OrchidSafari Presents

Orchid Flower Color

Marilyn H.S. Light
Copyright 2003

Wednesday April 9, 2003

(Prechat handouts by Marilyn Light and Rob Griesbach follw the discussion)

MarilyninOttawa
Orchid flowers are colored by various pigments which are either evenly or regionally distributed in the flower parts. The pigments may be roughly placed in two groups.

The flavonoids (anthocyanins and co-pigments) are water-soluble pigments found in the cell vacuoles. Genes controlling the production of flavonoids are found in the nucleus. Anthocyanin and co-pigments produce the red-purple-blue range and their shade can be affected by the pH in the vacuoles. The production and concentration can be influenced by temperature and other growing conditions.

Here we see the concentration of purple anthocyanin/co-pigment-containing cells beneath a tuft of epidermal papillae (hairs) on the sepal of Masd. naranjapatae.

Here is a part of the flower of Masd. naranjapatae

You should be able to relate the microscopic section to the warts on the sepaline surface.

Sometimes, anthocyanins produce yellowish color as we see in Coelogyne speciosa. picture

When we take a closer look at a section cut through the lip, we see that the pale yellow pigmentation is restricted to the surface cells only. picture

Here we see a microscopic view of a section of the sepaline tail of Masdevallia triangularis. Reddish purple flavonoid pigment is unevenly distributed toward the outer surface of the tail.

The other main group are the carotenoid pigments, oil-soluble substances found in the plastids which are tiny structures found within the cell. Plastids contain genetic material which is inherited separately to that in the nucleus. Plastid genetic code is inherited with the egg cell and separately to nuclear genes which can be acquired from either the seed or pollen parent. Carotenoid pigments range in color from yellow through orange into red.

Here we see a purple-pigmented cell within a layer of yellow plastid-containing cells. This is a section of the sepal of Masd. triangularis.

Here is the flower of the same Masd. triangularis used to produce the microscopic sections. Note that the color is basically orange to our eyes. picture

You will note that what we perceive and what is there on the microscopic level can be quite different.

Floral pigments can be found in both the cell sap and in the plastids. In other words, you can have purple pigment in the cell sap and yellow-pigmented plastids. To our eyes such flowers appear orange, red or brown.

Even seemingly dark red flowers such as this Masd. whiteana are revealing when we examine the flowers under the microscope. picture

Here we see the flavonoid-containing cells are restricted to the flower surface. Yellow plastid-containing cells lie beneath. picture

Ed_in_Sat
Marilyn is this common to all orchids or specific to Masdevalia?

MarilyninOttawa
It varies Ed but flavonoids are generally in the surface layers.

Glenis
Astonishing pictures!

MarilyninOttawa
Thank you. You really have to look inside to understand (a bit) what is going on.

Chloroplasts, which are a specialized type of plastid, produce the pigment chlorophyll which is green. Since green pigmentation can become more or less intense in response to light levels, we can influence green-flowered orchids by varying the quantity of light.

Flower color has been of horticultural interest for centuries but the process of color inheritance and expression in orchids is complex and still under investigation. Factors affecting color expression and lack thereof in Cattleya were reported by Hurst in 1925.

Growers of pure white-flowered Cattleya species had observed that the progeny of two alba species did not always produce the expected alba flowers but self-pollination of albas almost always resulted in white-flowered offspring. They wondered why.

Careful breeding experiments suggested that there were two genes necessary for the production of purple-colored flowers. If one or the other gene was absent, pigment production was impaired and the flowers would be white. But, if the parents each lacked a different gene, pigment production was not impaired and the progeny would have purple flowers.

Alba forms of Cattleya species segregate into two groups based on which gene is impaired.

 Group I

C. gaskelliana,
intermedia,
labiata,
loddigesii,
lueddemanniana,
mossiae,
skinneri,
speciosissima, and
warneri

 Group II

C. eldorado,
harrisoniana,
mendelii,
percevaliana,
schroederae,
warneri and
warscewiczii

Cattleya trianiae has alba clones in both groups therefore, when breeding for albas with this species, it helps to know which group beforehand.

Floriculture has prompted interest in the inheritance of flower color in other orchids including Cymbidium and Dendrobium. Following what was known with Cattleya, investigators have looked at the inheritance of semi-alba and alba in Dendrobium.

Kamemoto and Amore (1990) pointed out that a big challenge has been the time taken to raise large numbers of seedlings to maturity, the cost of maintaining these plants to maturity, and the difficulty of working with complex hybrids. Nevertheless, they were able to report that with Den. affine, dicuphum and phalaenopsis, the expression of purple color is controlled similarly to what we see in Cattleya. The major difference reported was with the inheritance of the semi-alba trait which in Cattleya is a recessive trait and in Dendrobium, is controlled by a dominant gene.

Ed_in_Sat
We ran line breeding trials with purple and white Den phalaenopsis and found cross breeding the lines did not interrupt the pure strains of either color - pure whites and pure purples showed up either way. I think we went down to F9 or so. Work by Richard Mizuta confirmed this.

MarilyninOttawa
So-called albino Cymbidiums lack red color in otherwise white, yellow or green flowers. Australian, Alvin Bryant, began a breeding program in the 50's which led to the production of pure color standard and miniature Cymbidiums. He used alba forms such as Cym. Durham Castle, bred in England to develop the initial breeding pool. These plants lack the ability to produce red pigment. Pure color flowers do not fade prematurely if the anther cap is dislodged which is a valuable trait for both show and commercial purposes.

N_Calif_Kathy
It must have taken decades to establish that program

MarilyninOttawa
My feelings exactly. Ten years per generation at least.

