JXB Advance Access originally published online on October 24, 2005
Journal of Experimental Botany 2005 56(422):3061-3070; doi:10.1093/jxb/eri303
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Published by Oxford University Press [2005] on behalf of the Society for Experimental Biology.
RESEARCH PAPER |
Root-ABA1, a major constitutive QTL, affects maize root architecture and leaf ABA concentration at different water regimes
Department of Agroenvironmental Science and Technology (DISTA), University of Bologna, Viale Fanin 44, I-40127 Bologna, Italy
* To whom correspondence should be addressed. Fax: +39 (0)51 2096241. E-mail: plandi{at}agrsci.unibo.it
Received 21 March 2005; Accepted 2 September 2005
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
Near-isogenic hybrids (NIHs), developed from crossing maize (Zea mays L.) backcross-derived lines (BDLs) differing for the parental alleles at a major QTL for leaf ABA concentration (L-ABA), were field-tested for 2 years under well-watered and water-stressed conditions. Differences among NIHs for L-ABA and other morpho-physiological traits were not affected by water regimes. On average, the QTL allele for high L-ABA markedly reduced stomatal conductance and root lodging. To elucidate the effects of the QTL on root architecture and L-ABA, root traits of two pairs of BDLs were measured in plants grown in soil columns at three water regimes. Differences among BDLs were not affected by water regimes. Across water regimes, the QTL confirmed its effect on L-ABA and showed a concurrent effect on root angle, branching, number, diameter, and dry weight. Based on these results, it is concluded that the QTL affects root lodging through a constitutive effect on root architecture. In addition, there is speculation that the QTL effects on root traits and L-ABA are probably due to pleiotropy rather than linkage and a model is proposed in which the QTL has a direct effect on root architecture, while indirectly affecting L-ABA.
Key words: Abscisic acid, root architecture, root lodging, stomatal conductance, water stress, Zea mays
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
The application of QTL (Quantitative Trait Locus) analysis and other genomics approaches provide unprecedented opportunities with which to identify the chromosome regions controlling variation in the adaptive response to water stress and, eventually, to clone the gene/s responsible for such variation. The positional cloning of a QTL takes advantage from the availability of NILs (Near Isogenic Lines) for the target QTL (Paran and Zamir, 2003
). In addition, the availability of NILs for a specific QTL allows an in-depth characterization of the QTL effects on a number of traits which, in turn, facilitates the elaboration of models and hypotheses on their cause–effect relationships (Tuberosa et al., 2002).
Among the plethora of quantitative traits affecting drought tolerance, particular attention has been devoted to the concentration of abscisic acid (ABA), in view of its pivotal role in regulating other molecular and morpho-physiological processes involved in the adaptive responses of crops to an insufficient water supply (Quarrie, 1991; Tuberosa et al., 1994; Sanguineti et al., 1996; Sharp et al., 2004), particularly during the reproductive stage (Saini and Westgate, 2000; Wang et al., 2002; Boyer and Westgate, 2004). Previous studies conducted in maize (Zea mays L.) by Lebreton et al. (1995) and by Tuberosa et al. (1998) indicated that the concentration of ABA in the leaf (L-ABA) is a complex trait controlled by several QTLs. In particular, Tuberosa et al. (1998) detected 16 QTLs for L-ABA by analysing 80 F3:4 families derived by the cross between inbred lines Os420 (high L-ABA parent) and IABO78 (low L-ABA). The most important QTL was identified by the RFLP marker csu133 on chromosome 2 (bin 2.04) and accounted for 32% of the phenotypic variation for the trait. The importance of this QTL was confirmed in a subsequent work of divergent selection for L-ABA conducted in the F2 population derived from the same cross (i.e. Os420xIABO78). In fact, the allele at the RFLP locus csu133 provided by the low L-ABA parent was fixed in the eight F3:4 lines selected for low L-ABA, while it was present in only one of the eight F3:4 lines selected for high L-ABA (Landi et al., 2005).
To gain more thorough information on the effects of the QTL in question on L-ABA and other traits, NILs were developed at this QTL following a marker-assisted backcross procedure. Backcross-derived lines (BDLs) were obtained for both parental inbreds Os420 and IABO78. These BDLs were then tested in the field under both water-stressed and well-watered regimes, which allowed the effect of the QTL on L-ABA to be validated (Landi et al., 2005). For a cross-pollinated species like maize, the agronomic performance of inbred lines per se is highly affected by inbreeding depression, especially for those traits (like grain yield) largely influenced by non-additive gene actions. Therefore, to achieve a more accurate evaluation of the effects of a particular QTL on heterotic traits the investigation should, preferably, be carried out using near-isogenic hybrids (NIHs), which can be obtained by crossing BDLs at the same target region introgressed in different genetic backgrounds. According to the BDLs used as parents, NIHs are either homozygous or heterozygous at the target QTL region, while being heterozygous for most of the remaining portion of the genome. In this case, the availability of BDLs for the QTL near csu133 in the Os420 and IABO78 backgrounds made it possible to obtain NIHs in order to evaluate the effects of the QTL on heterotic traits.
