Carotene Hydroxylase Activity Determines the Levels of Botha-Carotene and Total Carotenoids in Orange CarrotsW

Jacobo Arango,a,1 Matthieu Jourdan,b Emmanuel Geoffriau,b Peter Beyer,a and Ralf Welscha,2

a University of Freiburg, Faculty of Biology II, Freiburg, Germanyb Agrocampus Ouest, Unité Mixte de Recherche 1345 Institut de Recherche en Horticulture et Semences, Angers, France

ORCID IDs: 0000-0003-3039-818X (M.J.); 0000-0003-4384-3107 (E.G.); 0000-0002-2865-2743 (R.W.)

The typically intense carotenoid accumulation in cultivated orange-rooted carrots (Daucus carota) is determined by a high proteinabundance of the rate-limiting enzyme for carotenoid biosynthesis, phytoene synthase (PSY), as compared with white-rootedcultivars. However, in contrast to other carotenoid accumulating systems, orange carrots are characterized by unusually high levelsof a-carotene in addition to b-carotene. We found similarly increased a-carotene levels in leaves of orange carrots compared withwhite-rooted cultivars. This has also been observed in the Arabidopsis thaliana lut5 mutant carrying a defective carotenehydroxylase CYP97A3 gene. In fact, overexpression ofCYP97A3 in orange carrots restored leaf carotenoid patterns almost to thosefound in white-rooted cultivars and strongly reduced a-carotene levels in the roots. Unexpectedly, this was accompanied by a 30 to50% reduction in total root carotenoids and correlated with reduced PSY protein levels while PSY expression was unchanged. Thissuggests a negative feedback emerging from carotenoidmetabolites 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 mappinganalysis using a large carrot population revealed a significant association of this polymorphism with both a-carotene content andthe a-/b-carotene ratio and explained a large proportion of the observed variation in carrots.

INTRODUCTION

Carrot (Daucus carota) roots exhibit a panoply of colors rangingfrom yellow to orange and red (Surles et al., 2004; Arscott andTanumihardjo, 2010). This color spectrum is mostly determinedby the abundance of carotenoids, with the exception of purplecarrots, which also accumulate anthocyanins (Simon et al., 2008;Montilla et al., 2011). The earliest cultivated carrots were yellow orpurple and emerged in the Afghanistan region before the 900s, whileorange and white cultivars were bred in the 16th century (Banga,1957; Soufflet-Freslon et al., 2013). Despite significant progress inthe identification of genetic markers associated with root color inrecent years, knowledge of the molecular basis of diverse colorphenotypes is sparse (Santos and Simon, 2006; Just et al., 2009;Fuentes et al., 2012; Rodriguez-Concepcion and Stange, 2013).

In carrot roots, carotenoids accumulate as crystals within chro-moplasts (Maass et al., 2009; Kim et al., 2010; Wang et al., 2013). Inleaf chloroplasts, these pigments are mostly protein-bound, formingessential constituents of light-harvesting complexes and pho-tosynthetic reaction centers. Furthermore, carotenoids representultimate precursors for at least two classes of plant hormones,abscisic acid (ABA) and the strigolactones (Walter and Strack,2011; Alder et al., 2012).

The first carotenoid-specific reaction of plastid prenyllipid bio-synthesis is catalyzed by the enzyme phytoene synthase (PSY;Figure 1) and yields the C40 hydrocarbon phytoene (for a reviewon carotenoid biosynthesis, see Cazzonelli, 2011). Followingdesaturation and isomerization reactions, which are catalyzedby four enzymes in plants (phytoene desaturase, z-carotenedesaturase, z-carotene isomerase, and carotenoid isomerase),the red-colored all-trans-lycopene is formed (Isaacson et al.,2002; Park et al., 2002; Li et al., 2007). Two different cyclasesintroduce e-ionone (e-cyclase, LCYe) and/or b-ionone rings(b-cyclase, LCYb), yielding a-carotene (e-b-carotene) or b-carotene(b-b-carotene), respectively.The synthesis of the oxygen-containing xanthophylls from

either a- or b-carotene requires ring-specific hydroxylation reactions.In Arabidopsis thaliana, these reactions are catalyzed by a set offour enzymes (Kim et al., 2009). Two non-heme di-iron enzymes(b-carotene hydroxylase 1 [BCH1] and BCH2) are primarily re-sponsible for b-ring hydroxylation of b-carotene and producezeaxanthin, while two heme-containing cytochrome P450 enzymes(CYP97A3 and CYP97C1) preferentially hydroxylate the e- andb-ionone rings of a-carotene, yielding lutein. Zeaxanthin is furtherepoxidized by the enzyme zeaxanthin epoxidase, leading toantheraxanthin, violaxanthin, and neoxanthin, a reversible reactionrepresenting the xanthophyll cycle (Niyogi et al., 1998). The analysisof leaf carotenoid patterns from hydroxylase-deficient Arabidopsismutants revealed partially overlapping substrate specificities of thehydroxylases involved and reflect the requirement for the completeset of hydroxylases to generate a balanced pigment composition(Kim et al., 2009).High carotenoid levels in non-green plant tissues have frequently

been found to be dependent on PSY expression, indicating thisenzyme to be rate-limiting in carotenogenesis. The overexpressionof PSY, e.g., in tomato fruits (Solanum lycopersicum), cassava

1Current address: International Center for Tropical Agriculture, TropicalForages, AA 6713 Cali, Colombia.2 Address correspondence to ralf.welsch@biologie.uni-freiburg.de.The author responsible for distribution of materials integral to the findingspresented in this article in accordance with the policy described in theInstructions for Authors (www.plantcell.org) is: Ralf Welsch (ralf.welsch@biologie.uni-freiburg.de).W Online version contains Web-only data.www.plantcell.org/cgi/doi/10.1105/tpc.113.122127

This article is a Plant Cell Advance Online Publication. The date of its first appearance online is the official date of publication. The article has been

edited and the authors have corrected proofs, but minor changes could be made before the final version is published. Posting this version online

reduces the time to publication by several weeks.

The Plant Cell Preview, www.aspb.org ã 2014 American Society of Plant Biologists. All rights reserved. 1 of 11

roots (Manihot esculenta), canola seeds (Brassica napus), andrice endosperm (Oryza sativa), effectively increased the fluxthrough the biosynthetic pathway (Shewmaker et al., 1999; Yeet al., 2000; Fraser et al., 2007; Welsch et al., 2010). Fluxes canbe driven at levels where even b-carotene crystals form. Thiswas observed in transgenic white-rooted carrots overexpressingthe bacterial PSY (crtB; Maass et al., 2009). Accordingly, highPSY protein levels can be revealed by immunoblotting in or-ange-rooted cultivars while being undetectable in white-rootedcultivars. However, these large differences are not equivalentlyreflected in PSY mRNA levels, suggesting the involvement ofadditional regulatory mechanisms. Identification of the remain-ing genes involved in carotenoid biosynthesis permitted analysisof their expression during root development of various colorcultivars (Clotault et al., 2008). However, the results obtainedcan only partially explain the different carotenoid patterns andfail to account for the wide range of carotenoid levels found.

Compared with other carotenoid accumulating tissues, oneremarkable feature of orange carrots is the high abundance ofa-carotene, in addition to b-carotene. We found that unusuallyhigh a-carotene levels are also present in leaves of orange-rooted carrots. This is a distinctive characteristic of plants witha deficiency in one cytochrome P450-type carotene hydroxylase,as concluded from investigations of Arabidopsis mutants (Kimand DellaPenna, 2006). In this work, we took advantage of thisknowledge and investigated whether the high a-carotene phe-notype in carrots is attributed to such a defect by overexpressingthe corresponding Arabidopsis carotene hydroxylase gene inorange-rooted carrots. Our data show that this assumption iscorrect inasmuch as the carotenoid pattern is concerned andidentified a feedback mechanism on total carotenoid contentthat might be involved in carotenoid differences observed invarious carrot varieties.

RESULTS

a-Carotene Levels in Carrot Cultivars

Most cultivated orange carrots exhibit remarkably high a-carotenelevels, between 15 and 30% of the total carotenoid content, andan accordingly high a-/b-carotene ratio of 0.2 up to 0.5 (Surleset al., 2004; Clotault et al., 2008; Simon et al., 2008; Arscott andTanumihardjo, 2010). For instance, tomato fruits accumulate ;4%a-carotene with an a-/b-carotene ratio of 0.02 (Stigliani et al.,2011), and Arabidopsis leaves contain <1% a-carotene with ana-/b-carotene ratio of ;0.03 (Kim et al., 2009). Furthermore,root-specific overexpression of PSY in a white-rooted carrotbackground (cultivar Queen Anne’s Lace [QAL]) increases thetotal carotenoid amounts, but this is almost exclusively due toelevated b-carotene while a-carotene remains at low levels. Simi-larly, PSY overexpression in Arabidopsis roots results in significantcarotenoid accumulations with only 3% a-carotene and lowa-/b-carotene ratios (Maass et al., 2009).We compared the carotenoid profiles in leaves from two

orange-rooted cultivars (Chantenay Red Cored [CRC] andNantaise [NAN]) with two white-rooted cultivars (QAL and Küttinger[KUT]), which accumulate high and low PSY protein levels, re-spectively (Maass et al., 2009). Table 1 shows that irrespectiveof root color and PSY abundance, total leaf carotenoid levelsand the carotenoid/chlorophyll ratios were very similar and sowere the levels of the major xanthophyll lutein, an a-carotenederivative. The difference to be noted is that the proportion ofa-carotene is strongly elevated, up to 10% of the total carotenoids,in leaves of orange-rooted cultivars, compared with those of white-rooted cultivars where only;1% a-carotene was found. Based onprevious results where a-carotene levels were unchanged uponoverexpression of PSY (lines At12 and At22 in Table 1; Maasset al., 2009), we conclude that high levels of a-carotene in leaves,and supposedly also in roots of orange carrots, involve additionalmechanisms that are present in orange but absent from white-rooted cultivars.Alteration in e-ring hydroxylation pattern affecting a-carotene

levels are known from Arabidopsismutants with reduced carotenehydroxylase capacity (Kim et al., 2009). Among the cytochrome

Figure 1. The Carotenoid Biosynthesis Pathway in Plants.

Enzymes catalyzing specific reactions are indicated in boxes. Carotenehydroxylases exhibiting main activity for the reaction are shown in bold,while those with minor activities are shown in parentheses. The centralpart covering C9 to C9’ is replaced with “R” in cyclic carotenoids (adaptedfrom Kim et al., 2009).

2 of 11 The Plant Cell

P450 (CYP97C1 and CYP97A3) and non-heme hydroxylases(BCH1 and BCH2) involved, only the CYP97A3 mutant showeda high proportion of a-carotene (between 15 and 20%), exhibitingthe lut5 phenotype (Kim and DellaPenna, 2006). The increase ofa-carotene in lut5 is compensated for by an almost equivalent re-duction in b-carotene, resulting in wild-type total carotenoid levels(Table 1; HPLC chromatograms are provided in SupplementalFigure 1). This was also found in leaves of orange-rooted carrotcultivars (;14% b-carotene in CRC and NAN compared with;18% in QAL and KUT; Table 1). Furthermore, the (e-ring) mono-hydroxylated intermediate a-cryptoxanthin, which accumulatesdue to the CYP97A3 deficiency in Arabidopsis lut5, was also de-tected in CRC and NAN, while being absent in QAL and KUT.

