Do 3-Way Crosses Exist?

[EBHarvey]

I've heard people mention them offhand, but always doubted that 2 pollen grains could simultaneously enter and pollinate the same ovule. The reason I ask is that I have a very mysterious F1 in my garden right now. it was supposed to be a yellow brandywine X green zebra cross but it is red with delicate yellow striping, similar to beauty king but not as pronounced.

I have no idea where the red came from, or could have come from given the parents were both genetically yellows, and although I definitely had some donor pollen mixups, green zebra was the only thing in my garden with stripes. I'm stumped, and the only explanation seems to be that it's a yellow brandy X green zebra X some red tomato 3-way cross.....if such a thing actually exists.

Any ideas?

[Darren Abbey]

The yellow color from the B and the green color from the GZ are both recessive traits. When you cross the two, you get complementation and the wild-type alleles at both sites (in the other strain) are revealed. From a genetics perspective, it would be surprising to not get the red color in this cross.

The striping gene is apparently co-dominant, meaning that the trait is seen when either one or two copies of the gene are present... just to a lesser degree when only a single copy is present.
----

The plants that grow from seeds of this plant will have a wide range of colors/stripes/etc.

Roughly 9/16 will have red; 3/16 will have yellow; 4/16 will have green.
Roughly 1/4 of each of those categories will have strong striping; 1/2 of each category will have faint striping.
There will also be variations in size and flavor which are trickier to predict.
----

There are what people refer to as 3- or 4- way crosses, but those represent multi-generation hybridization experiments rather than a single cross as you described.

[EBHarvey]

Thanks Darren, but I'm still not following where the red color came from - is that what you mean by the wild type alleles being revealed?

[Darren Abbey]

Also, your cross sounds like what turned up on one of my plants that was supposed to be GZ.



Got any pictures of yours?

[Darren Abbey]

I'll try to graph out the genetics a little bit... It is somewhat complicated, but bear with me.
----

1) The gene responsible for red/yellow color comes in two versions. "R" is the wild-type, responsible for the last step in synthesizing the red lycopene. "r" is a mutant version which prevents the final stage of lycopene synthesis, so instead a yellow carotenoid builds up.

2) The gene responsible for the green-when-ripe trait comes in two versions. "Gr" is the wild-type, normal ripening version. "gr" is the mutant version, which prevents normal color development.

3) The gene responsible for the striped trait comes in two versions. "Gs" produces striping. "gs" produces no striping.
----

Each variety has two copies of one version or the other of each of the three genes. The genetics of each variety can be represented as follows:
Yellow Brandywine : r/r; Gr/Gr; gs/gs
Green Zebra: R/R; gr/gr; Gs/Gs

Tthe cross (referred to as F1, for 'filial 1') will get one version of each gene from each parent.
F1 : R/r; Gr/gr; Gs/gs
----

Generally, the version of the gene which is capitalized will be dominant, meaning that trait will appear if one or two copies of the gene are found.

1) The "R" (red) version of the first gene is dominant to the "r" (yellow) version, so the cross would expected to produce the red lycopene.

2) The "Gr" (normal ripening) version of the second gene is dominant to the "gr" (green-when-ripe) version, so the cross would be expected to develop the colors generated by the first gene.

The striped gene is one of the cases where the behavior is different.

3) The "Gs" (striped) and "gs" (no-stripes) versions of the third gene are co-dominant, meaning you get the average effect of the two parent (Gs/Gs & gs/gs) traits when you have one copy of each (Gs/gs).

---
[All 'Fs'/'fs' references changed to 'Gs'/'gs', as striping seen in Green Zebra is more consistent with gene with label 'Gs'. (http://kdcomm.net/~tomato/Tomato/mutant.html) - Thanks, Heritage]

[Darren Abbey]

There are probably other genes responsible for the stripes being yellow (in your cross) vs. green (in my cross) on the red background, but I don't have any idea how they behave.

I was very surprised to see the green stripes on the red tomato in my brother's garden.

[Fred Hempel]

3-way crosses (in the strict sense) don't exist. Simply put, pollen from one parent fuses with the egg from a second parent.

[travis]

Is it possible the genetics for the "yellow" in Yellow Brandywine is "t" (tangerine) rather than "r" yellow flesh?

[Heritage]

Originally Posted by Darren Abbey

3) The gene responsible for the striped trait comes in two versions. "Fs" produces striping. "fs" produces no striping.
----

Darren, I believe you are dealing with the 'gs' gene here, rather than 'Fs'. Otherwise, nice explanation, thanks!

Steve

[Doug9345]

Darren thank you for the explanation. I didn't realize that the GZ was basically a red tomato with the red development turned off.

[Darren Abbey]

Originally Posted by Heritage

Darren, I believe you are dealing with the 'gs' gene here, rather than 'Fs'. Otherwise, nice explanation, thanks!

I have no doubt that I may be doing so. ;-)

[Darren Abbey]

Originally Posted by travis

Is it possible the genetics for the "yellow" in Yellow Brandywine is "t" (tangerine) rather than "r" yellow flesh?

Definitely.

In this case, instead of the parents being...
Yellow Brandywine : r/r; Gr/Gr; gs/gs
Green Zebra : R/R; gr/gr; Gs/Gs

they would be...
Yellow Brandywine : t/t; Gr/Gr; gs/gs
Green Zebra : T/T; gr/gr; Gs/Gs

... and the logic of the genetics will be much the same.
----

If one parent was 'yellow' (r/r; T/T) and the other was 'tangerine' (R/R; t/t), the cross would be (R/r;T/t) and would look like a perfectly normal red tomato.