The study presented here was conducted using two water regimes on the NIHs among such BDLs in order (i) to obtain an evaluation unbiased by inbreeding depression of the effects of the QTL on L-ABA and other physiological and agronomic traits, and (ii) to assess whether these effects are influenced by the level of water stress experienced by the plants grown in the field. Two pairs of BDLs were also tested in the greenhouse using three water regimes, in order (iii) to gather further information on the QTL effects in plants grown under controlled conditions, and (iv) to evaluate the QTL effects on root characteristics. This latter objective was suggested by a preliminary observation (Landi et al., 2005) that a pair of NIHs differed markedly for root lodging resistance.
|
Materials and methods |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
Plant materials
The development of the BDLs for the target QTL close to csu133 is summarized here; a more detailed description has been provided by Landi et al. (2005)
. Two parallel backcross programmes were developed by the cross Os420xIABO78. In one programme, Os420 was the recurrent parent and IABO78 the donor of the decreasing allele (–) for L-ABA at the target QTL; in the other programme, IABO78 was the recurrent parent and Os420 the donor of the increasing allele (+) for L-ABA. Five backcross cycles were followed by two selfing cycles, so that the BDLs finally developed were BC5F3 lines. Throughout the two parallel backcross programmes, two molecular markers flanking the target QTL were used to identify the desired genotypes. Starting from BC2, two families were grown for each backcross programme; each family was developed by one ear harvested in BC1 and was identified by the suffix .1 and .2. Consequently, four couples of BDLs were finally produced, i.e. Os420.1 (+/+) and (–/–), Os420.2 (+/+) and (–/–), IABO78.1 (+/+) and (–/–), and IABO78.2 (+/+) and (–/–). The four Os420 BDLs were factorially crossed to the four IABO78 BDLs, thus producing 16 near-isogenic hybrids (NIHs). At the target QTL, four of these NIHs were homozygous (+/+), four were homozygous (–/–), and eight were heterozygous (+/–).
Field evaluation of NIHs
The 16 NIHs were tested at Cadriano (44° 33' N lat., E long., Po Valley, Northern Italy) for two years (2002 and 2003) and with two water regimes. In each year, two trials were conducted in the same field; they were separated by a 4-m-wide alley and by three border rows on each side. The two trials differed only for the irrigation volumes, which corresponded either to 120% or to 40% of the estimated evapotranspiration after accounting for rainfall. The crop evapotranspiration was calculated by following the procedure described in detail by Landi et al. (1995). For each trial, the field design was a randomized complete block with four replicates; plots were 3.06 m long single rows separated by 0.80 m from adjacent rows.
Trials were sown on 29 April 2002 and 28 April 2003; c. 40 d later plots were thinned leaving 11 plants per plot (corresponding to 4.5 plants m–2). Plants were grown following the field practices locally adopted. In particular, fertilizer rates were 200 kg ha–1 of N (half applied before sowing and half after thinning) and 44 kg ha–1 of P (all before sowing); K was not applied because of its high availability in the soil. Weeds were controlled mechanically and by hand. Irrigation was applied from the mid-end of stem elongation to the end of silking (i.e. from V14-15 to R1 stage, according to Ritchie et al., 1997). Total irrigation volumes corresponding to 120% of evapotranspiration were 54 mm and 90 mm in 2002 and 2003, respectively, while the irrigation volumes corresponding to 40% of evapotranspiration were 18 mm and 30 mm, respectively. Trials were hand-harvested on 3 September 2002 and 3 September 2003; then, ears were dried down in an aired storage room and shelled when a constant moisture was attained.
The following traits were measured: (i) L-ABA, on leaf samples (third leaf from the top) collected at the mid-end of pollen shedding (i.e. between VT and R1 stages). Leaf samples were collected from 09.00 h to 10.00 h, immediately frozen and stored at –20 °C until ABA analysis was carried out. L-ABA was measured on duplicate samples per plot using an ABA-specific monoclonal antibody, as described in Tuberosa et al. (1998); (ii) leaf relative water content (RWC), measured on the same leaf sample used for L-ABA and following the procedure described by Sanguineti et al. (1999). RWC was computed as (fresh weight–dry weight)/(turgid weight–dry weight)x100; (iii) stomatal conductance (SC), analysed on the second leaf from the top at VT-R1 stage. Leaves were analysed from 09.00 h to 11.00 h; a steady-state porometer (model LI-1600 Li-Cor) was utilized following the procedure described by Sanguineti et al. (1999); (iv) leaf water potential (WP) measured at the same time and on the tip of the same leaf blades used for measuring SC; a pressure chamber was used following the procedure described in Landi et al. (2001). SC and WP were investigated only on two pairs of NIHs homozygous at the target QTL, i.e. Os420.1 (+/+)xIABO78.2 (+/+) and the counterpart Os420.1 (–/–)xIABO78.2 (–/–); Os420.2 (+/+)xIABO78.2 (+/+) and the counterpart Os420.2 (–/–)xIABO78.2 (–/–); (v) pollen shedding date, assessed when 50% of plants per plot had extruded anthers; (vi) anthesis-silking interval (ASI), as the difference between silking date (assessed when 50% of plants per plot had extruded silks) and the pollen shedding date; (vii) plant height, measured at the flag leaf collar; (viii) root lodging, consequent upon strong windstorms which occurred c.10 d after silking (i.e. the beginning of the R2 stage) in 2002 and at mid-stem elongation (V10-V11 stage) in 2003. Plants leaning more than 30° from the vertical were counted as lodged; (ix) grain yield, and (x) kernel weight (as a mean of 200 kernels) both adjusted to 15.5% moisture; and (xi) number of kernels per plant, calculated as the ratio between grain yield per plant and kernel weight. In each plot, leaf samples and data were collected on the central plants; in particular, the number of sampled plants was three for SC and WP, five for L-ABA, RWC, and plant height, and seven for flowering dates, root lodging, grain yield, and its components.
