New Scientist 72(1008): 70-71 (8 July 1976)
Pathways strewn with plant pigments
New techniques of inducing the synthesis of a range of plant pigments provide a bright outlook
to a better understanding of the genetic and biosynthetic principles underlying plant pigmentation

Dr Brian Harrison and Dr Richard Stickland are researchers in the Department of Genetics at the John Innes Institute, Norwich

We recognise plant pigments most readily by the wide range of colours and patterns in the flowers of our gardens and hedgerows. In many plants the pigments can occur elsewhere but it is the coloured corollas that principally attract the attention of ourselves and pollinating insects. Most of these coloured end-products are substances known as anthocyanidins which are formed in higher plants from sugar (glucose) probably via the biosynthetic pathway illustrated in simplified form in Figure 1.

Plants may lack anthocyanidin pigment and this can be due to a total lack of the pathway to anthocyanidin, or to a block in a pre-existing pathway. In the first case none of the necessary enzymes is present. In the second case, however, a mutation in a single gene may mean that only one enzyme is missing in an otherwise complete sequence.

Gregor Mendel was the first person to detect the existence of such genetic blocks affecting flower colour, but it was Muriel Wheldale in 1914 who originally conceived the idea of studying the biochemical formation of pigments through the utilisation of such blocks. She was able to show that in Antirrhinum majus there was indeed a connection between the presence or absence of a pigment and a change in a single gene. In the 1930s Scott-Moncrieff, Lawrence, Price and Haldane at the John Innes Horticultural Institution (then at Merton Park) pushed this type of investigation still further and they achieved significant advances in this field. However, identification of the pigments was still incomplete. As so often in science, further progress had to wait for new technology. In this case it was the development of paper chromatography as a method of separating mixtures of chemicals. By identifying the chemical components in growing plants, it has been possible to elucidate the main steps by which many of these pigments are manufactured.

We have recently found that it is possible to feed pigment precursors to flowers either as pure chemicals or as homogenates of other flowers. Anthocyanidin is then formed only if the precursor occupies a position after the genetic block in the biochemical pathway. This has been a favourite trick in microbiology for a long time, but it is now possible to subject plants to similar analysis. For instance, we have identified a number of genetic blocks in Antirrhinum majus, the garden snapdragon (see Figure 1).

The first block, at A in the diagram, is controlled by the nivea gene: a plant homozygous for nivea (that is, one with a double dose of the recessive gene) has albino flowers, The gene incolorata is responsible for block B: flavones and aurones are found to be present but the pathway to anthocyanidins cannot be continued and the flowers are ivory and/or yellow. Blocks at C are controlled by pallida or, just in the corolla tube, by the delila gene. Flowers not blocked by one of these four genes can produce one of the two anthocyanidins found in the snapdragon: if the plant is homozygous for the recessive gene eosinea, pelargonidin (with one hydroxyl [OH) in the phenyl ring) is synthesised to give the pink flowered form. The dominant Eosinea gene enables an extra hydroxyl to be added to form the magenta cyanidin pigment.

We are now in a position to discuss some of the precursors. A dilute aqueous solution (0.1 per cent) of some flavanonols enables many acyanic flowers of particular Antirrhinum majus genotypes to complete the pathway and synthesise anthocyanidin. The precursor dihydroquercetin (which already has two OH groups [Figure 1]) will initiate a synthesis only of cyanidin (2 OHs) in an A. majus plant blocked at A and/or B, but dihydrokaempferol (only 1 OH) can initiate a synthesis of either pelargonidin (1 OH), if the genotype of the plant is homozygous eosinea, or cyanidin (2 OHs) if the plant has the dominant eosinea gene. Although pelargonidin and cyanidin are the only anthocyanidins synthesised normally in A. majus, if the precursor dihydromyricetin (3 OHs) is used then delphinidin (3 OHs) can be produced by the flower tissue. This pigment is common in some plants (for example, delphinium) but is completely foreign to the snapdragon.

The nature of four genes involved in pigment synthesis can now be easily determined in acyanic Antirrhinum majus flowers by the use of chemical precursors. The albino (double dose nivea gene) can be recognised by the absence of aurone (a yellow pigment) on the face area and the pure white appearance of the flowers. If an albino flower is fumed with ammonia no yellow coloration will appear: this indicates an absence of flavones. If the albino flowers are fed with a precursor that is intermediate in the pathway (namely a flavanone such as naringenin or eriodictyol) then if no pigment is produced, we can deduce that the pathway is blocked at B and/or C as well as at A. Again, if the flower is fed a flavanonol and no pigment is produced, then there is a block at C.