Expression of Arabidopsis CYP97A3 in Orange CarrotsRescues the Leaf Phenotype

The similarity in the leaf carotenoid patterns between Arabidopsislut5 and orange-rooted carrots suggests the presence of reducedCYP97A3 activity in orange cultivars. Aiming at rescuing the lowa-carotene “chemotype,”we transformed theorange carrot cultivarCRC with a construct harboring the CYP97A3 cDNA from Arabi-dopsis (At-CYP97A3; Kim and DellaPenna, 2006) under control ofthe DJ3S promoter from yams. This promoter is highly active incarrot roots and;100-fold lessactive in leaves (Arangoet al., 2010).In leaves of all three transgenic DJ3Spro:AtCYP97A3 events ge-nerated, a-carotene was found to be strongly reduced, resultingin an accordingly decreased a-/b-carotene ratio (Figure 2; seeSupplemental Table 1 for detailed carotenoid data). The changesin different DJ3Spro:AtCYP97A3 lines largely corresponded withtransgene expression levels (Supplemental Figure 2). Thus, leafcarotenoid patterns in the transgenic CRC gradually approachedthose fromwhite-rootedcarrot leavesupononly the small increasesinAt-CYP97A3expression allowedby thepromoter characteristics.

Root-Specific Overexpression of Arabidopsis CYP97A3 inOrange Carrots Strongly Reduces a-Carotene Amountsand Causes a Reduction in Total Carotenoid Levels

DJ3Spro:AtCYP97A3 CRC lines and wild-type CRC carrots weregrown in soil and their roots harvested after 8, 12, and 16 weeks.

Growth of transgenic lines was indistinguishable from non-transformed CRC and roots developed similar to the wild type(Supplemental Table 2). At-CYP97A3 mRNA was strongly expressedthroughout all selected stages (Figure 3A). Furthermore, immunoblotanalysisperformedwithroots fromtwotransgenicDJ3Spro:AtCYP97A3lines,Arabidopsiswild-type and lut5 leaves, confirmed the presenceof the mature At-CYP97A3 protein and mirrored overall transgeneexpression differences between these lines (Figure 3C).As with the leaves, a-carotene was found strongly reduced in

these roots throughout all stages, comprising only ;2% of totalcarotenoids compared with 20% in untransformed CRC controlplants (Figure 3B; Supplemental Figure 3). Accordingly, thea-/b-carotene ratio decreased from 0.5 in nontransformed CRCto 0.1 and 0.04 in average in lines Dc#1 and Dc#25, respectively;this decrease was;3-fold during root growth (Figure 3E). Youngroots of transgenic lines also accumulated CYP97A3-catalyzedmonohydroxylated intermediates, such as zeinoxanthin andb-cryptoxanthin, which were absent in the controls (SupplementalTable 2). These intermediates disappeared in older transgenicroots (12 and 16 weeks), suggesting a limited activity of othercarotene hydroxylases, e.g., the e-ring specific CYP97C1, duringearly root development. Surprisingly, the decrease in a-carotenefrom;300 µg in untransformed CRC to 10 µg in line Dc#25 did notentail an equivalent increase in CYP97A3-derived xanthophylls.These remained at only small total amounts in both young roots(100 µg in Dc#25 versus 71 µg in CRC) as well as older roots(150 µg in Dc#25 versus 100 µg in CRC).Moreover, the conceivable decrease in a-carotene was accom-

panied by an unexpected 30 to 50% reduction in total carotenoidlevels in transgenic DJ3Spro:AtCYP97A3 relative to untransformedcontrol roots of the same age (Figures 3B and 3D). This decreaseaffected not only b-carotene, but all other carotenoids, includingdesaturation intermediates such as phytoene, phytofluene, andz-carotene. For instance, b-carotene was reduced by 77%, whilenoncolored carotenes were reduced by 32% in 8-week-old rootsfrom line Dc#25 compared with the control. Expression of theputative carrot homolog of the ABA-responsive gene 9-cis-epoxycarotenoid dioxygenase 3 (NCED3) was largely unchangedin the transgenic lines (Supplemental Figure 4; Auldridge et al.,2006; Just et al., 2007). This suggests that the altered carotenoidlevels did not entail changes in ABA synthesis.

Table 1. Carotenoids in Arabidopsis and Carrot Leaves

Line Tot. Car Lutein a-Crypt a-Carotene b-Carotene VAZN chla/b

QAL 431.4 6 2 206.6 6 0 (47.9) ND 2.6 6 0 (0.6) 84.0 6 1 (19.5) 138.2 6 1 (32.0) 3.6 6 0.1KUT 458.1 6 36 213.4 6 3 (46.8) ND 3.4 6 1 (0.7) 81.5 6 17 (17.6) 159.8 6 15 (34.8) 3.4 6 0.4CRC 438.0 6 8 213.4 6 3 (48.7) 1.4 6 0.3 (0.3) 26.1 6 1 (6.0) 64.9 6 4 (14.8) 132.3 6 0 (30.2) 3.6 6 0.1NAN 458.3 6 2 222.8 6 5 (48.6) 6.9 6 1.1 (1.5) 45.7 6 5 (10.0) 59.6 6 8 (13.0) 123.2 6 5 (26.9) 3.4 6 0.1At-Wt 521.7 6 94 233.6 6 30 (45.0) ND 3.9 6 1 (0.7) 114.8 6 19 (22.0) 169.5 6 44 (32.3) 3.1 6 0.5At12 486.2 6 31 227.0 6 14 (46.7) ND 3.1 6 1 (0.6) 105.8 6 10 (21.7) 150.3 6 9 (30.9) 3.2 6 0.1At22 534.0 6 2 248.6 6 8 (46.6) ND 2.4 6 2 (0.4) 107.6 6 15 (20.2) 175.4 6 24 (32.9) 2.9 6 0.1lut5-1 503.8 6 18 248.1 6 6 (49.3) 8.9 6 0.9 (1.8) 78.9 6 4 (15.7) 45.5 6 7 (9.0) 122.4 6 8 (24.3) 3.1 6 0.1

Carotenoids were quantified by HPLC from leaves of white-rooted (QAL and KUT) and orange-rooted carrot cultivars (CRC and NAN). For comparison,leaf carotenoids are shown from twoCaMV35Spro:At-PSY Arabidopsis lines (At12 and At22), the corresponding wild type (At-Wt), and from the lut5-1mutantcarrying a T-DNA integration within CYP97A3. Carotenoids are expressed in mmol pigment mol21 chlorophyll. The relative percentage of each carotenoid isgiven in parentheses. Tot. Car, total carotenoids; a-Crypt, a-cryptoxanthin; VAZN, sum of violaxanthin, antheraxanthin, zeaxanthin, and neoxanthin; chla/b,ratio between chlorophyll a and b; ND, not detectable. Data represent the mean 6 SD of four (carrot) and three (Arabidopsis) replicates, respectively.

a-Carotene in Orange Carrots 3 of 11

Since PSY has frequently been reported as the rate-limitingstep in non-green plant tissues (Fraser et al., 2002; Paine et al.,2005; Rodríguez-Villalón et al., 2009; Welsch et al., 2010), reducedphytoene and total carotenoid levels may suggest an attenuatedphytoene synthesis capacity as a consequence of At-CYP97A3overexpression. However, transcript levels of both carrot PSYgenes were quite similar in all lines throughout all stages com-pared with nontransformed CRC carrots (Figure 4A). We thereforeinvestigated PSY protein levels in roots from DJ3Spro:AtCYP97A3lines and nontransformed CRC carrots. Protein gel blot analysisrevealed that, in fact, PSY protein amounts were very stronglyreduced, e.g., by;50% in line Dc#25 (Figures 4B and 4C). Levelsof a putative PSY degradation product did not increase, whichsuggests that PSY translation rather than its turnover rate isaffected. Furthermore, Arabidopsis wild-type and lut5 roots ac-cumulate barely detectable PSY protein levels, which appearsomewhat higher in the mutant (Supplemental Figure 5). Thismight indicate that the observed dependence of PSY proteinlevels on hydroxylation capacity is not only a distinct feature ofcarrots. In conclusion, the altered carotene hydroxylation uponAt-CYP97A3 overexpression in orange carrots negatively affectedPSY protein levels, explaining the reduced total carotenoid ac-cumulation observed.

Cloning of CYP97A3 from White Carrots

High a-carotene levels in CRC wild type and their strong re-duction upon Arabidopsis CYP97A3 overexpression suggest a de-ficient CYP97A3 activity in CRC. Since sequence information forcarrot cytochrome P450 hydroxylases is available for the e-ring-specific CYP97A1 (Just et al., 2007), but not for CYP97A3, we firstclonedCYP97A3 from QAL, assuming its intactness in white-rootedcarrots. Based on sequence conservation of CYP97A3 homologs,we amplified a 1.3-kb fragment and completed the contiguouscoding region by rapid amplification of cDNA ends (39end) andgenome walking (59end). A phylogenetic analysis with the de-duced amino acid sequence (616 amino acids) confirmed itsidentity as carrot CYP97A3 (Figure 5A). This was concluded

from the high sequence identity to several cytochrome P450enzymes, which were functionally confirmed as b-ring-specifica-carotene hydroxylases, e.g., from tomato (Sl-CYP97A29, 76%identity; Stigliani et al., 2011), rice (Os-CYP97A4, 69% identity;Quinlan et al., 2012), and Arabidopsis CYP97A3 (71% identity;alignment in Supplemental Figure 6). In contrast, the sequenceexhibited only 41% identity to e-ring-specific carotene hydrox-ylases, including carrot CYP97C1, and branched-off separatelyin phylogenetic analysis, excluding that the sequence obtainedrepresented an additional CYP97C1 copy (Figure 5A). DNA gelblot analysis with genomic DNA from QAL and CRC revealed thepresence of carrot CYP97A3 as single-copy gene in both culti-vars. Quantitative RT-PCR (qRT-PCR) conducted on root RNAfrom QAL and CRC revealed no large differences in expressionlevels (Supplemental Figure 7), so that the possibility of muta-tions was considered.

An Insertion Present in a CYP97A3 Allele in Orange Carrots

Total leaf RNA was isolated from CRC and NAN plants and used toamplifyCYP97A3with the same primer combination used to amplifyCYP97A3 from QAL. Remarkably, CYP97A3 cDNAs from both or-ange-rooted varieties were 8 nucleotides longer than from QAL,which was caused by a single insertion at position #1074 (Figure 5B;the electrophoretic mobility shift is shown in Supplemental Figure 7).The insertion results in a premature translational stop, encodinga truncated protein with only 382 instead of 616 amino acids(CYP97A3Ins, Figure 5C). The truncated protein lacks a C-terminalcysteine residue that is essential for the assembly of the hemecofactor; it is therefore nonfunctional.In order to confirm the involvement of the CYP97A3 gene in

a-carotene accumulation in carrot roots, the identified insertion/deletion was genotyped in a broad unstructured population of380 individuals, and an association mapping strategy was con-ducted on carotenoid content. Three associations were found to besignificant with the insertion/deletion tested. The a-carotenecontent was significantly associated with this polymorphism,as expressed in dry (P value = 3.53 3 1025) or fresh weight

Figure 2. a-Carotene and a-/b-Carotene Ratio in Leaves from DJ3Spro:AtCYP97A3 Carrots.

