The Plant Cell, Vol.

7 , 1027-1038, July 1995 O 1995 American Society of Plant Physiologists

Plant Carotenoids: Pigments for Photoprotection, Visual Attraction, and Human Health
Glenn E. Bartley and Pablo A. Scolnik
DuPont Experimental Station, Route 141 and Rising Sun Lane, Wilmington, Delaware 19880-0402

INTRODUCTION Plant carotenoids are red, orange, and yellow lipid-solublepigments found embedded in the membranes of chloroplasts and chromoplasts. Their color is masked by chlorophyll in photosynthetic tissues, but in late stages of plant development these pigments contributeto the bright colors of many flowers and fruits and the carrot root. Carotenoids protect photosynthetic organisms against potentially harmful photooxidative processes and are essential structuralcomponents of the photosynthetic antenna and reaction center complexes. In plants, some of these compounds are precursors of abscisic acid (ABA), a phytohormone that modulates developmental and stress processes (Koornneef, 1986). Carotenoids with provitamin A activity are essential components of the human diet, and there is considerable evidence that many carotenoids have anti-cancer activity. Carotenoids absorbed through the diet, and often metabolized into other compounds, are responsible for the color of familiar animals such as lobster, flamingo, and fish. Often unaware of the chemical nature of food colorants, humans consume large amounts of carotenoids added to foods primarily to please the eye (Klai and Bauernfeind, 1981). As with other natural pigments, carotenoids attracted the attention of 19th-centuryorganic chemists. BCarotene was isolated in 1817, and the pigments of autumn leaves were identified as xanthophyllsin 1837 (Isler, 1971). Classical geneticistsidentified carotenoid mutants in maize, tomato, and Arabidopsis during the early part of this century. However, difficulties with the biochemical properties of the enzymes of the pathway and the generally poor viability of mutants affected in carotenoid biosynthesisduring early stages of developmentconspired to delay development of the field. Genetic studies of carotenoid biosynthesis in bacteria (reviewed in Armstrong, 1994) have played a fundamental role in the cloning and characterization of plant carotenoid genes. Since 1991, when a bacterial probe was first used to clone a plant gene for an enzyme of the pathway, progress has been rapid. However, much remains to be learned about the environmental and developmental regulation of the pathway, and many classic mutants have not yet been characterized at the molecular level. Recent reviews (Armstrong, 1994; Bartley et al., 1994; Sandmann, 1994)and two books on methods (Packer, 1992a,
i

1992b)have covered many aspects of the chemistry, biochemistry, and biology of carotenoids in bacteria and plants, and the classic lsler (1971) monograph on carotenoidsis being updated (Britton et al., 1995). In this review, we summarize and update our knowledge of genes and enzymes for carotenoid biosynthesis in plants and focus on the current challenges for plant developmental biology and biotechnology.

THE CAROTENOID PATHWAY The structures of some common plant carotenoids and a simplified biosynthetic pathway are shown in Figure 1. Carotenoids are generally Ca terpenoid compounds formed by the condensation of eight isoprene units (for a review of terpenoid metabolism,see McGarvey and Croteau, 1995, this issue). At the center of the molecule, the linkage order is reversed, so the molecule as a whole is symmetrical. A set of conjugated double bonds is responsible for the absorption of light in the visible region of the spectrum. In the following sections we discuss the biosynthetic pathway, emphasizing those steps for which the corresponding genes have been cloned. The cloned plant genes are listed in Table 1. Genetic and molecular evidence indicate that all enzymes of the pathway are encoded by nuclear genes and imported post-translationally into chloroplasts. We mention genes from nonplant species only in those cases in which they served as probes for cloning plant genes or in which sequence comparisons revealed putative functions for the corresponding plant proteins.

Synthesis of Geranylgeranyl Pyrophosphate In the first step of the pathway, the enzyme geranylgeranyl pyrophosphatesynthase (GGPS) catalyzes three successive condensation reactions,the first of them being the condensation of isopentenyl pyrophosphate and dimethylallyl pyrophosphate (DMAPP) to form geranylgeranyl pyrophosphate(GGPP; see McGarvey and Croteau, 1995, this issue). GGPS from pepper has been purified to homogeneity and shown to catalyze the

To whom correspondence should be addressed.

1028

The Plant Cell

isopentenyl diphoephate dimethylallyl diphosphate geranyl diphosphate fame ey 1 diphosphate

x -^ ^-^^ ^-^^' %^^- *%-^ %^'^^>^ OH-^^+ violaxanthin ' '

><

<s

s^,-OH

C02H

Figure 1. Carotenoid Biosynthesis in Plants. (A) A simplified biosynthetic pathway. The colors indicate the prevalent color of the compounds in nature. Compounds before ^-carotene do not absorb light in the visible region of the spectrum. The color of compounds in nature does not necessarily correspond to the colors of purified compounds in solution due to interactions with other components of chromoplast membranes and to concentration effects. The first steps of the pathway are condensation reactions that result in the formation of geranylgeranyl diphosphate (geranylgeranyl pyrophosphate, GGPP). Phytoene synthase (PSY) catalyzes the condensation of two molecules of GGPP into prephytoene pyrophosphate (not shown) and then into phytoene. A series of desaturation reactions results in the synthesis of lycopene, which is then cyclized into (5-carotene. The circular arrow shows the interconversion of zeaxanthin and violaxanthin, the two main compounds of the xanthophyll cycle. The phytohormone ABA is derived from xanthophylls. (B) Four biological systems that have contributed to the development of the genetics of carotenoid biosynthesis in plants. Basic information provided by cyanobacteria (Synechocystis) and photosynthetic bacteria (Rhodobacter) was instrumental in identifying plant genes for carotenoid biosynthesis. Maize and soybean are two of the plants from which carotenoid biosynthesis genes have been cloned. (C) and (D) Plant carotenoid mutants affect plastid development. In the case of the tomato ghost mutant (C) and the Arabidopsis immutans mutant (D), a block in the desaturation of phytoene results in variegated green/white leaves. White sectors accumulate phytoene, and green sectors accumulate wild-type carotenoids.

Carotenoid Biosynthesis and Function

1029

Table 1. Cloned Plant Genes for Enzymes of the Carotenoid Biosynthesis Pathway

DNA Clone Typea Enzyme Geranylgeranyl pyrophosphate synthase Plant Gene Genomic cDNA dbestb

GenBank Accession Number

Arabidopsis Lupine Pepper

GGPS GGPSl GGPS GGPS GGPS PSY PSY PSY PSYl PSY PSY PSYl PSYl PSYl PSYl PS Y2 PS Y2 PDS Pds PDS PDS PDS PDS PDS PDS PDS PDS PDS

+ +

+ + + +

L22347 L25813 U 15778 X80267 P80042 L25812

-c

Phytoene synthase

Arabidopsis Maize Daffodil Pepper Rice Tomato

+ + + + + + + +
+

+ +

X78814 X68704 D48697 D48251 A21360 M84744 Y00521 X60441 L23424 X60440 LI6237 L39266 X78815 X68058 D23378 M64704 M88683 X59948 X71023 X78271 U 19262 L40176 22921O 229211 F13561 233983
-d
-d

