The Genetics of Flower Pigments in Daylilies
This paper is written in order to get hybridizers thinking about what might be happening in their flowers and to present a proposal for researchers to falsify. Researchers have finally gotten to the point that they understand the pathway, which leads to the production of anthocyanins. This understanding may help us to better understand the genetics of the flower colors of Hemerocallis. I fully expect present or future research will indicate that some of the presented information is incorrect but it will serve to get people thinking.
Some definitions of genetic terminology:
Enzyme: A protein made from a gene, which functions in mediating (catalyzing) specific chemical reaction in plants and animals. Each different enzyme is dependent on a specific gene for its production. If four enzymes are needed to change a compound from one form to another, then at least four different genes must be responsible for the change (some enzymes are actually composed of two proteins).
Phenotype: General appearance of the flower or plant, such as leaf color, height, width, hardiness, or flower color, size, tepal shape, etc. Some phenotypes, such as leaf color (green or yellow), are often produced by a single pair of genes. Multiple genes or the interaction of genes cause variations in many other characteristics such as height, flower color, size, shape, degree of ruffling, etc.
Genotype: The genetic makeup of the plant producing a particular phenotype or appearance. The genotype is indicated by the actual genes involved not the external characteristic.
Genes: For our purposes, a gene is a portion of DNA on the chromosome, which is responsible for determining the makeup of a protein. Many of these proteins act as enzymes, which may be responsible for determining which chemical reactions occur in the plant. The strands of DNA are at a particular location (locus) in the chromosome.
Alleles: Genes located on the same chromosome, in the same position and coding for the same or only slightly different protein/enzyme are considered to be alleles.
Dominant: Usually a strand of DNA, which is responsible for the production of an enzyme, which has a noticeable effect (influences phenotype). In daylilies1, yellow appears to be dominate to melon, drab flowers appear to be dominate to clear or not drab.
Recessive: Usually a strand of DNA, which is responsible for the production of a protein (enzyme), which has no noticeable effect (no change in phenotype). We often think of this as a non-functional portion of DNA. In daylilies, likewise, the allele for melon colored flowers would be recessive to the allele for yellow. In other words, the allele for melon is not changing lycopene, a reddish pigment, into beta-carotine to produce a yellow colored flower.
Homozygous: Referring to a genotype composed of either all dominant or all recessive alleles.
Heterozygous: Referring to a genotype with one or more dominant alleles and one or more recessive alleles.
Incomplete (Partial) Dominance: A situation where a single strand of functional DNA is responsible for the production of an enzyme, but that amount of enzyme is not sufficient to produce the maximum change in the characteristic.
Codominance: In diploids both alleles produce functional, but different, enzymes. In the heterozygous condition, the phenotype will show both characteristics.
Epistasis: A situation where one gene influences the action of a different gene. This commonly occurs when one gene is responsible for the production of a compound required for another gene/enzyme to function.
Multiple Gene Inheritance: Many different genes influence the phenotype of a particular characteristic such as height, flower color, flower size, etc.
Pleiotropy: When a single gene influences more than one phenotype, for example a gene responsible for chlorophyll production will influence the color of the leaves and the amount the plant grows.
Modifiers: Genes or possibly minerals (commonly iron and magnesium) which influence or change the flower color. The yellow genotype may be changed due to modifiers to near white, pale yellow, light yellow, medium yellow, gold and orange.
Gene Mutations: A change in the DNA of a gene. Gene mutations cause a change in the DNA, which causes 1) a non-functional gene to become functional, 2) a functional gene to change and function for some other purpose or 3) causes an function gene to produce a non-functional protein/enzyme (by far the most common). These gene mutations might be situations where parts of the DNA (bases) are deleted, duplicated, or moved in one way or another.
Chromosomal Mutations: The rearrangement of whole genes, where an entire gene section of a DNA chain is moved from one region of the chromosome to another, or even to another chromosome entirely is called a chromosomal mutation. When the DNA binding site of a regulatory gene moves to a new position, it can then control a new gene. These sorts of changes might lead to enzymes being produced at new or different times. This may influence the shapes of flowers (bigger, rounder, ruffled) or to new color combinations and patterns (white flowers, eyes and picotee edges, braided edges, etc.). In general, mutations occur naturally but may be increased by various environmental factors such as certain chemicals, UV light and radiation.
Dominant Mutations: A change in the DNA which produces an enzyme which is functional. Dominant mutations are very rare, but are sometimes very important in introducing brand new characters into the population. If a mutation occurs in the base pairs of a gene, the chances of any base pair being eliminated, replaced, doubled or changed in any way AND produce a working gene/enzyme is slim.