Greenhouse evaluation of BDLs
To obtain further information on the QTL effects on L-ABA concentration and morpho-physiological traits as well as to investigate the QTL effect on root characteristics in relation to the level of water stress, two pairs of BDLs [Os420.2 (+/+) and (–/–); IABO78.2 (+/+) and (–/–)] were tested in the greenhouse with three water regimes. Seeds were surface-sterilized and placed in Petri dishes at ambient temperature in the dark. After 48 h, seedlings at a similar germination stage were transferred in soil columns (diameter 44 cm, height 1 m) holding 27 kg of peat:sand sifted mixture 1:1 v/v for growing until flowering (VT-R1 stage). Twelve plants (i.e. four plantsxthree water regimes) were grown for each BDL under the following conditions, day: 16 h, 26–28 °C, with supplemental light 500 µE m–2 s–1 photosynthetic photon flux density; night: 16 °C. To achieve different levels of water stress, irrigation was interrupted at subsequent times (i.e. plants at the V8 stage to obtain the severe stress and at the V11 stage for the moderate stress) while irrigation was continued up to tassel appearance (i.e. a few days before VT) for the well-watered treatment. Plants were distributed according to a completely randomized design and their position was periodically changed.
At flowering (VT-R1), data were taken at the single plant level for: (i) L-ABA concentration and (ii) RWC on the central part of the third leaf from the top; (iii) WP on the tip of the third leaf from the top; (iv) plant height; and (v) shoot dry weight. Traits (i) to (iv) were measured as described for the NIHs in the field. To carry out root analyses, the soil level was marked on the stem, the plants were then removed from the pots, and the soil mix was gently removed from the roots by washing. For both the first node above the soil level (A) and the first node below the soil level (B) data were taken for (vi) root number per node (for node A, only elongated and branched roots were taken into account); (vii) root diameter at 5 mm from the stem; (viii) root length; (ix) root angle of growth from the vertical (0°) using a protractor; (x) root branching, scoring for the lateral roots emerging from the whole crown roots, from 0 (no branching) to 9 (maximum branching); (xi) dry weight of the whole root system (i.e. considering roots from all the above- and below-soil nodes); and (xii) root dry weight/shoot dry weight. Root diameter, length, and angle were measured on three random roots per node.
Statistical analysis
For field experiment on NIHs, the analysis of variance (ANOVA) was conducted on the plot mean values of each trial and then combined across trials. Prior to ANOVA, the data of root lodging percentage were subjected to angular transformation (Steel and Torrie, 1980). The variation among the four trials was partitioned as between years, between water regimes (i.e. 120 versus 40% irrigation volumes) and water regimexyear interaction. The variation among the 16 NIHs was analysed by following two alternative procedures providing complementary information. The first procedure was followed in order to obtain information on the average additive and dominance effects of the QTL in question. In particular, the variation among NIHs was partitioned into variation among the three groups of NIHs defined on the basis of their genotype at the target QTL [i.e. (+/+), (–/–) and (+/–)] and residual. The variation among the three groups was further subdivided into (+/+) versus (–/–), estimating the average additive effect of the QTL, and (+/–) versus [mean value of (+/+) and (–/–)], estimating the average dominance effect. The second procedure was followed in order to obtain information on the contribution of each BDL family to the average additive effect. In particular, the variation among the 16 NIHs was partitioned into effects of general combining ability (g.c.a.) due to the four parental BDLs derived by Os420, effects of g.c.a. due to the four parental BDLs derived by IABO78, and effects of specific combining ability (s.c.a.). It should be noted that the g.c.a. of a line is related to its mean performance across its hybrids, while the s.c.a. of a hybrid is related to the deviation between the observed and the expected performance based on the g.c.a. of the two parental lines (Falconer and Mackay, 1996). Effects of g.c.a. are mainly due to additive gene action while effects of s.c.a. are due to non-additive gene actions (mainly dominance). For both Os420 and IABO78 backgrounds, the g.c.a. effects were then partitioned into (+/+) versus (–/–) BDLs within the first family (thus estimating the additive effect of the QTL in the first family), (+/+) versus (–/–) BDLs within the second family (additive effect of the second family), and between families. All the interactions involving hybrids were partitioned following the two alternative procedures. The effects of years were considered as random and the effects due to irrigation volumes and hybrids as fixed; therefore, the F-test for irrigation volumes and hybrids was made by using, as denominator, the corresponding interactions with years in case such interactions were significant.
For SC and WP, which were investigated on only two pairs of NIHs, the variation among hybrids was partitioned following the first procedure only, i.e. as between (+/+) and (–/–) groups and residual.
With respect to the greenhouse experiment on BDLs, the ANOVA was conducted on the data collected on the 48 plants (i.e. on the single shoot or root data or the mean value of the three roots per node). The variation among the 12 entries (i.e. three water regimesxfour BDLs) was partitioned into among water regimes, among BDLs, and water regimexBDL interaction. The among BDLs was then partitioned into between families of BDLs, i.e. Os420.2 versus IABO78.2, between the two groups of BDLs differing for the QTL [i.e. (+/+) versus (–/–)], and the interaction (Os420.2 versus IABO78.2)x[(+/+) versus (–/–)]. For the F-test a fixed model was adopted and the among plants within entry was used as the error term.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
Field evaluation of NIHs
The ANOVA (not shown) indicated that the first and second order interactions involving years were significant only in a few instances; for this reason and for the sake of conciseness only mean values across years are presented and discussed. The comparison between the mean values of the two water regimes across years and hybrids was highly significant (P
0.01) for most traits, with the exception of RWC, pollen shedding date, and kernel weight (Table 1). As compared with the well-watered treatment, the water-stressed treatment led, as expected, to an increase in L-ABA and ASI as well as to a decrease in stomatal conductance (SC), leaf water potential (WP), plant height, root lodging, grain yield, and number of kernels per plant.
