To gain information on the genetic control of L-ABA in maize, we analyzed the segregating populations derived from the cross Os420 x IABO78; parental inbred Os420 was chosen because it shows high L-ABA, especially under drought, while inbred IABO78 shows low L-ABA (Conti, S et al., Euphytica 78:81-89, 1994). The studies on such populations allowed us to identify an important QTL affecting L-ABA located on chromosome 2 between the RFLP loci umc34 and csu4a (Salvi, S et al., MNL 71:15-16, 1997; Tuberosa, R et al., Theor Appl Genet 97:744-755, 1998). To investigate the effects of this QTL on L-ABA and on related physiological traits, we undertook a backcross breeding program aimed at developing near-isogenic lines for this QTL. As emphasized by Prioul, JL et al. (J Exp Bot 48:1151-1163, 1997), near-isogenic lines are a useful "tool" to study the physiological basis of plant responses across environments.
In this note, we summarize the development of the near-isogenic lines and provide results of their preliminary evaluation.
As source material, we used the cross between Os420, providing the allele with increasing effect (+) on L-ABA at the target QTL, and IABO78, providing the allele with decreasing effect (-) on L-ABA. To develop the near-isogenic lines for both parents two backcross procedures were followed. In the first backcross procedure, Os420 was used as recurrent parent and IABO78 as the donor, while in the second one the role of the parents was reversed. The backcross procedure was aided by using the two closely linked RFLP loci umc34 and csu4a (ca. 8 cM apart) flanking the target QTL. If for simplicity we indicate umc34 as A and csu4a as B, then the genotype of Os420 can be indicated as A1B1/A1B1 and the genotype of IABO78 can be indicated as A2B2/A2B2. Plants heterozygous A1B1/A2B2 were also considered to be heterozygous for the two parental chromosome segments flanked by the two RFLP loci, assuming that no double crossover occurred between the two segments. These heterozygous plants were backcrossed for five generations and then selfed for one generation. Four BC5F2 families were developed throughout the breeding process because, starting from the BC2 generation, two different families were grown for each recurrent parent (the two families were identified as .1 and .2).
In 2000, the four BC5F2 families were grown near Bologna, under mild drought stress conditions in a randomized block design with three replications. All plants were assayed for their genotype at the two RFLP loci umc34 and csu4a. Plants that proved to be homozygous for the marker alleles provided by Os420 (i.e., A1B1/A1B1) were assumed to be homozygous (+/+) at the target QTL controlling L-ABA; analogously, plants homozygous for the two marker alleles provided by IABO78 (i.e., A2B2/A2B2) were assumed to be homozygous (-/-). Such BC5F2 homozygous plants were selfed to finally produce the BC5F3 near-isogenic lines. These were analyzed for L-ABA in order to obtain preliminary information on the effectiveness of the backcross-assisted procedure and on the additive effect of the QTL. In each plot, leaf samples for L-ABA analysis were collected at tassel appearance from four plants of the (+/+) near-isogenic line and from four plants of the (-/-) near-isogenic line.
The mean values for L-ABA of the four near-isogenic lines derived from Os420 as recurrent parent were higher than the mean values of the four near-isogenic lines derived from IABO78 (Table 1). Within each family, the (+/+) line showed a higher mean value than the corresponding (-/-) line; on a family mean basis, the additive effect was 14.8 and 14.0% for the two Os420 families and 8.2 and 10.2% for the two IABO78 families. Even if the additive effect of the target QTL was significant (P < 0.05) only for the two Os420 families, the interaction "family x additive effect" was not significant. Across the four families, the additive effect was highly significant (P < 0.01) and equal to 31 ng ABA g-1 DW, corresponding to 12.3% of the overall mean. It is worth mentioning that in our previous investigation conducted on F4 families derived from the same single cross (Tuberosa, R et al., Theor Appl Genet 97:744-755, 1998), the additive effect at this QTL was 49 ng ABA g-1 DW, corresponding to 12.1% of the overall mean.
The results of this preliminary investigation suggest that the target QTL was successfully transferred by the marker-assisted backcross and tend to confirm the relative importance of its additive effect in controlling L-ABA. As the next step, the near-isogenic BC5F3 lines (+/+) and (-/-) will be tested in field trials conducted at different levels of water stress. In addition, crosses between the lines derived from Os420 and from IABO78 will be produced (according to the factorial scheme), in order to investigate, in the absence of inbreeding depression, the QTL effects on physiological and agronomic traits related to drought tolerance.
Table 1. Mean values for L-ABA (ng ABA g-1 DW) at tassel appearance
of the near-isogenic lines (+/+) and (-/-) of the BC5F2 families derived
from Os420 and IABO78 as recurrent parents.
| Family |
|
|
|||
|
|
|
|
|
|
|
| Os420.1 |
|
|
|
|
|
| Os420.2 |
|
|
|
|
|
| IABO78.1 |
|
|
|
|
|
| IABO78.2 |
|
|
|
|
|
| Mean |
|
|
|
|
|
aCalculated as [(+/+) - (-/-)]/2.
bReferred to the family mean.
c* and ** indicate significance levels of P <
0.05 and P < 0.01, respectively; ns = not significant.
Return to the MNL 76 On-Line Index
Return to the Maize Newsletter Index
Return to the MaizeGDB Homepage