Temperature can affect the production and accumulation of anthocyanin pigments. We can see this with red-colored Phalaenopsis and Sarcochilus, where flower buds exposed to different temperatures can produce differently colored flowers, even on the same plant or on the same inflorescence. Nancy Mountford recently showed a picture of a Phal with two spikes of flowers. The spikes had developed out of phase under different temperature conditions. The flowers were distinctly different. I have seen the same arise with red Sarcochilus where the flower buds exposed to warm temperatures produced pink mottled flowers where they normally would be cherry red.

Ed_in_Sat
It works exactly as you describe with Rhynchostylis gigantea and its hybrids.

Glenis
It can be noticed when you bring a plant indoors with only half of it open too! Catts and Odonts and Phals - I've noticed

MarilyninOttawa
The higher the temperature and the longer the exposure to the critical temperature, the paler the flower color. The cooler the growing condition, the more intense the color. A grower would have to experiment with their particular plants to learn just what is the critical temperature, length of exposure, and at which stage of development, flower color could be affected.

A study with roses suggests that transient high temperature may not have an effect but several days of sustained high temperature could reduce flower color and affect quality (Dela et al, 2003).

We used to breed tomato-red Sophronitis coccinea with purple Cattleyas to get redder Sophrocattleya/Sophrolaeliocattleya. A more recent approach has been to marry yellow/orange Cattleya with purple-colored forms to get red. And there has been considerable success.

When we breed for specific colors, we should take into account the relative value of using a particular parent to get seed. One classic approach to get red is to breed a yellow carotenoid-pigmented type with a mauve-purple anthocyanin-pigmented kind.

N_Calif_Kathy
Would an example of marrying a yellow/orange to a purple be something like Goldenzelle with a purple?

Ed_in_Sat
My suspicion is Goldenzelle will prove to be more of a color intensifier than modifier. Not enough experience to tell at this point - just a feeling from too few samples. Nigrella was a great intensifier, you may recall.

MarilyninOttawa
Blc. Goldezelle is a fairly complex hybrid with lots of C. dowiana and trianiae in the parentage. The more complex the parent, the more difficult the prediction. I would use the Goldenzelle as the seed parent.

Oncidium Vera Arthurs has a bright yellow lip and mottled petals and sepals where the liver-colored spots are derived from a combination of purple red flavonoid and yellow carotenoid. picture

When Onc. Vera Arthurs was crossed with mauve-flowered Onc. ornithorhynchum, the hope was for the coveted red bloom. picture

Note that the callus in each parent flower is bright yellow and differently pigmented to the yellow petals.

The additive, intensifying or selective effects are more difficult to predict when blending already complex hybrids. Once a particular parent demonstrates that it tends to produce particular effects in the progeny then a breeder can predict outcomes more accurately.

Ed_in_Sat
AMEN!!!!!!!!!!!!

MarilyninOttawa
Onc. Remembering Carmen (Vera Arthurs x ornithorhynchum) has provided a range of colors from concolor bordeaux red to red-purple blotches on a white ground. One might hypothesize that flavonoid production has somehow been disrupted where we see the white color. In the sepals and petals, all yellow pigmentation has been suppressed except where the flavonoid pigment is also present in the tissue. In the pictured specimen, we can see intensified purple spotting on the otherwise white lip. The callus remains yellow. picture

I have remade Onc. Vera Arthurs with three different Onc. gheisbrechtianum and have three different plant types, flowering times and flower color patterns! Now I plan to compare the second generation with each of these as a seed parent. Stay tuned!

N_Calif_Kathy
And you'll probably get your least favorite being the best parent! LOL!!

LosChristabel
Kinda stands to reason, sometimes, doesn't it? P. Mahalo 'Carmela' took an FCC, but its sibling has proven superior for breeding purposes. Expression isn't always a clue to how its genes will recombine.

jim4eq
LOL, and wait years to find out what tendencies it has.

LosChristabel
The track record can be predicted fairly, based on breeding trends. I have seen seedlings from Blc. Goldenzelle.

jim4eq
I meant for the first time breeders, not the established ones.

MarilyninOttawa
We may admire the wide range of flower color and pattern we see but we are merely bystanders as far as an orchid is concerned.

Orchid colors and patterns have evolved to attract pollinators. Bees especially, see color differently to humans. To a bee's eyes, yellow flowers float like bright spots in a sea of dark, almost invisible foliage. Once the bee gets close to the bloom, there may be lines of contrasting color (nectar guides) to guide the bee to the nectar reward.

There may be ultra-violet light-reflecting dots around and on the staminode of certain Paphiopedilum species. These dots, which are located between the warts visible to our eyes, may suggest the presence of aphids to predatory hover (syrphid) flies looking for a place to lay their eggs.

Whatever the visual cue, if it attracts the attention of potential pollinator and pollination happens, then the investment in flower color and pattern has served its procreative purpose.

So you see, the experiments needed to predict color expression and pattern in orchid flowers takes time. Even with molecular techniques, we are still dealing with populations of individuals each with their particular traits and inheritance patterns. Once breeders locate a breeder with proven potential, they treasure it.

Ed_in_Sat
It takes a representative blooming, too. Many times the best trait will appear very early or very late. We need to bloom 100 or so of any cross if we can.

MarilyninOttawa
I agree Ed. I have found that the splash-petaled nature of C. aquinii turns up in the seedlings that bloom in the middle of the batch. The later ones were all concolor. The earlier had a lip splash. This of course has more to do with co-inherited traits than simply flower color pattern.

LosChristabel
Gene linkage?

MarilyninOttawa
Possibly, probably but what I have seen is more than a time course of splash petal expression. It is a complex of growth rate, flowering season, etc.

MarilyninOttawa
What can we do to manage color within our orchid collections? Well, for those of us that treasure pink, red and other flavonoid shades, we can manage the temperature to produce the most intense pigmentation. Of course, there are also nutritional components influencing plant health and behaviour so it is not only temperature that counts.