|
The water regimexNIH interaction was significant (P
0.05) only for L-ABA and due to magnitude effects rather than changes in rank; in fact, when the ANOVA was conducted on the L-ABA data previously subjected to logarithmic transformation (data not shown), the interaction was not significant. The significance of the water regimexNIH interaction of the L-ABA original data was solely due to the component water regimes x [(+/+) versus (–/–)]. Under well-watered conditions the mean values for the (+/+) and (–/–) groups of NIHs were 373 and 290 ng ABA g–1 DW, respectively, while under water-stressed conditions the corresponding values were 554 and 405 ng ABA g–1 DW, respectively (Table 2). Therefore, the average additive effects of the investigated QTL [calculated as half the difference between the means of the (+/+) and of the (–/–) groups] were 42 ng ABA g–1 DW for the well-watered and 75 ng ABA g–1 DW for the water-stressed conditions. These additive effects correspond to 12.5% (well-watered) and to 15.5% (water-stressed) as referred to the mean value of the two groups of NIHs in each water regime.
|
The differences among NIHs across years and water regimes were highly significant (P
0.01) for L-ABA, SC, and kernel weight, and significant (P 0.05) for root lodging; by contrast, for traits like plant height and grain yield differences were not significant. The comparison between the (+/+) and (–/–) groups largely accounted for most of the variation among NIHs for L-ABA (91.5%), for SC (87.8%), and for root lodging (73.7%), while the (+/+) and (–/–) groups had a much smaller effect on kernel weight (3.7%). The mean values of the (+/+) and (–/–) groups are presented in Table 2. The average additive effects calculated from these mean values were 58 ng g–1 DW for L-ABA [corresponding to 14.3% of the (+/+) and (–/–) mean], –0.03 cm s–1 for SC, –12.9% for root lodging, and –4 mg for kernel weight. For all traits, the mean value of the (+/–) group of NIHs did not significantly differ from the mean of the (+/+) and (–/–) groups (Table 2), thus indicating negligible dominance effects for the QTL in question.
The alternative procedure followed in order to analyse the differences among the 16 NIHs, pointed out that the significance of such differences was due only to g.c.a. effects of the parental BDLs, because the s.c.a. effects were not significant for any trait. This latter finding further suggests that the target QTL does not exert important dominance effects for the investigated traits, consistent with the results seen previously. Because of the large prevalence of g.c.a. effects, the mean values of each NIH are not reported and only the mean values of the parental BDLs across their hybrids are reported in Table 3 for those traits which showed significant differences among NIHs (BDLs' mean values for SC are not reported in Table 3, despite the highly significant differences among NIHs, because only four NIHs were tested). For L-ABA, the differences between the (+/+) and (–/–) BDL mean values within each family (representing the additive effects) were always highly significant and positive, thus indicating that the four families consistently contributed to the average additive effect (58 ng ABA g–1 DW) previously reported.
|
For root lodging, the (+/+) versus (–/–) BDLs within each family was significant or highly significant for all families except for IABO78.2; however, all the additive effects were negative and of rather similar values, indicating that, analogously to L-ABA, the four families consistently contributed to the average additive effect (–12.9%) previously reported.
For kernel weight, the difference between the (+/+) and (–/–) BDLs within family was positive and statistically negligible for the two Os420 families, negative and statistically negligible for the IABO78.1 family, and negative and highly significant for the IABO78.2 family. These findings indicate a lack of consistency among such differences and that the average additive effect previously seen (i.e. –4 mg) was mainly due to the contribution of only one family (IABO78.2).
For L-ABA, the relationship was investigated between the 16 NIHs tested here and the corresponding parental means of the BDLs tested per se in our previous study (Landi et al., 2005). The correlation coefficient was highly significant under well-watered (r=0.65) and water-stressed conditions (r=0.82), as well as across the two water regimes (r=0.77). These data thus indicate that the capacity to predict the NIHs' performance, based on the performance of their parental BDLs, is quite satisfactory (especially under the water-stressed conditions; r2=67.2%); this is consistent with the prevalence of additive gene action already mentioned for the target QTL.
The relationship (across the two water regimes) between the L-ABA mean value of each hybrid and the corresponding root lodging mean value was also investigated. The correlation coefficient was sizeable and highly significant (r= –0.88), indicating that a large proportion (r2=76.9%) of the variability among NIHs for one trait was accounted for by its linear relationship with the other.
Greenhouse evaluation of BDLs
The ANOVA (not shown) indicated that the comparison between the mean values of the three water regimes was highly significant (P 0.01) or significant (P 0.05) for several traits. As expected, the water-stressed treatments in comparison with the well-watered treatment (Table 4) led to an increase in L-ABA (i.e. 228, 756, 1078 ng g–1 DW in the well-watered, moderate stress, and severe stress conditions, respectively) and to a decline in RWC, WP, and plant height. The level of water stress also affected the number of roots per node A (i.e. first node above the soil level) and the root length for node A, leading to a decline for both traits. The interaction of water regimexBDL was not significant for any trait, with the exception of L-ABA for which an interaction due to magnitude effects was observed, analogous to what was found for NIHs in the field. In fact, differences among BDLs for L-ABA were larger with the two water-stressed treatments than with the well-watered treatment (for conciseness data are not shown).
|
The comparison between the families of BDLs, i.e. Os420.2 versus IABO78.2, was highly significant for L-ABA (792 versus 582 ng g–1 DW, respectively), diameter of the roots for both node A (4.9 versus 3.5 mm, respectively), and B (3.3 versus 2.1 mm, respectively) and root branch (5.3 versus 4.3). For all the other traits the differences between the two families of BDLs were not significant.