Orange-rooted carrots (CRC) were transformed with At-CYP97A3 under control of the yam DJ3S promoter (DJ3Spro). Both leaf a-carotene levels (A) aswell as a-/b-carotene ratios (B) are strongly reduced in DJ3Spro:AtCYP97A3 lines (Dc#1, Dc#25, and Dc#7) and approached levels determined in leavesof white-rooted carrots (QAL). a-Carotene content is given in mmol mol21 chlorophyll. Data are mean of four (controls) and three (transgenic lines)biological replicates + SD, respectively.

4 of 11 The Plant Cell

(P value = 5 3 1025) and explained 12 and 14% of the observedvariation, respectively (Figure 5D; Supplemental Figure 8). Finally, thisinsertion was significantly associated with the a-/b-carotene ratio(P value = 3.72 3 1024), but explained a small proportion of theobserved variation (r2 = 0.04). Associations with any other carotenoidwere not significant. These results confirm the involvement ofCYP97A3 in a-carotene accumulation and explained a relatively highproportion of the variability of a-carotene content in carrot roots.

DISCUSSION

Cytochrome P450 Carotene Hydroxylase Deficiencyin Orange Carrots

a-Carotene is the second most abundant carotene present inorange carrot cultivars. This is not limited to roots; a-caroteneaccumulates to unusually high amounts in leaves and is increasedup to 10-fold compared with leaves of white-rooted cultivars. Inthis work, we addressed this widespread feature of orange-rootedcarrots that is absent from white-rooted cultivars.

Leaf a-carotene content varies between different plant speciesand is also increased in response toward low light intensities(Matsubara et al., 2009). However, the constitutively high a-carotenelevels in orange carrots point to a genetic alteration like inthe Arabidopsis lut5 mutant, carrying a disruption within thecytochrome P450 carotene hydroxylase CYP97A3 (Kim andDellaPenna, 2006). In fact, the expression of the ArabidopsisCYP97A3 gene in CRC largely rescued leaf carotenoid patterns,approaching those of white-rooted carrots, regarding their lowa-carotene content, low a-/b-carotene ratio, and the completeabsence of a-cryptoxanthin. Similarly, roots of CYP97A3-overexpressing CRC lines revealed carotenoid patterns matchingoverexpressing a bacterial phytoene synthase (Maass et al.,2009) as well as of PSY-overexpressing tissues from othercrops, such as potato tubers (Solanum tuberosum) or cas-sava roots (Ducreux et al., 2005; Welsch et al., 2010). Theseobservations corroborate that a-carotene abundance in or-ange carrots is not a concomitant of increased pathway fluxby high PSY levels, but rather represents a particular prop-erty of orange carrots.

Figure 3. Characterization of Roots from DJ3Spro:AtCYP97A3 CRC Lines.

(A) At-CYP97A3 expression in roots from 8-, 12-, and 16-week-old DJ3Spro: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 DJ3Spro: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 andDJ3Spro: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 DJS3pro:AtCYP97A3 CRC line Dc#25 (right) and nontransformed CRC as control (left).(E) Ratio of a-carotene to b-carotene.Data represent the mean 6 SD of three biological replicates, except Dc#1/16 weeks (one sample; two technical replicates were used in [A]). For detailedcarotenoid data, see Supplemental Table 2.

a-Carotene in Orange Carrots 5 of 11

These findings are in agreement with our identification of aCYP97A3 allele in various orange carrot cultivars that carries an 8-nucleotide insertion and encodes a truncated protein that is mostlikely dysfunctional. An association study conducted on a largeunstructured carrot population revealed that this polymorphism isassociated with both a-carotene content and a-/b-carotene ratio inroots. More specifically, 88% of the individuals homozygous forCYP97A3Ins contained high a-carotene levels (above 2 mg g21

fresh weight), while this applied for only 47% of those individualscarrying the wild-type allele. Therefore, the CYP97A3 polymorphismidentified explains a large proportion of the variation observed,even though there are other currently unknown factors affectinga-carotene levels in orange carrots. This is in agreement witha quantitative trait loci study, which estimated four genes as theminimum number of genes involved in a-carotene inheritance(Santos and Simon, 2006). Apparently, the CYP97A3Ins allele rep-resents one of these genes, and it remains to be determined whetherother candidate genes cause similar CYP97A3 deficiencies throughalternative mechanisms, e.g., by altered CYP97A3 expression.

Considering the remarkably high a-carotene levels coincidingwith high b-carotene levels in various carrot cultivars, wequestioned whether CYP97A3 activity also modulates the levelof total carotenoids in roots. On one side, this is supported byour experimental data: The defective enzyme present in theorange-rooted varieties investigated correlated with high totalcarotenoids, while the overexpression of the intact version bringsthem down again, apparently by downregulation of PSY protein.However, this notion is not corroborated in the association analysis,as the deficient CYP97A3 allele was associated with a-carotenecontent, while an association with total carotenoid levels was notfound. It must be concluded that this is because the associationpanel does not provide the situation observed with the over-expression of a functional CYP97A3. One can expect one half orone dosage of functional CYP97A3 but not more. The transgenic

carrots are therefore thought to reveal a phenomenon that cannotbe discovered in a biparental cross.However, the principle discovered is that downstream carotenoids

can regulate the total pathway flux, which might contribute tothe explanation of several observations. For instance, a recentquantitative trait loci study identified two major interacting loci thatdetermine much of the variability of carotenoid amounts in carrots(Just et al., 2009). Markers for the e-ring specific a-carotenehydroxylase CYP97C1 were found in close proximity to one ofthe two loci (Just et al., 2007). While it was difficult to explainreductions in total carotene content by a deficiency of an enzymeacting downstream of b-carotene synthesis, our findings open theway for possible explanations.

Feedback Regulation of Carotenoid Biosynthesis

Similar to other nuclear-encoded plastid-localized pathways, theregulation of carotenogenic enzymes requires retrograde sig-naling to coordinate (at the level of gene expression) metabolicflux with product requirements. Molecules from various path-ways are currently considered as likely candidates for plastid tonucleus communication, e.g., the chlorophyll intermediate Mgprotoporphyrin IX for photosynthesis-related or methylerythritolcyclodiphosphate (MEP pathway) for stress-responsive geneexpression (Chi et al., 2013). It is currently unknown how thesemolecules shuttle between compartments traversing the plastidenvelope membranes. Carotenoids, especially, appear very unsuit-able to function as immediate signaling metabolites in the regulationof carotenogenic gene expression because of their lipophilicity.Nevertheless, there are several published examples documenting

an impact of altered xanthophyll patterns on total carotenoidamounts. For instance, silencing of zeaxanthin epoxidase andBCH in potatoes resulted in 6- and 4-fold higher total carotenoidamounts, respectively (Römer et al., 2002; Diretto et al., 2007).

Figure 4. PSY Expression in Roots of DJ3Spro: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 areexpressed relative to 8-week-old CRC roots.(B) PSY protein levels were determined in roots from CRC and DJ3Spro:AtCYP97A3 line Dc#1 and Dc#25 by immunoblot analysis using antibodies directedagainst Arabidopsis PSY. Along with the imported form (imp, 42 kD) a putative degradation product (deg, 27 kD) is detected. An immunoblot usingArabidopsis 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 6 SD of three biological replicates, except Dc#1/16 weeks (two technical replicates).

6 of 11 The Plant Cell

Likewise, both LCYb-deficient maize mutants as well as LCYe-silenced potato tubers accumulated higher total carotenoidlevels (Diretto et al., 2006; Bai et al., 2009). Furthermore, enhancedlycopene accumulation in the tomato ogcmutant is considered tobe due to increased activity of earlier enzymes of the pathway inaddition to reduced lycopene cyclization caused by a deficientfruit-specific lycopene cyclase (Ronen et al., 2000; Bramley,2002). The induced changes in metabolite abundance entailedaltered expression of carotenogenic enzymes, which in mostcases included elevated PSY transcript levels. Furthermore, recentfindings suggest the involvement of cis-carotene intermediates in theregulation of tomato fruit-specific PSY1 expression (Kachanovskyet al., 2012).

PSY levels/activity are key to carotenoid accumulation in non-green plant tissues (Ye et al., 2000; Lindgren et al., 2003; Ducreuxet al., 2005; Welsch et al., 2010), including carrot roots (Santoset al., 2005; Maass et al., 2009). The context described here be-tween CYP97A3 overexpression leading to reduced PSY proteinlevels suggests the presence of a negative feedback regulation,stemming from carotenoids downstream of a-carotene or metabo-lites thereof, acting on PSY levels. The difference with respect to theexamples given above is that the expression of carrot PSYgenes remained unchanged upon At-CYP97A3 overexpression

in our study, while PSY protein levels were reduced. We there-fore present an example of the modulation of PSY protein levelsby carotenoid metabolites through mechanisms acting beyondtranscription. Whether the feedback regulation at this level af-fects additional enzymes of the pathway cannot be conclusivelyanswered as yet due to the lack of suitable antibodies. However,there are several examples in non-green tissues in which varyingthe rate of phytoene synthesis affects the total carotenoid con-tent strongly while the carotenoid pattern is hardly affected (Paineet al., 2005; Maass et al., 2009). It is therefore conceivable thatfeedback-inhibited phytoene synthesis is sufficient to explain thereduced carotene amounts in At-CYP97A3–overexpressing carrots.

Metabolite-Induced Feedback Regulation of PSYProtein Abundance

In contrast to carotenoids, apocarotenoids are known to function assignaling molecules. In fact, it is reasonable to assume that stronglyreduced a-carotene levels in At-CYP97A3 overexpressing CRCroots are caused by enhanced metabolization. However, this iscontrasted by the lack of accordingly elevated b-ring hydroxylateda-carotene derivatives, i.e., zeinoxanthin or lutein, especiallyin older roots. The levels of these xanthophylls increased only

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 b-ringcarotene hydroxylases. Bootstrap values are reported next to the branches (only bootstraps above 50% are shown).(B) SectionsofsequencingchromatogramsofRT-PCRfragmentsfromDc-CYP97A3 (top)andDc-CYP97A3Ins (bottom); the8-nucleotide insertion ismarkedwithabox.(C) Schematic map of proteins encoded by Dc-CYP97A3 and Dc-CYP97A3Ins. The insertion results in a premature translational termination prior to thesequence 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 a-carotene content (in mg 100 g21 FW). The box plots show the quartile divisionwith the median indicated by lines within the boxes. The two whiskers correspond to the first and third quartile.

a-Carotene in Orange Carrots 7 of 11

slightly, and total xanthophyll levels remained at as low levels as innontransformed CRC control plants. This suggests increasedxanthophyll turnover through their enhanced cleavage by caro-tenoid cleavage oxygenases (Simkin et al., 2004; Rodrigo et al.,2013). In fact, several carotenoid-derived cleavage products areimportant signaling molecules involved in developmental pro-cesses (strigolactones) and mediate responses toward abioticstress (ABA; Walter and Strack, 2011; Alder et al., 2012). However,the expression of the ABA-responsive carrot NCED3 was un-changed upon carotenoid changes in the transgenics. It thereforeappears unlikely that ABA is involved in the observed response.

A putative function of other metabolites might be to triggertotal carotenoid flux via the regulation of the rate-limiting stepsof the pathway, most notably PSY. A dose dependence of PSYprotein levels from carotenoid-derived metabolites might in factbe concluded from our data. a-Carotene levels in line Dc#25 aremuch lower than in line Dc#1, which might correlate with higherlevels of cleavage products derived from e-ring xanthophylls.Interestingly, this correlates with a more pronounced negativefeedback of Dc#25 over Dc#1 since PSY protein levels are lowerin Dc#25, corresponding with lower total carotenoid levels.