Phytoene desaturase

Arabidopsis Maize Daffodil Pepper Rice Soybean Tomato

+
+

Wild tobacco Lycopene cyclase Arabidopsis

L YC LYC LYC LYC L YC

Tomato Tobacco Capsanthin-capsorubin synthase

a

LYC
LYC

+ + + + + + + + +
+ +

+ + +

ccs ccs ccs

X77289 X76165 X78030

+ indicates a sequence has been identified in the expressed sequence tags data base. B. Buckner (personal communication). J. Hirschberg (personal communication).
corresponding reactions in vitro (Dogbo and Camara, 1987). An antibody to the purified enzymewas used to isolate a cDNA, which was shown to encode a protein with GGPS enzymatic activity (Kuntz et al., 1992). An Arabidopsis GGPS cDNA was cloned using information derived from the pepper protein sequence (Scolnik and Bartley, 1994). GGPS proteins from different organisms, including plants, share conserved regions that are likely to correspond to catalytic centers (Carattoli et al., 1991; Chen and Poulter, 1994). In addition to its role in carotenoid biosynthesis, GGPP is a precursor of other plant compounds, such as chlorophyll phytol chains, phylloquinones, plastoquinones, taxol, tocopherols, abietic acid, gibberellic acid, diterpenoid phytoalexins, and diterpenes (see McGarvey and Croteau, 1995, this issue).

+ indicates the type of clone (genomic or cDNA) that has been isolated.

1030

The Plant Cell

Therefore, it is likely that the synthesis o f this branch compound is subject to complex regulation and, possibly, compartmentalization. There is evidence for multiple GGPS genes in Arabidopsis. In addition to the aforementioned cDNA, we have reported the DNA sequence of a genomic clone (GenBank accession number L22347)for which a cDNA has not yet been identified, two cDNAs for additional GGPS genes (G.E. Bartley and P.A. Scolnik, unpublished data), and a fourth cDNA that may encode a more divergent GGPS (GenBank accession number L40577; Scolnik and Bartley, 1995a). One possibility is that these genes encode dedicated enzymes for different branches of the pathway.

Desaturation of Phytoene
A series of desaturationreactionsconverts the colorless compound phytoene into yellow, orange, and red carotenoids by creating the conjugated double bonds that form the chromophore. There are four desaturation steps between phytoene and lycopene (Figure 1). A higher plant gene for phytoene desaturase (PDS) was cloned from soybean (Bartley et al., 1991) and subsequently from severa1 other species (Table 1). Expressionof plant PDS genes in bacteria leads to the formation of (carotene (Bartleyet al., 1991; Pecker et al., 1992).This result, plus the existenceof mutants of maize and tomato that accumulate (-carotene and of herbicides that primarily block the conversion of (-carotene into lycopene, suggests that, like some bacteria(Linden et al., 1993b; Armstrong, 1994), plants have two carotenoid desaturases. The gene for the plant (carotene desaturase, however, has not yet been cloned, and it is possible that interactions with other proteins andlor modifications of PDS alter the specificity of the enzyme to accept (-carotene as a substrate. Phytoene is the first lipophilic compound in the carotenoid pathway. Accordingly, whereas PSY is peripherally associated with plastid membranes (Bartley et al., 1992), PDS is an intrinsic membrane protein in both chloroplasts and chromoplasts (Bartley et al., 1991; Hugueney et al., 1992; Linden et al., 1993a). Bacterial, fungal, and plant PDS proteins contain a typical pyridine dinucleotide binding motif generally associated with the binding of FAD (Armstrong et al., 1990; Bartley et al., 1990), and this cofactor has been found to bind to purified pepper PDS (Hugueney et al., 1992). Other studies with purified protein have demonstrated that, depending on the species, either NAD(P) or FAD is sufficient to catalyze the desaturation reaction, supporting the idea of an enzymatic mechanism involving a dehydrogenationreaction with an oxidized dinucleotide as electron and hydrogen mediator (Sandmann, 1994). As in other redox reactions, the desaturation of phytoene requires ancillary redox proteins. The mode of action of the bleachingherbicide sulcotrione supportsthe concept that PDS is part of a multiproteinredox complex in which quinones play an important role. Plants treated with sulcotrione are blocked in the desaturation of phytoene, but this herbicide is not an inhibitor of PDS (Schulz et al., 1993). Further characterization of sulcotrione revealed that it is a potent inhibitor of the enzyme p-hydroxyphenylpyruvatedioxygenase, which converts p-hydroxyphenylpyruvateto homogentisic acid, an intermediate in the biosynthesis of plastoquinones, tocopherols, and related compounds. The most likely interpretation of this result is that sulcotrione inhibits PDS indirectly through a block in the biosynthesis of quinones required for PDS activity (Schulz et al., 1993). Supporting this biochemical evidence, two Arabidopsis phytoene-accumulatingmutants that map to genetic loci other than PDS were found blocked at two different steps of the plastoquinone biosynthesis pathway (D. DellaPenna, personal communication). In photosynthetically inactive daffodil (Narcissuspseudonarcissus) chromoplasts,O2appears to be the final acceptor for

Synthesis of Phytoene In the first dedicated step of carotenoid biosynthesis, the enzyme phytoene synthase (PSY) catalyzes the two-step conversion of two molecules of GGPP into prephytoene pyrophosphate(PPPP) and then into phytoene (Dogbo et al., 1988). The possible mechanism of this enzyme, including regions of substrate recognition, have been reviewed recently (Sandmann, 1994). Two molecules of GGPP are joined in a condensation reaction with the loss of hydrogen and the diphosphate group from C-lof the same molecule. Cleavage of the C-1 diphosphate group of the resulting PPPP, followed by a 1-1 rearrangement, results in the formation of phytoene, which, depending on the stereochemistryof the reaction, may be 15-cis or all-trans. Phytoene has the basic Ca skeleton of carotenoids, and all subsequent reactions in the pathway involve chemical conversions of this basic structure. Clone pTOM5, a cDNA for a tomato ripening-specificgene of unknown function (Ray et al., 1987), was later shown by genetic approaches to correspond to a PSY gene. The first evidence for this assignment was the observationthat the predicted protein encoded by pTOM5 is more than 20% identical to PSY enzymes encoded by the crtB genes of the carotenoidproducing bacteria Ewinia herbicola and Rhodobacter capsulatus (Armstrong et al., 1990). The function suggested by this structuralsimilarity was confirmed by studies in plants and bacteria. Fruit and flower carotenoids are greatly reduced in transgenic tomato plants carrying a pTOM5 antisense construct (Bird et al., 1991), and the block is in the conversion of GGPP to phytoene (Bramley et al., 1992). When driven by bacterial transcription and translation signals, the pTOM5 cDNA complements R. capsulatus crtB mutants (Bartley and Scolnik, 1992). A cDNA for a second PSY gene, now named PSY2, has been identified in tomato leaves (Bartley and Scolnik, 1993). The gene corresponding to the pTOM5 cDNA has been renamed PSY1. Transgenic tobacco plants expressing high levels of PSY7 show a reduction in size and in levels of gibberellic acid (GA) and chlorophyll, suggestingthat overproduction of PSY diverts GGPP from a pool common to the GA and phytol biosynthetic pathways (R. Fray, personal communication).

Carotenoid Biosynthesis and Function

1031

electrons derived from double-bondformation, and an NAD(P)Hdependent quinone oxidoreductase regulates PDS activity (Mayer et al., 1992). The gene for this oxidoreductase has been cloned, and the deduced sequence shows a high leve1of conservation with the 23-kD protein of plant oxygen-evolving complexes (I? Beyer, personal communication).

