Printer Friendly

Characterization of a yield quantitative trait locus on chromosome five of maize by fine mapping.

The concept of using genetic markers to explain quantitative trait variation is not a new idea (Sax, 1923; Lindstrom, 1926; Thoday, 1961). However, in those earlier studies, very few useful markers were available. The fairly recent advancements in the development of DNA-based genetic markers [RFLPs, randomly amplified polymorphic DNAs (RAPDs), simple sequence repeats (SSRs), amplified fragment length polymorphisms (AFLPs), etc.] now provide for marker saturation of nearly any experimental genome. Perhaps the biggest advantage of such comprehensive coverage is the ability to generate experimental lines with the desired genomic constitutions, which can be used to test or explain previous observations.

Stuber et al. (1992) reported the results of a study to identify heterotic factors in a maize hybrid. In this study, they derived 264 [F.sub.3] lines from the cross of the two elite inbred lines, B73 and Mo17. Each [F.sub.3] line was backcrossed to the two parents to create 528 backcross families that were evaluated in six diverse environments. One of the major results of this study was the identification of a large QTL for grain yield near the marker Amp3 on chromosome 5. In the backcross to B73, this chromosomal region accounted for about 20% of the phenotypic variation in yield. This was associated with a 680 kg [ha.sup.-1] effect measured as the difference between homozygous and heterozygous genotypic classes in the backcross families. In the backcross to Mo17, this region accounted for 13% of the total variation, a 609 kg [ha.sup.-1] effect. Also, in 16 of more than 20 populations of maize studied in our program at Raleigh, NC, the region between Pgm2 and Amp3 on chromosome 5 has showed significant associations with grain yield (Abler et al., 1991; Guffy et al., 1988; Ragot et al., 1995; Stuber et al., 1987, 1992; and unpublished data, 1994). The importance of this region in determining yield in most populations evaluated, and in both backcrosses in the Stuber et al. (1992) study, provided the impetus to further investigate this chromosomal segment.

Because phenotypes were measured on backcross families in the Stuber et al. (1992) study described above, QTLs were declared present if the effect of substituting one allele for another was significant. In the backcross to Mo17, for example, a significant difference between the marker genotypes MM and MB, where M and B are alleles contributed by Mo17 and B73 at some specific locus, respectively, suggested that a QTL was present in that chromosomal region. Likewise, in the backcross to B73, a significant difference between BB and BM genotypes at a particular marker locus indicated a significant allelic substitution effect. As previously mentioned, a QTL near Amp3 on chromosome 5 was significantly associated with grain yield in both backcrosses, and in both cases the heterozygote marker class (MB or BM) was superior to either homozygote (MM or BB) class. In the backcross to Mo17, the mean grain yield of backcross families with a MB genotype at Amp3 was greater than the yield of the MM families. Similarly, in the B73 backcross, the yield of the BM families was greater than that of the BB families. Therefore, this region acts in a truly heterotic manner.

Two possible explanations appear plausible for the above observations, both of which would be consistent with the current heterosis models. The simplest model would include one locus with overdominant gene action, with the heterozygote always being phenotypically superior to either homozygote. The second model would include two partially or completely dominant loci in repulsion phase linkage (pseudo-overdominance as defined by Crow, 1952). In this model, at one locus the Mo17 allele is dominant and favorable; at the other linked locus, the B73 allele is dominant and favorable. In the backcross to Mo17, only the locus where the B73 allele is dominant would be detected because the tester (Mo17) would mask the phenotypic expression of the second locus (where the Mo17 allele is dominant) by always providing a Mo17 allele. Consequently, the genotype at the second locus would not be relevant in this cross. The opposite situation would be true in the B73 backcross. More detailed resolution of QTLs should allow for better manipulation in marker-assisted selection (MAS) programs and, more importantly, may provide insight into the inheritance and expression of such loci.

The relative importance of dominance versus overdominance has been discussed extensively since the discovery of heterosis. Shull (1908) proposed that physiological stimulation is a consequence of heterozygosity resulting in superior phenotypes. This idea later became a central point of Hull's (1945) overdominance theory. The dominance theory of heterosis as proposed by Bruce (1910) and Keeble and Pellew (1910) is based on the complementary effect of dominant factors introduced from each parent, resulting in progeny with higher phenotypic values than either parent. Jones (1917) extended this concept of dominant favorable factors to include linkage, which probably is still the most popular explanation of the genetic basis of heterosis.