The two messages of this discussion are: 1 - pay careful attention to temperature 2 - use or look for carotenoid-pigmented (yellow) seed parents when wanting to identify likely carriers of yellow genes.

jim4eq
I know some of the equitants in Jamaica will bloom pink when growing on the limestone, then bloom white in culture. (dang it!!)

Susan-from-Oregon
Does it need a little chalk in it's medium?

N_Calif_Kathy
You mentioned the pH of the vacuoles earlier in the presentation, Marilyn. Is that subject to our efforts to change pH? Like Hydrangeas?

MarilyninOttawa
pH - Most of us know how we can influence the inflorescence color of Hydrangea. Blue/pink depending on whether you use acidic compost or not. Although the exact chemical composition of the pigments may be different between Cattleya and Hydrangea, the same process probably holds true. However, it may be more difficult to manipulate the cell sap pH of Cattleya.

Ed_in_Sat
Aluminum sulphate isn't very reliable with Catts - perhaps because the application date is tricky to figure. Kodachrome film for the catalog shots is cheaper.

LosChristabel
I was told, by someone I respect highly, that a researcher managed to manipulate the sap of a 'blue' cattleya. The result was that the flower became an amazing bright REAL blue, just before the flower collapsed, due to cellular breakdown. So I suppose it might be possible.

MarilyninOttawa
pH - I feel that if it was easy to manipulate color in Cattleya it would have been done. It likely is not easy because the cell pH needed to produce the color shift is not tolerated by the orchid cell. These are living tissues with buffering mechanisms to control wild shift of pH. Still, if someone is able to bioengineer a plant to have a more amenable pigment then maybe it is possible.

LosChristabel
Thanks to modern science, we have glow-in-the-dark mice. I expect that anything is possible.

barbara_in_no._CA
Marilyn, will controlling the temperature apply to all types of orchids or just certain kind?

MarilyninOttawa
I expect that any flavonoid pigmented orchid in the Vandaceous group might respond similarly to temperature. This goes for Sarcochilus, Phalaenopsis, Doritis/Doritaenopsis, Aerides, Rhynchostylis (thanks to Ed). Where we have species, and hybrids with flavonoid-colored or influenced blooms (C., SLC, etc), we can expect that warm temperatures may negatively influence pigmentation.

LosChristabel
Seems to cut across the spectrum, then. It is a major factor in roses.

Jade
That explains lots of questions I had about color variations in Phals. when summer was cooler.

Ed_in_Sat
Throw in a shot of philippinensis - they'll all open the same color.

N_Calif_Kathy
Marilyn speaking a bit about genetics and alba forms... I know that pelorism vs splash petal is caused by the same gene, its just a matter of when its turned on as to whether you get a splash of color or full pelorism. (Griesbach) Um...A fellow from the Orchid Digest fell in love with a splash petal phal-type Dendrobium at the last Oakland society show. I take it pelorism isn't common in Dends? I see it in phals, Cyms and Catts....

MarilyninOttawa
There are marked petals and there is pelorism. There is a line of Dendrobium phalaenopsis type (DE Bush) comes to mind. These have three rounded petals. Pelorism is likely a very late genetic expression in the flower bud. Flowers tend to develop quite late in orchids.

Ed_in_Sat
We can also induce peloric forms chemically with fair reliability. In the past, we have put peloric and non-peloric flowers on the same Phal stem.

N_Calif_Kathy
Ed, I'll ask about chemical induced pelorism (both on the same inflorescence) What is used and when /how do you apply it? At Santa Barbara Norman's Orchids was showing 2 peloric phals that I never thought had pelorism in their parentage. Someone commented that it was a process of mericloning. No one could answer if the pelorism would continue in future generations. I just went to Norman's webpage hoping he'd have them listed there... no luck. Otherwise I'd show you the ones I'm talking about.

Ed_in_Sat
The technique is based upon BAP: Benzylamio purine family - very like 'keiki grow'. There are some aspects to use that we consider trade secrets. I'll think a bit on whether we want to make them public. If you want to play with this use a lanolin carrier.

N_Calif_Kathy
Well, I don't want you to divulge secrets! I'm just glad to know not to be impressed if I see this again, since it can be chemically induced.

Ed_in_Sat
Be impressed. The induction rate isn't all that good.

jim4eq
I found a peloric equitant that happened in the cloning for Carmela, reblooms peloric too. Onc. Midas 'Carmela' is the plant. They can grow to be a huge plant, too.

MarilyninOttawa
Back to pelorism and any related developmental events in flower formation. Orchid flowers mature shortly before they open. The processes are quite sensitive to events that could interfere with developmental switches. Water even, can exert an effect. temperature and light also. In Catasetum, the development of the female flower is decided a few weeks before opening and is likely influenced by one or several coincident events. Pelorism where all flowers on a spike are equally affected is likely a stable anomaly.

MarilyninOttawa
I will put in a word for all those anonymous hybridizers working hard to produce new orchids for our collections. Their job is still not easy. It takes time to develop new hybrids and time to evaluate them. As for the source of new genes and ideas, we must all support conservation efforts of species and populations. Without diversity, our possibilities are limited.

Jade
And getting more so each day.

Ed_in_Sat
Marilyn, we need a place where hybridizers can deposit their notes. Maybe sealed during their lifetimes. It is tragic to lose the hybrid nose of Joe Ozzella, Ben Bracey, Ernest Hetherington, for that matter. Priceless resources are being lost every year.

Jade
Ed, I had had that exact thought earlier. Those notes are priceless

MarilyninOttawa
Ed, would you like to send me a formal proposal/suggested approach? This is one area of conservation after all.

Ed_in_Sat
Marilyn, you know hybridizers can't write! Its hard to write down what you feel and suspect. We'll see what we can do. I have no idea how many hybrids we have done but I know our percentage of success. I could figure the number from the tag book but can't count very high.