The comparison between the (+/+) and (–/–) groups of BDLs across the Os420 and IABO78 genetic backgrounds was significant or highly significant for L-ABA and several root traits (Table 5). As to L-ABA, the (+/+) group of BDLs showed, as expected, a higher mean value than the (–/–) group, i.e. 783 versus 591 ng g–1 DW, with a QTL additive effect of 96 ng ABA g–1 DW, or 14.0% of the overall mean. The (+/+) group also exhibited a higher mean value for the number of roots per node B (4.8 versus 3.8), root diameter for both node A (4.5 versus 3.9 mm) and B (3.0 versus 2.5 mm), root angle for node A (48.0 versus 41.9°), root branching score (5.2 versus 4.4), whole-root dry weight (8.7 versus 5.8 g), and ratio between this latter trait and shoot dry weight (0.26 versus 0.19); therefore, the additive effect for all these traits was always positive, consistently with the positive effect found for L-ABA. Although differences between the two groups of BDLs were not significant for the number of roots per node A and the root angle for node B, it is noteworthy that the (+/+) group of BDLs showed higher values for both traits, thus strengthening the positive additive effect detected for the number of roots per node B and the root angle for node A. The only root trait showing non-significant differences for both node A and B was root length; however, it should be mentioned that the coefficients of variation for root length of both nodes were close to 20%, suggesting that the lack of significant differences (especially for node A) could be due to the high level of environmental variation affecting such traits. The superiority for the whole root mass of the (+/+) group compared with the (–/–) group can also be appreciated from Fig. 1 as it shows the root system at flowering of each of the four BDLs grown with moderate water stress.
|
|
The interaction (Os420.2 versus IABO78.2)x[(+/+) versus (–/–)] was not significant for any trait, indicating that the QTL effect did not substantially change from one genetic background to the other.
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
QTL effects on L-ABA and physiological traits
Consistently with the well-known effects of water stress on maize, most of the investigated traits were significantly affected by the water regimes applied in the field and in the greenhouse, thus indicating the adequacy of such water regimes for the objectives of the present work. Interestingly, for both NIHs and BDLs the interaction water regimexgenotype and its component water regimex[(+/+) versus (–/–)] were always negligible, the only exception being L-ABA, for which a change in magnitude of the differences among genotypes, or between (+/+) and (–/–) groups of genotypes, was detected. The negligible role of water stress on the QTL additive effect for L-ABA and associated traits was also pointed out when the eight BDLs were tested per se in the field (Landi et al., 2005
). Altogether, these results indicate that the QTL additive effect on L-ABA and on the associated physiological traits is not much affected by the intensity of water stress, at least within the levels attained in this study. Therefore, in accordance with what was discussed by Blum (1996) concerning the distinction between constitutive and adaptive traits, the low level of the water regimexgenotype interaction detected here indicates that the effects of this QTL on the investigated traits are generally constitutive rather than adaptive. On a breeding basis, a QTL with constitutive effects on a number of traits influencing the water balance and water use of the plant may contribute alleles that provide, more predictably, advantages under varying environmental conditions and rainfall patterns/water regimes, as recognized by Blum (1996, 2002). Recently, Tardieu and co-workers have applied a modelling approach in maize to investigate the constitutive versus the adaptive effects of QTLs on morphological traits in relation to environmental parameters (Reymond et al., 2003; Tardieu, 2003). In this respect, it will be valuable to follow a similar approach to ascertain, in greater detail, the influence of environmental parameters on the QTL effects for morpho-physiological traits (e.g. root and leaf elongation, pollen sterility, ovary abortion, etc) affected to varying degrees by ABA and which have been shown to influence final grain yield (Saini and Westgate, 2000; Tuberosa et al., 2002; Reymond et al., 2003).
The QTL additive effect for L-ABA averaged across water regimes was always sizeable and, when referred to the genotypes' mean, was c. 14% for both the NIHs in the field and the BDLs in the greenhouse. The average additive effect of the QTL was c. 11% for the BDLs grown in the field in a previous study (Landi et al., 2005) and 12% in the study conducted on 80 random F3:4 lines obtained from the same cross (i.e. Os420xIABO78; Tuberosa et al., 1998). All these findings consistently indicate that the QTL additive effect for L-ABA is stable, irrespective of the vigour of the tested materials, i.e. of their inbreeding coefficients (which should be equal to 0 for NIHs, 0.88 for F3:4 lines, and should be close to 1 for the BDLs). Conversely, the dominance effect (as pointed out by the study on NIHs) was negligible. The much greater importance of the additive versus the dominance effect fully accounts for the marked changes of the allelic frequencies at the QTL following a divergent selection for L-ABA on the source F2 of the cross Os420xIABO78 (Landi et al., 2005).