Similarly, an involvement of metabolites derived from e-ring hy-droxylated xanthophylls might also be concluded from transgenictomato plants fruit-specifically overexpressing the bacterial desa-turase crtI, which show an unexpected strong reduction in totalcarotenoid levels including phytoene (Römer et al., 2000). Whilea-carotene is almost absent in control fruits, crtI-overexpressingfruits accumulate small amounts of a-carotene and e-ring xantho-phylls, which correlate with PSY enzyme activity reduced by half.

As shown previously, high abundance of carotenes results in theircrystallization, a mode of sequestration that reduces their accessi-bility toward degradation, either enzymatically or nonenzymatically(Maass et al., 2009; Nogueira et al., 2013). Therefore, the sug-gested negative feedback inhibition might also be part of a regu-latory network that prevents critically high carotene concentrationsin order to maintain carotenoid homeostasis.

It is currently unclear which of the two putative carrot PSYparalogs is affected by the negative feedback, as the antibodyused cannot differentiate between these variants. Whether re-duced PSY translation or increased protein turnover results indecreased steady state PSY protein levels cannot be conclu-sively answered yet. However, lower PSY protein levels in At-CYP97A3–overexpressing carrots do not entail higher levels ofputative PSY breakdown products. Furthermore, PSY proteincan be very strongly expressed in Arabidopsis (Maass et al.,2009), which may point to protein turnover being less relevantfor flux adjustments. Metabolite-dependent regulation of trans-lation required for the fine-tuning of biosynthetic efficiency ap-pears more likely. Investigations to further discriminate betweenthese possibilities are currently underway.

METHODS

Plant Material and Growth

Arabidopsis thaliana (ecotype Wassilewskija) seeds were grown asepti-cally on Murashige and Skoog medium under long-day conditions for14 d. Seedlings were ground in liquid nitrogen immediately after harvest

and stored at 270°C for further analysis. Arabidopsis roots were harvestedfrom seedlings grown according to Hétu et al. (2005). Seeds from carrots(Daucus carota) were obtained as follows:QueenAnne’s Lace, RichtersHerbs;Küttiger, Dreschflegel Saatgut; Nantaise, Freya; Chantenay Red Cored, B&TWorld Seeds. Carrot plants were grown in soil under long-day conditions.Roots of 8-, 12-, and 16-week-old carrot plants were removed from the soil,ground in liquid nitrogen, and stored at 270°C for further analysis.

Carrot Transformation

The At-CYP97A3 coding region was amplified by PCR with primers 59-GTGGCGTTCCATGGCTATGGCCTTTCCTCTTTCTTATACT-39 and 59-AAAGATGAACACGTGAGAAAGAGCAGATGAAACTTCATCC-39 from theRIKEN full-length cDNA clone RAFL09-10-L12, thereby introducing NcoIand PmlI restriction sites, respectively. The fragment was digested withNcoI and PmlI and ligated into the vector backbone obtained from theaccordingly digested vector pCAMBIA-DJ3S-GUS (Arango et al., 2010),revealing pCAMBIA-DJ3S-AtCYP97A3. This vector was used to generatetransgenicCRC lines as described byMaass et al. (2009), with the exceptionthat 10 mg mL21 hygromycin was used for selection of transgenic calli.

Cloning of Dc-CYP97A3

Sequence alignments with CYP97A3 sequences from various taxa wereperformed to identify conserved regions. Degenerated oligonucleotideswith wobble nucleotides were deduced: forward, ATT GCT/G TCT/CGGT/A/G/C GA/C G/T/ATT CAC T/C/GGT, and reverse, TGA AAG TT/CTAA/CC GT/A C/G A/G A/T CCT CGA/T GG. Total QAL leaf RNA wasreverse transcribed using the reverse primer and the verso cDNA kit(Thermo), followed by a PCR using both primers and Phusion polymerase(Thermo). Specific primers were deduced from the partial CYP97A3 se-quence obtained and the 39 end of themRNAwas retrieved using the 59/39RACE kit (Roche Diagnostics) and the forward primer 59-ACATCGC-TGTCCTCAAAGGTGGGA-39 according to the manufacturer’s instructions.Sequence information of the 59mRNA end was retrieved by genome walking(GenomeWalker Universal Kit; Clontech) followed by exon predictions usingthe FGENESH algorithm available in Softberry (www.softberry.com). Full-length CYP97A3 coding regions were amplified using the primers forward59-CCCAGCCAACGTCTCCAAATGATAGCC-39 and reverse 59-GAACA-TTGCTGCACATACAAAGAT-39 and cloned in the vector pCR2.1 (InvitrogenLife Technologies). For CYP97A3 RT-PCR, 200 ng total RNA was reversetranscribed with primer reverse 59-CATCTCCAGATGCCAACAAAAAAT-39followed by PCR upon the addition of primer forward 59-ATTCCCATT-TGGAAAGACATTTCG-39 (1 min 93°C, 1 min 60°C, 1 min 72°C, 28 cycles).Amplification products were separated by agarose gel electrophoresis.

DNA Gel Blot Analysis

Genomic DNA isolation and DNA gel blot analysis was performed as de-scribed by Arango et al. (2010) using the DIG DNA labeling and detection kit(Roche Diagnostics). A probe covering 676 bp of the Dc-CYP97A3 cDNAwas used (positions 318 to 994; accession number JQ655297).

Phylogenetic Analysis

Sequences used for phylogenetic analysis were selected according toStigliani et al. (2011). Sequences were aligned with T-Coffee (Notredameet al., 2000), and the phylogenetic tree was reconstructed using theneighbor-joining method in MEGA5 (Tamura et al., 2011). The evolutionarydistances were computed using the Poisson correction method andbootstrap test was selected with 1000 replicates. A member of family 86 ofthe cytochrome superfamily, the fatty acid hydroxylase At-CYP86A1, waschosen as an outgroup. For GenBank accession numbers in Figure 5A.

8 of 11 The Plant Cell

TaqMan Real-Time RT-PCR Assay

Total RNA was isolated using the plant RNA purification reagent (Invi-trogen Life Technologies). RNA purification, DNaseI digestion, and real-time RT-PCR assays were performed as described (Welsch et al., 2008).Carrot PSY1, PSY2, CYP97A3, and nos 39UTR (present in the At-CYP973transgene mRNA) expression was detected with 6FAM-labeled probes,while SYBR green was used for NCED3 expression analysis. For primersand probes, see Supplemental Table 3. For 18S rRNA quantification, theeukaryotic 18S rRNA endogenous control kit (Life Technologies) wasused. The relative quantity of the transcripts was calculated using thecomparative threshold cycle method (Livak and Schmittgen, 2001). Datawere normalized first to the corresponding 18S rRNA levels and thenexpressed relative to a selected sample indicated.

Carotenoid Extraction and Quantification

Carotenoids were extracted using lyophilized plant materials (5 mg forleaves; 20 mg for carrot roots) and analyzed by HPLC as described(Welsch et al., 2008).

Immunoblot Analysis

Generation and affinity purification of antibodies directed against Arab-idopsis PSY is described by Maass et al. (2009). Serum containing anti-At-CYP97A3 antibodies was used in 1:5000 dilutions. Proteins wereextracted with phenol as described (Welsch et al., 2007). After SDS-PAGE, blotting onto polyvinylidene fluoride membranes (Carl Roth), andtreatment with blocking solution (TBS containing 5% [w/v] milk powder),membranes were incubated with antibodies in PBS containing 0.1% (v/v)Tween 20 and 1% (w/v) milk powder. For detection, the ECL system (GEHealthcare) was used. Protein gel blots were stripped and reprobed withanti-actin antibodies (Sigma-Aldrich). Quantification of band intensitieswas performed with ImageJ (Abramoff et al., 2004).

Validation through Association Mapping Analysis

Polymorphism in CYP97A3was tested in association mapping analysis ina broad unstructured population obtained by three intercrossing gen-erations of a diversified panel of 67 carrot cultivars. A total of 380 in-dividuals were randomly chosen at the third generation and were grown infield (Agrocampus Ouest, Angers, France) following standard agronomicpractices. Carotenoid contents were quantified by HPLC with a modifiedprocedure adapted from Clotault et al. (2008). CYP97A3 polymorphismwas genotyped by KASPar Assay technology (LGCGenomics), along with92 single nucleotide polymorphisms (SNPs) spread all over the genome.Raw data on carotenoid content and CYP97A3 polymorphism are givenin Supplemental Data Set 1. Association analysis (Hall et al., 2010) wasperformed using the TASSEL software (Bradbury et al., 2007). As re-latedness between individuals can lead to false positive detection, theassociation tests were based on a mixed linear model in which a kinshipmatrix, estimated by identity by states calculated on SNP data set, wasadded as a covariable (Kang et al., 2008).

Accession Numbers

Sequence data from this article can be found in the GenBank/EMBL datalibraries under the following accession numbers: JQ655297 (Dc-CYP97A3mRNA), JQ655298 (Dc-CYP97A3 gene, partial), DQ192196 (Dc-CYP97C1),EEC67393.1 (Os-CYP97C2), EEC74248.1 (Os-CYP97A4), NP_190881(At-CYP97C1), At1g31800 (At-CYP97A3), ABC59096 (Mt-CYP97C10),ABD28565.1 (Mt-CYP97A10), ACJ25968 (Sl-CYP97C11), ACJ25969(Sl-CYP97A29), At5g58860 (At-CYP86A1), XP_002512609.1 (Rc-CYP97A),XP_002519427.1 (Rc-CYP97C), and DQ192202 (Dc-NCED3).

Supplemental Data

The following materials are available in the online version of this article.

Supplemental Figure 1. HPLC Chromatograms from Carrot andArabidopsis Leaves.

Supplemental Figure 2. At-CYP97A3 Expression in Transgenic CRC.

Supplemental Figure 3. Carotenoid Patterns of DJ3Spro:AtCYP97A3CRC Lines.

Supplemental Figure 4. Dc-NCED3 Expression in Roots of DJ3Spro:AtCYP97A3 Lines.

Supplemental Figure 5. Carotenoid and PSY Protein Levels inArabidopsis and Carrot Roots.

Supplemental Figure 6. Alignment of Carrot CYP97A3 with OtherCytochrome p450 Proteins.

Supplemental Figure 7. Gene Copy Number and Expression of CarrotCYP97A3.

Supplemental Figure 8. Allelic Effects of CYP97A3 Polymorphism ona-Carotene Content and a-/b-Carotene Ratio.

Supplemental Table 1. Carotenoids in Leaves from DJ3Spro:AtCYP97A3Carrots.

Supplemental Table 2. Detailed Carotenoid Amounts in Roots fromAt-CYP97A3-Overexpressing CRC Lines.

Supplemental Table 3. Primers and Probes Used for qRT-PCR.

Supplemental Data Set 1. Carotenoid Contents and CYP97A3Genotype.

ACKNOWLEDGMENTS

This work was supported by the HarvestPlus research consortium anda Deutsche Forschungsgemeinschaft Research Grant to R.W. (GrantWE4731/2-1). We thank Dean DellaPenna (Michigan State University) forproviding lut5 seeds and the At-CYP97A3 cDNA. We also thank StefanoCazzaniga and Roberto Bassi (Università di Verona, Italy) for providingantibodies against At-CYP97A3. We thank Jérémy Clotault (Universitéd’Angers, France) for initiating the collaboration between R.W. and E.G.M.J. was supported by the French Ministry of Research. We are indebtedto Carmen Schubert for her skillful contribution to carrot transformationand analysis.