N_Calif_Kathy
Ernest Hetherington is very big on this idea too. He's trying to get hybridizing notes donated to the Orchid Digest Library.

barbara_in_no._CA
Maybe oral notes? Very much like the VA all over the country going to nursing homes to get the recording of old war stories?

MarilyninOttawa
Interviews, recorded, by those they trust would likely be the best approach. There was a series more than 20 years ago in the Bulletin about Paph parents and breeding outcomes. I think that that is the sort of synthesis we might be able to conserve. And where to conserve it? We would have to come to some mutually agreeable, secure location.

jim4eq
A legal office? Just like wills, make it out and then forget for 40 years (or in hollywood,
wait 40 days for the divorce, LOL)

N_Calif_Kathy
Have any idea of what genes were affected to produce those Harlequin Phals? I've noticed their spots are more warty textured the darker the spots become.

Ed_in_Sat
Golden Peoker exhibits the same quality.

LosChristabel
Mutated, I would assume?

MarilyninOttawa
I would suggest that one of you with a harlequin phal and a microscope, sacrifice a flower, cut a thin section, mount in water and observe what structures and pigments you see.

I suspect, repeat, suspect, that the pattern seen in the harlequin phals is influenced similarly to what we saw with the Onc. Remembering Carmen. The yellow pigmentation had all disappeared from the flower. We had variable amounts of white with markings. The flavonoid pigmentation was restricted to where there was coincidentally also carotenoid pigment. I suspect in the harlequins that you will find both plastids and flavonoids layed where there are spots.

Jade
Would doing a cross section cut show that placement?

MarilyninOttawa
Yes Jade, very clearly. I do not grow this so someone else will have to do it. Try to cut through a blotch and not necessarily through a very thick one. You should see purple pigmented cells and yellow-pigmented plastids in other cells. These two types may be mixed or in layers. The plastids may be orange-red colored. Cut through from the top to the bottom. Use a safety razor blade or a scalpel. Have a glass slide ready with a drop of water. Do not let the section dry out. Cover with a thin coverslip.

N_Calif_Kathy
Well I'm plumb out of other questions. Other than to comment that I read Gary Baker's article on Masd coccinea color production. If you ever get a chance to read it do so. AOS Bulletin vol 59 pg 382 Glamboyant Queens of the Andes: Color Forms of Masd coccinea.

MarilyninOttawa
I have been looking at the inheritance of pattern in Masdevallias. Interesting what we find. Look for notes eventually in the Pleurothallid Alliance newsletter.

Well, that's all folks! Next month, I plan to do Aerides and Aerangis. Other possible subjects are always welcomed. Good night!

Jade
Marilyn, this was an absolutely fantastic topic. I think it is going to generate lots of further discussion. Thank you so very much.

end

Orchid Flower Color

Marilyn H. S. Light
Copyright March 2003

Orchid flowers are colored by various pigments which are either evenly or regionally distributed in their flower parts. The pigments may be roughly placed in two groups. The anthocyanins are water-soluble pigments found in the cell vacuoles. Genes controlling the production of anthocyanins are found in the nucleus. Anthocyanin pigments are in the red-purple-blue range and their shade can be affected by the pH in the vacuoles. The production and concentration can be influenced by temperature and other growing conditions. The other main group are the carotenoid pigments, oil-soluble substances found in the plastids which are tiny structures found within the cell. Plastids contain genetic material which is inherited separately to that in the nucleus. Plastid genetic code is inherited with the egg cell and separately to nuclear genes which can be acquired from either the seed or pollen parent. Carotenoid pigments range in color from yellow through orange into red. Chloroplasts, which are a specialized type of plastid, produce the pigment chlorophyll which is green. Floral pigments can be found in both the cell sap and in the plastids. In other words, you can have purple pigment in the cell sap and yellow-pigmented plastids. To our eyes such flowers appear orange, red or brown.

Flower color has been of horticultural interest for centuries but the process of color inheritance and expression in orchids is complex and still under investigation. Factors affecting color expression and lack thereof in Cattleya were reported by Hurst in 1925. Growers of pure white-flowered Cattleya species had observed that the progeny of two alba species did not always produce the expected alba flowers but self-pollination of albas almost always resulted in white-flowered offspring. They wondered why. Careful breeding experiments suggested that there were two genes necessary for the production of purple-colored flowers. If one or the other gene was absent, pigment production was impaired and the flowers would be white. But, if the parents each lacked a different gene, pigment production was not impaired and the progeny would have purple flowers. Alba forms of Cattleya species segregate into two groups based on which gene is impaired. Group I contains C. gaskelliana, intermedia, labiata, loddigesii, lueddemanniana, mossiae, skinneri, speciosissima, and warneri, Group II contains C. eldorado, harrisoniana, mendelii, percevaliana, schroederae, warneri and warscewiczii. Cattleya trianiae has alba clones in both Groups therefore, when breeding for albas with this species, it helps to know which group beforehand.

Floriculture has prompted interest in the inheritance of flower color in other orchids including Cymbidium and Dendrobium. Following what was known with Cattleya, investigators have looked at the inheritance of semi-alba and alba in Dendrobium. Kamemoto and Amore (1990) pointed out that a big challenge has been the time taken to raise large numbers of seedlings to maturity, the cost of maintaining these plants to maturity, and the difficulty of working with complex hybrids. Nevertheless, they were able to report that with Den. affine, dicuphum and phalaenopsis, the expression of purple color is controlled similarly to what we see in Cattleya. The major difference reported was with the inheritance of the semi-alba trait which in Cattleya is a recessive trait and in Dendrobium, is controlled by a dominant gene. So-called albino Cymbidiums lack red color in otherwise white, yellow or green flowers. Australian, Alvin Bryant, began a breeding program in the 50's which led to the production of pure color standard and miniature Cymbidiums. These plants lack the ability to produce red pigment. Pure color flowers do not fade prematurely if the anther cap is dislodged which is a valuable trait for both show and commercial purposes.