Cycllzation of Lycopene

have not yet been purified to homogeneity (Yamamoto, 1979; Gruszecki and Krupa, 1993; Hager and Holocher, 1994). The enzyme for the conversionof the 5,6spoxycarotenoids antheraxanthinand violaxanthin into the pepper carotenoids capsanthin and capsorubin has been purifiedto homogeneity, and antibodies raised against this enzyme were used to isolate a full-length cDNA. The monomeric molecular mass of this enzyme is 50 kD, and the primary structure reveals the presence of a consensus dinucleotide-bindingsite (Bouvier et al., 1994).

Lycopene, the product of the sequential desaturationsof phytoene, is converted into p-carotene by the enzyme lycopene cyclase (LYC). This enzyme can accept either of the acyclic compounds neurosporeneor lycopene as substrates to form cyclic carotenoids such as a- and p-carotene. The biochemistry of carotenoidcyclization has been reviewed by Sandmann (1991). Synthesis of p-carotenefrom lycopene by extracts from spinach chloroplasts or tomato chromoplasts requires FAD (Kushwahaet al., 1969). The cyanobacterial gene for LYC (crtL) has recently been cloned by genetic complementation from a mutant strain resistant to the LYC inhibitior 2-(4-methylphen0xy)triethylamine (Cunningham et al., 1994). A computer search of the data base of expressed sequence tags (dbest) revealed the existence of an Arabidopsis cDNA with extensive sequence conservationto the cyanobacterialgene (GenBank accession number L40176; Scolnik and Bartley, 1995b). The predicted protein product of this gene contains an apparent plastid transit peptide and a conserved pyridine dinucleotide binding domain at the N terminus of the mature protein. Expression of this cDNA in a lycopene-producing Escherichia coli strain confirms that it encodes a protein with LYC activity (Scolnik and Bartley, 1995b). The LYC gene has also been cloned from tobacco and tomato based on sequence similarity with the cyanobacterialcrtL gene (J. Hirschberg, personal communication).

Synthesis of Abscisic Acid

Synthesis of Xanthophylls

Xanthophylls are hydroxy, epoxy, furanoxy, and oxy derivatives of the carotenes formed in the late stages of the pathway. One of the many different plant xanthophylls, violaxanthin, is shown in Figure 1. The first xanthophylls are formed from cyclic carotenoids such as p-caroteneby the introduction of hydroxy groups in positions C3 and CYof the ionone rings, followed by epoxidation at the 5,6and 5:6' positions. Little is known about the mechanisms of the enzymes for xanthophyll biosynthesis, and the corresponding genes have not yet been cloned. Xanthophylls undergo light-dependentepoxidation/de-epoxidation cycles, represented in Figure 1 by the interconversion of zeaxanthin and violaxanthin.As discussed later, the xanthophyll cycle is likely to play an important role in photoprotection under conditions of excess excitation energy. The epoxidase and de-epoxidase enzymes responsible for the xanthophyll cycle have been detected in chloroplast fractions, but they

The plant hormone ABA participates in the control of water relations in plants and is a general inhibitor of growth and, in some species, seed dormancy. According to the currently accepted hypothesis, ABA forms from products of the oxidative cleavage of xanthophylls (Parry and Horgan, 1992; Parry et al., 1992; for review, se8 Zeevaart et al., 1991). Therefore, all steps involving C4 compounds are common to both carotenoids and ABA, whereas steps after xanthophyll cleavage are specific for ABA biosynthesis. Mutants in carotenoid biosynthetic steps up to zeaxanthin have low levels of ABA; they also have sharply reduced levels of chlorophyll, dueto the alterations in chloroplast differentiation induced by carotenoid deficiency. The connection between carotenoids and ABA was supported by the analysis of maize mutantswith reduced levels of colored carotenoids and ABA in seedlings and endosperms (Fong et al., 1983) and the observation that the mutant phenotype can be mimicked by application of the PDS inhibitorfluridone (Smith et al., 1983). BecauseABA is required for dormancy, these mutants often show vivipary, that is, premature germination of the embryo while it is still attached to the plant. The Arabidopsis aba mutant has greatly reduced levels of violaxanthin and neoxanthin and accumulates zeaxanthin at levels higher than wild type, thus contributing to the biochemical evidence suggesting that epoxycarotenoids are precursors of ABA (Rock and Zeevaart, 1991; Rock e t al., 1992). The tomato mutants flacca and sitiens, thought to carry genetic blocks at the leve1of ABA-aldehyde oxidation, have contributed to our understanding of the late stages of ABA biosynthesis (for review, see Zeevaart et al., 1991).

FUNCTIONS OF CAROTENOIDS

Plant carotenoids play fundamental roles as accessory pigments for photosynthesis, as protection against photooxidation, and as structuraldeterminants in plastid pigment-protein complexes. The role of these pigments is determined primarily by whether the tissue is photosyntheticor nonphotosynthetic. In photosynthetic tissues, photoprotection against harmful oxygen species is their most important function. In nonphotosynthetic

1032

The Plant Cell

tissues, carotenoids determine or contribute to the color of flowers and fruits.

Carotenoids in Chloroplasts Carotenoids, chlorophylls, and various apoproteins form the photosystem complexes of the chloroplast thylakoid membranes (Thornber et al., 1987; Herrin et al., 1992; see von Wettstein et al., 1995, this issue). Carotenoids absorb light in the blue region of the spectrum (400 to 600 nm), and the energy absorbed can be transferred to chlorophylls. Therefore, carotenoids serve as accessory pigments by harvesting radiant light in a region of the spectrum not covered by the chlorophylls. In addition to their light-harvesting functions, carotenoids are essential for photoprotection. In the absence of colored carotenoids, plants suffer severe photooxidativedamage, which generally results in the death of the organism. Based on work with photosynthetic bacteria and with in vitro assays containing porphyrins,the likely mechanismfor photoprotectionis the quenching, by colored carotenoids,of chlorophylltriplets that would otherwise lead to the generation of oxygen singlets that can react with lipids, proteins, and other macromolecules, causing irreparabledamage (Krinsky, 1979; Davidson and Cogdell, 1981). The light-harvesting and photoprotective functions of carotenoids have been reviewed by Koyama (1991). Unless treated with herbicides that inhibit carotenoid biosynthesis, plants are not likely to suffer severe depletion of carotenoids under field conditions.A more physiologicallyrelevant situation, particularly for plants growing in full sunlight, is when light levels exceed the maximum that can be used productively by the photosynthetic apparatus. There is considerable evidence in support of a photoprotectiverole of the xanthophyll cycle in the removal of excess excitation energy from the photosyntheticantennae (reviewed in Demmig-Adams and Adams, 1993). Under conditions of excess excitation energy, zeaxanthin accumulates from violaxanthin via a twostep enzymatic de-epoxidation, and this accumulation correlates with rapid nonphotochemical quenching of chlorophyll fluorescence. Recent studies on the xanthophyll cycle point to a complex light-mediated process that takes place in the chloroplast membranes. An analysis of the pigment stoichiometry of barley pigment-protein complexes revealed that the xanthophyll cycle intermediates occur mainly in the light-harvesting complexes of photosystem I and photosystem II, but the amounts of carotenoid per polypeptide are nonstoichiometric (lue-chih and Thornber, 1995). The sites of xanthophyll epoxidation and de-epoxidation within the thylakoids appear to be different (Gruszecki and Krupa, 1993), and recent results suggest that the de-epoxidase is mobile within the thylakoid lumen at the neutra1 pH values expected to occur in vivo in the dark but becomes attached to the membranes at the low pH values induced by strong illumination, thus gaining access to its substrate, violaxanthin (Hager and Holocher, 1994). A working hypothesis postulates that excess light captured by the

photosystem II antenna system triggers an acidification of the lumen that in turn activates the xanthophyll cycle de-epoxidase and may alter the conformation of antenna pigment-protein complexes to facilitate interaction with xanthophylls, resulting in nonphotochemical quenching(Gilmore et al., 1995). Clearly, much remains to be learned about the mechanistic aspects of the xanthophyll cycle, and, as pointed out by lue-chih et al. (1995), mutants lacking parts of the photosyntheticapparatus may help elucidate the arrangement of carotenoids within their protein complexes and their role in photoprotection.