AUTHOR CONTRIBUTIONS

R.W., P.B., and E.G. conceived this project and designed all experi-ments. R.W., J.A., and M.J. performed experiments. R.W., P.B., J.A.,E.G., and M.J. analyzed data. R.W., P.B., and E.G. wrote the article.

Received December 19, 2013; revised March 14, 2014; accepted April16, 2014; published May 23, 2014.

REFERENCES

Abramoff, M.D., Magelhaes, P.J., and Ram, S.J. (2004). ImageProcessing with ImageJ. Biophotonics International 11: 36–42.

Alder, A., Jamil, M., Marzorati, M., Bruno, M., Vermathen, M., Bigler, P.,Ghisla, S., Bouwmeester, H., Beyer, P., and Al-Babili, S. (2012). Thepath from b-carotene to carlactone, a strigolactone-like plant hormone.Science 335: 1348–1351.

a-Carotene in Orange Carrots 9 of 11

Arango, J., Salazar, B., Welsch, R., Sarmiento, F., Beyer, P., andAl-Babili, S. (2010). Putative storage root specific promoters fromcassava and yam: cloning and evaluation in transgenic carrots asa model system. Plant Cell Rep. 29: 651–659.

Arscott, S.A., and Tanumihardjo, S.A. (2010). Carrots of many colorsprovide basic nutrition and bioavailable phytochemicals acting asa functional food. Compr. Rev. Food Sci. Food Saf. 9: 223–239.

Auldridge, M.E., Block, A., Vogel, J.T., Dabney-Smith, C., Mila, I.,Bouzayen, M., Magallanes-Lundback, M., DellaPenna, D., McCarty, D.R., and Klee, H.J. (2006). Characterization of three members of theArabidopsis carotenoid cleavage dioxygenase family demonstrates thedivergent roles of this multifunctional enzyme family. Plant J. 45: 982–993.

Bai, L., Kim, E.-H., DellaPenna, D., and Brutnell, T.P. (2009). Novellycopene epsilon cyclase activities in maize revealed through perturbationof carotenoid biosynthesis. Plant J. 59: 588–599.

Banga, O. (1957). The development of the original European carrotmaterial. Euphytica 6: 64–76.

Bradbury, P.J., Zhang, Z., Kroon, D.E., Casstevens, T.M., Ramdoss, Y.,and Buckler, E.S. (2007). TASSEL: software for association mapping ofcomplex traits in diverse samples. Bioinformatics 23: 2633–2635.

Bramley, P.M. (2002). Regulation of carotenoid formation during tomatofruit ripening and development. J. Exp. Bot. 53: 2107–2113.

Cazzonelli, C. (2011). Goldacre Review: Carotenoids in nature: insightsfrom plants and beyond. Funct. Plant Biol. 38: 833–847.

Chi, W., Sun, X., and Zhang, L. (2013). Intracellular signaling fromplastid to nucleus. Annu. Rev. Plant Biol. 64: 559–582.

Clotault, J., Peltier, D., Berruyer, R., Thomas, M., Briard, M., andGeoffriau, E. (2008). Expression of carotenoid biosynthesis genesduring carrot root development. J. Exp. Bot. 59: 3563–3573.

Diretto, G., Tavazza, R., Welsch, R., Pizzichini, D., Mourgues, F.,Papacchioli, V., Beyer, P., and Giuliano, G. (2006). Metabolicengineering of potato tuber carotenoids through tuber-specific silencingof lycopene epsilon cyclase. BMC Plant Biol. 6: 13.

Diretto, G., Welsch, R., Tavazza, R., Mourgues, F., Pizzichini, D.,Beyer, P., and Giuliano, G. (2007). Silencing of beta-carotene hydroxylaseincreases total carotenoid and beta-carotene levels in potato tubers. BMCPlant Biol. 7: 11.

Ducreux, L.J.M., Morris, W.L., Hedley, P.E., Shepherd, T., Davies,H.V., Millam, S., and Taylor, M.A. (2005). Metabolic engineering ofhigh carotenoid potato tubers containing enhanced levels of beta-carotene and lutein. J. Exp. Bot. 56: 81–89.

Fraser, P.D., Enfissi, E.M.A., Halket, J.M., Truesdale, M.R., Yu, D.,Gerrish, C., and Bramley, P.M. (2007). Manipulation of phytoenelevels in tomato fruit: effects on isoprenoids, plastids, and intermediarymetabolism. Plant Cell 19: 3194–3211.

Fraser, P.D., Römer, S., Shipton, C.A., Mills, P.B., Kiano, J.W., Misawa,N., Drake, R.G., Schuch, W., and Bramley, P.M. (2002). Evaluation oftransgenic tomato plants expressing an additional phytoene synthase ina fruit-specific manner. Proc. Natl. Acad. Sci. USA 99: 1092–1097.

Fuentes, P., Pizarro, L., Moreno, J.C., Handford, M., Rodríguez-Concepcion, M., and Stange, C. (2012). Light-dependent changesin plastid differentiation influence carotenoid gene expression andaccumulation in carrot roots. Plant Mol. Biol. 79: 47–59.

Hall, D., Tegström, C., and Ingvarsson, P.K. (2010). Using associationmapping to dissect the genetic basis of complex traits in plants. BriefFunct. Genomics 9: 157–165.

Hétu, M.-F., Tremblay, L.J., and Lefebvre, D.D. (2005). High rootbiomass production in anchored Arabidopsis plants grown in axenicsucrose supplemented liquid culture. Biotechniques 39: 345–349.

Isaacson, T., Ronen, G., Zamir, D., and Hirschberg, J. (2002).Cloning of tangerine from tomato reveals a carotenoid isomeraseessential for the production of beta-carotene and xanthophylls inplants. Plant Cell 14: 333–342.

Just, B.J., Santos, C.A., Yandell, B.S., and Simon, P.W. (2009). Major QTLfor carrot color are positionally associated with carotenoid biosyntheticgenes and interact epistatically in a domesticated x wild carrot cross.Theor. Appl. Genet. 119: 1155–1169.

Just, B.J., Santos, C.A.F., Fonseca, M.E.N., Boiteux, L.S., Oloizia,B.B., and Simon, P.W. (2007). Carotenoid biosynthesis structuralgenes in carrot (Daucus carota): isolation, sequence-characterization,single nucleotide polymorphism (SNP) markers and genome mapping.Theor. Appl. Genet. 114: 693–704.

Kachanovsky, D.E., Filler, S., Isaacson, T., and Hirschberg, J.(2012). Epistasis in tomato color mutations involves regulation ofphytoene synthase 1 expression by cis-carotenoids. Proc. Natl. Acad.Sci. USA 109: 19021–19026.

Kang, H.M., Zaitlen, N.A., Wade, C.M., Kirby, A., Heckerman, D.,Daly, M.J., and Eskin, E. (2008). Efficient control of populationstructure in model organism association mapping. Genetics 178:1709–1723.

Kim, J., and DellaPenna, D. (2006). Defining the primary route for luteinsynthesis in plants: the role of Arabidopsis carotenoid beta-ring hydroxylaseCYP97A3. Proc. Natl. Acad. Sci. USA 103: 3474–3479.

Kim, J., Smith, J.J., Tian, L., and Dellapenna, D. (2009). Theevolution and function of carotenoid hydroxylases in Arabidopsis.Plant Cell Physiol. 50: 463–479.

Kim, J.E., Rensing, K.H., Douglas, C.J., and Cheng, K.M. (2010).Chromoplasts ultrastructure and estimated carotene content in rootsecondary phloem of different carrot varieties. Planta 231: 549–558.

Li, F., Murillo, C., and Wurtzel, E.T. (2007). Maize Y9 encodesa product essential for 15-cis-zeta-carotene isomerization. PlantPhysiol. 144: 1181–1189.

Lindgren, L.O., Stålberg, K.G., and Höglund, A.-S. (2003). Seed-specific overexpression of an endogenous Arabidopsis phytoenesynthase gene results in delayed germination and increased levels ofcarotenoids, chlorophyll, and abscisic acid. Plant Physiol. 132: 779–785.

Livak, K.J., and Schmittgen, T.D. (2001). Analysis of relative geneexpression data using real-time quantitative PCR and the 2(-DeltaDelta C(T)) method. Methods 25: 402–408.

Maass, D., Arango, J., Wüst, F., Beyer, P., and Welsch, R. (2009).Carotenoid crystal formation in Arabidopsis and carrot roots causedby increased phytoene synthase protein levels. PLoS ONE 4: e6373.

Matsubara, S., Krause, G.H., Aranda, J., Virgo, A., Beisel, K.G.,Jahns, P., and Winter, K. (2009). Sun-shade patterns of leaf carotenoidcomposition in 86 species of neotropical forest plants. Funct. Plant Biol.36: 20–36.

Montilla, E.C., Arzaba, M.R., Hillebrand, S., and Winterhalter, P.(2011). Anthocyanin composition of black carrot (Daucus carota ssp.sativus var. atrorubens Alef.) cultivars Antonina, Beta Sweet, Deep Purple,and Purple Haze. J. Agric. Food Chem. 59: 3385–3390.

Niyogi, K.K., Grossman, A.R., and Björkman, O. (1998). Arabidopsismutants define a central role for the xanthophyll cycle in the regulation ofphotosynthetic energy conversion. Plant Cell 10: 1121–1134.

Nogueira, M., Mora, L., Enfissi, E.M.A., Bramley, P.M., and Fraser,P.D. (2013). Subchromoplast sequestration of carotenoids affectsregulatory mechanisms in tomato lines expressing different carotenoidgene combinations. Plant Cell 25: 4560–4579.

Notredame, C., Higgins, D.G., and Heringa, J. (2000). T-Coffee: Anovel method for fast and accurate multiple sequence alignment. J.Mol. Biol. 302: 205–217.

Paine, J.A., Shipton, C.A., Chaggar, S., Howells, R.M., Kennedy, M.J.,Vernon, G., Wright, S.Y., Hinchliffe, E., Adams, J.L., Silverstone, A.L.,and Drake, R. (2005). Improving the nutritional value of Golden Ricethrough increased pro-vitamin A content. Nat. Biotechnol. 23: 482–487.

Park, H., Kreunen, S.S., Cuttriss, A.J., DellaPenna, D., and Pogson,B.J. (2002). Identification of the carotenoid isomerase provides

10 of 11 The Plant Cell

insight into carotenoid biosynthesis, prolamellar body formation,and photomorphogenesis. Plant Cell 14: 321–332.

Quinlan, R.F., Shumskaya, M., Bradbury, L.M.T., Beltrán, J., Ma, C.,Kennelly, E.J., and Wurtzel, E.T. (2012). Synergistic interactions betweencarotene ring hydroxylases drive lutein formation in plant carotenoidbiosynthesis. Plant Physiol. 160: 204–214.

Rodrigo, M.J., Alquézar, B., Alós, E., Medina, V., Carmona, L., Bruno,M., Al-Babili, S., and Zacarías, L. (2013). A novel carotenoid cleavageactivity involved in the biosynthesis of citrus fruit-specific apocarotenoidpigments. J. Exp. Bot. 64: 4461–4478.

Rodriguez-Concepcion, M., and Stange, C. (2013). Biosynthesis ofcarotenoids in carrot: an underground story comes to light. Arch.Biochem. Biophys. 539: 110–116.

Rodríguez-Villalón, A., Gas, E., and Rodríguez-Concepción, M.(2009). Phytoene synthase activity controls the biosynthesis of carotenoidsand the supply of their metabolic precursors in dark-grown Arabidopsisseedlings. Plant J. 60: 424–435.