Temperature can affect the production and accumulation of anthocyanin pigments. We can see this with red-colored Phalaenopsis and Sarcochilus, where flower buds exposed to different temperatures can produce differently colored flowers, even on the same plant or on the same inflorescence. The higher the temperature and the longer the exposure to the critical temperature, the paler the flower color. The cooler the growing condition, the more intense the color. A grower would have to experiment with their particular plants to learn just what is the critical temperature, length of exposure, and at which stage of development, flower color could be affected. A study with roses suggests that transient high temperature may not have an effect but several days of sustained high temperature could reduce flower color and affect quality (Dela et al, 2003).

When we breed for specific colors, we should take into account the relative value of using a particular parent to get seed. One classic approach to get red is to breed a yellow carotenoid-pigmented type with a mauve-purple anthocyanin-pigmented kind. Oncidium Vera Arthurs has a bright yellow lip and mottled petals and sepals where the mahogany-colored spots are derived from a combination of purple red and yellow. When Onc. Vera Arthurs was crossed with mauve-flowered Onc. ornithorhynchum, the hope was for the coveted red bloom. The grex has provided a range of colors from concolor bordeaux red to red-purple blotches on a white ground. Except for the yellow callus, all yellow pigmentation has been suppressed except where anthocyanin pigment is also present in the tissue.

We may admire the wide range of flower color and pattern we see but we are merely bystanders as far as an orchid is concerned. Orchid colors and patterns have evolved to attract pollinators. Bees especially, see color differently to humans. To a bee's eyes, yellow flowers float like bright spots in a sea of dark, almost invisible foliage. Once the bee gets close to the bloom, there may be lines of contrasting color (nectar guides) to guide the bee to the nectar reward. There may be ultra-violet light-reflecting dots around and on the staminode of certain Paphiopedilum species. These dots, which are located between the warts visible to our eyes, may suggest the presence of aphids to predatory hover (syrphid) flies looking for a place to lay their eggs. Whatever the visual cue, if it attracts the attention of potential pollinator and pollination happens, then the investment in flower color and pattern has served its procreative purpose.

 CREATING CUSTOM COLORED ORCHID FLOWERS

R. J. Griesbach, Floral and Nursery Plant Research, U.S. National Arboretum, USDA, ARS, Beltsville, MD 20705-2350 USA

Classical breeding, as well as genetic engineering, can be used to create novel flower colors. However, a thorough understanding of the chemistry, biochemistry and genetics of flower color is necessary for creating flowers in an endless array of colors.

Flower color is due to three different pigments - chlorophyll, flavonoids, and carotenoids (Griesbach, 1984). The chlorophyll pigments are located within chloroplasts which are found in the cytoplasm of the cell. These pigments are responsible for green color. The carotenoid pigments are located within chromoplasts which are found in the cytoplasm. These pigments are responsible for yellow through orange colors. The flavonoid pigments, unlike the other two pigments, are located within the cellular vacuole and are responsible for red through blue colors.

Each pigment is the result of a different sequence of biochemical reactions. The production of each pigment is independent of the other pigments. For example, a defect in the flavonoid pathway has no effect on the carotenoid and chlorophyll pathways. This can be clearly seen in the white-flowered form of Phalaenopsis lueddemanniana and Paphiopedilum bellatulum where the flavonoids which are usually present in the spots are absent. The carotenoids, however, are unaffected. Therefore, the spots are yellow, not purple.

Flower color is the result of mixing the three pigments (flavonoids, chlorophyll and carotenoids) in different proportions. By mixing and matching the three pigments, a wide array of different colors can be created. For example, the red color of Sophronitis (Matsui & Nakamura, 1988), Phalaenopsis (Griesbach, 1984) and Disa (Volgelpoel, 1986) hybrids is the result of mixing orange carotenoids with magenta flavonoids.

Very little information is available on the chemistry and genetics of carotenoid pigmentation. A large amount of information has been collected on the chemisrty and genetics of flavonoid pigmentation. The remainder of the paper will focus on flavonoid pigmentation.

Flavonoid Chemistry

The flavonoid pigments can be artificially subdivided into two groups - the anthocyanins and the co-pigments. At physiological pHs, the anthocyanins are not very stable and are nearly colorless. The addition of co-pigments to the anthocyanins increases both the stability and intensity of the anthocyanin's color. For example, a solution containing 10 mM of the anthocyanin cyanidin 3,5-diglucoside and 30 mM of the co-pigment quercetin has an absorption eight times greater than a solution containing only the anthocyanin (Asen et al., 1972). This effect is called co-pigmentation.

Within the cell, the anthocyanins and co-pigments occur bound together as a chemical complex. In Commelina communis, the anthocyanin/co-pigment complex contains six anthocyanin, six co-pigment and two magnesium molecules (Kondo et al., 1992). To date, this is only anthocyanin/co-pigment complex in which the exact structure has been elucidated. In the anthocyanin\co-pigment complex, hydrophobic interactions between the aromatic rings of the anthocyanin and co-pigment molecules result in visible color (Brouillard, 1988).

It is generally assumed that red flowers contain predominantly cyanidin and blue flowers mostly delphinidin. Although this is usually true, there are many exceptions. For example, flowers that contain cyanidin can either be red as in Rosa (Asen et al., 1971) or blue as in Meconopsis (Takeda et al., 1996). One of the major reasons why flowers containing the same anthocyanin can be different colors is pH (Stewart et al., 1975). As the pH becomes more alkaline, the color of a specific anthocyanin/co-pigment complex becomes more blue. All the anthocyanins except pelargonidin have the capability of producing blue flowers (Asen, 1976).