With respect to the associated effects on physiological traits, the additive effect of the QTL was negative for SC, consistent with the well-known role of ABA in reducing SC (Quarrie, 1991). Interestingly, the additive effect estimated on four NIHs (–0.03 cm s–1) was the same as that estimated on 80 random F3:4 lines investigated by Sanguineti et al. (1999), thus indicating the stability of the QTL effect across generations, not only for L-ABA but also for SC.
Contrary to SC, no associated effects of the QTL were found for RWC and WP. As to RWC, the lack of associated effect is in contrast with the findings of the study on the eight BDLs in the field in which a significant and negative additive effect of the QTL was detected (Landi et al., 2005). At least in part, such a discrepancy might be due to differences among the dynamics of drought episodes experienced by the plants from one environment to the other, especially with respect to the intensity of the stress and/or to the plants' growing stage in which it occurred. Additional factors possibly involved are represented by the level of the soil water table and the intensity of evapotranspiration during the RWC measurements. For WP, the results presented here are consistent with those reported by Pekic et al. (1995) and by Landi et al. (2001), who found that materials differing in L-ABA did not significantly differ for WP. This lack of relationship between L-ABA and WP is not surprising because maize, an isohydric species (Tardieu, 1996), maintains a stable water status at varying stress conditions by regulating SC.
QTL effects on agronomic traits
With respect to the agronomic traits evaluated on the NIHs, the QTL showed a strong negative additive effect on root lodging, as the (+/+) group of NIHs was much less susceptible than the (–/–). This finding is consistent with what has been reported previously in a preliminary study on a pair of NIHs (Landi et al., 2005), thus suggesting that this associated effect is also quite stable across environments. It is well-known that root lodging in maize depends on both root architecture affecting the strength of root anchorage to the soil and the leverage effect exerted by the wind force on the plant. Because no significant differences between the (+/+) and (–/–) NIHs were detected in both the preliminary and the present study for plant height, it seems plausible that such differences in root lodging were mainly, if not solely, due to differences in root architecture.
In contrast to root lodging, the QTL had no significant effect on grain yield. This finding is consistent with the preliminary results of Landi et al. (2005) and would suggest that grain yield is not much affected by the QTL in question. However, considerable caution should be exerted in this regard because, in all the field trials considered here and by Landi et al. (2005), root lodging occurred before flowering or just a few weeks after flowering, i.e. when grain yield potential was still not fully determined. Therefore, the (–/–) NIHs, more heavily affected by root lodging than the (+/+) NIHs, were probably more penalized in terms of grain yield because of the negative effect that can be exerted on this latter trait by root lodging (Carter and Hudelson, 1988). This implies that, in the absence of root lodging, the (–/–) NIHs could attain a higher grain yield than the (+/+). This hypothesis is supported by the findings of several studies (reviewed by Saini and Westgate, 2000) reporting a negative association between ABA levels in cereals with seed set and, hence, with grain yield. In addition, a causal role of ABA has also been suggested for ovary abortion, the most important factor determining final grain yield in maize exposed to a water deficit during the reproductive phase, i.e. when the ovary is particularly sensitive to a decline in the supply of sugars (Boyle et al., 1991; Zinselmeier et al., 1999; Boyer and Westgate, 2004).
QTL effects on root traits
The analysis conducted on the two pairs of BDLs in the greenhouse revealed that the investigated QTL also controls several root characteristics, with similar effects in both genetic backgrounds of Os420 and IABO78. Across the two backgrounds, the QTL positively affected root number, diameter, angle, branching, dry weight, and the ratio between root and shoot dry weight. This substantiates the hypothesis formulated on the basis of field data, i.e. that differences among NIHs for root lodging are mainly due to differences in root architecture affecting the strength of plant anchorage. In fact, several studies have shown that root strength is positively affected by characteristics such as root number, diameter, angle, and weight (Jenison et al., 1981; Ennos et al., 1993; Guingo and Hébert, 1997; Bruce et al., 2001). Moreover, the higher root branching shown in the greenhouse experiment by the (+/+) group versus the (–/–) group coupled with a wider insertion angle of the roots on the main stalk may have been paralleled by a greater root density in the shallower soil layers under field conditions, further increasing the anchorage strength. In this regard, it is worth mentioning that Bolaños et al. (1993) found that maize populations with higher root density in the shallow soil layers were also characterized by a higher root strength. The involvement of the QTL near csu133 in the control of both L-ABA and root strength was also revealed by Lebreton et al. (1995), based on the analysis of F2 plants derived from the cross Polj17xF-2. Interestingly also in that study, the additive effects of the QTL on L-ABA and anchoring strength were concurrent. On a broader scale and exploiting syntenic information, Landi et al. (2005) compared the QTL results for root traits of four maize populations with the root QTL data of seven rice (Oryza sativa, L.) populations. The highest frequency of QTLs for root characteristics in rice was observed in the region syntenic to the maize bin 2.04, further supporting the involvement of this chromosome region in the control of root characteristics. Moreover, this finding suggests that rice could be used as a model species to facilitate the positional cloning of the gene(s) responsible for this maize QTL.
