Origin of the biological pathway

does anyone know if the biological pathway in white flowers the produce yellow in the stamens be used in white flowers as long as they do not have the carotenoid cleaving gene? I assume that the initiation would have to have an enhancer and promoter for expression in petals.

I am thinking to a classic cross (Rosa altaica X Persian Yellow) How many of the progeny were dark yellow? I have heard that there are 7 genes that need be recovered for this pathway to work. is this true?

I would love to hear from those working with Hazeldean.


I assume that the initiation would have to have an enhancer and promoter for expression in petals.

Would there be separate copies of the entire carotenoid gene complex for each differentiation pattern, different promoter and enhancer sets only, or one set that gets switched upstream of differentiation between petals and stamens - or something else?

A more basic question is whether stamens (in their entirety) are analogous to petals and petioles tous ensemble or, instead, are filaments analogous to petioles, say, and petals analogous to anthers? If something like the latter is true then the corresponding genetic control relationships would be a highly complex matrix.

I am thinking to a classic cross (Rosa altaica X Persian Yellow)…I have heard that there are 7 genes that need be recovered for this pathway to work. is this true?

Saturation of yellow is a function of cyclization of the carotenoids where the non-aromatic precursors are less densely colored than the later stage, cyclized deriviatives. See Eugster. The rose-specific pathways are outlined here where the numbers in red denote corresponding compounds in Eugster. Each red dot probably represents a single enzyme. You can guess at the minimum number of genes from the number of enzymes in the rose-specific pathway.

These papers link 1, link 2 are the most recent I have on the subject of carotenoid biosynthesis.

Maybe you or Larry Davis have something more comprehensive or more recent on heritability and expression of carotenoids?

The whole list of possible carotenes is mind boggling. And this is only for the yellow types. Some Rosa species produce lycopene in their hips - technically part of the floral corolla.

A given tissue may or may not have plastids in its cells. If the plastids are present, there may be few to many. Also, the plastids may be small or medium sized or large. These factors are probably under genetic control, though I have not seen detailed studies of segregation and reassortment. But if the petal-cells contain few, small plastids, it really doesn’t matter which version of yellow carotene they contain - they will not be deep yellow.

Certainly, it is possible to “borrow” a more complex carotene from a non-yellow (but fragrant) species, and combine it with non-cleaving. That’s probably why some of the Tea-Noisettes are substantially darker yellow than any of the “pure” Teas. The terminal carotene ring that provides much of the structure of violet-scented ionone, also deepens the yellow when the carotene is not degraded.

BTW. Anthers are analogous to petals, filaments to the “claw” of the petal. And the hypanthium (enlarged to form the hip) is composed of the sepal bases fused into a single structure.

That’s probably why some of the Tea-Noisettes are substantially darker yellow than any of the “pure” Teas.

It doesn’t take much beta-ionone to produce a scent so theoretically, at least, we should be able to have our yellows and smell them too. I think that Johannes is asking the right questions, it comes down to regulatory genes.

Anthers are analogous to petals, filaments to the “claw” of the petal. And the hypanthium (enlarged to form the hip) is composed of the sepal bases fused into a single structure.

Reasonable, and I thought as much. Any citation?

think the key thing here is that what happens in flower petals and sepals so far as color, may have very little to do with enzymes in pathways directly. More likely it is some specific regulation sequences in DNA that are linked to the whole batch of enzymes that go into making a petal or sepal, or stamens or stigmas. So for instance the gene may be turned on a certain amount in leaves, and a whole lot more in petals, and perhaps not at all in sepals On the other hand, there might be specific forms of enzyme controlled by DNA (genes) that are unique to petals or stamens. We simply don’t know.

What is known from the work that Don linked to, is that in some species of plants there is only one copy of certain key genes for some carotenoid synthesis pathways. so if there is a mutation in that gene, the pathway is shut off. For instance a tomato flower is white, instead of yellow, in a mutant. But so many pathways in plants have multiple copies of genes for their enzymes, that we can’t really guess, even in diploids whether there are multiple active genes or not. Only if we find a mutant can we know. Failure to find one proves nothing.