There are several examples of the effects of pH on the flower color of living cells. In Rosa x 'Better Times', the pH of the petal cells increased as the flower aged from 3.7 to 4.4 and color of the cyanidin/quercetin complex changed from a 8max of 539 to 543 (Asen et al., 1971). In Consolida ambigua L. 'Dark Blue Supreme', the pH increased from 5.6 to 6.6 and the color of the delphinidin/kaempferol complex changed from a 8max of 537 to 570 (Asen et al., 1975). In Anthurium andraenum Andre 'Ozakie Red', the pH increased from 5.2 to 5.6 and the color changed from red to blue (Paull et al., 1985). In Petunia hybrida, a controlling element was used to change the pH of the corolla from 5.4 to 5.8. The 0.4 increase in pH changed the color of the petunidin/quercetin complex from red to blue (Griesbach, 1998).

Flavonoid Structural Genes

Flavonoids are the pigments responsible for flower color in most plants. The flavonoid biosynthetic pathway (fig. 2) is very well understood in Petunia (Holton and Cornish, 1995; Mol et al., 1998; Winkel-Shirley, 2001). All of the enzymes and their corresponding genes have been studied in detail. The first studies identified genes that were involved in the inheritance of flower color. As biochemical data became available, these genes were assigned specific functions (Wiering and deVlaming, 1984).

In Petunia, three genes (Chs, Chi and An3) are involved in creating the basic flavonoid ring. Chs encodes the chalcone synthase multigene family with eight complete (ChsA, B, D, F, G, H, J and L) and four incomplete (ChsC, E, I and K) copies per haploid genome (Koes et al., 1987 and 1989). ChsA gene is the only gene transcribed to a significant extent in flower tissue. Each complete Chs gene consists of two exons separated by an intron of variable size and sequence (Koes et al., 1989). The incomplete Chs genes do not contain an intron.

Chi encodes the chalcone flavanone isomerase multigene family with two copies (ChiA and ChiB) per haploid genome (van Tunen et al., 1988). ChiA is expressed in all floral tissue and contains no introns; while ChiB is only expressed in anthers and contains three introns. The Po mutation is the result of a mutation in the regulatory region of ChiA abolishing promoter activity in anthers but not in corollas (van Tunen et. al., 1991). The last gene involved in creating the basic flavonoid ring structure is An3. An3 encodes a 2-oxoglutarate-dependent dioxygense (Britsch, 1990).

In corolla cells, three different genes (Ht1, Hf1 and Hf2) are responsible for hydroxylating the flavonoid ring to create a dihydroflavonol (Stotz et al., 1985). The Ht genes encoded encode a cytochrome P450-dependent monooxygenase which hydroxylates the carbon at the 3' position (Brugliera et al.,1999). The Hf genes also encode a cytochrome P450-dependent monooxygenase which hydroxylates the carbon at the 5' position instead of the 3' position (de Vetten et al., 1999; Shimada et al., 2001). The 5' hydroxylase requires the presence of an additional protein (cytochrome b5) encoded by DifF. Cytochrome b5 acts as the electron donor between NADPH and cytochrome P450-dependent monooxygenase.

The conversion of dihydroflavonols into anthocyanins requires the concerted action of three enzymes (Nakajima et al., 2001; Saito et. al., 1999; Turnbull et. al., 2000). This a very complex step in the pathway which involves two different reactions --- the reduction of the double bonded oxygen on the carbon at the 4 position and the glucosylation of the hydroxyl group at the 3 position. The first enzyme (dihydroflavonol reductase) is encoded by An6 and catalyzes the conversion of dihydroflavonols to leucoanthocyanins (Huitts et. al., 1994). An6 contains five introns. Besides An6, thereare two other dihydroflavonol reductase genes (Beld et al., 1989). The second enzyme in this complex reaction (anthocyanidin synthase, a 2-oxoglutarate-dependant oxygenase) is encoded by An17 and converts leucoanthocyanins into 3-flaven-2,3-diols (Weiss et al., 1993). The last enzyme in the reaction (UDP-glucose : anthocyanin 3-O-glucosyltransferase) creates the anthocyanin-3-glucoside (Kho et. al., 1978). The 3-O-glucosyltransferase gene has not yet been identified in Petunia.

Dihydroflavovols can also be converted in flavonol glycosides. Fl encodes flavonol synthase which is a 2-oxoglutarate-dependant oxygenase (Holton et al., 1993). The Petunia flavonol synthase has a greater Km for dihydrokaempferol and dihdroquercetin than for dihydromyricetin (Gerats et. al., 1982; Forkman and Ruhnau, 1987). In addition, this enzyme has a greater Km for dihydrokaempferol and dihdroquercetin than dihydroflavonol reductase. Therefore in Fl+ genotypes, quercetin glycosides accumulate at the expense of cyanidin-based anthocyanins. To a less extent, myricetin glycosides accumulate at the expense of delphinidin-based anthocyanins in Fl+ genotypes.

The 3-glucosyl anthocyanin is the substrate for the Rt encoded enzyme which adds a rhamnose to the glucose at the 3 position to create a rutinoside (Brugliera et. al., 1994; Kroon et al., 1994). The 3-rutinoside is now the substrate for Gf which encodes an enzyme which attaches acylates the 3-rutinoside with either caffeic acid or coumaric acid (Jonsson et al., 1984a). Once the acyl group is attached, UDP-glucose : anthocyanin 5-O-glucosyltransferase adds a glucose at the 5 position (Jonsson et al., 1984a). The 5-O-glucosyltransferase gene has not yet been identified in Petunia. The last steps in the pathway involve the methylation of the acylated rutinoside.