Using mathematical proofs, Crow (1952) showed that dominance alone could not account for observed heterosis; this led many to believe that overdominance was the .cause. Early studies in maize, using the North Carolina Design III (Comstock and Robinson, 1952), suggested that overdominance was a major factor in the expression of heterosis. However, further studies conducted after random mating the populations used in the Design III studies showed that overdominance was not important, and that this psuedo-overdominance effect could be attributed to multiple favorable dominant factors in repulsion phase linkage (Gardner and Lonnquist, 1959). Hallauer and Miranda (1981) summarized the overdominance versus dominance controversy by concluding " seems that evidence supports the hypothesis that heterosis results from an accumulation of dominant favorable growth factors."

With the use of properly designed genetic materials, and marker technology, it is now possible to dissect regions showing significant QTLs into small segments, which may provide a means to distinguish between the two heterosis models outlined above. For example, Paterson et al. (1990) described a method in which they were able to define a QTL to a region as small as 3 cM. This technique is similar to deletion analysis except that it uses chromosomal recombinants rather than deletions to map the trait of interest. The method involves creating a set of near isogenic lines (NILs) containing overlapping introgressed donor segments at the putative QTL position. These lines are then evaluated in replicated field trials. The targeted QTL can then be located with respect to the shared introgregsed chromosomal segment associated with extreme phenotypes.

With two primary goals in mind, we adapted the Paterson et al. (1990) method to our genetic stocks and created two sets of [BC.sub.2][S.sub.1] NILs focusing on recombination events in the region on chromosome 5 associated with Amp3. The first goal was to evaluate the targeted region on chromosome 5 in more detail, which included identifying and characterizing the independent genetic factors in this region. The second was to define QTLs to the smallest possible genomic segment. Achieving these goals was limited by the number of markers available in the targeted region and the number of NILs generated.


Genetic Material

The crossing and selfing procedures for developing the [BC.sub.2][S.sub.1] lines for this study are diagrammed in Fig. 1 for a line in the Mo17 background. Two [F.sub.3] lines from the Stuber et al. (1992) study were identified with a high proportion (approximately 60% based on 76 RFLP and isozyme markers) of Mo17 background and contained the B73 chromosomal segment between Amp3 and Pgm2 on chromosome 5. As shown in Fig. 1, these lines were backcrossed twice to Mo17 and then self-pollinated to create 50 [BC.sub.2][S.sub .1] NILs. During these crossing and selfing procedures, 109 lines were genotyped with Amp3 and Pgm2 to select for appropriate recombinant types of the introgressed donor segment (B73) in the targeted region on chromosome 5; six other isozyme markers on chromosomes 1, 3, 6, and 8 were also used to assist in recovering the recurrent parent (Mo17). The Mo17 background [BC.sub.2][S.sub.1] lines with an introgressed B73 segment will be referred to as Mo17(B73)[BC.sub.2][S.sub.1] lines (abbreviated M/Bseg). A similar procedure was used to create a set of 35 lines in a B73 background containing appropriate recombinant types with the introgressed Mo17 segment. These lines will be referred to as the B73(Mo17)[BC.sub.2][S.sub.1] lines (abbreviated B/Mseg) and were derived from one [F.sub.3] line with an approximately 60% B73 background. Both sets of [BC.sub.2][S.sub.1] lines were testcrossed to both B73 and Mo17. The original [F.sub.3] lines selected contained a high proportion (about 60%) of the genome of the recurrent parent, therefore the proportion of the recurrent genome in the [BC.sub.2][S.sub.1] lines was expected to average 90%. Because each line was developed independently, donor segments not at the targeted region on chromosome 5 would be expected to occur randomly. Consequently, the chance occurrence of a common donor segment (at a location other than the targeted region) in a significant number of [BC.sub.2][S.sub.1] lines is extremely small. If this rare event did occur, and if the unidentified donor fragment also contained a QTL affecting the trait being evaluated, the effect of the QTL would be attributed to the chromosome 5 location.


For comparisons of genotypic means among marker classes in the targeted region on chromosome 5, it should be noted that the number of lines included in each class varied with the position of the marker. As expected, the number of lines with the donor class genotype was less near the ends of the introgressed segment than near the center. For the BNL5.40 marker (Fig. 2), there were only three homozygous donor and seven heterozygous [BC.sub.2][S.sub.1] lines across all backgrounds. However, at all other loci, there was a minimum of 15 homozygous donor and 14 heterozygous [BC.sub.2][S.sub.1] lines for use in making appropriate comparisons.


