more on untagged mutants

FELDMANN at CCIT.ARIZONA.EDU FELDMANN at CCIT.ARIZONA.EDU
Tue May 25 09:27:55 EST 1993


This is in response to Tim Casper's memo pertaining to mutants isolated
from the transformed population (24 May 93). Tim wonders how we show
cosegregation and indicates that the 13 mutants that he has isolated
may not be tagged.

Cosegregation can be tested in several different ways depending on the
phenotype that is of interest.  If it is an early expressing phenotype,
even if it is selectable in some way, one can grow an F2 family, 
containing a single, on medium lacking kanamycin.  After the mutants
 appear (1/4 of the seedlings) they can be transferred to medium
containing kanamycin.  If the KanR marker and the mutation are unlinked
there is just a 75% probability that the mutant will be KanR and survive
on Kan medium.  In this way, if one transfers 11 mutants, (0.75 to the 
11th, and all are KanR, linkage between the KanR marker and the mutation
has been established.  Additional mutants are transferred to show 
tighter linkage.  For example, we have transferred several hundred dwarfs
from each line to establish linkage tight enough that we are cloning.  I
should point out that this data is in addition to a substantial amount
of pedigree data over several generation for each line.  I would strongly
argue that this is sufficient data to begin cloning. From my experience,
if one pursues large numbers of families the probability of making a
mistake, i.e., transferring the wrong plant, pollen or seed contamination,
misscoring, etc., increases substantially. Based on the increasing
number of genes that have been cloned and published from this population,
if one has one or mutants which show "tight linkage", what is the 
probability that the insert is simply close to the mutation? The data
indicates, NOT VERY PROBABLE.

For late expressing mutations, a similar strategy could probably be 
followed by pipetting a kan solution onto the leaves and inflorescences
of the mutants once they are identified. A better way, as Tim suggests,
is to identify KanR mutants (or KanR wt) in the F2 and follow their
segregation patterns in the F3.  This is a simlar stategy to that
pursued by Bertrand Lemieux and myself of the epicuticular wax mutants.
We both 1) identified KanR plants and scored the phenotype in the F2,
followed by segregation analyses in the F3 and 2) did not do Kan
selections but instead scored the phenotype in the F2, collected
seeds from single plants and scored them for Kan segregations. In the
first experiment we did 30-40 plants and in the second 20-30 plants
for each mutant line.  These data combined, indicate linkage tight
enough that we have begun to clone these genes as well (<2cM).

Unfortunately, a common way in which cosegregation analyses are 
performed cannot be applied to every phenotype nor to every lab. For
labs which possess a great deal of growth space and technical support,
the analyses of dozens of families amy be the way to go but for other
labs, interested in a mutant which is mapped, a small amount of
genetics, showing linkage, followed by isolation of the plant flanking
DNA and mapping to an RFLP map may the the most optimal manner to
establish tight linkage.

Concerning the 13 mutant lines that Tim has isolated that do not
appear to be tagged, it is possible, but it runs contrary to all of
the other data I know about. For Kevin's three mutants there is a
reasonable probability that they may not be tagged (20-30%) but for
13 random, independent? mutants not to be tagged has a probability
of <1%.  It is more likely that deletions of translocations are
occurring in one or more of these mutants generating ratios that
indicate that the two phenotypes are not tagged. If all of the ratios
that you have seen are normal Mendelian, Kan and mutant, then I do
not know what to suggest. However, if you have seen ratios that do
not fall with an acceptable chi-square you should pursue these families
more vigorously.  I hope this helps, ken



More information about the Arab-gen mailing list