesis Pathway in Plants. Enzymes catalyzing specific reactions are indicated in boxes. Carotene hydroxylases exhibiting main activity for the reaction are shown in bold, while those with minor activities are shown in parentheses. The central part covering C9 to C9’ is replaced with “R” in cyclic carotenoids (adapted from Kim et al., 2009).

Figure 2.

α-Carotene and α-/β-Carotene Ratio in Leaves from DJ3S_pro:AtCYP97A3 Carrots. Orange-rooted carrots (CRC) were transformed with At-CYP97A3 under control of the yam DJ3S promoter (DJ3S_pro). Both leaf α-carotene levels (A) as well as α-/β-carotene ratios (B) are strongly reduced in DJ3S_pro:AtCYP97A3 lines (Dc#1, Dc#25, and Dc#7) and approached levels determined in leaves of white-rooted carrots (QAL). α-Carotene content is given in mmol mol⁻¹ chlorophyll. Data are mean of four (controls) and three (transgenic lines) biological replicates + sd, respectively.

Figure 3.

Characterization of Roots from DJ3S_pro:AtCYP97A3 CRC Lines. (A) At-CYP97A3 expression in roots from 8-, 12-, and 16-week-old DJ3S_pro:AtCYP97A3 carrots (cultivar CRC). Transcript levels determined by qRT-PCR, normalized to 18S rRNA levels, and expressed relative to one selected sample from line Dc#1 (8 weeks). (B) Carotenoids in roots from DJ3S_pro:AtCYP97A3 lines and nontransformed CRC. (C) Immunoblot analysis using 60 µg leaf protein from Arabidopsis wild type and lut5-1 (right) and 40 µg carrot root protein from wild-type CRC and DJ3S_pro:AtCYP97A3 lines (left). Antiserum directed against At-CYP97A3 was used; actin signals are shown as loading control. (D) Transverse section of 8-week-old roots from DJS3_pro:AtCYP97A3 CRC line Dc#25 (right) and nontransformed CRC as control (left). (E) Ratio of α-carotene to β-carotene. Data represent the mean ± sd of three biological replicates, except Dc#1/16 weeks (one sample; two technical replicates were used in [A]). For detailed carotenoid data, see Supplemental Table 2.

Figure 4.

PSY Expression in Roots of DJ3S_pro:AtCYP97A3 Lines. (A) Transcript levels of both putative carrot PSY paralogs (Dc-PSY1 and Dc-PSY2) were determined by qRT-PCR, normalized to 18S rRNA, and are expressed relative to 8-week-old CRC roots. (B) PSY protein levels were determined in roots from CRC and DJ3S_pro:AtCYP97A3 line Dc#1 and Dc#25 by immunoblot analysis using antibodies directed against Arabidopsis PSY. Along with the imported form (imp, 42 kD) a putative degradation product (deg, 27 kD) is detected. An immunoblot using Arabidopsis leaf protein (At-lvs) is shown for comparison. Actin levels were used as loading control. The plant age given refers to weeks after transfer to soil. (C) PSY protein levels were normalized to actin of the corresponding sample and expressed relative to the ratio detected in 8-week-old CRC roots. Data represent the mean ± sd of three biological replicates, except Dc#1/16 weeks (two technical replicates).

Figure 5.

The CYP97A3 Gene from Carrots. (A) Phylogenetic tree of cytochrome P450 subfamily 97A and C proteins. At, A. thaliana, Dc, D. carota, Os, O. sativa, Rc, R. communis, Sl, S. lycopersicum, Mt, M. truncatula. Arabidopsis CYP86A1 was included as the outgroup sequence. The identified carrot CYP97A3 sequence (framed) groups with other β-ring carotene hydroxylases. Bootstrap values are reported next to the branches (only bootstraps above 50% are shown). (B) Sections of sequencing chromatograms of RT-PCR fragments from Dc-CYP97A3 (top) and Dc-CYP97A3Ins (bottom); the 8-nucleotide insertion is marked with a box. (C) Schematic map of proteins encoded by Dc-CYP97A3 and Dc-CYP97A3Ins. The insertion results in a premature translational termination prior to the sequence encoding the heme binding cysteine residue (P450 cys). Primers used for RT-PCR in (B) are indicated by arrows. (D) Allelic effect of the insertion/deletion polymorphism in CYP97A3 on α-carotene content (in mg 100 g⁻¹ FW). The box plots show the quartile division with the median indicated by lines within the boxes. The two whiskers correspond to the first and third quartile.