Genotyping Procedures

All [BC.sub.2][S.sub.1] lines were genotyped in the targeted region on chromosome 5 by means of 16 RFLP probes and two isozyme markers. To ensure proper order and distance estimates, these markers were placed on a set of 192 recombinant inbred lines developed from the cross of B73 x Mo17 (Senior et al., 1996) and were mapped with Mapmaker EXP 3.0 (Lincoln et al., 1993). All DNA extractions, gel electrophoresis, Southern blotting, oligolabelling, hybridizations, and isozyme analyses were done according to Sambrook et al. (1989), Saghai-Maroof et al. (1984), and Stuber et al. (1988).

Experimental Design and Statistical Analyses

The 85 testcrosses and appropriate checks were evaluated at experiment stations near Clayton, Plymouth, and Lewiston, NC. Experimental plots consisted of two rows, 4.6 m long. Interrow spacing was 0.97 m at Clayton and Plymouth and 0.91 m at Lewiston. To provide a broader range of test environments, two planting densities were used at each location, 72 000 and 36 000 plants [ha.sup.-1]. This effectively increased the number of test environments to six, two densities at each of three locations.

The 50 M/Bseg x Mo17 and 35 B/Mseg x B73 testcross progenies plus 11 checks were arranged in a sets-in-replications design, with one replication of each density at each location. Set 1 contained 15 M/Bseg x Mo17 testcrosses and one Mo17 inbred check. Sets 2 and 3 each had 14 M/Bseg x Mo1.7 testcrosses and two Mo17 checks. Set 4 was a mixture of M/Bseg x Mo17 and B/Mseg x B73 testcrosses containing seven of each, one Mo17 check, and one B73 check. Sets 5 and 6 each contained 14 B/Mseg x B73 testcrosses and two B73 inbred checks. These testcross progenies were evaluated together because they were presumably inbred throughout the genome except at the targeted chromosome 5 region. The cheeks in each set corresponded to the tester line used.

The experimental design for the 50 M/BSeg X B73 testcrosses and the 35 B/Mseg X Mo17 testcrosses plus 11 checks was identical to that described above, except that the check used was the B73 x Mo17 [F.sub.1] hybrid in all sets. This hybrid check was used in the evaluation of these testcrosses because all of the experimental materials should be heterozygous at most loci (except the targeted chromosome 5 region) and were expected to yield similarly to the B73 x Mo17 hybrid.

Traits measured included grain yield (kg [ha.sup.-1], adjusted to 150 g [kg.sup.-1] grain moisture), ear height (m), plant height (m), leaf area ([cm.sup.2]), ear length (cm), number of kernels per row, kernel weight (g), kernel length (mm), and kernel width (ram). Prior to machine harvest, five plants from each plot were harvested by hand to measure ear length, number of kernels per row, kernel weight, kernel length, and kernel width. After the remainder of each plot was machine harvested, the weight of the grain from the five hand-harvested ears was added to the machine harvested grain weight to determine the total plot weight. Leaf area was calculated from measurements of the top ear leaf on five plants in each plot. Ear height was taken as the height to the ear node of the top ear and was measured on 10 plants in each plot. Plant height was measured from the soil to the tip of the tassel. Ear height, plant height, kernel weight, kernel length, and kernel width did not show any consistent association with the targeted region on chromosome 5 and will not be discussed further.

Statistical analyses were done with SAS (1989). Because the set effect was not a significant source of variance, it was dropped from the analyses and the data were analyzed as a randomized complete block (RCB). Two models were used to analyze the marker data. The first model used single factor analysis of variance including only the marker effect to explain variation among entry means. The second model included both marker and location effects. A significance level of [Alpha] = 0.05 was used for the marker analyses. Although we undoubtedly committed type I errors (falsely rejecting the null hypothesis) only those QTLs that were significant over both backcrosses and planting densities were accepted, and this should minimize false positives.

QTL Detection and Fine Mapping

The first method used to map QTLs to smaller genomic regions was done by graphing the significance of each marker against the marker position. Significance on the y-axis is reported as the log [P.sup.-1] so that lower (more significant) P-values are presented as higher values. Peaks therefore represent the most significant markers in a specific region. On this scale, a P-value of 0.05 = 1.3, 0.01 = 2, and 0.001 = 3. With this procedure, putative positions of QTLs were identified and targeted for fine mapping.

Fine mapping was done by comparing the extent at which introgressed segments overlapped. [BC.sub.2][S.sub.1] lines were first graphically represented using the program Hypergene (Young and Tanksley, 1989). Based on this information and that provided by the graphs, [BC.sub.2][S.sub.1] lines were selected that contained overlapping genomic segments. Line means were adjusted using the least squares (LS) means (lsmeans) option in SAS 6.09 (SAS, 1989) to correct for possible biases due to missing data. Next, pairwise comparisons of the selected lines were made using the location x line mean square as the error term. Assuming an isogenic background, significant differences among lines were attributed to differences in their genotypes in the targeted chromosome 5 region. Comparisons among lines containing specific chromosomal segments were used to determine if their presence or absence was required for a specific phenotype.