Römer, S., Fraser, P.D., Kiano, J.W., Shipton, C.A., Misawa, N.,Schuch, W., and Bramley, P.M. (2000). Elevation of the provitamin Acontent of transgenic tomato plants. Nat. Biotechnol. 18: 666–669.

Römer, S., Lübeck, J., Kauder, F., Steiger, S., Adomat, C., andSandmann, G. (2002). Genetic engineering of a zeaxanthin-rich potato byantisense inactivation and co-suppression of carotenoid epoxidation.Metab. Eng. 4: 263–272.

Ronen, G., Carmel-Goren, L., Zamir, D., and Hirschberg, J. (2000).An alternative pathway to beta -carotene formation in plant chromoplastsdiscovered by map-based cloning of beta and old-gold color mutations intomato. Proc. Natl. Acad. Sci. USA 97: 11102–11107.

Santos, C.A.F., and Simon, P. (2006). Heritabilities and minimumgene number estimates of carrot carotenoids. Euphytica 151: 79–86.

Santos, C.A.F., Senalik, D., and Simon, P. (2005). Path analysis suggestsphytoene accumulation is the key step limiting the carotenoid pathwayin white carrot roots. Genet. Mol. Biol. 28: 287–293.

Shewmaker, C.K., Sheehy, J.A., Daley, M., Colburn, S., and Ke,D.Y. (1999). Seed-specific overexpression of phytoene synthase:increase in carotenoids and other metabolic effects. Plant J. 20: 401–412X.

Simkin, A.J., Schwartz, S.H., Auldridge, M.E., Taylor, M.G., andKlee, H.J. (2004). The tomato carotenoid cleavage dioxygenase 1genes contribute to the formation of the flavor volatiles beta-ionone,pseudoionone, and geranylacetone. Plant J. 40: 882–892.

Simon, P.W., Tanumihardjo, S.A., Clevidence, B.A., and NovotnyDura, J. (2008). Role of color and pigments in breeding, genetics,

and nutritional improvement of carrots. In Color Quality of Fresh andProcessed Foods, C.A. Culver and R.E. Wrolstad, eds (Washington,DC: Oxford University Press), pp. 151–165.

Soufflet-Freslon, V., Jourdan, M., Clotault, J., Huet, S., Briard, M.,Peltier, D., and Geoffriau, E. (2013). Functional gene polymorphism toreveal species history: the case of the CRTISO gene in cultivated carrots.PLoS ONE 8: e70801.

Stigliani, A.L., Giorio, G., and D’Ambrosio, C. (2011). Characterization ofP450 carotenoid b- and e-hydroxylases of tomato and transcriptionalregulation of xanthophyll biosynthesis in root, leaf, petal and fruit. PlantCell Physiol. 52: 851–865.

Surles, R.L., Weng, N., Simon, P.W., and Tanumihardjo, S.A. (2004).Carotenoid profiles and consumer sensory evaluation of specialtycarrots (Daucus carota, L.) of various colors. J. Agric. Food Chem.52: 3417–3421.

Tamura, K., Peterson, D., Peterson, N., Stecher, G., Nei, M., andKumar, S. (2011). MEGA5: molecular evolutionary genetics analysisusingmaximum likelihood, evolutionary distance, andmaximum parsimonymethods. Mol. Biol. Evol. 28: 2731–2739.

Walter, M.H., and Strack, D. (2011). Carotenoids and their cleavageproducts: biosynthesis and functions. Nat. Prod. Rep. 28: 663–692.

Wang, Y.-Q.Q., Yang, Y., Fei, Z., Yuan, H., Fish, T., Thannhauser,T.W., Mazourek, M., Kochian, L.V., Wang, X., and Li, L. (2013).Proteomic analysis of chromoplasts from six crop species revealsinsights into chromoplast function and development. J. Exp. Bot.64: 949–961.

Welsch, R., Arango, J., Bär, C., Salazar, B., Al-Babili, S., Beltrán, J.,Chavarriaga, P., Ceballos, H., Tohme, J., and Beyer, P. (2010).Provitamin A accumulation in cassava (Manihot esculenta) rootsdriven by a single nucleotide polymorphism in a phytoene synthasegene. Plant Cell 22: 3348–3356.

Welsch, R., Maass, D., Voegel, T., Dellapenna, D., and Beyer, P.(2007). Transcription factor RAP2.2 and its interacting partner SINAT2:stable elements in the carotenogenesis of Arabidopsis leaves. PlantPhysiol. 145: 1073–1085.

Welsch, R., Wüst, F., Bär, C., Al-Babili, S., and Beyer, P. (2008). Athird phytoene synthase is devoted to abiotic stress-induced abscisicacid formation in rice and defines functional diversification of phytoenesynthase genes. Plant Physiol. 147: 367–380.

Ye, X., Al-Babili, S., Klöti, A., Zhang, J., Lucca, P., Beyer, P., andPotrykus, I. (2000).Engineering theprovitaminA (beta-carotene)biosyntheticpathway into (carotenoid-free) rice endosperm. Science 287: 303–305.

a-Carotene in Orange Carrots 11 of 11

12 3

4 6

5

QAL

CRC

At-Wt

lut5-1

20 40 60min 20 40 60min

6

5

Supplemental Figure 1

Supplemental Data. Arango et al. Plant Cell (2014) 10.1105/tpc.113.122127

Supplemental Figure 1. HPLC chromatograms from carrot and Arabidopsis leaves.Representative HPLC chromatograms at 450 nm of leaf extracts from white-rooted (Queen Anne‘sLace, QAL) and orange-rooted carrot cultivars (Chantenay Red Cored, CRC), Arabidopsis wild type(At-Wt) and from the Arabidopsis lut5-1mutant carrying a T-DNA integration within CYP97A3.1,violaxanthin/neoxanthin, 2; lutein; 3 chl b; 4, chl a; 5,α-carotene; 6,β-carotene.

rel.

Dc#

25

0

0.5

1.0

1.5

Dc#25 Dc#7Dc#1

At-CYP97A3

Supplemental Figure 2

Supplemental Data. Arango et al. Plant Cell (2014) 10.1105/tpc.113.122127

Supplemental Figure 2. At-CYP97A3 expression in transgenic CRC.At-CYP97A3 transcript levels were determined by qRT-PCR. Transgene levels were normalized to18SrRNA and are expressed relative to one selected sample of line Dc#25. Data represent the mean±SE of three biological replicates.

CRC

other xanthos 4%

lutein 7%

α-carotene22%

β-carotene41%

phytoene 12%phytofluene 7%

ζ-carotene 7%

Dc#25

vio/neo 2%other xanthos 2%

lutein 18% α-carotene 2%

β-carotene40%

phytoene 17%phytofluene

10%

ζ-carotene 9%

vio/neoother xanthoslutein

α-caroteneβ-carotene

phytoenephytoflueneζ-carotene

Supplemental Figure 3

Supplemental Data. Arango et al. Plant Cell (2014) 10.1105/tpc.113.122127

Supplemental Figure 3. Carotenoid patterns of DJ3Spro:At-CYP97A3 CRC lines.Chantenay Red Cored carrots (CRC) were transformed with At-CYP97A3 under control of theroot-specific DJ3S promoter from yams. Root carotenoids patterns of 12-week-old plants fromline Dc#25 were quantified by HPLC and compared with those determined for non-transformedcontrols (CRC). Data represent the mean ± SE of three biological replicates. Vio/neo,violaxanthin/neoxanthin; other xanthos, other xanthophylls.

0

0.5

1

1.5

Dc#25

8 12

CRC Dc#1

Dc-NCED3 expression

8 12 8 12

Dc-

NC

ED

3re

l. C

RC

age (weeks)

line

Supplemental Figure 4

Supplemental Data. Arango et al. Plant Cell (2014) 10.1105/tpc.113.122127

Supplemental Figure 4. Dc-NCED3 expression in roots of DJ3Spro:At-CYP97A3 lines.Dc-NCED3 transcripts were determined by qRT-PCR, normalized to 18SrRNA and are expressedrelative to 8-week-old CRC roots. Data represent the mean + SD of two biological replicates.

PSYactin

wt lut5

At-rts

CRC

Dc-rts

rec

0

1

2

3

4

5

6

7

wt lut5

carotenoid contentβ-caroteneα-caroteneluteinβ/β-xanthophylls

BA

µg g

-1D

W

PSY protein levels

Supplemental Figure 5

Supplemental Data. Arango et al. Plant Cell (2014) 10.1105/tpc.113.122127

Supplemental Figure 5. Carotenoid and PSY protein levels in Arabidopsis and carrot roots.Arabidopsis wild-type and lut5-1 roots were harvested from hydroponically grown seedlings aftertwo weeks of growth and used for carotenoid (A) and immunoblot analysis with antibodies directedagainst Arabidopsis PSY (B). Carrot CRC roots and1.25 ng recombinant purified His-tagged At-PSY(rec) was included for comparison. Actin levels were determined as loading control. 40 µg rootprotein was used. Data in A represent the mean + SD of three biological replicates.

Dc-CYP97A3 MATTLNLHL--F-PAQQ-----FLQSKIQHN---RRNPT---STF------------TKLNGSYRFSGIKCSVSNGKLPN 54Dc-CYP97A3Ins MATTLNLHL--F-PAQQ-----FLQSKIQHN---RRNPT---STF------------TKLNGSYRFSGIKCSVSNGKLPN 54Os-CYP97A4 MAATSSAAAAAAPPPCR-----LLGSGQAHL---RLPPA-------------------AASGRRRRLLIRCSASGGNNGK 53At-CYP97A3 MAMAFPLSY--T-PTIT------------------VKPV--------------------TYSRRSNFVVFSSSSNGRDPL 39Mt-CYP97A10 MASHLTLLH--A-PPPL-----SLQTKTFHSKYITIKPLKPTTTFSSSCSLFPCSLKTSHRGSC-SSFIACSSSNGRSPN 71Sl-CYP97A29 MASSLPLFQ--F-PTHH-----YSKSRL------TLSPK--------------------FKGSVSNFTIRCSNSNGKQPE 46Rc-CYP97A MAANFALLQ--V-PSSISAKHHCLQTKFQVR---RVKPI---NLS-----SFPP-TQNGVLGKRKFAVISCASSKGREPE 65

Dc-CYP97A3 S----TEEDE----EEEKKKRIRAELSARIASGEFTVEK-SGFQSQLLDGLAKLGAPSEVLDVLSKWIGASEDYPKIPEA 125Dc-CYP97A3Ins S----TEEDE----EEEKKKRIRAELSARIASGEFTVEK-SGFQSQLLDGLAKLGAPSEVLDVLSKWIGASEDYPKIPEA 125Os-CYP97A4 GGGGSGSDPI----LEERRRRRQAELAARIASGEFTAQG-PAWIAPLAAGLAKLGPPGELAAALLTKVA-GGGGPEIPQA 127At-CYP97A3 EEN-SVPNGV-KSLEKLQEEKRRAELSARIASGAFTVRK-SSFPSTVKNGLSKIGIPSNVLDFMFDWTGSDQDYPKVPEA 116Mt-CYP97A10 D---SVDDGVVKSADQLLEEKRRAELSAKIASGEFTVKQESGLPSILKKSLSNLGVSNEILEFLFG------LYPKIPEA 142Sl-CYP97A29 S----VDEGV-KKVEKLLDEKRRAELSARIASGEFTVEQ-SGFPSLLKNGLSKLGVPKEFLEFFSRRTG---NYPRIPEA 117Rc-CYP97A S----EEDPV-KSVERILEEKRRAELSAKIASGEFTVQQ-SGFPSILRNGLSKLGVPNETLEFLFKWVDAGEGYPKIPEA 139