Carotenoids in Chromoplasts Chromoplastsare carotenoid-containingplastids responsible for the yellow, orange, and red colors of many flower petals and fruits and of some roots. In the absence of chlorophyll, the photoprotective function of carotenoids is no longer essential, and it is likely that the main, and perhaps only, function of chromoplast carotenoids is the attraction of pollinating insects and animals. One possible exception is the proposed structural role of chromoplast carotenoids in the tapetum during pollen development (Brooks and Shaw, 1968). To provide the intense color of flowers and fruits, carotenoids have to accumulate in chromoplasts at very high levels. Chromoplasts contain specializedstructurescomposed of carotenoids, lipids, and proteins that function to sequester large amounts of carotenoids. These structures can be classified as globular, membranous, fibrillar, crystalline, and tubular (Emter et al., 1990). The isolation and identification of the individual components of these structures, together with the development of in vitro reconstitutionsystems, are required for understanding the molecular interactions governing their assembly. Severa1 carotenoid binding chromoplast proteins have been identified (Hansmann and Sitte, 1984; Bathgate et al., 1985; Cervantes-Cervantes et al., 1990; Smirra et al., 1993), but detailed structural information is still lacking. In addition to high levels of the most common apolar carotenoids and xanthophylls, chromoplastsaccumulate secondary carotenoids such as acylated xanthophylls that are thought to be required for the formation of chromoplast structures. A study with chemical inhibitors of protein and carotenoid biosythesis confirmed the essential role of these components in chromoplast formation (Emter et al., 1990). Changes in carotenoid esterification during pepper fruit ripening have been reported recently, and the fatty acids esterifying yellow and red xanthophylls have been characterized (Mnguez-Mosquera and Hornero-Mendez, 1994). Recently, Derure et al. (1994a) studied the fibril assembly process in bell pepper chromoplasts. The fibrils contain carotenoids, galactolipids, phospholipids, and fibrillin, a 32-kD carotenoidbinding protein.A cDNA for fibrillin was cloned and shown to encode a protein that is post-translationallyimported into plastids (Derureet al., 1994b). Fibrils were reconstituted in vitro by mixing purified fibrillin, carotenoids, and polar lipids in the same stoichiometric ratio found in vivo. Therefore, fibril

Carotenoid Biosynthesis and Function

1033

assembly is another example of cellular processes that prevent potential destabilizing or toxic effects by targeting excess lipids to storage structures (Deruere et al., 1994a).

REGULATION OF CAROTENOIDS

Carotenoid biosyntheticenzymes and carotenoid binding proteins are encoded by nuclear genes, and the corresponding precursor proteins are post-translationally imported into plastids. Therefore, transcriptional and post-transcriptional steps are potential targets for regulationof carotenoid biosynthesis. At present, some information is available on the steady state mRNA levels for genes encoding apoproteinsand some biosynthetic enzymes. The mRNA levels for tomato PSYl and PDS are highest in flowers and ripening fruits. Transcript levels are elevated in seedlings treated with herbicidesthat inhibit carotenoid biosynthesisand in leaves of the tomato ghosf mutant, which is blocked in phytoene desaturation. Therefore, expression of PSYl and PDS in tomato is controlled by both environmental and developmental signals (Maunders et al., 1987; Pecker et al., 1992; Giuliano et al., 1993). In contrast to PSY1, whose transcripts accumulate primarily in flowers and fruits, PSY2 transcripts accumulate in seedlings, leaves, and fruits, and the expression of this gene is not significantly affected by fruit ripening (Bartley and Scolnik, 1993). The existence of more than one'active PSY gene in tomato is consistent with results obtained with PSYl antisense tomato plants, which show a sharp decline in carotenoid levels in fruits but not in leaves (Bird et al., 1991; Bramley et al., 1992). In contrast to tornato, expression of PSY is not induced by ripening in pepper fruits (Romer et al., 1993). As indicated later, this result may point to differences in the Control of gene expression between climacteric and nonclimacteric fruits.

and are highly vacuolated. In addition to these unpigmented plastids, the white sectors of the immufans mutant also contain some small green plastids and even some normal chloroplasts, indicating that this nuclear mutation does not affect equally all plastids in white sectors (Wetzel et al., 1994). The nuclear-encoded functions affected in ghost and immufans offer the opportunity to explore unknown areas of nucleo-cytoplasmic interactions. Until these mutants are characterized at the molecular level, we can only speculate about the possible functions of the corresponding genes. It is important to consider that although these mutants accumulate phytoene, the primaryeffect may be unrelatedto carotenoid biosynthesis. For example, because phytoene synthase is a peripheral membrane protein but subsequent enzymes are part of a multiprotein integral membrane complex, disruptions in the development of plastid membranes may lead to an accumulation of phytoene. Alternatively, the ghosf or immutans gene product may regulate the expressionof PDS, the activity of PDS, or the assembly of complexes required for PDS activity. The basis for variegation is still unknown, but, as discussed by Wetzel et al. (1994), immufans (and by extensionghosf) are clear examples of the metabolic heterogeneityof cells within an organ.

Chromoplast Differentiation in Flowers, Fruits, and Endosperms

DEVELOPMENTAL BIOLOGY

Chloroplast Differentiation and Leaf Variegation

Loss of function mutations in genes critical for carotenoid biosynthesis generally result in conditional albino or pale green mutants that accumulate some chlorophyll only at very low light. Exceptions to this general rule are the tomato mutant ghost and the Arabidopsis mutant immufans, in which genetic blocks in the desaturation of phytoene result in variegated greenhhite leaf phenotypes. Severa1similaritiesin the phenotypes of these mutants suggest a possible common basis, although definite molecular evidence is still lacking. 60th ghost and immutans accumulate phytoene in their white leaf sectors but map genetically to loci other than PDS, suggesting that the structural gene is not affected (Giuliano et al., 1993; Wetzel et al., 1994). Plastids in the green sectors of both mutants are phenotypically wild type, whereas in white sectors they lack thylakoids

As discussed previously, carotenoids accumulate in chromoplasts of flowers and fruits, old leaves, and some roots. Color development in flowers, which is important for attracting insect pollinators, is part of the floral morphogenesisprogram. Flower carotenoids are generally yellow, but the combination of carotenoids and anthocyaninsmay result in a different final color. Nove1 colors of orchid flowers have been obtained by genetically altering the composition of carotenoids and anthocyanins (Griesbach, 1984; see also Holton and Cornish, 1995, this issue). During the development of tomato flowers, transcripts for PSY and PDS reach maximum levels right before anthesis. In organs of mature flowers, expression of both genes is maximal in petals and anthers (Giuliano et al., 1993). A PDS promoter fragment has been shown to confer chromoplastspecific expression to a reporter gene (G. Giuliano, personal communication). In a cucumber in vitro bud culture system, gibberellinAS (GA3)was shown to promote the rapid induction of both carotenoids and a carotenoid binding protein (Vainstein et al., 1994). Although transcript levels for carotenoid biosynthesis enzymes were not measured, this work opens the possibility that GA3 participates in the modulation of carotenoid biosynthesis and accumulation during flower development. Interestingly, GA3 also induces anthocyanin biosynthesis during peta1 development (Weiss et al., 1992), suggesting a possible link between the induction of carotenoids and anthocyanins during flower development.