Recessive Mutations: A change in the DNA, which causes the protein ultimately produced from that DNA to no longer function correctly; for example a gene responsible for a pigment might be changed so that the pigment is no longer produced. Well over ninety-five percent of all mutations are recessive
The genetics of diploid daylilies has been examined to a small degree. I suspect that little has been done with tetraploid genetics for reasons that should be obvious. The genes are the same but it would take much longer to determine their function. We can assume that the diploid study results will hold for tetraploid daylilies, as well. The study by Norton and Arisumi2, while old, gives us a starting point. While they were not concerned with such characters as eye color or size and ruffling and throat color, their study of petal color is of interest (Table 1.).
Flower pigments in the genus Hemerocallis appear to be due basically to four pigment groups.
Two groups are composed of chloroplast pigments. The green pigment usually associated with the throat, and sometimes the flower edge, is composed of chlorophyll molecules located on the membranes within the chloroplasts. Chloroplasts also contain other pigments. Yellow or orange pigments associated with the chloroplast are called carotenoids. While found in chloroplasts, these pigments become more noticeable when the chlorophyll breaks down giving the chloroplast, or what would now be called chromoplasts, a yellow, orange or reddish color. Each step in the production3 is mediated or regulated by one or more enzymes (Figure 1.). The production of these enzymes is determined or specified by DNA of a gene. If a particular allele of a gene fails to produce a correctly functioning enzyme, then the reaction stops. For example, if lycopene is present and the correct enzyme is present, then lycopene might be changed to beta-carotene. If both the alleles responsible for the production of the particular cyclase enzyme are recessive (non-functional) then lycopene will build up in amounts and not be changed to beta-carotene. When one allele is functional and the other is not, then some of the cyclase enzyme will be produced and beta-carotene will be found in the cells. If both alleles produce a functional enzyme then there would be more functional cyclase enzyme present with the potential of producing an even greater concentration of beta-carotene.
Bisset, in his 1976 dissertation proposes a number of daylily genotypes and possible phenotypes as summarized by Norton and Arisumi. A few of these will be discussed here now that we know more about the chemical pathway leading to the carotenoids (See Figure 1.). His first was that melon (yy) flowers we recessive to yellow (Yy or YY) flowers. This makes perfect sense because lycopene (the reddish orange pigment in tomatoes) would give the chloroplasts in the flower parts a reddish cast. Hence, their Y gene might be the gene for the cyclase enzyme, which changes lycopene to beta- carotene4. The elimination of some or all of the yellow pigment, along with a build up of lycopene would produce a melon color. A non-functional cyclase enzyme gene would be recessive.
Modifier genes, which influence the phenotypes produced by other genes, might effect the colors by making them less sharp. These might effect the drabness or muddiness of the flower color. The most common explanation for this is the presence of carotenoids as well as anthocyanin in the cells5. It is well know that daylilies such as Great White and Clarification seem to cause their offspring to be sharper in color. This could easily be explained if these daylilies were homozygous recessive for that character. We still do not know what particular compound is influencing this characteristic but we suspect that carotenoids in the flower tend to make the tepals drab or muddy. If this is true, the lack the production of the colorful carotenoids is probably due to a non-functioning enzyme somewhere prior to the production of lycopene in the carotenoid pathway. Thus clearness (lack of carotenoid) would most likely be due to a recessive gene. While two genes might very well be involved, there is probably not enough evidence, at this point, to hypothesis clarity regulation by more that one gene.
Carotenoid Pathway
Figure 1.
Table 1: Daylily Genetics - Genes as suggested by Joanne Norton and Toru Arisumi
Y – yellow y - melon
P – pink or lavender
p – not pink or lavender
R – red
r – not red
IP – pink influencing modification IL – lavender-influencing modification
D – drab
d – clear, not drab
M – muddy (same as D?)
m – not muddy (same as d?)
The other most important two pigments found in daylilies are water-soluble anthocyanins. Anthocyanins are located in the cell sap, not in the cytoplasm or plastids. Cyanidin and Delphinidin have been found in Hemerocallis. The colors of both will vary depending on the pH or acidity of the cell sap. In a test tube they both tend to be red under acid conditions and blue under alkaline conditions. Cyanidin, in daylilies tends to be red in color while Delphinidin appears purple or violet. Some daylilies have only Cyanindin, some have only Delphinidin and some have both. The enzyme responsible for changing Dihydrokaempferol to Pelargonidin is not found in the genus Hemerocallis.