Dc-CYP97A3 KGAISAIRSEAFFIPLYELYLTYGGVFRLTFGPKSFLIVSDPTVAKHILRDNSKAYSKGILAEILEFVMGKGLIPADGEI 205Dc-CYP97A3Ins KGAISAIRSEAFFIPLYELYLTYGGVFRLTFGPKSFLIVSDPTVAKHILRDNSKAYSKGILAEILEFVMGKGLIPADGEI 205Os-CYP97A4 VGSMSAVTGQAFFIPLYDLFLTYGGIFRLNFGPKSFLIVSDPAIAKHILRDNSKAYSKGILAEILEFVMGTGLIPADGEI 207At-CYP97A3 KGSIQAVRNEAFFIPLYELFLTYGGIFRLTFGPKSFLIVSDPSIAKHILKDNAKAYSKGILAEILDFVMGKGLIPADGEI 196Mt-CYP97A10 KGSISAIRSEAFFIPLYELYITYGGIFRLNFGPKSFLIVSDPAIAKHILKDNSKAYSKGILAEILDFVMGKGLIPADGEI 222Sl-CYP97A29 KGSISAIRDEPFFMPLYELYLTYGGIFRLIFGPKSFLIVSDPSIAKHILKDNSKAYSKGILAEILDFVMGKGLIPADGEI 197Rc-CYP97A KGAISAIRSEAFFIPLYELYLTYGGIFRLTFGPKSFLIVSDPSIAKHILRDNSKAYSKGILAEILDFVMGKGLIPADGEI 219

Dc-CYP97A3 WRVRRRAIVPALHQKYVTAMISMFGQATDRLCAKLDAAASDAEDVEMESLFSRLTLDIIGKAVFNYDFDSLTNDTGIVEA 285Dc-CYP97A3Ins WRVRRRAIVPALHQKYVTAMISMFGQATDRLCAKLDAAASDAEDVEMESLFSRLTLDIIGKAVFNYDFDSLTNDTGIVEA 285Os-CYP97A4 WRVRRRAIVPAMHQKYVTAMISLFGEASDRLCQKLDKAASDGEDVEMESLFSRLTLDVIGKAVFNYDFDSLSYDNGIVEA 287At-CYP97A3 WRRRRRAIVPALHQKYVAAMISLFGEASDRLCQKLDAAALKGEEVEMESLFSRLTLDIIGKAVFNYDFDSLTNDTGVIEA 276Mt-CYP97A10 WRVRRRTIVPALHLKFVAAMIGLFGQATDRLCQKLDTAASDGEDVEMESLFSRLTLDVIGKAVFNYDFDSLSNDTGIIEA 302Sl-CYP97A29 WRVRRRAIVPALHQKYVAAMIGLFGKATDRLCKKLDVAATDGEDVEMESLFSRLTLDIIGKAVFNYDFDSLTVDTGIVEA 277Rc-CYP97A WRVRRRAIVPAFHQKYVAAMIGLFGQATDRLCKKLDAAASDGEDVEMESLFSRLTLDIIGKAVFNYEFDSLANDTGIVEA 299

Dc-CYP97A3 VYTVLREAEDRSVSPIPFWEIPIWKDISPKLKKVNSALELINGTLDDLIAICKRMVDEEELQFHEEYMNETDPSILHFLL 365Dc-CYP97A3Ins VYTVLREAEDRSVSPIPFWEIPIWKDISPKLKKVNSALELINGTLDDLIAICKRMVDEEELQFHEEYMNETDPTILVFSI 365Os-CYP97A4 VYVTLREAEMRSTSPIPTWEIPIWKDISPRQRKVNEALALINKTLDELIDICKRLVEEEDLQFHEEYMNEQDPSILHFLL 367At-CYP97A3 VYTVLREAEDRSVSPIPVWDIPIWKDISPRQRKVATSLKLINDTLDDLIATCKRMVEEEELQFHEEYMNERDPSILHFLL 356Mt-CYP97A10 VYTVLREAEDRSISPIPVWDLPIWKDISPRQRKVTAALKLVNDTLNNLIAICKRMVDEEELQFHEEYMNEQDPSILHFLL 382Sl-CYP97A29 VYTVLREAEDRSVAPIPVWELPIWKDISPKLKKVNAALKLINDTLDDLIAICKRMVDEEELQFHEEYMNEKDPSILHFLL 357Rc-CYP97A VYTVLREAEDRSVAPIPVWEIPIWKDISPRQRKVSAALKLINDILDDLIALCKRMVDEEELQFHDEYMNEQDPSILHFLL 379

Dc-CYP97A3 ASGDDVSSKQLRDDLMTMLIAGHETTAAVLTWTFYLLSKEPSVMLKLQNEVDSVLGDRIPTIEDMKKLKYTTRVINESLR 445Dc-CYP97A3Ins FCWHLEMTSQVSSSVMI--------------------------------------------------------------- 375Os-CYP97A4 ASGDDVSSKQLRDDLMTMLIAGHETSAAVLTWTFYLLSKYPNVMAKLQDEADTVLGDRLPTIEDVKKLKYTTRVINESLR 447At-CYP97A3 ASGDDVSSKQLRDDLMTMLIAGHETSAAVLTWTFYLLTTEPSVVAKLQEEVDSVIGDRFPTIQDMKKLKYTTRVMNESLR 436Mt-CYP97A10 ASGDDVTSKQLRDDLMTMLIAGHETSAAVLTWTFYLLSKEPSVMSKLQEEVDSVLGDRFPTIEDMKKLKYTTRVINESLR 462Sl-CYP97A29 ASGDEVSSKQLRDDLMTMLIAGHETSAAVLTWTFYLLSKEPSVMAKLQDEVDSVLGDRLPTIEDLKKLRYTTRVINESLR 437Rc-CYP97A ASGDDVSSKQLRDDLMTMLIAGHETSAAVLTWTFYLLSKEPSVLSKLQNEVDTILGDRFPTIEDVKKLKYTTRVINESLR 459

Dc-CYP97A3 LYPQPPVLIRRSLEDDKLGEYPIKRNEDIFISIWNLHRCPQRWEDADKFNPERWPLDGPNPNETNQAFSYLPFGGGPRKC 525Os-CYP97A4 LYPQPPVLIRRSIEEDMLGGYPIGRGEDIFISVWNLHHCPKHWDGADVFNPERWPLDGPNPNETNQNFSYLPFGGGPRKC 527At-CYP97A3 LYPQPPVLIRRSIDNDILGEYPIKRGEDIFISVWNLHRSPLHWDDAEKFNPERWPLDGPNPNETNQNFSYLPFGGGPRKC 516Mt-CYP97A10 LYPQPPVLIRRSIEDDVLGEYPIKRGEDIFISVWNLHRSPTLWNDADKFEPERWPLDGPNPNETNQGFKYLPFGGGPRKC 542Sl-CYP97A29 LYPQPPVLIRRSIEEDVVGGYPIKRGEDIFISVWNLHRCPNHWEEADRFNPERWPLDGPNPNETNQNFSYLPFGGGPRKC 517Rc-CYP97A LYPQPPVLIRRSLQDDMLGKYPIKRGEDIFISVWNLHRSPHLWDDAEKFNPERWPLDGPNPNETNQNFCYLPFGGGPRKC 539

Dc-CYP97A3 VGDMFASFEAIVAVAMLVRRFNFQMALGAPPVKMTTGATIHTTEGLNMTVTKRITPPAVTA-ETSSLKDDSSVNM-SR– 601Os-CYP97A4 VGDMFATFETVVATAMLVRRFDFQMAPGAPPVEMTTGATIHTTEGLKMTVTRRTKPPVIPNLEMKVISD-SPENMSTTTS 606At-CYP97A3 IGDMFASFENVVAIAMLIRRFNFQIAPGAPPVKMTTGATIHTTEGLKLTVTKRTKPLDIPSVPILPMDT----------- 585Mt-CYP97A10 IGDMFASYEVVVALAMLVRRFNFQMAVGAPPVVMTTGATIHTTQGLNMTVTRRIKPPIVPSLQMSTLEVDPSVSISDK-- 620Sl-CYP97A29 VGDMFATFENLVAVAMLVQRFDFQMALGAPPVKMTTGATIHTTEGLKMTVTRRSRPPIVPNLEMATLEVDVNS-VSSD-- 594Rc-CYP97A VGDMFASFETVVATAMLVRRFNFQLALGAPPVKMTTGATIHTTEGLTMTVTRRIQPPIMPMLDMPAMKGDAPGSVPSG-- 617

Dc-CYP97A3 --------EAVADQKGEVS-VART 616Os-CYP97A4 MPISAASIASGEDQQGQVS-ATRI 629At-CYP97A3 -------------SRDEVS-SALS 595Mt-CYP97A10 --------TEEIGQKDQVYQAQKS 636Sl-CYP97A29 ------------RAEAEAS-TVRP 605Rc-CYP97A --------ESQLGQKGEVS-PAHS 632

Supplemental Data. Arango et al. Plant Cell (2014) 10.1105/tpc.113.122127

Supplemental Figure 6

Supplemental Figure 6. Alignment of carrot CYP97A3 with other cytochrome p450 proteins.Amino acid alignment of carrot CYP97A3 (Dc-CYP97A3) and the protein encoded by the Dc-CYP97A3Ins allele containing an 8 nt insertion with other members of cytochrome P450 subfamily97A . At, A. thaliana, Os, O. sativa, Rc, R. communis, Sl, S. lycopersicum, Mt, M. truncatula.

B21

53.5

21.51.4

0.9

QALCRCH E/B H E/BA

0

1

1.8

QAL CRC

Dc-

CY

P97

A3

rel.

QA

L

193 201

C Stop

Dc-CYP97A3

Dc-CYP97A3Ins

8 bp insertionStop

ATG

ATG

qPCR RT-PCR

Supplemental Figure 7

Supplemental Data. Arango et al. Plant Cell (2014) 10.1105/tpc.113.122127

Supplemental Figure 7. Gene copy number and expression of carrot CYP97A3.(A)Southern Blot analysis using Dc-CYP97A3 as a probe and genomic DNA from QAL and CRC,digested with HindIII (H) and EcoRI/BamHI (E/B), respectively; fragment sizes of EcoRI/HindIIIdigested lambda DNA are shown on the right (in kb).(B)Dc-CYP97A3 expression in roots of 12-week-old plants. Transcript levels were normalized to18S rRNA levels and are expressed relative to one selected sample from QAL (mean + SD of fivebiological replicates), primer positions are indicated below.(C)RT-PCR amplification products from QAL (193 bp) and CRC (201 bp), primers are indicated inthe schematic maps of Dc-CYP97A3 and Dc-CYP97A3Ins on the right side..

mg

100

g-1

DW

ratio

α-t

o β-

caro

tene

α-carotene ratio α-/β-carotene

.

CYP:CYP CYP:CYPInsCYPIns:CYPIns CYP:CYP CYP:CYPInsCYPIns:CYPIns

Supplemental Figure 8

Supplemental Data. Arango et al. Plant Cell (2014) 10.1105/tpc.113.122127

Supplemental Figure 8. Allelic effects of CYP97A3 polymorphism on α-carotene content and α-/β-carotene ratio.Allelic effect of the insertion/deletion polymorphism in CYP97A3 on α-carotene content (in mg 100g-1DW, left) and the α-/β-carotene ratio (right). The box plots show the quartile division with themedian indicated by lines in the boxes. The two whiskers correspond to the first and third quartile.