Figure 2 shows the genetic map of the markers used to genotype the targeted segment on chromosome 5 covering 105.4 cM (Haldane map units). The region from Pgm2 to NPI449 spanned 56.0 cM. This distance compared favorably with the 55.7 cM computed using [F.sub.3] lines from the Stuber et al. (1992) study.

Single factor analysis of variance, which included only markers in the model, was compared with analysis of variance that included both locations and markers. Although the significance level for all QTLs was greater when both locations and markers were included in the model, both models detected the same QTLs. No marker by location interaction effects were detected, thus the data were analyzed by the model that included only markers.

For comparison with the results from this investigation, the original LOD score distribution for the Stuber et al. (1992) study is shown in Fig. 3. Note that the same region is significant in both backcrosses suggesting overdominance or pseudo-overdominance.


The QTL showing the largest effect on grain yield in the current study was near NPI449, adjacent to Amp3, and was significant only in crosses to the Mo17 tester. In the M/Bseg X Mo17 testcrosses, substituting a B73 allele for a Mo17 allele increased yield by 1038 kg [ha.sup.-1] at high planting density and 746.2 kg [ha.sup.-1] at low density (Table 1). This region near NPI449 also showed a significant association with yield in the B73 background, i.e., in the B/Mseg X Mo17 testcrosses (Table 1). As in the Mo17 background, the B73 allele was favorable and increased yield by approximately 750 kg [ha.sup.-1] at both high and low densities in the B/Mseg X Mo17 testcrosses. Although the B73 inbred ear is shorter and has fewer kernels/row than the Mo17 ear, there is considerable evidence for an association of the increased yield with an increase in ear length and kernels/row because of the B73 allele (Table 1). No significant effect on grain yield was associated with NPI449 in the testcrosses to B73 for either background (Table 2).
Table 1. Genotypic means for grain yield and three other
quantitative traits.


                              ([dagger])MB              MM

                                     M/Bseg x Mo17 testcross
Low density
  Yield (kg [ha.sup.-1)             2847.0          2100.8(***)
  Leaf area ([cm.sup.2])              68.6            66.8
  Ear length (cm)                     19.9            19.0
  Kernel/row                          35.3            31.6(**)
High density
  Yield (kg [ha.sup.-1])             5321.1          4283.1(***)
  Leaf area ([cm.sup.2])              62.9            59.7(*)
  Ear length (em)                     19.5            18.3(***)
  Kernels/row                         37.2            33.3(***)

                                      B/Mseg x Mo17 testcross

Low density
  Yield (kg [ha.sup.-1])          5850.8          5104.6(***)
  Led area ([cm.sup.2])             97.6            93.4
  Ear length (cm)                   22.0            21.1(**)
  Kernels/row                       49.5            47.3(*)
High density
  Yield (kg [ha.sup.-1])          9368.9          8616.4(*)
  Leaf area ([cm.sup.2])            90.2            86.3
  Ear length (cm)                   19.2            19.0
  Kernels/row                       44.5            45.0

                                     MB             MM

                                 M/Bseg x Mo17 testcross
Low density
  Yield (kg [ha.sup.-1)            2796.9         2370.4
  Leaf am ([cm.sup.2])               69.1           66.9
  Ear length (cm)                    20.1           19.2
  Kernel/row                         36.4           33.0
High density
  Yield (kg [ha.sup.-1])           5505.9         4646.8(*)
  Leaf area ([cm.sup.2])             63.1           61.3
  Ear length (em)                    19.9           18.7(**)
  Kernels/row                        38.6           34.3(**)

                                    B/Mseg x Mo17 testcross
Low density
  Yield (kg [ha.sup.-1])           5788.1         5054.4(***)
  Led area ([cm.sup.2])              97.5           92.3(*)
  Ear length (cm)                    21.9           21.0(*)
  Kernels/row                        49.4           47.3(*)
High density
  Yield (kg [ha.sup.-1])           9230.9         8873.5
  Leaf area ([cm.sup.2])             89.7           84.7
  Ear length (cm)                    19.2           18.6
  Kernels/row                        44.7           44.6

(*,**,***) Refer to significant differences between the two marker classes at the 0.05, 0.01, and 0.001 level, respectively.

([dagger]) MB and MM refer to the marker genotype of the testcross material.
Table 2. Genotypic means for grain yield and three other
quantitative traits.