1034

The Plant Cell

The developmental process of tomato fruit ripeninghas been well characterized genetically and biochemically. It involves tissue softening, conversion of chloroplasts into chromoplasts, and metabolic changes that lead to the accumulation of sugars and acids, and it results in fruits with appealing color and taste (Gillaspy et al., 1993). In contrast to mutationsthat affect cttloroplast development, genetic blocks in fruit chromoplast differentiation do not affect the viability of the organism and are therefore easy to propagate and analyze (Rick, 1991). Severa1 tomato mutants affected in chromoplast development during fruit ripening appear to result from defects in carotenoid r review se8 Bartleyet al., 1994). To date, only own as the R tocus) has been shown gene (PSY7;Fray and Grierson, 1993). As indicated previously, expression of this gene is induced more than 25-fold during ripening (Bartley and Scolnik, 1993; Giuliano et al., 1993). The molecular basis for the remaining mutants is not currently known. Ethylene promotes ripening in climacteric fruits such as tomato (Theologis, 1992). Expressionof PSYand other ripeningspecific genes was previously reported as induced by ethylene (Grierson, 1986), but more recent results indicatethat the control of expression is not at the leve1of transcription (reviewed in Bartley et al., 1994). Fruit ripening in pepper, unlike that in tomato, is a nonclimacteric process and is thus useful for elucidating the roles in ripening of signals other than ethylene. Pepper PSY in contrast with the corresponding tomato gene, is minimally induced by ripening (Romer et al., 1993). However, a cDNA clone (P541) for a gene that was shown to encode a plastid protein is expressed at high levels in late stages of pepper fruit ripening (HoulnB et al., 1994). The fruits of a mutant pepper Kme homozygous for the y locus are yellow due b the accujnulationof violaxanthin, a precursor of the red ketocarotenoidscapsanthin and capsorubin, and the P541 transcript is absent. A computer data base search (Bouvier et al., 1994; G.E. Bartley and PA. Scolnik, unpublished data) revealed that P541 encodes the capsanthin-capsorubin synthase reported by Derubre et al. (1994b), a possibility contemplated by Houln6 et al. (1994) based on the sequence of the protein and the timing of expression. Therefore, pepper fruit ripening involves the inductionof a gene encoding an enzyme that catalyzes the synthesis of the ketocarotenoids characteristic of pepper fruit. The expression of this gene is under genetic control by the y gene, which also controls the synthesis or accumulation of the carotenoid binding protein ChrA (OrenShamir et al., 1993). Thus, y may be a general regulator of keto-carotenoidcomplexes in pepper fruits. Although carotenoids accumulate during maize endosperm development, little work has been done on the developmental characterization of chromoplasts in this organism. Severa1carotenoid mutants of maize have been identified (Robertson, 1975), many of which exhibit vivipary due to the lack of ABA (Fong et al., 1983). The yellow-7 (Y7) locus was cloned by transposon tagging and shown to correspondto PSY(Buckner

et al., 1990, 1993). The vivipemus-5 (vp5) locus has been recently shown by genetic mapping to correspond to PDS (W. Hable and K. Oishi, unpublished data; GenBank accession number L39266; E. Wurtzel, personal communication).

BIOTECHNOLOGY
Some carotenoids are precursors of vitamin A, an essential component of the mammalian diet. In the Western diet there is no lack of pro-vitamin A compounds, but in some underdeveloped countries, particularly those whose diets are based primarily on rice, inadequate supplies of vitamin A can lead to night blindness and, in extreme cases, to xerophthalmia, an eye disease that primarily affects children (Toenniessen, 1991). The availability of cloned genes for carotenoid biosynthesis opens the possibility of genetically engineering rice to produce more p-carotene in the endosperm. A consortium of plant biologists sponsored by the Rockefeller Foundation is attempting to modify the carotene content of rice without altering the useful agronomical traits of this staple crop. Epidemiological studies have shown that carotenoid-rich diets are correlated with a significant reduction in the risk for lung cancer, other forms of cancer, coronary heart disease, and some degenerative diseases (Bendich, 1994). Cancer prevention is generally attributed to the antioxidant activities of carotenoids (see, for example, Seis et al., 1992). However, even in developed countries, dietary choices result in a 4.5mg difference betweenthe actual and recommendeddaily intake of p-carotene (Lachance, 1988). This deficiency can be compensated for by either taking vitamin supplements or ingesting more carotenoid-rich foods. Carotenoids used as vitamin supplements are currently extracted from plants or produced by organic synthesis (Pfander, 1992). More recently, a biotechnologyapproach has been used to engineer carotenoidoverproducing yeast strains, which can now be used as a source of natural carotenoids (Ausich, 1994). It is likely that the genetic manipulation of plant carotenoid biosynthesis genes could be used to increase the carotenoid content of widely consumed fruits and vegetables. In addition to the genetic engineering of plants with improved carotenoidcontent, genes for carotenoid biosynthesis can be used to generate plants resistant to herbicides that inhibit carotenoid biosynthesis. lntroduction of a recombinant DNA construct containingthe E. uredovora gene for PDS (Crfl)fused to a chloroplast transit peptide and under control of the cauliflower mosaic virus 3SS promoter into tobacco plants resulted in plants resistant to severa1 herbicidesthat target carotenoid biosynthesis and show altered metabolism of xanthophylls (Misawa et al., 1993,1994). Expressionof bacterial carotenoid biosynthesisgenes in plants may therefore extend the use of carotenoid inhibitors as herbicides for crops that are currently sensitive to these compounds.

