Bean Genes

The mysterious black beans continued to play on my mind last week.  Two things still needed more clarification.  First was the question of origin of my purported Gramma Walters variety, and the second was how bean seed colors work when you do a cross.

I talked with Andrew Still a bit more last weekend, and he is quite sure my pretty beans are ones he brought to that seed swap where I got them a few years ago, and that they came from somewhere in central Europe.  I’m waiting for the details…  In the mean time I’ll continue to call them Gramma Walters, but you might expect a correction in the near future.

I little research gave light to the origins of modern common beans.  There are two major ancestral gene pools  for the domesticated common bean coming from the Mesoamerica, (Mexico and Central America) and Andean (Peru and Chile) regions.  Most common bean varieties still can trace their roots (or genetic markers) to one region or the other.  For instance, what we call kidney beans (red, dark red, and white varieties) and cranberry beans derive from the Andean center of domestication (COD).  Modern varieties such as navy, small white, black beans, pinto and great northern all derive from the Mesoamerica COD.  Kentucky Wonder is also descendant from the Mesoamerica beans.  Curiously, most (70%) of traditional European bean varieties are descendant from Andean stock.  Apparently, the early explorers brought bean seed back to the continent largely from South America first to Spain and then disseminating to the rest of Europe.  Beans are one of the most easily saved seeds, so it is not surprising that there are hundreds of varieties surviving on family farms where traditional farming is still practiced.  I’m guessing that my pretty maroon and white beans come from this European heritage, and are probably descendant from the Andean COD.

So how could marbled maroon and white beans crossed with brown beans end up black?  Turns out that bean seed color has been investigated by many researchers over the past century, and at this point much is known about how the genetics of bean color works.  Considering the diversity of colors, it shouldn’t come as a surprise that there are several genes that determine the seed coat color.  Here is a partial list of bean genes, symbols and variations, and what they are thought to do that came from some of the references below.

Gene Symbol	Description
P	Main color gene.  Without P, other color genes are inactive.
p	Lack of P, seeds are colorless (white)
pgri	Allele of P, seeds are gray-white.   dominance is P > pgri  > p
C	Color gene required for full expression of color modifying genes. C usually  closely linked to R, dominant red color gene, and coded [C r]
c	Lack of C
J	Coat colors fully developed in immature seeds
j	Produces  pale coats colors in immature seeds
G	Color modifying gene for Greenish Brown
g	Lack of G
B	Color modifying gene for Yellow Brown
b	Lack of B
R	Red color dominant gene
r	Lack of R
rk	Red Kidney color – recessive gene
V	Violet  to black color gene – Intensifier for other color genes.
v	Lack of V
Asp	Glossy or shiny seed coat
asp	Dull seed coat
T	Total color coverage
t	Patterned color coverage; t/t produces white flowers without color,  except with V purple flowers.

A couple of things become apparent when we look at this list and try to write down the genotypes of the beans in my garden.  Here is what I get…

Gramma Walters:    Has color, so must have P; the maroon could be red, [C r].  Immature seeds are pale, so j likely. Must have V, otherwise where did the black cross come from.  Possibly [C r] and V get us maroon. Must have t to have patterned seeds.

Proposed genotype for Gramma Walters:  PPggbb[C r][C r]jjVVtt

Oregon Giant:  Seeds are gray-white with black markings.  So perhaps the main color gene is p^gis   rather than P. Likely other genes would be V for the dark black color markings, and t for the partial color trait. No browns or greens or reds, so g,b, and [c r].  Markings more apparent on immature seeds, so perhaps J.

Proposed genotype for Oregon Giant: p^gisp^gisggbb[c r][c r]JJVVtt

Kentucky Wonder:   Brown seeds have main color gene P; perhaps color genes G and B make up the brown.  Immature seeds are pale, so j likely.  No black or violet, so v.  Solid seed color, so T.