                          ([dagger])BM              BB

                                M/Bseg x B73 testcross

Low density
  Yield (kg [ha.sup.-1])          5813.2          5825.8
  Leaf area ([cm.sup.2])           101.6           104.2
  Ear length (cm)                   21.8            21.3(*)
  Kernels/row                       48.7            47.7
High density
  Yield (Kg [ha.sup.-1])          9111.8          8923.6
  Leaf area ([cm.sup.2])            95.2            95.9
  Ear length (cm)                   18.5            18.3
  Kernels/row                       41.9            40.6

                                  B/Mseg x B73 testcross

Low density
  Yield (kg [ha.sup.-1])           272.9          2615.0
  Leaf area ([cm.sup.2])            90.5             87.1
  Ear length (cm)                   14.2             14.6
  Kernels/row                       27.5             25.6
High density
  Yield (kg [ha.sup.-1])          4421.1           3982.1
  Leaf area ([cm.sup.2])            83.5             80.3
  Ear length (cm)                   14.4             14.3
  Kernels/row                       27.2             25.8

                                      BM            BB
                                    M/Bseg x B73 testcross

Low density
  Yield (kg [ha.sup.-1])          5894.7          5643.9(*)
  Leaf area ([cm.sup.2])           101.9           103.9
  Ear length (cm)                   21.7            20.8(**)
  Kernels/row                       48.8            45.9(**)
High density
  Yield (Kg [ha.sup.-1])          9030.2          8647.2(**)
  Leaf area ([cm.sup.2])            95.1            96.2
  Ear length (cm)                   18.6            18.0
  Kernels/row                       41.5             8.7(**)

                                    B/Mseg x B73 testcross

Low density
  Yield (kg [ha.sup.-1])          2959.9          2633.8(*)
  Leaf area ([cm.sup.2])            90.8            87.4
  Ear length (cm)                   14.4            14.6
  Kernels/row                       29.3            25.99(**)
High density
  Yield (kg [ha.sup.-1])          4759.7          3975.8(**)
  Leaf area ([cm.sup.2])            82.8            80.6
  Ear length (cm)                   14.6            14.3
  Kernels/row                       28.5            25.8(*)

(*,***,***) Refer to significant differences between the two marker classes at the 0.05, 0.01, and 0.001 level, respectively.

([dagger]) BM and BB refer to the marker genotype of the testcross material.

If overdominance is invoked to explain the effect associated with this marker, a significant effect should be detected regardless of the tester used because the comparison is always made between heterozygous and homozygous, marker classes. In this case, the grain yield QTL associated with NPI449 was detected only in the testcrosses to Mo17, indicating that the B73 allele is dominant. Consequently, in any cross to the B73 tester, the genotype of the [BC.sub.2S.sub.1] lines at this locus will not affect the results, so this QTL will not be detected.

This dominant effect can be more easily seen graphically. In Fig. 4, the significance levels of grain yield effects associated with each marker are plotted over the targeted chromosome 5 region for the M/Bseg X Mo17 testcrosses (upper graph) and M/Bseg X B73 testcrosses (lower graph). Figure 5 is analogous except that it is for the B/Mseg X Mo17 and B/Mseg X B73 testcrosses. In the upper graph of Fig. 4, the peak shown for the testcrosses to Mo17 is centered at marker NPI449; in the lower graph reflecting the testcrosses to B73, no effect is detected at this marker. A similar pattern of expression is shown in Fig. 5 for the B/Mseg X Mo17 testcrosses (upper graph) and the B/Mseg X B73 testcrosses (lower graph).


Another QTL associated with grain yield was detected approximately 19 cM from NPI449, near NRZ5, and was significant in all of the crosses to the B73 tester. When averaged over the M/Bseg X B73 and the B/Mseg X B73 testcrosses, substituting a Mo17 allele for a B73 allele increased yield by 288.5 and 583.5 kg [ha.sup.-1] at low and high planting densities, respectively (Table 2). The possible cause for this increase in yield, associated with the favorable Mo17 allele, is an associated increase in ear length and number of kernels/row. The Mo17 inbred ear is longer and has more kernels/ row than the B73 inbred ear, so it is not surprising that an allele from Mo17 would increase these traits. In the B/Mseg X B73 testcross, the Mo17 allele increased kernels/row by 3.4 and 2.7 at low and high densities, respectively (Table 2). Likewise, in the testcross of M/Bseg to B73, the Mo17 allele increased kernels/row by 2.8 at both planting densities. Because the QTL associated with NRZ5 was detected consistently in the testcrosses to B73, it appears that the Mo17 allele is dominant.