Carotenoid Biosynthesis and Function

1035

CONCWSIONS
A knowledge of the biology of plant carotenoids is required both to understand such basic plant processes as photoprotection and plastid differentiation during the development of reproductive organs and to develop biotechnology approaches to improve the nutritional value of crops. Recent advances in the molecular biology of the carotenoid biosynthesis pathway in plants have been fueled by previous breakthroughs in the biochemistry and genetics of the pathway in carotenogenic bacteria. Although there are examples of purification of carotenoid biosynthetic enzymes to the degree of homogeneity requiredto identify the correspondinggenes, mostof the cloned plant genes were isolated either by using DNA probes of bacteria1 origin or by identifying entries in sequence data bases by searching with bacterial queries. Although we can anticipate that this approach will still yield interesting genes, many components of the pathway will not be accessible by this methodology. Thus, to isolate the remaining genes for carotenoid biosynthetic enzymes and ancillary proteins, including redox proteins and regulators, it is also necessary to pursue the purification of the corresponding proteins and the study of mutants disrupted at specific biosynthetic steps. In planning these studies, it will be important to take into account the fact that a holistic approach to the study of the biology of carotenoids in plants will require parallel examinations of the genes and proteins for biosynthetic enzymes, apoproteins, and redox components and will regard complexes, and not just carotenoids, as final products of the pathway. With the isolation of plant biosynthetic genes, it has been possible to study their regulation in response to environmental and developmental factors. Based on the few plant carotenoid genes that have been cloned, a picture has begun to emerge in which this process is divided into two major areas: biosynthesis in photosynthetic tissues, which appears to be under the control of environmental signals, primarily light, and biosynthesis in nonphotosynthetic tissues, which is controlled by yet-unidentified developmental signals. The genes for carotenoid binding proteins are also likely to be regulated by environmental signals in green tissues and by developmental signals in reproductive organs. In addition to understanding the control of gene expression, it is important to elucidate the molecular interactionsthat govern the assembly of carotenoid complexes. Carotenoids accumulate in plastids bound to specific proteins and in complexes that include other components, such as chlorophylls and various lipids. Although there is evidence for binding specificity, the basis for molecular recognition between carotenoids and carotenoid binding proteins is poorly understood. Combined biochemical and genetic approaches to understanding the mechanistic links among morphogenesis, plastid differentiation, carotenoid biosynthesis, and complex assembly are likely to dominate the next decade of research in this field.

ACKNOWLEDGMENTS

We apologize to many colleagues in the field who were not cited in this article due to space limitations and the need to reach a nonspecialized audience. We thank Steve Rodermel (University of lowa, Ames, IA) for providing the photograph of immutans shown in Figure 1. We are also indebted to those colleagues who communicated unpublished results and to Rebecca Chasan, Giwanni Giuliano, and Peter Beyer for critical reading of the manuscript.

REFERENCES

Armstrong, G.A. (1994). Eubacteria show their true colors: Genetics of carotenoid pigment biosynthesis from microbes to plants. J. Bacteriol. 176, 4795-4802. Armstrong, G.A., Albertl, M., and Hearst, J.E. (1990). Conserved enzymes mediate the early reactions of carotenoid biosynthesis in nonphotosyntheticand photosynthetic procaryotes. Proc. Natl. Acad. Sci. USA 67, 9975-9979. Ausich, R.L. (1994). Production of carotenoids by recombinant DNA technology. Pure Appl. Chem. 66, 1057-1068. Bartley, G.E., and Scolnlk, P.A. (1993). cDNA cloning, expression f PSY2, a second tomato during development and genome mapping o gene encoding phytoene synthase. J. Biol. Chem. 268,25718-25721. Bartley, G.E., Schmldhauser, T.J., Yanofsky, C., and Scolnlk, P.A. (1990). Carotenoid desaturases from Rhodobacter cepsulafus and Neurospora crassa are structurally and functionally conserved and contain domains homologous to flavoprotein disulfide oxidoreductases. J. Biol. Chem. 265, 16020-16024. Bartley, G.E., Viltanen, P.V., Pecker, I.,Chamovltz, D., Hlmchberg, J., and Scolnlk, P . A. (1991). Molecular cloning and expression in photosynthetic bacteria of a soybean cDNA coding for phytoene desaturase, an enzyme of the carotenoid biosynthesis pathway. Proc. Natl. Acad. Sci. USA 88, 6532-6536. Bartley, G.E., Vlltanen, P.V., Bacot, K.O., and Scolnlk, RA. (1992). A tomato gene expressed during fruit ripening encodes an enzyme of the carotenoid biosynthesis pathway. J. Biol. Chem. 267, 5036-5039. Bartley, G.E., Scolnlk, P.A., and Glullano, G. (1994). Molecular biology of carotenoid biosynthesis in plants. Annu. Rev. Plant Physiol. Plant MOI. Biol. 45, 287-301. Bathgate, B., Purton, M.E., Grlerson, D., and Goodenough, P.W. (1985). Plastid changes during the convemion of chloroplasts to chromoplasts in ripening tomatoes. Planta 165, 197-204. Bendich, A. (1994). Recent advances in clinical research involving carotenoids. Pure Appl. Chem. 66, 1017-1024. Blrd, C.R., Ray, J.A., Fletcher, J.D., Boniwell, J.M., Blrd, A.S., Teulleres, C , Blaln, I., Bramley, RM., and Schuch, W. (1991). Using antisense RNA to study gene function: lnhibition of carotenoid bio, 835-639. synthesis in transgenic tomatoes. BicfTechnology 9 Bouvler, F., Hugueney, P., dHarllngue, A., Kuntz, M., and Camara, B. (1994). Xanthophyll biosynthesis in chromoplasts: lsolation and molecular cloning of an enzyme catalyzing the convemion of 5,6-epoxycarotenoids into ketocarotenoid. Plant J. 8, 45-54.

1036

The Plant Cell

Bramley, P., Teulleres, C., Blaln, I.,Blrd, C., and Schuch, W. (1992). Biochemical characterization of transgenic tomato plants in which carotenoid synthesis has been inhibited through the expression of antisense RNA to pTOM5. Plant J. 2, 343-349. Brltton, G., LlaaenJensen,S., and Pfander, H.P., eds (1995). Carotenoids, Vols. I and II. (Basel: Birkheuser Verlag). Brooks, J., and Shaw, G. (1968). Chemical structure of the exine of pollen walls and a new function for carotenoids in nature. Nature 219, 532-533. Buckner, B., Kelson, T.L., and Robertson, D.S. (1990). Cloning of the y7 locus of maize, a gene involved in the biosynthesis of carotenoids. Plant Cell 2, 867-876. Buckner, B., San Mlguel, P., and Bennetzen, J.L. (1993). The Y7 gene codes for phytoene synthase. Maize Genet. Coop. Newslett. 67, 65. Carattoli, A., Romano, N., Ballarlo, P., Morelli, O., and Macino, G. (1991). The Neurospom crassa carotenoid biosyntheticgene (Albino 3)reveals highly conserved regions among prenyltransferases.J. Biol. Chem. 266, 5854-5859. Cewantes-Cewantes, M., Nadjeb, N., Newman, L.A., and Prke, C.A. (1990). ChrA 1 s a carotenoid-binding protein in chromoplasts of Capsicum annuum. Plant Physiol. 1990, 1241-1243. Chen, A., and Poulter, C.D. (1994). lsolation and characterization of idsA: The gene for the short chain isoprenyl diphosphate synthase from Methanobacterium thermoautotrvphicum. Arch. Biochem. Biophys. 314, 399-404. Cunningham, F.X., Sun, Z., Chamovitz, D., Hlrschberg, J., and Gantt, E. (1994). Molecular structure and enzymatic function of lycopene cyclase from the cyanobacterium Synechococcus sp strain PCC7942. Plant Cell 6, 1107-1121. Davldson, E., and Cogdell, R.J. (1981). Reconstitutionof carotenoids into the light-hamsting pigment-ptuteincomplex from the carotenoidless mutant of Rhodopseudomonas sphemides R26. Biochim. Biophys. Acta 635, 295-303. Demmig-Adams, B., and Adams, W. W., 111. (1993). The xanthophyll cycle. In Antioxidants in Higher Plants, R.G. Alscher and J.L. Hess, eds (London: CRC Press), pp. 91-110. Derubre, J., RBmer, S., d'Harllngue, A., Backhaus, R.A., Kuntz, M., and Camara, 8. (1994a). Fibril assembly and carotenoid overaccumulation in chromoplasts: A model for supramolecular lipoprotein structures. Plant Cell 6, 119-133. Deruhre, J., Bouvier, F., Steppuhn, J., Kleln, A., Camara, B., and Kuntz, M. (1994b). Structure and expression of two plant genes encoding chromoplast-specific proteins: Occurrence of partially spliced transcripts. Biochem. Biophys. Res. Commun. 199, 1144-1150. Dogbo, O., and Camara, B. (1987). Purification of isopentenyl pyrophosphateisomerase and geranylgeranyl pyrophosphatesynthase from Capsicum chromoplasts by affinity chromatography. Biochim. Biophys. Acta 920, 140-148. Dogbo, O., Laferrlere, A.D., d'Harlingue, A., and Camara, 8. (1988). Carotenoid biosynthesis: lsolation of a bifunctional enzyme catalyzing the synthesis of phytoene. Proc. Natl. Acad. Sci. USA85,7054-7058. Emter, O., Falk, H., and Sltte, P . (1990). Specific carotenoidsand proteins as prerequisites for chromoplast tubule formation. Protoplasma 157, 128-135. Fong, F., Smlth, J.D., and Koehler, D.E. (1983). Early events in maize seed development: l-Methyl-3-phenyl-5-(3[trifluoromethyl]phenyl)4-(1H)-pyridinone induction of vivipary. Plant Physiol. 73, 899-901.

