180
constant, the size of the screened population determines the precision of a transfer.
The calculation is simple (metaphase I pairing rates for each rye chromosome in
wheat are known, Naranjo and Fernandez-Rueda 1996 ): with MI pairing rate of 2 %
(recombination rate of 1 %) and with no more than 1 cM of alien chromatin to
remain on either side of the targeted locus, ca. 30,000 progeny would have to be
screened for a 95 % probability of success (the formula is n = Ln(1 − p )/Ln(1 − x )
where n is the populations size, p the desired probability of success and x the event
frequency). Care needs to be taken to use the recombination rate of the donor and
recipient arms, and not the listed general rates for the donor arms. Recombination
fi delity of the donor arms is never perfect; quite often they recombine with non-
targeted arms. This reduces the yield of primary recombinant chromosomes by as
much as one-third and the screening populations need to be increased accordingly.
The general perception is that chromosome engineering by homoeologous
recombination requires very large populations and in most cases is not feasible.
This may not be entirely true (see above). Several such transfers have been success-
fully made in wheat; more so with more closely related species than rye. With rye,
the fi rst demonstration of feasibility was by Koebner and Shepherd ( 1985 ) using the
chromosome arm 1RL. This was followed by engineering of the 1RS arm in the
1RS.1DL translocation. The 1RS arm in this translocation originates from Imperial
rye and the translocation was produced to introduce a very potent locus for stem
rust resistance. Judging by the publication dates, the effort took over 20 years but
after three rounds of chromosome engineering (Koebner and Shepherd 1986 ;
Rogovsky et al. 1991 ; Anugrahwati et al. 2008 ) the stem rust locus ( Sr50 ) was sepa-
rated from the Sec - 1 locus and so it is now available without the penalty to bread
making quality.
A somewhat different approach to and engineering the 1RS arm was taken for
the 1RS.1BL and 1RS.1DL translocations (Lukaszewski 2000 , 2006 ). The presence
of the rye arm was deemed benefi cial; the only identifi ed problem of the transloca-
tion is the quality defect (see discussion above). Rather than to extract a specifi c
locus from 1RS for a transfer to wheat, the starting concept was to replace Sec - 1
with Gli - B1 / Glu - B3 of wheat, leaving as much rye chromatin present as possible.
To do this as quickly as possible, 103 primary breakpoints involving 1RS and 1BS
were isolated (and another 40+ involving 1AS and 1DS). Tests of these recombinant
chromosomes quickly showed that wheat and rye storage protein loci were located
in non-corresponding positions and were separated by a block of resistance genes.
This dictated creation of a four translocation breakpoint chromosome arm 1RS,
with two small inserts from wheat: the proximal one to remove the Sec - 1 locus, and
the distal one inserting the block of wheat storage protein loci. The two are sepa-
rated by a segment of rye chromatin with four disease resistant loci: Pm8 , Lr26 , Yr9 ,
and Sr31. It took screening of ca. 17,000 progeny to identify all necessary primary
breakpoints to accomplish the intended task and another ca. 8000 to produce the
secondary and tertiary recombinants 1RS-1BS (Lukaszewski 2000 ). The irony of
the entire exercise was that when it was all done and fi nished (Fig. 7.4 ), it became
apparent that a locus responsible for increased root biomass in the 1RS.1BL wheats
is closely linked to the Sec - 1 locus (Sharma et al. 2009 , 2011 ), and it was removed
A.J. Lukaszewski