Again, this dominant effect can be more easily seen graphically (Fig. 6 and 7). Figures 6 and 7 are comparable with Fig. 4 and 5; however, the trait shown for Fig. 6 and 7 is number of kernels/row. In both backgrounds and in both planting densities, the testcrosses to B73 showed a significant effect associated with a QTL near NRZ5, with the Mo17 allele increasing both grain yield and number of kernels/row.


Exceptions to the results which suggested that the QTL associated with NRZ5 was detected only in the testcrosses to B73 can be found in the M/Bseg X Mo17 testcross evaluated at high planting density and the B/Mseg X Mo17 testcross evaluated at low density (Table 1). In these testcrosses, the pattern of expression associated with NRZ5 was similar to that associated with NPI449. We believe this result is due to close linkage of these two markers, and in these testcrosses to Mo17, the QTL near NPI449 produces such a large effect that NRZ5, only 19 cM away, also reflects an association but at a lower significance level. The intervening marker, BNL5.71, shows an intermediate P-value in most cases (Fig. 4-7), which would be consistent with moving away from the true location of a gene. Taking this into account, it seems plausible that the effect associated with the NRZ5 marker can be attributed to a dominant favorable Mo17 allele.

Based on the above data, we believe that the presumed overdominance effect associated with the major QTL detected on chromosome 5 by Stuber et al. (1992) can be explained by two separate loci in repulsion phase linkage, each showing dominant gene action (pseudo-overdominance). The most important evidence supporting this interpretation is that each of the two QTLs mapped to a different location and each exhibited different genetic effects. Also, it is unlikely that a single overdominant locus could have given the phenotypic pattern observed. There are three possible orders for one overdominant locus and two genetic markers: (i) overdominant locus - NPI449 - NRZ5, (ii) NPI449 - overdominant locus - NRZ5, and (iii) NPI449 - NRZ5 - overdominant locus. If the first order were true, then the effect associated with NPI449 should always be more significant than the effect associated with NRZ5 because NPI449 is the marker closest to the overdominant locus. Because this is not true (see Fig. 4-7, lower) order 1 can be eliminated. The second order, with the markers bracketing the overdominant locus, might produce the observed pattern. However, in this case, an intervening marker (BNL5.71) should also be. more consistently significant in all background X tester combinations. This again is not true, so order 2 is not valid. Order 3 can be excluded for the same reason as order 1. Therefore, because a single overdominant locus could not reasonably account for the separate positions of the genetic effects, we believe these data support the dominance theory of heterosis, at least for this region on chromosome 5.

Another effect near UMCI was detected but it did not show a consistent mode of expression as did the effects near NPI449 and NRZ5, although this effect did occur mostly in the testcrosses to Mo17. This could be a spurious observation because there is another phenomenon that agrees with the observed data. Based on information from a separate study, we investigated the possibility of segregation distortion in the targeted region on chromosome 5. Using the marker Pgm2 to evaluate segregation ratios, we detected significant distortion in the Mo17 backcross and [F.sub.2] generations where the B73 allele was favored. It is of interest to note that two gametophyte factors, ga2 and ga10 (Burnham 1936, Gonella and Peterson, 1975) lie on chromosome 5 and Sari-Gorla et al. (1992) found factors affecting pollen tube growth rate and pollen germinability in this region. Significance of yield expression associated with this locus could be due to the increased frequency of B73 alleles in Mo17 testcross progeny. Increasing the frequency of B73 genomic segments in this testcross would boost yield because of the linked effect associated with NPI449.

Comparison of the Stuber et al. (1992) data (Fig. 3) with that from the current study shows a slight difference in the position of the QTL near the marker Amp3. It shifted from a position slightly proximal to Amp3 to a position slightly distal to Amp3 in the current study. One reason for this difference could be attributed to sampling. Another could result from peak artifacts that are known to occur when using the interval mapping procedure (Haley and Knott, 1992) that was used in the 1992 study. However, both studies show that the NPI449 marker has the highest level of significance based on single factor analysis of variance.

Another goal of this study was to define the QTLs on chromosome 5 to smaller intervals than found in the Stuber et al. (1992) study. We were able to do this only for the largest effect near marker NPI449. By comparing the extent at which the introgressed segments overlapped in the M/Bseg X Mo17 testcrosses, we have placed this QTL between NPI275 and BNL5.71 (Fig. 8), which mapped 27.5 cM apart in the recombinant lines (Fig. 2). This distance might be suitable for MAS, provided selection is based on flanking markers. However, for other purposes, such as map-based cloning, this is clearly too large.