Fray, R.G., and Grlerson, D. (1993). identification and genetic analysis of normal and mutant phytoene synthase genes of tomato by sequencing, complementation and cesuppression. Plant MOI.Biol. 22, 589-602. Glllaspy, G., Ben-Davld, H., and Grulssem, W. (1993). Fruits: A developmental perspective. Plant Cell 5, 1439-1451. Gilmore, A.M., Hazlett, T.L., and Govlndjee(1995). Xanthophyll cycledependent quenching of photosystem II chlorophyll afluorescence: Formation of a quenching complex with short fluorescence lifetime. Proc. Natl. Acad. Sci. USA 92, 2273-2277. Glullano, G., Bartley, G.E., and Scolnlk, P.A. (1993). Regulation of carotenoid biosynthesis during tomato development. Plant Cell 5,

379-387.
Grlerson, D. (1986). Gene expression in ripening tomato fruit. CRC Crit. Rev. Plant Sci. 3, 113-132. Griesbach, R.J. (1984). Effects of carotenoid-anthocyanin combinations on flower color. J. Hered. 75, 145-147. Gruswckl, W., and Krupa, Z. (1993). LHCII, the major light-harvesting proteincomplex, is a zeaxanthin epoxidase. Biochim. Biophys. Acta 1144, 97-101. Hager, A., and Holocher, K. (1994). Localization of the xanthophyll cycle enzyme violaxanthin de-epoxidase within the thylakoid lumen and abolition of its mobility by a(light-dependent) pH decrease. Planta 192, 581-589. Hansmann, P., and Sltte, P. (1984). Comparison of polypeptide complement of different plastid types and mitochondria of Narcissus pseudonarcissus. 2. Naturforsch. 39, 758-766. Herrln, D.L., Battey, J.F., Greer, K., and Schmldt, G.W. (1992). Regulationof chlorophyllapoprotein expression and accumulation. J. Biol. Chem. 267, 8260-8269. Holton, T.A., and Cornlsh, E.C. (1995). Genetics and biochemistry of anthocyanin biosynthesis. Plant Cell 7, 1071-1083. Houlnb, G., Schantz, M.-L., Meyer, B., Pozueta-Romero, J., and Schantz, R. (1994). A chromoplast-specific protein in Capsicum anuum: Characterization and expression of the corresponding gene. Curr. Genet. 26, 524-527. Hugueney, P., RBmer, S., Kuntz, M., and Camara, 8. (1992). Characterization and molecular cloning of a flavoprotein catalyzing the synthesis of phytoflueneand L-carotene in Capsicum chromoplasts. Eur. J. Biochem. 209, 399-407. Isier, O., ed (1971). Carotenoids. (Basel: Birkhaiiser Verlag). lue-chlh, A., and Thornber, J.P. (1995). Analysis of the pigment stoichiometry of pigment-proteincomplexes from barley (Hordeum vulgare). Plant Physiol. 107, 565-574. Klaiii, H., and Bauernfelnd, J.C. (1981). Carotenoids as food color. In Carotenoids as Colorants and Vitamin A Precursors, J.C. Bauernfeind, ed (New York: Academic Press), pp. 48-317. Koornneet, M. (1986). Genetic aspects of abscisic acid. In A Genetic Approach to Plant Biochemistry, A.D. Blonstein and P.J. King, eds (New York: Springer-Verlag), pp. 35-54. Koyama, Y. (1991). Structures and functions of carotenoids in photosynthetic systems. J. Photochem. Photobiol. B 9, 265-280. Krlnsky, N.I. (1979). Catutenoid protection against phdooxidation. Pure Appl. Chem. 51, 649-660. Kuntz, M., RBmer, S., Suire, C., Hugueney, P., Well, J.H., Schantz, R., and Camara, B. (1992). ldentification of a cDNA for the plastidlocated geranylgeranyl pyrophosphate synthase from Capsicum

Carotenoid Biosynthesis and Function

1037

annum: Correlative increase in enzyme activity and transcript leve1 during fruit ripening. Plant J. 2, 25-34.