Daylily Pigments - Anthocyanins: Cyanidin, Delphinidin
Daylily Co-pigments
Daylily Flavones: Luteolin, Tricetin
Daylily Flavonols: Quercitin, Myricetin
The general pathway from Chalcone to Cyanidin and Delphinidin is shown in Figure 2.
Enzymes are responsible for each change of one compound to another compound. The DNA of a gene determines the structure and therefore, function of each enzyme. The enzymes and compounds involved the flavonoid pathway6 are indicated by abbreviations.
Key to identification of the chemical compounds:
Flavonoids: THC - tetrahydroxychalcone; NAR - naringenin; ERI – eriodictyol; PHF – pentahydroxyflavanone; LU – luteolin; TR – tricetin; DHK – dihydrokaempferol; DHQ – dihydroquercetin; DHM – dihydromyricetin; KM – kaempferol; QU quercetin; MY – myricetin; CY3g – cyaninidin 3-glycoside; PG3g – pelargonidin 3-glycoside; DP3g – delphinidin 3-glycoside
Enzymes: CHI – chalcone isomerase; FHT – flavanone 3 ß-hydroxylase; FNS I/II – flavanone synthase I and II; F3’H – flavonoid 3’ hydroxylase; F3’5’H – flavonoid 3’5’H; DFR – dihydroflavonol 4-reductase; FLS flavonol synthase; ANS - anthocyanin synthase; GT – glycosyl transferase.
The important things to notice in this somewhat confusing diagram is that starting with the compound chalcone there are three possible anthocyanins that may be produced: 3 MPA, 3,4 DPA and 3,4,5 TPA. Only the last two, 3,4 DPA and 3,4,5 TPA are found in daylilies. Many of the same enzymes are involved in the production of more than one compound. For example, if enzyme non-functional ANS were to be produced from a mutated gene, all the anthocyanins would be effected and none would be produced. Daylily enthusiasts are probably most interested in the two enzymes: F3’H and F3’5’H. If F3’H is present, cyanidin will be produced. If F3’5’H is functioning the delphinidin will be made. As one might expect, if both enzymes are present then both pigments may be produced.
When examining the flavonoid pathway as presented by Martens one might logically ask why pelargonidin has not been found in daylilies. Eder7 states that without the activity of either F3’H or F3’5’H then pelargonidin may be produced. It might be possible that if both F3’H and F3’5’H were non-functional in a daylily then pelargonidin might be produced instead of the normal pigments.
Using the proposal as given in Bisset, one possible enzyme responsible for the production of Cyanidin could be considered to be the Red (R) gene. The R gene would be responsible for the production of an enzyme Flavonoid 3’hydroxylase (F3’H), which changes mostly colorless DHK to Cyanidin.
Remember that by saying that the enzyme is functioning we are saying that the gene that produced the enzyme, at least one of the alleles, is dominant. A recessive is one that produces a non-functioning enzyme. If the genes show incomplete dominance, like I suspect many daylily genes do, the more functional genes present the more color will be present.
In daylilies that have only delphinidin the most likely pathway to delphinidin would be through DHK. In this case F3’5’H must be present but F3’H would have to be absent. If F3’H was present, then cyanidin would also be produced. If DHQ were produce from ERI or DHK then cyanidin would be produced.
In cyanidin only flowers, F3’5’H must be lacking and F3’H must be present. This would prevent DHM from being made and, therefore, there would be no delphinidin produced.
If both colored pigments were lacking then there are a number of options depending on which flavonones or flavonols are present.
Table 2: Suggested genotypes for the production of flower color as suggested by Joanne Norton and Toru Arisumi
Melon – yy pp rr (lycopene and no anthocyanins)
Yellow – Y_ pp rr (beta-carotene and no anthocyanins)
Clear Pink – yy P_ IP_ dd rr (lycopene and delphinidin and co-dominant for pink)?????????