The first goal of this study was to further characterize a major QTL on chromosome 5 affecting grain yield in maize. This QTL was chosen because of its large phenotypic effect, and our data show that it has a complex mode of expression. In this study we dissected the region encompassing this QTL into two different significant areas. The regions near NPI449 and NRZ5 show two QTLs, each with dominant effects. This supports the dominance theory of heterosis. The effect near NPI449 is present in the testcrosses of the [BC.sub.2][S.sub.1] lines to Mo17 but not in the testcrosses to B73, with the B73 allele being superior. The QTL near NRZ5 acts in a dominant manner in testcrosses to B73 but not to Mo17, with the Mo17 allele being favorable. This suggests that the large QTL mapped in the Stuber et al. (1992) study resulted from the expression of two dominant genes in repulsion phase linkage, and that overdominance probably was not the type of gene action being expressed. These results agree with the Cockerham and Zeng (1996) reanalysis of the original F3 data from the Stuber et al. (1992) study. Using a North Carolina Design III, modified to accommodate markers, they determined that most effects acted in a dominant manner.

The second objective of this study was to fine map this region. Only the largest effect, near NPI449, was mapped to a smaller (27.5 cM) region. There are several reasons why the effect could not be defined to an even smaller region. The first is the lack of polymorphic markers in the immediate area. Additional markers might allow for an enhanced examination. Finer mapping could also be accomplished by increasing the number of recombination events in the desired region. Because only 85 lines were used in this study, we did not obtain many "desirable" recombinants. Increasing the number of lines could assist with additional resolution. However, the most likely reason that this effect, or that of any of the other QTLs, could not be defined to a smaller area was due to the complexity of this region.

Choosing previously identified QTLs that show effects in only one testcross might allow for more precise fine mapping. If QTLs are as complex as indicated by this analysis, MAS based on early generation populations (backcross, [F.sub.2], or [F.sub.3]) could be difficult. To characterize any genomic segment containing a putative QTL in the manner described above requires extensive resources in terms of labor and time, and is probably not amenable to most breeding projects using MAS. However, with a QTL possessing simple expression patterns, MAS should increase the efficiency of accumulating desired genomic segments. With this in mind, we are characterizing a QTL on chromosome 4 which was detected only in the backcross to Mo17.

Abbreviations: AFLP, amplified fragment length polymorphism; cM, centimorgan; MAS, marker-assisted selection; NIL, near isogenic line; QTL, quantitative trait locus; RAPD, randomly amplified polymorphic DNA; RFLP, restriction fragment length polymorphism; SSR, simple sequence repeat; M/Bseg, Mo17(B73)[BC.sub.2][S.sub.1] line; B/Mseg, B73(Mo17)[BC.sub.2][S.sub.1] line.


The technical assistance of Wayne Dillard, M. Lynn Senior, David Rhyne, and Dianne Beattie, and the constructive review by John LeDeaux are gratefully appreciated. The research was a joint contribution from the USDA, ARS, and North Carolina State University, Raleigh, NC, and supported in part by the USDA competitive Research Grants 86-CRCR-1-2030 and 89-37140-449. The NC Biotechnology Center also supported this research through the Institutional Development Grant 9110-IDG-1022. The authors also thank the McKnight Foundation for the support provided.


Abler, B.S.B, M.D. Edwards, and C.W. Stuber. 1991. Isoenzymatic identification of quantitative trait loci in crosses of elite maize hybrids. Crop. Sci. 31:267-274.

Bruce, A.B. 1910. The Mendelian theory of heredity and the augmentation of vigor. Science 32:627-628.

Burnham, C., 1936. Differential fertilization in the bt-pr linkage group of maize. J. Am. Soc. Agron. 28:968-975.

Cockerham, C.C., and Z.B. Zeng. 1996. Design III with marker loci. Genetics 143:1437-1456.

Comstock, R.E., and H.F. Robinson. 1952. Estimation of average dominance of genes. p. 495-516. In J.W. Gowen (ed.) Heterosis. Iowa State College Press, Ames, IA.

Crow, J.F. 1952. Dominance and overdominance, p. 282-297. In J.W. Gowen (ed.) Heterosis. Iowa State College Press, Ames, IA.

Gardner, C.O., and J.H. Lonnquist. 1959. Linkage and degree of dominance of genes controlling quantitative characters in maize. Agron. J. 45:186-191.

Guffy, R.D., C.W. Stuber, and T. Helentjaris. 1988. Molecular markers for evaluating quantitative traits across varying genetic backgrounds and environments in maize. p. 82. In Agronomy abstracts. ASA, Madison, WI.

Gonella, J., and P.A. Peterson. 1975. Gametophytic factor (ga10) on chromosome 5 distal to A2. Maize Genetic Coop. Newsl. 49:71-73.

Hallauer, A.R., and J.B. Miranda, Fo. 1981. Quantitative genetics in maize breeding. Iowa State Univ. Press, Ames.