Rapid identification of genes

The delila gene restricts anthocyanidin production to the lobes of the flower, the corolla tube remaining acyanic; if pigment is only synthesised in the lobes of a fed flower then the delila gene must be present in the genotype in double dose. The eosinea gene that controls the synthesis of the pink pelargonidin, as distinct from the magenta cyanidin, can be assessed if dihydrokaempferol is given to the plant, for this will enable either of the two anthocyanidins to be synthesised according to the genotype. The pelargonidin or cyanidin can sometimes be identified by inspection but for certainty a chromatographic analysis readily identifies the pigment synthesised. Several unexpressed genes involved in pigment production in Antirrhinum majus can thus be identified within 24 hours, obviating the lengthy crossing programme previously needed for genotype analysis.


Formation of anthocyanidins can also be initiated in acyanic flowers if an homogenate prepared from flowers of another suitable genotype is given to the first plant. If an Antirrhinum majus is acyanic because of a block at A and the pathway is otherwise open, then an acyanic plant blocked at B should be able to provide the necessary natural precursor and initiate synthesis of anthocyanidin. This works. For example, an homogenate of macerated ivory (incolorota/incolorata) Antirrhinum flowers, containing all likely anthocyanidin precursors. given to an albino Antirrhinum (nivea/nivea) allows it to synthesise anthocyanidin. The reciprocal, as expected, does not work; there being no precursor (flavanonol) in the albino, no synthesis can be initiated. Thus, from two acyanic types a little "cannibalism" can elicit plant pigment formation.

Figure 1 Pathway to anthocyanidin synthesis. A, B and C are genetic blocks exemplified in Antirrhinum majus

Figure 2 Production of various anthocyanids by feeding homogenates to acyanic flowers

Figure 3 Flowers of white Primula obconica before and after feeding with dihydroquercetin

Figure 4 Flowers of Streptocarpus (Maasen's White) before and after imbibition of dihydroquercetin

Extracts from different genera can also provide natural precursors (Figure 2). An homogenate from white Primula obconica fed to ivory Antirrhinum majus flowers initiated anthocyanidin synthesis. An extract prepared from an homogenate of pink (eosinea/eosinea) Antirrhinum flowers enabled pelargonidin to be produced in an albino (but potentially pink) Antirrhinum; this same extract produced cyanidin in another albino, potentially magenta Antirrhinum, and, malvidin in a white Streptocarpus (Maassen's White).

The pink Antirrhinum majus extract presumably has a precursor similar or identical to dihydrokaempferol. In an acyanic plant (which would be capable of producing pelargonidin [1 OH] if the genetic block were removed) the pigment pelargonidin is synthesised from the homogenate; similarly, in a blocked plant otherwise capable of producing cyanidin [2 OHs] the flower can produce this pigment. Again, from the same extract the white Streptocarpus synthesises malvidin, an anthocyanidin found in many varieties of Streptocarpus, which has three OH groups, two of them methylated.

The precursor dihydroquercetin (DHQ) is readily soluble in water and is easily available. We have tested many different. flowering plants with DHQ, particularly those which were acyanic. We found that pigmented flowers often synthesised extra pigment or developed it in acyanic parts of the flower. Failure of a particular species to respond to the imbibition of DHQ does not necessarily mean that there is no pathway available but that the particular plant tested had a genetic block later than flavanonol production; other genetic variants could be responsive. Although DHQ initiated pigment production in a wide range of plants other precursors could probably initiate synthesis when DHQ failed.

Addition of selected precursor can be used as a method of screening plant populations for genetic variants in pigment production. We did a preliminary survey of white bluebells and found that there is a variation in response to DHQ. The "whites" could therefore originate from different genetic causes: that is, mutations in different genes could produce dissimilar blocks; some could contain a late block and be unresponsive, while others might have an early block and be able to synthesise pigment. Variants in Silene populations have also been identified.

These new methods have opened the way for advances in this field, and should provide a fresh stimulus for the biochemical and genetic analysis of plant pigments.