There are four different anthocyanin-O-methyltransferase genes in Petunia (Mt1, Mt2, Mf1, and Mf2) (Jonsson et al., 1983). Each gene controls a distinct and independent enzyme which is capable of methylating both the 3' and 5' positions on the anthocyanin molecule (fig.1). Each enzyme, however, had a distinct substrate specificity. The Mf1 (8 :M) and Mt2 (6 :M) encoded enzymes had an approximately three-fold lower Km values for cyanidin and petunidin as the substrate than the Mf2 (21 :M) and Mt1 (25 :M) encoded enzymes. Each enzyme also had a different efficiency in methylating delphinidin - Mf1 (175 pkatCmg protein-1 ), Mf2 (100 pkatCmg protein-1 ), Mt1 (60 pkatCmg protein-1 ), and Mt2 (30 pkatCmg protein-1). The Mt encoded enzymes preferred cyanidin as a substrate, instead of either petunidin or delphinidin . When delphinidin was the substrate, the Mt encoded enzymes produced mainly petunidin; while the Mf encoded enzymes produced mainly malvidin. There was, however, a differential effect on substrate inhibition. High concentrations of delphinidin reduced the amount of malvidin produced, but not the amount of petunidin produced. In addition, a dosage effect was suggest for Mf / Mt gene expression. The greater the number of Mf+ / Mt+ genes, the higher the relative concentration of malvidin. High concentrations of petunidin coupled with low concentrations of delphinidin promoted malvidin synthesis (Jonsson et al., 1984b).

Flavonoid Regulatory Genes

In Petunia flowers, the genes encoding the enzymes that are expressed early in the anthocyanin biosynthetic pathway (chalcone synthase, chalcone-flavone isomerase, flavanone 3-hydroxylase, etc.) are control by a different set of regulatory genes that those genes encoding the enzymes expressed late in the pathway (dihydroflavonol reductase, anthocyanin rhamnosyltransferase, anthocyanin methyltransferase, etc.) (Quattrocchio et al., 1993). At least four regulatory genes (An1, An2, An4, and An11) are required for the transcription of the genes expressed late in the pathway. An1 encodes a basic helix-loop-helix (bHLH) transcription factor that is active in all parts of the flower (Spelt et al., 2000). An2 and An4 encode MYB-domain transcription factors (Quattrocchio et al., 1999). An2 is only active within the limb, while An4 is only active within the anthers. An11 encodes a regulatory protein with five WD-repeat units that is active in all parts of the flower (de Vetten et al., 1997).

These regulatory genes operate in a complex regulatory hierarchy which is still not completely understood. An11 encodes a cytoplasmic protein that regulates the expression of An2, as well as, other non-anthocyanin related genes (de Vetten et al., 1997). It appears that An11 links cellular and/or environmental signals with transcription of An2. However, An2 does not directly regulate the transcription of any anthocyanin structural gene. An2 controls the expression of An1 which directly activates the transcription of the structural genes within the limb and tube (Spelt et al., 2000). Besides regulating anthocyanin biosynthesis, An1, An2, and An11 also control vacuolar pH (Mol et al., 1998). An1 (previously studied as Ph6 ) regulates the expression of Ph1 and Ph2 (Griesbach, 1998).