If this seems odd, think of red flowers. Obviously the pathways for red are present in every cell. But they are sometimes active in leaves giving red young foliage as in Mr. Lincoln, but not really very active in some other plants that have red flowers and green leaves, like General Jaq. Somehow, the switching on of the red color to full strength happens only in certain tissues. Carotenoids accumulate in tissues with chromoplasts (plastids that accumulate color). Examples include red and yellow pepper fruits.Yellow is more difficult to see in leaves but I expect it is there in leaves of something like Carefree Sunshine. the chloroplasts need the carotenoids to procect them from photodamage in excess sunshine.

Mapping out the regulatory regions of some key carotenoid enzymes would make a great PhD project, but maybe roses wouldn’t be the best model system to start on. I guess the tomato folks will get it done.


Reasonable, and I thought as much. Any citation?[/quote]

In many cases it is easy to see the homology when one of the organs is half of one and half of the other. For instance, I first observed in the daylily ‘Kwanzo’ an occasional petal would be malformed - with half of an anther on the side.

Goethe made a particular study of these “monstrous” confusions of organs in his attempt to understand the metamorphosis of plants. O. F. Cook (1926) continued the study. I have photographed some examples on my Freaky Flowers page. Note the camellia blossom with petal-like growths from the anthers.

Finally, the ‘Vegetativnaya’ pear is a novel example that shows the homology of petioles and sepal-bases.


I never heard of Kwanzo. Is it a double? Doubling is commonly produced by converting anthers to petaloids, rather than truly doubling the number of cycles of petals are produced. Somewhere perhaps in a newsletter last year, I discussed the ABC model for organ determination. Leaves=> sepals=>petals=>anthers=> stigmas/styles. Depending on when these regulatory genes switch on and off you get the various parts made in consecutive order, or redundantly in some cases. There is some data suggesting that the sum total of stamens + stigmas is sort of constant, and when you make more petaloids/petals, you tend to deplete the stamens, then the stigmas. So very double flowers end up with no reproductive parts left.

I just looked at my Austrian Copper flowers and noticed they have red stigmas. Carefree Copper, the offspring does not. But it has the color in petals. I need to look at other offspring when they open and make a note of how that trait goes. Spring is so late this year I’ve only had AC flowers for a week, and Therese Bugnet since Monday. CC started Weds and I hope something else useful gets going pretty soon so I can do some pollinations.

A couple other bushes of CC offspring opened flowers today with severely truncated petals. We had rather bad frost, probably just as they were forming. But stamens are abundant. Whether fertile I don’t know.


Yes, ‘Kwanzo’ is the very double form of Hemerocallis fulva var disticha. In Kansas it was quite stable. A swirl of petals, some with partial anthers, but no style or ovary. I dissected quite a few and found nothing inside. And I never got any seedlings using its pollen.

I came across a bed of ‘Kwanzo’ in Palo Alto, CA that was remarkably variable. One had six tepals with a proliferation of stamens on a stalk growing from the center of the bloom. Others had various numbers of petals and stamens. One had what appeared to be a thin, white style … but I wasn’t sure.

I have some pictures of ‘Kwanzo’ on my Freaky Flowers page.

I’m attaching a picture of a “normal” flower of ‘Kwanzo’.

[attachment 1657 Kwanzo.jpg]

Thanks Karl for the photo. Looks like our double ditch lily.

I went out and looked at a couple CC offspring that have started blooming. One has red filaments, another pure gold. Both have very similar brilliant red petals that fade to dark pink. I’ll have to do a scoring on the rest when they come into bloom. Then check on the other parent of each.

Take a look at Captain Thomas on HMF, the photo by jannorcal. Very interesting pattern of red filaments, and two petaloids with a red streak from base to tip right up the center. That clearly shows the homology, and conversion pattern. I can’t tell from any of the photos if the styles are red, though the tops of stigmas definitely are not. There is another photo by calif sue with red on tips of petals, perhaps after some stress or damage judging by the shape. Clearly we have genes in lots of places, being regulated by several different signals.

