T-DNA tagged mutants

caspart at esvax.dnet.dupont.com caspart at esvax.dnet.dupont.com
Mon May 24 11:28:29 EST 1993

As discussed previously by Kevin Pyke and Ken Feldmann, there 
seems to be some discrepancy between the estimates of the 
frequencies of tagged genes identified in the T-DNA mutant 
collections.  Ken cited two papers in press which deal with 
large numbers of T-DNA mutants and show 35-40% cosegregation.  
I would be interested in how these studies were done.  
Depending on how the experiments are set up, cosegregation 
can imply as little as linkage on the same chromosome arm or 
very tight genetic linkage.  For example, one common method 
for determining cosegregation is to screen for mutants in a 
segregating F2 populations and then determine whether they 
are all kan resistant (or make nopaline).  However, since 
both the kan and nopaline markers are dominant, this approach 
is not robust in differentiating between truly tagged genes 
and just a closely linked T-DNA.  For example, if the T-DNA 
is 2 cM from the gene of interest, Kan sensitive mutants 
would be present in the F2 population at a frequency of only 
1/10,000.  (For 10 cM, the frequency rises to 25/10,000.)  
Obviously, it is very difficult to establish tight linkage 
using this approach for most mutants.  A better way to 
differentiate between tight linkage and tagging is to 
progeny-test mutant individuals from the F2 population and 
determine whether they are all 100% kan resistant (i.e. 
showing that the F2 was homozygous for the Kan gene).  For 
example, if the T-DNA is 2 cM away from the mutated gene, and 
mutants are identified in the F2 and their F3 progeny tested 
for kan resistance, then about 1 out of 24 of the F3 lines 
will segregate kan sensitive individuals, correctly 
indicating that the gene is not tagged.

I used to think that 2 cM was so close that the gene would 
definitely be tagged if the T-DNA was that close.  However, 
one of the mutants we have identified appears to be separated 
from the T-DNA by about 1.2 cM.  (We found 3 out of 122 
mutant F2 plants which when progeny tested segregated 25% kan 
sensitives.)  These results could also be due to a low 
frequency of inactivation or deletion of the kan and nopaline 
genes.  Has anyone else seen this kind of instability in a 
mutant which is definitely tagged or this kind of apparently 
very tight linkage between an untagged mutant and the T-DNA?  
We are currently testing whether the T-DNA in the apparent 
segregants has simply been inactivated (e.g. by methylation).  

Our overall experience with identifying tagged mutants is 
similar to what Kevin Pyke described.  From the T3 lines Ken 
Feldmann developed at Dupont we have identified 13 mutants 
with heritable phenotypes.  Of these 13, 8 lines segregated 
with normal ratios in the T3 populations (i.e. about 25% 
mutant), 2 of the other lines were all mutant in the T3 and 
the other 3 lines had very low numbers of mutants.  Of these 
13 lines, 8 are clearly not tagged with active T-DNA's (i.e. 
giving nopaline production or Kan resistance) and the other 5 
are probably not tagged, but we are still working on them.

Tim Caspar
Dupont Central Research

More information about the Arab-gen mailing list