Supplemental Data. Arango et al. Plant Cell (2014) 10.1105/tpc.113.122127

Supplemental Table 1

Line tot car. lutein α-carotene β-carotene VAZN

CRC 554.7±46 256.1±23 (46.2) 27.3±6 (5.0) 103.6±11 (20.7) 155.7±10 (28.1)

QAL 557.6±3 241.9±3 (43.4) 2.6±0 (0.5) 137.3±1 (24.6) 175.8±7 (31.5)

Dc#1 549.3±41 252.4±34 (45.8) 6.3±4 (1.2) 124.0±15 (22.5) 166.6±8 (30.5)

Dc#7 528.9±32 242.1±23 (45.7) 6.1±4 (1.2) 123.0±14 (23.2) 157.7±1 (29.9)

Dc#25 551.6±56 267.1±41 (48.3) 5.0±3 (1.0) 125.9±9 (22.9) 153.6±12 (27.9)

Supplemental Table 1. Carotenoids in leaves from DJ3Spro:At-CYP97A3 carrots.

Chantenay Red Cored carrots (CRC) were transformed with At-CYP97A3 under control of the DJ3S promoter from yam (DJ3Spro). Leaf carotenoid

patterns of DJ3Spro:At-CYP97A3 lines (Dc#1, Dc#7 and Dc#25) were compared with those determined for white-rooted (QAL) and non-transformed CRC carrots. Carotenoids are expressed in mmol mol-1 chlorophyll. The relative percentage of each carotenoid is given in parentheses. tot car., total

carotenoids; VAZN, sum of violaxanthin, antheraxanthin, zeaxanthin and neoxanthin. Data represent the mean ± SD of four (controls) and three

(transgenic lines) biological replicates, respectively.

Supplemental Data. Arango et al. Plant Cell (2014) 10.1105/tpc.113.122127 Supplemental Table 2

carotenoid CRC-8 wks CRC4-12 wks CRC-16 wks Dc#1-8 wks Dc#1-12 wks Dc#1-16 wks Dc#25-8 wks Dc#25-12 wks Dc#25-16 wks Dc#7-12 wks

viola/neox 1 4.30±0.92 7.51±1.09 2.81±1.09 0.00±0.00 2.86±0.83 3.25 0.00±0.00 2.42±0.26 2.43±0.30 3.52±0.01

viola/neox 2 9.38±2.50 16.03±1.79 4.12±0.87 0.00±0.00 7.82±1.89 10.67 0.00±0.00 14.14±0.43 13.13±0.81 6.20±0.22

luteoxanthin 1 0.00±0.00 0.00±0.00 0.00±0.00 1.90±1.90 0.00±0.00 1.09 1.06±0.33 0.00±0.00 0.66±0.18 0.00±0.00

luteoxanthin 2 0.00±0.00 0.00±0.00 0.00±0.00 1.01±0.29 0.00±0.00 2.06 0.00±0.00 0.00±0.00 1.23±0.19 0.00±0.00

cryptochrome 1 0.00±0.00 0.00±0.00 0.00±0.00 4.16±1.20 0.00±0.00 0.00 5.81±0.52 0.00±0.00 0.00±0.00 0.00±0.00

cryptochrome 2 0.00±0.00 0.00±0.00 0.00±0.00 4.33±1.63 0.00±0.00 0.00 6.00±0.51 0.00±0.00 0.00±0.00 0.00±0.00

antheraxanthin 1.09±0.79 1.98±0.78 0.00±0.00 0.00±0.00 2.00±0.89 2.77 0.00±0.00 3.59±0.39 2.73±0.57 1.53±0.00

lutein 70.55±8.91 145.83±13.41 36.07±7.13 14.40±12.16 136.89±57.82 237.58 7.29±7.05 152.15±10.97 154.79±25.09 99.49±2.66

lutein-E1 0.00±0.00 0.00±0.00 0.00±0.00 30.61±3.83 0.00±0.00 0.00 31.13±3.30 0.00±0.00 0.00±0.00 1.68±0.35

lutein-E2 0.00±0.00 0.00±0.00 0.00±0.00 40.88±6.33 1.45±0.26 0.00 37.75±2.47 0.00±0.00 0.00±0.00 0.94±0.07

zeinoxanthin 0.00±0.00 0.00±0.00 0.00±0.00 6.81±0.06 0.00±0.00 0.00 7.05±0.87 0.00±0.00 0.00±0.00 0.00±0.00

β-cryptoxanthin 0.00±0.00 0.00±0.00 11.85±2.69 29.46±7.48 0.00±0.00 0.00 17.71±4.53 0.00±0.00 0.00±0.00 0.00±0.00

α-carotene 277.39±37.02 366.89±100.01 336.20±31.43 76.54±40.44 73.73±50.58 42.80 10.90±2.42 13.16±1.53 8.70±2.28 27.81±0.44

13-cis-β-carotene 13.06±3.78 0.00±0.00 18.77±5.1 65.53±24.46 12.56±1.06 0.00 21.77±2.14 10.47±0.38 0.00±0.00 13.19±1.79

trans-β-carotene 598.51±161.96 736.54±98.71 612.57±103.44 289.99±122.28 348.85±77.73 597.53 86.42±13.87 321.48±4.96 349.70±54.82 380.57±9.27

9-cis-β-carotene 0.00±0.00 0.00±0.00 0.00±0.00 34.74±7.27 0.00±0.00 0.00 11.62±2.65 0.00±0.00 0.00±0.00 0.00±0.00

phytoene 210.83±8.45 193.89±57.97 363.52±49.08 106.38±11.64 94.47±1.48 150.60 123.02±3.57 145.44±3.59 170.73±38.48 102.40±6.86

phytofluene 1 107.37±4.26 108.22±30.74 186.83±31.43 78.39±10.92 55.00±6.18 97.49 73.19±1.48 83.15±0.98 100.35±22.35 57.59±3.64

phytofluene 3 9.45±1.39 4.30±1.06 15.77±2.24 0.00±0.00 5.13±0.05 3.48 3.09±0.31 5.56±0.28 3.40±0.91 5.83±0.82

ζ-carotene 1 54.08±17.19 98.93±47.53 138.87±63.11 75.85±8.92 52.23±24.64 133.13 53.96±9.8 82.71±10.74 92.33±3.19 26.71±2.50

ζ-carotene 2 27.66±4.81 28.43±11.59 30.07±10.41 17.51±0.87 44.68±7.38 14.67 15.18±2.15 28.79±2.56 23.82±7.30 40.12±14.70

ζ-carotene 3 5.62±2.44 6.13±1.24 4.47±1.29 8.60±1.78 3.52±0.83 7.06 46.71±26.96 3.84±0.48 4.54±2.18 4.75±1.89

ζ-carotene 4 0.00±0.00 0.00±0.00 0.00±0.00 5.59±2.21 0.00±0.00 0.00 8.98±1.32 0.00±0.00 0.00±0.00 0.00±0.00

δ-carotene 3.41±0.35 6.03±2.15 3.82±0.65 0.00±0.00 1.15±0.12 0.00 0.46±0.65 0.00±0.00 0.00±0.00 0.00±0.00

γ-carotene 6.31±1.54 6.12±3.26 3.53±1.16 2.25±0.29 1.72±0.14 1.59 0.66±0.55 2.43±0.60 1.73±0.53 1.76±0.30

xanthophyll ester 0.00±0.00 1.57±1.37 0.00±0.00 6.26±0.34 2.88±1.39 10.34 3.29±0.72 1.87±0.07 5.58±2.45 0.96±0.23

total carotenoids 1399.01±237.39 1728.39±331.16 1769.29±241.22 901.20±260.64 846.92±131.24 1316.13 573.06±32.38 871.23±21.26 935.84±115.36 775.05±19.63

length 3.5±0.7 3.3±0.5 4.8±1.3 2±0.5 2.3±0.3 3.8 3.7±0.8 4.5±0.5 5.5±1.8 3.8±0.3

diameter 2.4±0.1 2.8±0.2 4.6±0.4 2.0±0 2.8±0.3 3.3 3.3±0.5 3.8±0.3 4.5±0.5 2.3±0.3

Supplemental Table 2. Detailed carotenoid amounts and root sizes in roots from At-CYP97A3-overexpressing CRC lines. Carotenoid amounts are given in µg g-1 DW, root length and diameter are given in cm. Numbers behind compound names indicate different isomers, E, ester. Data are the average ± SD of three biological replicates, except for Dc#1-16 weeks (1 sample).

Supplemental Data. Arango et al. Plant Cell (2014) 10.1105/tpc.113.122127

Supplemental Table 3

gene name primer/probe sequence (5‘ to 3‘)

Dc-PSY1 forward GAGCTAGTAGACGGGCCTAATGC

(DQ192186) reverse TTCAGCCTCTTCTCCCATCTG

probe TCCCATATCACGCCCAAGGCTCTTG

Dc-PSY2 forward ACTCCTATTTGTGCTTCATCGAAA

(DQ192187) reverse GAAGGTTACCGATAAATGGAGGAAT

probe ACATCCTTGCCCTCTTAATTTGCTTCTTCATGA

Dc-CYP97A3 forward TATAGTCCCCGCATTGCATCA

(JQ655297) reverse TGCGCACAAGCGATCTGT

probe TGTAACTGCAATGATAAGCATGTTTGGACAAGC

nos 3’UTR forward GAATTACAGGTGACCAGCTCGAA

reverse CGGCAACAGGATTCAATCTTAAG

probe TTCCCCGATCGTTCAAACATTT

Dc-NCED3 forward CGTTTCGGTGTATTGGACAAGTAT

(DQ192202) reverse GCAGAAACAGTCTGGAACCTCAA

Supplemental Table 3. Primers and probes used for qRT-PCR.

Probes for Dc-PSY1, Dc-PSY2, Dc-CYP97A3 and nos 3’UTR were labeled with 6FAM (5’) and TAMRA (3’); Dc-NCED3 expression was

determined using SYBR green. Detection of nos 3’UTR was used to determine At-CYP97A3 expression levels in DJ3Spro:At-CYP97A3 carrots as

this sequence is encoded by the nos terminator present in the transgene. GenBank accession numbers are given in brackets below gene names.

DOI 10.1105/tpc.113.122127; originally published online May 23, 2014;Plant Cell

Jacobo Arango, Matthieu Jourdan, Emmanuel Geoffriau, Peter Beyer and Ralf Welschin Orange Carrots

-Carotene and Total CarotenoidsαCarotene Hydroxylase Activity Determines the Levels of Both

 This information is current as of June 23, 2014

 

Supplemental Data http://www.plantcell.org/content/suppl/2014/04/18/tpc.113.122127.DC1.html

Permissions https://www.copyright.com/ccc/openurl.do?sid=pd_hw1532298X&issn=1532298X&WT.mc_id=pd_hw1532298X

eTOCs http://www.plantcell.org/cgi/alerts/ctmain

Sign up for eTOCs at:

CiteTrack Alerts http://www.plantcell.org/cgi/alerts/ctmain

Sign up for CiteTrack Alerts at:

Subscription Information http://www.aspb.org/publications/subscriptions.cfm

is available at:Plant Physiology and The Plant CellSubscription Information for

ADVANCING THE SCIENCE OF PLANT BIOLOGY © American Society of Plant Biologists