Fristensky Lab Protocols

Finding lines homozygous for a transgene

(update 12/2/98)

When a transgenic line contains a T-DNA insertion at a single locus, the initial transformant plant T0, is hemizygous at that locus.  For example, given a gene R present in hemizygous state in the T0 generation,  selfing of any individual will give progeny segregating 3:1  for the transgene. (The null allele is represented by "0").
  
   T0               R0
                     |
                     v
   T1          RR    R0     00
                1  :  2  :   1
              ----------
                   3     :   1

The 3:1 ratio should obtain for all subsequent generations if random mating is allowed.

To identify individuals homozygous for a transgene,

  1. Grow 10 - 20 plants until several true leaves are available, but well before flowering.
  2. Extract DNA from each individual, and test for presence of the transgene by PCR. Use primers specific for the transgene, not the vector.
  3. Discard null segregants, which do not give a band,  and bag positive plants at flowering to prevent cross-pollination.
  4. Collect seed from each plant and keep in separate envelopes.
  5. For 5 positive parents, grow at least 15* plants until several true leaves are available.
  6. Extract DNA from each individual, and test for the presence of the transgene by PCR.  (You may save some work by freezing leaf tissue for 5 sets of progeny, but only testing one individual at a time, until you get a true homozygote). If any line tests positive in all individuals, that line must be homozygous. If some individuals test negative, it may be because of bad DNA preps. It is sometimes worth re-doing the DNA preps and doing PCR on the new DNA.
All seed from that parent  for which no null segregants were seen should be homozygous, and it should be possible to propogate this line indefinitely with no loss of the transgene due to segregation.
 

*How many plants need to be tested to ensure homozygosity?

We want to eliminate the possibility that a parent testing positive for the transgene is heterozygous. When a single heterozygous plant is tested,  the probability of a false positive (ie. seeing a + band rather than a - band)  is 0.75. Therefore, when N progeny are tested the probability P that all N progeny will test positive is 0.75N. If 2/3 of the parents that test positive are heterozygous,  the probability P that a false positive will occur given N progeny testing positive is

P(heterozygous parent | all N progeny test positive) = 2/3 (0.75N)

The table illustrates the probability of a parent being heterozygous given N progeny testing  positive.
 

N P
1 0.500
5 0.158
10 0.037
15 0.009
20 0.002

Therefore,  about 2 times out of 100 trials, if  10 progeny  all test positive, the parent was a heterozygote.  If 15 progeny test positive, there is less than 1% probability that the parent was heterozygous. If 20 progeny test positive, there is only 0.2% probability that the parent was heterozygous.