Kushwaha, S.C., Subbarayan, C., Beeler, D.A., and Porter, J.W. (1969). The conversion of lycopene-15-153H to cyclic carotenesby soluble extracts of higher plant plastids. J. Biol. Chem. 244, 3635-3642. Lachance, P . (1988). Dietary intake of carotenes and the carotene gap. Clin. Nutr. 7 , 118-122. Linden, H., Lucas, M.M., Rosario de Felipe, M., and Sandmann, G. (1993a). lmmunogold iocalization of phytoene desaturase in higher plant chloroplasts. Physiol. Plant. 88, 229-236. Linden, H., Vloque, A., and Sandmann, G. (1993b). lsolation of a carotenoid biosynthesis gene coding for C-carotene desaturase from Anabaena PCC 7120 by heterologous complementation. FEMS Microbiol. Lett. 106, 99-104. Maunders, M.J., Holdsworth, M.J., Slater, A., Knapp, J.E., Bird, C.R., Schuch, W., and Grierson, D. (1987). Ethylene stimulates the accumulationof ripening-relatedmRNAs in tomatoes. Plant Cell Environ. 10, 177-184. Mayer, M.P., Nievelstein, V., and Beyer, P . (1992). Purification and characterizationof a NADPH dependem oxidoreductase from chromoplastsof Narcissuspseudonarcissus:A redox-mediator possibly involved in carotene desaturation. Plant Physiol. Biochem. 30, 389-398. McGatvey,D.J., and Croteau, R. (1995). Terpenoid metabolism. Plant Cell 7, 1015-1026. Mnguez-Mosquera, M., and Hornero-M6ndez, D. (1994). Changes in carotenoid esterification during the fruit ripening of Capsicum annuum Cv. Bola. J. Agric. Food Chem. 42, 640-644. Mlsawa, N., Yamano, S., Llnden, H., de Felipe, M.R., Lucas, M., Ikenaga, H., and Sandmann, G. (1993). Functional expression of the Erwinie uredovora carotenoid biosynthesis gene crtl in transgenic plants showing an increase of p-carotene biosynthesisactivity and resistance to the bleaching herbicide norflurazon. Plant J. 4, 833-840. Misawa, N., Masamoto, K., Hori, T., de Lorenzo, C., Chtani, T., Boger, P., and Sandmann, G. (1994). Expression of an Erwinia phytoene desaturase gene confers not only multiple resistance to herbicides interfering with carotenoid biosynthesis but also alters xanthophyll metabolism in transgenic plants. Plant J. 4, 481-490. Oren-Shamir, M., Hadjeb, N., Newman, L.A., and Price, C A . (1993). Occurrence of the chromoplast protein ChrA correlates with a fruitcolor gene in Capsicum annuum. Plant MOI. Biol. 21, 549-554. Packer, L., ed (1992a). Carotenoids, part A. Methods Enzymol. 213, 1-538. Packer, L., ed (1992b). Carotenoids, part 6. Methods Enzymol. 214, 1-468. Parry, A.D., and Horgan, R. (1992). Abscisic acid biosynthesis in roots. I. The identification of potential abscisic acid precursors, and other carotenoids. Planta 187, 185-191. Parry, A.D., Grlffiths, A., and Horgan, R. (1992). Abscisic acid biosynthesis in roots. II. The effects of water-stress in wild-type and abscisic-acid-deficient mutant (notabilis) plant of Lycopersicon esculentum Mill. Planta 187, 192-197. pecker, I.,Chamcnrltr, O., Llnden, H.,Sandmann, G., and Hirschberg, J. (1992). A single polypeptide catalyzing the conversion of phytoene to c-carotene is transcriptionally regulatedduring tomato fruit ripening. Proc. Natl. Acad. Sci. USA 89, 4962-4966.

Pfander, H. (1992). Carotenoids. Methods. Enzymol. 213, 3-13. Ray, J., Bird, C.R., Maunders, M., Grierson, D., and Schuch, W. (1987). Sequence of pTOM5, a ripening related cDNA from tomato. Nucleic Acids Res. 24, 10587. Rick, CM. (1991). Tomato paste: A concentrated review of genetic highlights from the beginnings to the advent of molecular genetics. Genetics 128, 1-5. Robertson, D.S. (1975). Survey of the albino and white-endosperm mutants of maize. J. Hered. 66, 67-74. Rock, CD., and Zeevaart, J.A.D. (1991). The aba mutant o f Arabidopss thaliana is impaired in epoxy-carotenoid biosynthesis. Proc. Natl. Acad. Sci. USA 88, 7496-7499. Rock, C.D., Bowlby, N.R., Hoftmann-Bennlng, S., and Zeevart, J.A.D. (1992). The ABA mutant of Arabidopsis thaliana L. Heynh. has reduced chlorophyll fluorescence yields and reduced thylakoid stacking. Plant Physiol. 100, 1796-1801. Romer, S., Hugueney, P., Bouvler, F., Camara, B., and Kuntz, M. (1993). Expression of the genes encoding the early carotenoid biosynthetic enzymes in Capsicum annuum. Biochem. Biophys. Res. Commun. 196, 1414-1421. Sandmann, G. (1991). Biosynthesisof cyclic carotenoids: Biochemistry and molecular genetics of the reaction sequence. Physiol. Plant. 83, 186-193. Sandmann, G. (1994). Carotenoid biosynthesis in microorganisms and plants. Eur. J. Biochem. 223, 7-24. Schulz, A., Ort, O., Beyer, I?,and Klelning, H. (1993). SC-0051, a 2-benzoylcyclohexane-1,3-dione bleaching herbicide, is a potent inhibitor of the enzyme p-hydroxyphenylpyruvate dioxygenase. FEBS Lett. 318, 162-166. Scolnik, P.A., and Bartley, G.E. (1994). Nucleotide sequence of an Arabidopsis cDNA for geranylgeranyl pyrophosphatesynthase. Plant Physiol. 104, 1469-1470. Scolnik, P.A., and Bartley, G.E. (1995a). Nucleotide sequence of a putative geranylgeranyl pyrophosphatesynthase (GenBank L40577) from Arabidopsis. Plant Physiol. 108, 1342. Scolnik, P.A., and Bartley, G.E. (1995b). Nucleotide sequence of lycopene cyclase (GenBank L40176) from Arabidopsis. Plant Physiol. 108, 1342. Seis, H., Stahl, W., and Sundquist, A.R. (1992). Antioxidant functions of vitamins. Vitamins E and C, beta-carotene, and other carotenoids. Ann. NY Acad. Sci. 669, 7-20. Smirra, I.,Halevy, A.H., and Valnsteln, A. (1993). lsolation and characterization of a chromoplast-specificcarotenoid-associated protein from Cucumis sativus corollas. Plant Physiol. 102, 491-496. Smith, J.D., Fong, F., Magill, C.W., and Herlick, S. (1983). Fluridoneinduced phenocopies of vp-5 in Zea mays seed. Genetics 104 (suppl.), 66. Theologis, A. (1992). One rotten apple spoils the whole bushel: The role of ethylene in fruit ripening. Cell 70, 181-184. Thornber, J.P., Peter, G.F., and Nechustai, R. (1987). Biochemical composition and structure of photosyntheticpigment-proteinsfrom higher plants. Physiol. Plant. 7l, 236-240. Toenniessen, G.H. (1991). Potentially useful genes for rice genetic engineering. In Rice Biotechnology, G.S. Khush and G.H. Toenniessen, eds (Wallingford, UK CAB International), pp. 253-280. Vainstein, A., Halevy, A.H., Smirra, I.,and Vlshnevetsky, M. (1994). Chromoplast biogenesis in Cucumic sativus corollas. Plant Physiol. 104, 321-326.

1038

The Plant Cell

von Wettsteln, D., Gough, S., and Kannangara, C.G. (1995). Chlorophyll biosynthesis. Plant Cell 7 , 1039-1057. Welss, D., van Blokland, R., Kooter, J.M., MOI, J.N.M., and van Tlrnen, A.J.M. (1992). Gibberellic acid regulates chalcone synthase gene transcription in the corolla of Petunia hybrida. Plant Physiol. 98, 191-197. Wetzel, CM., Jlang, C.Z., Meehan, L.J., Voytas, D.F., and Rodermel, S.R. (1994). Nuclear-organelleinteractions:The immutens variegation

mutant of Arabidopsis is plastid autonomousand impaired in carotenoid biosynthesis. Plant J. 6, 161-175.

Yamamoto, H.Y. (1979). Biochemistry of the violaxanthin cycle in higher plants. Pure Appl. Chem. 51, 639-648. Zeevaart, J.A.D., Rock, C.D., Fanteuuo, F., Heath, T.G., and Gege, D.A. (1991). Metabolism of ABA and its physiological implications. In Abscisic Acid, W.J. Davies and H.G. Jones, eds (Oxford: Bios Scientific), pp. 39-51.