Proposed genotype for Kentucky Wonder: PPGGBB[c r][c r]jjvvTT

Now we are ready to make some crosses!  First, consider Gramma Walters with Oregon Giant.  The recessive t/t genes mean that these crosses would also produce patterned seed, not solid black.  Hence, this can’t be the cross that made the black beans.

The Gramma Walters / Kentucky Wonder cross should look something like: PPGgBb[C r][c r]jjVvTt with all kinds of heterozygous genes since almost nothing is the same!  But if we look at the dominant genes, we see we have genes for several color genes, including dominant V which makes things black.  We also have T which produces self-colored – non-patterned seed coats.  That’s our black bean!  Mystery solved!

If I should plant those beans what should I get?  In the F2 there will be all kinds of combinations of the dominant and recessive pairs. Just consider the Tt with  Tt possibilities.  Namely, TT  Tt  Tt and tt would produce 1/4 patterned beans, and 3/4 solid color beans.  Two thirds of the solid beans would still carry the recessive t pattern-color gene.  It works the same way for Vv so we would expect 3/4 black or violet beans and 1/4 lightly colored beans.  The other color genes for the Kentucky Wonder and Gramma Walters are probably all different.  No sign of greens or browns in the Gramma Walters, and no sign of red in Kentucky Wonder, so for the quarter of the beans that are light-colored, we might expect just about any color you can imagine to show up!

If I plant 100 of my F1 black beans, I can expect about 75 plants that produce solid color, and of those 75 about 56 will produce black or dark-colored beans. Of those 56 only about 18 will be homozygous in VV and produce black beans reliably in future generations. Of those 18, only about 6 plants will be homozygous in TT, and only produce non-patterned seed in future generation.

Note how much easier it is to select for a recessive trait you wish to propagate. Any of the patterned beans will have the tt homozygous recessive genes, as will all of the self pollinated offspring.  A single selection breeds true with a recessive trait in a self-pollinator.  If instead we are trying to get rid of the t patterned-color trait, which is what I would have to do to get a true black bean, then I have to remove the patterned-color beans each generation to weed out the homozygous tt genes as they show up, but I can’t do anything about the Tt carriers since I don’t know which ones those are.  If you do the math, after you remove the ones showing the homozygous recessive trait, you will have 1/3 of the remaining seed homozygous in the dominant trait in the F2.  After selecting in the F3 you will have 60% true seed; 78% true seed after F4; 88% true after F5; and 94% true seed after the sixth generation of selection to remove the obvious recessive genes.  You may never get a 100% pure dominant trait since those recessive genes can hide for a long time!

We’ve only considered bean seed coat color, but much more important for a new bean variety would be characteristics such as taste, disease resistance, earliness of harvest, size of pods, yield, etc.  These traits are controlled by hundreds of genes, all of which need to be selected for desired qualities.  The parents could easily be representatives of the Andean and Mesoamerican races, so in some sense, the hybrid is the mixing of long-lost relatives.  The plants in the F2 generation will generate many new combinations for traits from these diverse parents.  Clearly the selection at the F2 generation for the traits we want is most crucial.  I could just as easily have as a goal a bean with the dry-bean qualities of Gramma Walters with pod shape of the Kentucky Wonder, and I would have started with the same F1 cross with the same black beans.  Of all the possible combinations of genes possible between the two lines, I am only sampling the hundred or so possibilities that happen to be in the seeds I sow.

So should I plant those seeds and see what comes?

Here are a few references for those of you who need to know more.

Mark J. Bassett and Phil McClean,  A Brief Review of the Genetics of Partly Colored Seed Coats in Common Bean

P. E. McClean, R .K. Lee, C. Otto, P. Gepts, and M. J. Bassett, Molecular and Phenotypic Mapping of Genes Controlling Seed Coat Pattern and Color in Common Bean (Phaseolus vulgaris L.)

James D. Kelly, The Story of Bean Breeding in the US.