Haley, C.S., and S.A. Knott. 1992. A simple regression method for mapping quantitative trait loci in line crosses using flanking markers. Hered. 69:315-324.

Hull, F.H. 1945. Recurrent selection for specific combining ability. J. Am. Soc. Agron. 37:134-145.

Jones, D. F. 1917. Dominance of linked factors as a means of accounting for heterosis. Genetics 2:466-479.

Keeble, F., and C. Pellew. 1910. The mode of inheritance of stature and of time of flowering in peas (Pisum sativum). J. Genetics 1:47-56.

Lincoln, S.E., M. Daly, and E.S. Lander. 1993. Constructing genetic linkage maps with Mapmaker/EXP version 3.0: A tutorial and reference manual. Whitehead Institute for Biomedical Research Technical Report, Cambridge, MA.

Lindstrom, E.W. 1926. Hereditary correlation of size and color characters in tomatoes. Res. Bull. Iowa Agric. Exp. Stn. 93:99-128.

Paterson, A.H., J.W. DeVerna, B. Lanini, and S.D. Tanksley. 1990. Fine mapping of quantitative trait loci using selected overlapping recombinant chromosomes, in an interspecies cross of tomato. Genetics 124:735-742.

Ragot, M., P.H. Sisco, D.A. Hoisington, and C.W. Stuber. 1995. Molecular-marker-mediated characterization of favorable exotic alleles at quantitative trait loci in maize. Crop. Sci. 35:1306-1315.

Saghai-Maroof, M., K. Soliman, R. Jorgensen, and R. Allard. 1984. Ribosomal DNA spacer length polymorphism. in barley: Mendelian inheritance, chromosomal location, and population dynamics. Proc. Natl. Acad. Sci. (USA) 91:8014-8018.

Sari-Gorla, M., M. Pe, D. Mulcahy, and E. Ottaviano. 1992. Genetic dissection of pollen competitive ability in maize. Hered. 69:423-430.

Sambrook, J., E.F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York.

SAS Institute Inc. 1989. SAS language and procedures: Usage, Version 6, 1st ed. SAS Institute Inc., Cary, NC.

Shull, G.H. 1908. A composition of a field of maize. Am. Breeders Assoc. Rep. 4:296-301.

Senior, M.L., E.C.L. Chin, M. Lee, J.S.C. Smith, and C.W. Stuber. 1996. Simple sequence repeat markers developed from maize sequences found in GENBANK Database: Map construction. Crop Sci. 36:1676-1683.

Sax, K. 1923. The association of size differences with seed coat pattern and pigmentation in Phaseolus vulgaris. Genetics 8:552-560.

Stuber, C.W., M.D. Edwards, and J.F. Wendel. 1987. Mecular marker facilitated investigations of quantitative trait loci in maize: II. Factors influencing yield and its component traits. Crop Sci. 27: 639-648.

Stuber, C.W., S.E. Lincoln, D.W. Wolff, T. Helentjaris, and E.S. Lander. 1992. Identification of genetic factors contributing to heterosis in a hybrid from two elite maize inbred lines using molecular markers. Genetics 132:823-839.

Stuber, C.W., J.F. Wendel, M.M. Goodman, and J.S.C. Smith. 1988. Techniques and scoring procedures for starch gel electrophoresis of enzymes from maize (Zea mays L.). North Carolina Agric. Res. Ser. Tech. Bull. 286.

Thoday, J.M. 1961. Location of polygenes. Nature 191:368-370.

Young, N.D., and S.D. Tanksley. 1989. Restriction fragment length polymorphism maps and the concept of graphical genotypes. Theor. Appl. Genet. 77:95-101.

G.I. Graham, Asgrow Seed Co., 205 N. Michigan, Oxford, IN 47971-8505; D.W. Wolff, Texas Agric. Exp. Stn, Texas A&M Univ., Weslaco, TX 78596; C.W. Stuber, USDA-ARS, Dep. of Genetics, North Carolina State Univ., Raleigh, NC 27695-7614. Received 15 Nov. 1995. Charles W. Stuber, Corresponding author (
COPYRIGHT 1997 Crop Science Society of America
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 1997 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Graham, Geoffrey I.; Wolff, David W.; Stuber, Charles W.
Publication:Crop Science
Date:Sep 1, 1997
Previous Article:Agronomic performance of soybean plant introductions and their genetic similarity to elite lines.
Next Article:Two microsatellite markers that flank the major soybean cyst nematode resistance locus.

Terms of use | Privacy policy | Copyright © 2020 Farlex, Inc. | Feedback | For webmasters