I don't have any idea how the two genes would interact and thus what a (r/r; t/t) would look like.

[Darren Abbey]

Originally Posted by Doug9345

Darren thank you for the explanation. I didn't realize that the GZ was basically a red tomato with the red development turned off.

It could have been a yellow tomato with yellow development turned off. ;-) It is only through 'experiments' like this one that such can be determined.

[travis]

Actually, I was suggesting that Yellow Brandywine's deep yellow/orange color is attributed to t/t rather than r/r, and Green Zebra's interior flesh color is r/r stained by gf/gf, or however that should be correctly expressed.

Ignoring the gs for the moment, and figuring what happens to flesh color when it's t/t x r/r, with gf suppressed in the F1, possibly that results in red R/r or whatever, with the skin stripes being incidentally independent in the mix.

Not being a geneticist, I'm just throwing that out there ... t x r = ?

[Darren Abbey]

Originally Posted by travis

Actually, I was suggesting that Yellow Brandywine's deep yellow/orange color is attributed to t/t rather than r/r, and Green Zebra's interior flesh color is r/r stained by gf/gf, or however that should be correctly expressed.

Ignoring the gs for the moment, and figuring what happens to flesh color when it's t/t x r/r, with gf suppressed in the F1, possibly that results in red R/r or whatever, with the skin stripes being incidentally independent in the mix.

Not being a geneticist, I'm just throwing that out there ... t x r = ?

[my initial model.]
YB : r/r;T/T;Gf/Gf ('yellow' color.)
GZ : R/R;T/T;gf/gF (green color.)
F1 : R/r;T/T;Gf/gf (red color.)

[Travis's model.]
YB : R/R;t/t;Gf/Gf ('tangerine' color).
GZ : r/r;T/T;gf/gf (green color).
F1 : R/r;T/t;Gf/gf (red color).

You're right, there are multiple models which are consistent with the observed F1.
As long as you have a geneticist's intuition, you're doing alright.
----

So, lets dive in a bit deeper and examine the possibilities.

The models which would produce the yellow/tangerine color in YB and green in GZ, while ignoring the stripes, are...
YB : (R/R;t/t;Gf/Gf) or (r/r;t/t;Gf/Gf) or (r/r;T/T;Gf/Gf) => yellow or tangerine.
GZ : (r/r;t/t;gf/gf) or (R/R;t/t;gf/gf) or (r/r;T/T;gf/gf) or (R/R;T/T;gf/gf) => green.

So, we can examine what all 12 possible crosses for these models would produce as an F1...
Each line shows a possible genetic model for YB crossed to a possible genetic model for GZ.
Any combinations which predicts a red F1 are bolded.
1) YB x GZ = (R/R;t/t;Gf/Gf) x (r/r;t/t;gf/gf) => (R/r;t/t;Gf/gf) (tangerine color.)
2) YB x GZ = (R/R;t/t;Gf/Gf) x (R/R;t/t;gf/gf) => (R/R;t/t;Gf/gf) (tangerine color.)
3) YB x GZ = (R/R;t/t;Gf/Gf) x (r/r;T/T;gf/gf) => (R/r;T/t;Gf/gf) (red color.) (Travis's model.)
4) YB x GZ = (R/R;t/t;Gf/Gf) x (R/R;T/T;gf/gf) => (R/R;T/t;Gf/gf) (red color.)
5) YB x GZ = (r/r;t/t;Gf/Gf) x (r/r;t/t;gf/gf) => (r/r;t/t;Gf/gf) (yellow/tangerine color.)
6) YB x GZ = (r/r;t/t;Gf/Gf) x (R/R;t/t;gf/gf) => (R/r;t/t;Gf/gf) (tangerine color.)
7) YB x GZ = (r/r;t/t;Gf/Gf) x (r/r;T/T;gf/gf) => (r/r;T/t;Gf/gf) (yellow color.)
8) YB x GZ = (r/r;t/t;Gf/Gf) x (R/R;T/T;gf/gf) => (R/r;T/t;Gf/gf) (red color.)
9) YB x GZ = (r/r;T/T;Gf/Gf) x (r/r;t/t;gf/gf) => (r/r;T/t;Gf/gf) (yellow color.)
10) YB x GZ = (r/r;T/T;Gf/Gf) x (R/R;t/t;gf/gf) => (R/r;T/t;Gf/gf) (red color.)
11) YB x GZ = (r/r;T/T;Gf/Gf) x (r/r;T/T;gf/gf) => (r/r;T/T;Gf/gf) (yellow color.)
12) YB x GZ = (r/r;T/T;Gf/Gf) x (R/R;T/T;gf/gf) => (R/r;T/T;Gf/gf) (red color.) (my first model.)

The F1 hybrid for half of the possible model crosses match what was observed. Discriminating between the models would require looking at the distribution of traits in the F2 progeny.

Predictions for F2 progeny sets per model cross...
#3,8,10) 16/64 green; 27/64 red; 9/64 yellow; 9/64 tangerine; 3/64 yellow&tangerine.
#12) 16/64 green; 36/64 red; 12/64 yellow.
#4) 16/64 green; 36/64 red; 12/64 tangerine.

If you combine 'yellow' & 'tangerine' into 'yellowish'...
#3,8,10) 16/64 green; 27/64 red; 21/64 yellowish.
#4,12) 16/64 green; 36/64 red; 12/64 yellowish.

You'd need to grow out quite a few plants (for the size of my garden) to confidently observe the difference between 12/64 and 21/64 in the F2 plants, and then you'd still have alternate models that would be consistent with the data collected.