(Could be Y_ or yy)
Muddy Pink – yy P_ IP _ D- rr
Peach, apricot, copper - Y_ P_ IP_ dd rr
Duff, Tan, Brown - Y_P_IP_D_ rr
Most important daylily flavonoid color genes:
Dom. Gene Enzyme End product Rec. Gene R F3’H red color(cyanidin) r P F3’5’H purple color (delphinidin) p E FHT flavones production e L FLS flavonol production l
According to this pathway. Anthocyanin pigments and their associated flavonols can be produced by only particular combinations of dominant alleles
Delphinidin (no Quercitin) P_ E_ rr L_ or P_ E_ R_ ll
Delphinidin and Quercitin P_ E_ R_ L_
Delphinidin (no Cyanidin) P_ E_ rr
Cyanidin (no Delphinidin) pp E_ R_
Delphinidin and Cyanidin P_ E_ R_
Other possibilities
White (flavones and flavonols present) = enzymes DFR, ANS or 3GT are non-functional
White (both quercetin, kaempferol and myrcetin absent)= enzyme FHT non-functional
White (flavones and flavonols absent) = enzyme CHI or earlier is non-functioning
Further variations may affect either of the anthocyanins because of the different number or kinds of sugars, which might be attached. Further work must be done to determine the presence of the nearly colorless to yellow flavones and flavanols, which are present in the different flowers. Notice that some of the enzymes may be used to do the same thing but on different compounds.
Joanne Norton and Toru Arisumi suggest in Bisset’s paper that whether a flower becomes pink or lavender is dependent on a gene possessing co-dominant alleles and the presence of the P gene for "lavender". One allele would produce pink flowers and the other allele will produce lavender. This would mean that the presence of some other compound is modifying the color of delphinidin, possibly by complexing or combining with it. The particular compound involved is unknown at this point. It might be as simple as a dominant/recessive flavone combining with delphinidin and causing a change in color. The enzymes FNS I or FNS II might be involved in such a situation.
Much discussion has occurred about delphinidin and why it doesn’t show a blue color in the flower like it does in a test tube at the same pH. This, most likely, again, is due to modifiers influencing the color. These could be particular compounds in the cell sap, which are preventing the blue color or a flavonol or flavone that is in to low a concentration to have much of an influence in the blue color. The compound Quercitin is often stated to by an important co-pigment with delphinidin to produce a blue color. Kasha (???) states that the way to get a blue colored daylily is to cross a delphinidin only daylily with a white daylily high in flavonols and flavones. If quercitin is important in the complex, to impart a blue color quercetin must be present. As DHQ is only produced along with cyanidin, a delphinidin only daylily will never be blue.
| It is interesting to note that
particular daylilies like Stamile’s "Got the
Blues" and Modavan’s "Piece of the
Sky" have a definite blue color to the eye. What is
different about the daylily eye? What must be done to
transfer the blue color to the rest of the flower? Is it
purely a matter of pigment concentration or is there more
to it. Note, every daylily produced so far which has a
definite blue eye produces both cyanidin and delphinidin
flavonoids .We already have delphinidin and a blue color in part of some of the daylilies. Getting a blue daylily will be different from getting a blue rose. With roses it was a simple matter of transferring a gene for delphinidin into the rose. That produced a blue color. With daylilies, we are dealing with compounds, which modify the color of delphinidin. |
 |
It has also been proposed that particular minerals such as magnesium, iron, calcium and possibly aluminum may influence the color by complexing with the anthocyanins, thus allowing them to give off a more blue color. This would not be surprising but I doubt that the solution is quite so simple. I suspect that with daylilies growing in so many soils today somebody would have already solved the problem.
1. Norton, J. 1972. Hemerocallis Journal 26 (3) in Bisset, K. 1976. Spectrophotometry, Chromatography and Genetics of Hemerocallis Pigments. Dissertation, Florida State Univ.
2. Norton, J. 1972. Hemerocallis Journal 26 (3) in Bisset, K. 1976. Spectrophotometry, Chromatography and Genetics of Hemerocallis Pigments. Dissertation, Florida State Univ.
3. Francis X. Cunningham, Jr. 2002. Regulation of carotenoid synthesis and accumulation in plants. Pure Appl. Chem., Vol. 74, No. 8, pp. 1409–1417.
4. Assuming that lutein (orange pigment in marigolds) may also occur in daylilies, the gene (Y) could be involved in the pathway from lycopene to lutein, as well as, or instead of the pathway to beta-carotene.
5. Kasha, Michael. 1978. New Insights for the Amateur and Professional Hemerocallis Hybridizer. Presented at American Hemerocallis Society National Meeting, Pittsburgh, PA July 13-15, 1978.
6. Martens, S. et.al. 2003. Biochemical Engineering Journal 14: 227-235.
7. Eder, Christian. 2001. Cloning and characterization of the flavonoid 3’-hydroxylase and the flavonoid 3’5’-hydroxylase. PhD. Dissertation. Technische Universität of München, Germany.