That’s the one. var disticha is the source of delphinin in daylilies, responsible for the purples and blues.

I went through my pictures, but don’t have one of a strange Agapanthus I saw in Palo Alto. Normally, the spathes are green. But in one clone I saw the spathes are sometimes colored just like the flowers, indicating that the “genes for pigment” can be activated (sometimes) in the spathes. It is not necessary to have a different set of genes for each organ, just a different “program” of regulation. In Haemanthus coccineus the spathes are always colored, and take over the role of “petals”.

Interesting thread Johannes, some good info has come from the other contributors. From what I have read from your post, are you after white flowers with yellow stamens particularly in your breeding ?

Though the following quotation is from a book on iris breeding, the principles would be the same for roses.

Breeding Red Irises (2005)

Dan H. Meckenstock

“Other factors that may play a role in determining pigment concentrations that have not been discussed include the rate of reaction of an enzyme and the number and size of chromoplasts in the cell. As for the rate of reaction, we can expect dominant alleles from different sources to produce similar enzymes with different reaction rates. However, determining their relative efficiencies is beyond the capabilities of most hobby breeders. Also, selecting for more chromoplasts per cell and larger chromoplasts which ostensibly would provide more structures on which carotenoids can be synthesized and stored are other factors that effect [sic] carotenoid content. Thus, selection for these traits should result in elevated levels of carotenoids. Since plastids are maternally inherited, reciprocal crosses would reveal any inherent differences in chromoplasts. We can expect some of these subtle differences to be discernible by the naked eye; there the astute observer can select them.”

Conversely, selecting for fewer, smaller chromoplasts in the petals - while selecting for larger and more numerous chromoplasts in the stamens - should also be possible.

I’m gonna try to get a picture with xanthina and its hybrid with rugosa in the picture too, but for now here’s one with ‘Hazeldean’ and a hybrid of it with rugosa to show how much difference there is in color saturation.

picture of Hazeldean and rugosa hybrid

If I’d grown more seedlings, it’s possible I could have found better “yellows”. But even so, I think that a lot of the color usually gets lost in the first generation cross with rugosa. Maybe using an impure rugosa would help preserve more yellow? Because species-type regular rugosa that I used, seems to always REALLY dilute the yellow.

Do all the seedlings have at least this dilute yellow or are there white ones as well? Johannes

The majority of our yellow garden roses derived from Rosa foetida have not received the foetida chromoplasts, since that species was commonly used as the pollen parent in the initial crosses. Of course, the Foetidas are not abundantly female fertile, but there have been reports …

A Foetida or Foetida-hybrid may contribute “genes for yellow”, but we should also consider the ability of the chromoplasts to respond.

Something similar occasionally when some hybrid progeny segregate for albinism (lack of chlorophyll). None of the ancestral parents are albinos, or they would not have survived. So, there may be a “failure to communicate” between the nuclear genes from one species and the chloroplasts from another.

Well, I missed the rugosa X xanthina blooms but found a late bloom on xanthina itself and got some comparison picture with Hazeldean and the rugosa X Hazeldean hybrid

yellow rose comparisons

Tom, is Rosa xanthina easy to work with, I have not got it ‘yet’. Yellow is the color I am going to work with.

Hi David,

I found xanthina to be relatively cooperative. I got it’s pollen to work on rugosa and glauca (although the glauca hybrid was a non-flowering runt). Pollen from the similar looking, Rosa primula, would not set seed on rugosa for me. I had some difficulties with hugonis too if I remember right, but can’t remember the details. So if I had to choose one of the early-blooming yellow species to work with I’d go with xanthina for better color and general “cooperative-ness”.

Seems like I recall an old American Rose Society article about a breeding program for hardiness (in China maybe???) where they were using these early yellow species as pollen parents on more modern repeat blooming types. Does anyone else remember something like that?

Thanks Tom.