Hypotheses on the genetic associations among traits
Until the cloning of the QTL on bin 2.04 is completed, it will not be possible to ascertain to what extent the effects on all the above-mentioned traits are due to linkage between the gene(s) for L-ABA and the gene(s) for the associated traits and/or to the pleiotropic action of one or more genes. However, the consistency of all such effects and their similarity with those detected in materials unrelated to those considered here (Lebreton et al., 1995) suggest that pleiotropy is probably involved in this respect. The assumption could be that the QTL marked by csu133 directly controls one trait only, i.e. root architecture, while affecting L-ABA and other traits through a sequence of causally related events, according to the general model described by Lebreton et al. (1995) and Tuberosa et al. (2002). In this respect, it should be mentioned that a preliminary investigation on the two pairs of BDLs tested in the greenhouse showed that the QTL had no significant effects on L-ABA before the V13 stage (S Giuliani, unpublished data). Conversely, in 2003, differences among NIHs for root lodging were already noticeable at an earlier stage (i.e. V10-V11). Therefore, based on these observations, this model hypothesizes that the (+) allele provided by Os420, as compared to the (–) allele of IABO78, may determine a larger root system with a wider insertion angle (i.e. a more horizontal root development). This, in turn, may increase root density in the superficial soil layers, thus accounting for the greater root lodging resistance provided by the Os420 allele. In addition, because the more superficial soil layers dehydrate more quickly, even under irrigated conditions, a larger and more horizontal root system implies a greater flux of xylem ABA towards the leaf, thus accounting for the higher L-ABA and lower SC of the (+/+) NIHs. Therefore, it is suggested that the QTL on bin 2.04 should be named root-ABA1, to emphasize its involvement in the control of both root architecture and L-ABA.
|
Conclusions |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
These results indicate that root-ABA1 exerts an important effect on L-ABA and that this effect is stable across various levels of water stress and of plant vigour of the materials tested. Moreover, this QTL exerts important effects on root architecture, root lodging, and stomatal conductance. Such associated effects are probably due to pleiotropy, with root-ABA1 directly influencing root architecture and growth, particularly in the more superficial soil layers; the effects on L-ABA and the other traits then follow as a sequence of causally related events. To elucidate the genetic basis of these associated effects, the fine mapping of the QTL will be undertaken using as base materials crosses between (+/+) and (–/–) BDLs in the Os420 background as well as in the IABO78 background. The fine mapping of root-ABA1 is an essential prerequisite to undertake its positional cloning, which would represent an important contribution towards a more effective and accurate manipulation of root architecture, a complex trait involved in crops' adaptation to drought and other abiotic stresses.
|
Acknowledgements |
|---|
This work was supported by MIUR, COFIN 40%, Project ‘Analysis of the molecular and phenotypic effects associated with the variation in abscisic acid concentration in leaves of near isogenic lines’. This paper is a contribution of the Interdepartmental Centre of Biotechnology of the University of Bologna. Thanks are due to Sandra Stefanelli and Stefano Vecchi for skilful technical assistance.
|
Footnotes |
|---|
Abbreviations: ASI, anthesis-silking interval; BDL, backcross-derived line; g.c.a., general combining ability; L-ABA, leaf abscisic acid concentration; NIH, near-isogenic hybrid; NIL, near-isogenic line; node A, first node above the soil level; node B, first node below the soil level; QTL, quantitative trait locus; RFLP, restriction fragment length polymorphism; RWC, relative water content; SC, stomatal conductance; s.c.a., specific combining ability; WP, leaf water potential.
|
References |
|---|
|
TopAbstractIntroductionMaterials and methodsResultsDiscussionConclusionsReferences |
|---|
Blum A. 1996. Crop responses to drought and the interpretation of adaptation. Plant Growth Regulation 20, 135–148.
Blum A. 2002. Drought tolerance: is it a complex trait? In: Saxena NP, O'Toole JC, eds. Field screening for drought tolerance in crop plants with emphasis on rice. International Workshop on Field Screening for Drought Tolerance in Rice. ICRISAT, Patancheru, India, 17–22.
Bolaños J, Edmeades GO, Martinez L. 1993. Eight cycles of selection for drought tolerance in lowland tropical maize. III. Responses in drought adaptive physiological and morphological traits. Field Crops Research 31, 269–286.[CrossRef]
Boyer JS, Westgate ME. 2004. Grain yields with limited water. Journal of Experimental Botany 55, 2385–2394.
Boyle MG, Boyer JS, Morgan JW. 1991. Stem infusion of liquid culture medium prevents reproductive failure of maize at low water potential. Crop Science 31, 1246–1252.
Bruce W, Desbons P, Crast O, Folkerts O. 2001. Gene expression profiling of two related maize inbred lines with contrasting root-lodging traits. Journal of Experimental Botany 52, 459–468.
Carter PR, Hudelson KD. 1988. Influence of simulated wind lodging on corn growth and grain yield. Journal of Productive Agriculture 1, 295–299.
Ennos AR, Crook MJ, Grimshaw C. 1993. The anchorage mechanics of maize, Zea mays. Journal of Experimental Botany 44, 147–153.
Falconer DS, Mackay TFC. 1996. Introduction to quantitative genetics, 4th edn. London, UK: Longman.
Guingo E, Hébert Y. 1997. Relationship between mechanical resistance of the maize root system and root morphology, and their genotypic and environmental variations. Maydica 42, 265–274.
Jenison JR, Shank DB, Penny LH. 1981. Root characteristics of 44 maize inbreds evaluated in four environments. Crop Science 21, 233–237.
Landi P, Conti S, Gherardi F, Sanguineti MC, Tuberosa R. 1995. Genetic analysis of leaf ABA concentration and agronomic traits in maize hybrids grown under different water regime. Maydica 40, 179–186.
Landi P, Sanguineti MC, Conti S, Tuberosa R. 2001. Direct and correlated responses to divergent selection for leaf abscisic acid concentration in two maize populations. Crop Science 41, 335–344.