Literature Cited

Arditti, J. 1992. Fundamentals of Orchid Biology. John Wiley and Sons, New York. pges. 244-256.
Asen, S., K.H. Norris and R.N. Stewart, 1971. Affect of pH and concentration of the
anthocyanin-flavonol co-pigment complex on the color of 'Better Times' roses. J. Amer. Soc. Hort. Sci. 96:770-773.
Asen, S., R.N. Stewart and K.H. Norris, 1972. Co-pigmentation of anthocyanins in plant
tissues and its affect on color. Phytochem. 11:1139-1144.
Asen, S., R.N. Stewart, and K.H. Norris, 1975. Anthocyanin, flavonol, copigments, and pH
responsible for larkspur flower color. Phytochem. 14:2677-2682. Asen, S. 1976. Known factors responsible for the infinite flower color variations. Acta Hort.
63:217-223.
Brouillard, R. 1988. Flavonoids and flower color. In: The Flavonoids: Advances in Research, ed. J.B. Harborne. Chapman & Hall, London. pges. 525-538.
Griesbach, R.J. 1984. Effects of carotenoid anthocyanin combinations on flower color J.
Heredity 75:145-147.
Griesbach, R.J. 1990. Flavonoid co-pigments and anthocyanin of Phalaenopsis schilleriana.
Lindleyana 5:231-234.
Griesbach, R.J. and T.M. Klein, 1993. In situ genetic complementation of a flower color mutant in Doritis pulcherrima. Lindleyana 8:223-226.
Griesbach, R.J. 1996. The inheritance of flower color in Petunia hybrida. J. Heredity 87:241-
245.
Griesbach, R.J. 1997. The biochemical basis for the blue flower color mutations in Doritis
pulcherrima and Phalaenopsis violacea. Lindleyana 12:64-71.
Griesbach, R.J., 1998. The affect of the Ph 6 gene on the color of Petunia hybrida Vilm.
flowers. J. Amer. Soc. Hort. Sci. (In Press).
Holton, T.A. and E.C. Cornish. 1995. Genetics and biochemistry of anthocyanin biosynthesis.
Plant Cell 7:1071-1083.
Huitts,H.S.M., A.G.M. Gerats, M.M. Kreike, J.N.M. Mol and R. Koes. 1994. Genetic control of dihydroflavonol 4-reductase gene expression in Petunia hybrida. Plant J. 6:295-310.
Hurst, C.C., 1903. Mendel's principles applied to orchid hybrids. J. Royal Hort. Soc., 27:614-624.
Hurst, C.C., 1913. The application of genetics to orchid breeding. J. Royal Hort. Soc. 38:412-
429.
Jonsson, L.M.V., P. deVlaming, H. Wiering, M.E.G. Aarsman and A.W. Schram. 1983.
Genetic control of anthocyanin-0-methyl-transferase activity in flowers of Petunia hybrida. Theor. Appl. Genet. 66:349-355.
Jonsson, L.M.V., M.E.G. Aarsman, J. van Diepen, P. deVlaming, N. Smit and A.W. Schram.
1984. Properties and genetic control of anthocyanin 5-0-glucosyltransferase in flowers of Petunia hybrida. Planta 160:341-347.
Kondo, T., K. Yoshida, A. Nakagawa, T. Kawai, H. Tamura, and T. Goto. 1992. Structural
basis of blue-color development in flower petals from Commelina communis. Nature 358:515-518.
Kuehnle, A.R., D.H. Lewis, K.R. Markham, K.A. Mitchell, K.M. Davis and B.R. Jordan, 1997. Floral flavonoids and pH in Dendrobium orchid species and hybrids. Euphytica 95:187-194.
Kroon, J., E. Souer, A. deGraaff, Y. Xue, J. Mol and R. Koes. 1994. Cloning and structural
analysis of the anthocyanin pigmentation locus Rt of Petunia hybrida. Plant J. 5:69-80.
Martin, C., R. Carpenter, E.S. Coen and T. Gerats. 1987. The control of floral pigmentation in Antirrhinum majus. In: Developmental Mutants in Higher Plants, eds. H. Thomas and D. Grierson. Cambridge Univ. Press, London. pges. 20-52.
Matsui, S. and M. Nakamura. 1988. Distribution of flower pigments in perianth of Cattleya and allied species. J. Japan. Soc. Hort. Sci. 57:222-232.
Mol, J., E. Grotewold and R. Koes. 1998. How genes paint flowers and seeds. Trends in Plant
Science 3:212-217.
Quattrocchio, F.M., J.F. Wing, H.T. Leppen, J.N.M. Mol, and R. Koes. 1993. Regulatory genes controlling anthocyanin pigmentation are functional conserved among plant species and have distinct set of target genes. Plant Cell 5:1497-1512.
Quattrocchio, F., J. Wing, K. Woude, H. Leppen, J. Mol and R. Koes. 1993. Regulatory genes
controlling anthocyanin pigmentation are functionally conserved among plant species and have distinct sets of target genes. Plant Cell 5:1497-1512.
Quattrocchio, F., J. Wing, K. Woude, E. Souer, N. de Vetten, J. Mol and R. Koes. 1999.
Molecular anlaysis of the anthocyanin 2 gene of Petunia and its role in the evolution of flower color. Plant Cell 11:1433-1444.
Saito, N., K. Toki, K. Uesato, A. Shigihara and T. Honda. 1994. An acylated cyanidin glycoside from the red-purple flowers of Dendrobium. Phytochem. 37:245-248.
Saito, N., M. Ku, F. Tatsuzawa, T. Lu, M. Yokoi, A. Shigihara and T. Honds. 1995. Acylated
cyanidin glycosides in the purple-red flowers Bletilla striata. Phytochem. 40:1523-1529.
Spelt, C., F. Quattrocchio, J. Mol and R. Koes. 2000. Anthocyanin 1 of Petunia encodes a
basic helix-loop-helix protein that directly activates transcription of structural anthocyanin genes. Plant Cell 12:1619-1631.
Stewart, R.N., K.H. Norris and S. Asen, 1975. Microspectrophotometric measurement of pH
and pH effects on the color of petal epidermal cells. Phytochem. 14:937-942.
Stotz, G., P. deVlaming, H. Wiering, A.W. Schram and G. Forkman. 1985. Genetic and
biochemical studies on flavonoid 3'-hydroxylation in flowers of Petunia hybrida. Theor. Appl. Genet. 70:300-305.
Takeda, K., S. Yamaguchi, K. Iwata, Y. Tsujino, T. Fujimor and S. Husain, 1996. A
malonylated anthocyanin and flavonols in the blue flowers of Meconopsis. Phytochem. 42:863-865.
Tatsuzawa, F., N. Saito, M. Yokoi, A. Shigihara and T. Honda. 1996. Acylated cyanidin 3,7,3'-triglucosides in flowers of xLaeliocattleya cv. Mini Purple and its relatives. Phtyochem. 41:635-64.
Tatsuzawa, F., N. Saito, H. Seki, R. Hara, M. Yokoi and T. Honda. 1997. Acylated cyanidin
glycosides in the red-purple flowers of Phalaenopsis. Phytochem. 45:173-177.
de Vetten, N., F. Quattrocchio, J. Mol and R. Koes. 1997. The an11 locus controlling
flower pigmentation in Petunia encodes a novel WD-repeat protein conserved in yeast, plants and animals. Genes & Develop, 11:1422-1434.
Vogelpoel, L. 1986. Flower colour- an appreciation. S. Afr. Orchid J.17: 109-113.
Weiss, D., A.H. can der Luit, J.T.M. Kroon, J.N.M. Mol and J.M. Kooter. 1993. The Petunia
homologue of the Antirrhinum majus candi and Zea mays A2 flavonoid genes; homology to flavanone 3-hydroxylase and ethylene forming enzyme. Plant Mol. Biol. 22:893-897.
de Vlaming, P., A.W. Schram and H. Wiering, 1983. Genes affecting flower colour and pH of
flower limb homogenates in Petunia hybrida. Theor. Appl. Genet. 66:2171-278.
Wiering, H. and P. deVlaming. 1984. Genetics of flower color and pollen colours. In: Petunia,
ed. K.C. Sink. Springer-Verlag, Berlin. pges. 49-75.
Yamaguchi, T., S. Fukada-Tanaka, Y. Inagaki, N. Saito, K. Yonekura-Sakakibara, Y. Tanaka,
T. Kusumi, and S. Iida. 2001. Genes encoding the vacuolar Na+/H+ exchanger and flower coloration. Plant Cell Physiol. 42:451-461.
Yoshitama, K. and K. Abe. 1976. Chromatographic analysis and spectral characterization of
3'-glycosylation in anthocyanins. Phytochem. 16:591-593.