Landi P, Sanguineti MC, Salvi S, Giuliani S, Bellotti M, Maccaferri M, Conti S, Tuberosa R. 2005. Validation and characterization of a major QTL affecting leaf ABA concentration in maize. Molecular Breeding 15, 291–303.[CrossRef]
Lebreton C, Lazic-Jancic V, Steed A, Pekic S, Quarrie SA. 1995. Identification of QTL for drought responses in maize and their use in testing causal relationships between traits. Journal of Experimental Botany 46, 853–865.
Paran I, Zamir D. 2003. Quantitative traits in plants: beyond the QTL. Trends in Genetics 19, 303–306.[CrossRef][Web of Science][Medline]
Pekic S, Stikic R, Tomljanovic L, Andjelkovic V, Ivanovic M, Quarrie SA. 1995. Characterization of maize lines differing in leaf abscisic acid content in the field. 1. Abscisic acid physiology. Annals of Botany 75, 67–73.[CrossRef]
Quarrie SA. 1991. Implications of genetic differences in ABA accumulation for crop production. In: Davies WJ, Jones HG, eds. Abscisic acid: physiology and biochemistry. Oxford: Bios Scientific Publishers, 227–243.
Reymond M, Muller B, Leonardi A, Charcosset A, Tardieu F. 2003. Combining quantitative trait loci analysis and an ecophysiological model to analyze the genetic variability of the responses of maize leaf growth to temperature and water deficit. Plant Physiology 131, 664–675.
Ritchie SW, Hanway JJ, Benson GO. 1997. How a corn plant develops. Iowa State University of Science and Technology Cooperative Extension Service, Ames, Iowa. Special Report 48, 1–21 (reprinted).
Saini HS, Westgate ME. 2000. Reproductive development in grain crops during drought. Advances in Agronomy 68, 59–96.[Web of Science]
Sanguineti MC, Conti S, Landi P, Tuberosa R. 1996. Abscisic acid concentration in maize leaves: genetic control and response to divergent selection in two populations. Maydica 41, 193–203.
Sanguineti M, Tuberosa R, Landi P, Salvi S, Maccaferri M, Casarini E, Conti S. 1999. QTL analysis of drought-related traits and grain yield in relation to genetic variation for leaf abscisic acid concentration in field-grown maize. Journal of Experimental Botany 50, 1289–1297.
Sharp RE, Poroyko V, Hejlek L, Spollen WG, Springer GK, Bohnert H J, Nguyen HT. 2004. Root growth maintenance during water deficits: physiology to functional genomics. Journal of Experimental Botany 55, 2343–2351.
Steel RGD, Torrie JH. 1980. Principles and procedures of statistics a biometrical approach, 2nd edn. New York: McGraw-Hill Book Company.
Tardieu F. 1996. Drought perception by plants. Do cells of droughted plants experience water stress? Plant Growth Regulation 20, 93–104.
Tardieu F. 2003. Virtual plants: modelling as a tool for the genomics of tolerance to water deficit. Trends in Plant Science 8, 9–14.[CrossRef][Web of Science][Medline]
Tuberosa R, Sanguineti MC, Landi P. 1994. Abscisic acid concentration in the leaf and xylem sap, leaf water potential, and stomatal conductance in drought-stressed maize. Crop Science 34, 1557–1563.
Tuberosa R, Sanguineti MC, Landi P, Salvi S, Casarini E, Conti S. 1998. RFLP mapping of quantitative trait loci controlling abscisic acid concentration in leaves of drought-stressed maize (Zea mays L.). Theoretical and Applied Genetics 97, 744–755.[CrossRef]
Tuberosa R, Salvi S, Sanguineti MC, Landi P, Maccaferri M, Conti S. 2002. Mapping QTLs regulating morpho-physiological traits and yield in drought-stressed maize: case studies, shortcomings and perspectives. Annals of Botany 89, 941–963.
Wang Z, Mambelli S, Setter TL. 2002. Abscisic acid catabolism in maize kernels in response to water deficit at early endosperm development. Annals of Botany 90, 623–630.
Zinselmeier C, Jeong BR, Boyer JS. 1999. Starch and the control of kernel number in maize at low water potentials. Plant Physiology 121, 25–36.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  G. L. Hammer, Z. Dong, G. McLean, A. Doherty, C. Messina, J. Schussler, C. Zinselmeier, S. Paszkiewicz, and M. Cooper Can Changes in Canopy and/or Root System Architecture Explain Historical Maize Yield Trends in the U.S. Corn Belt? Crop Sci., January 28, 2009; 49(1): 299 - 312. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 N. C. Collins, F. Tardieu, and R. Tuberosa Quantitative Trait Loci and Crop Performance under Abiotic Stress: Where Do We Stand? Plant Physiology, June 1, 2008; 147(2): 469 - 486. [Full Text] [PDF]
|
|
|
|
|
|
R. Tuberosa, S. Salvi, S. Giuliani, M. C. Sanguineti, M. Bellotti, S. Conti, and P. Landi Genome-wide Approaches to Investigate and Improve Maize Response to Drought Crop Sci., December 18, 2007; 47(Supplement_3): S-120 - S-141. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P Landi, M. Sanguineti, C Liu, Y Li, T. Wang, S Giuliani, M Bellotti, S Salvi, and R Tuberosa Root-ABA1 QTL affects root lodging, grain yield, and other agronomic traits in maize grown under well-watered and water-stressed conditions J. Exp. Bot., January 1, 2007; 58(2): 319 - 326. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||