46. Interval mapping of QTLs for yield and other related characters in rice

H-X Lin1 , H-R Qian1 , J-Y Zhuang1 , J Lu, S-K Min1 , Z-M Xiong1 , N Huang2 , K-L Zhnenu2

China National Rice Research Institute, Hangzhou 310006, P. R. China International
Rice Research Institute, P.O.Box 933, 1099 Manila, Philippines

     Two F2 populations were produced from two indica/indica crosses. These were Tesanai 2/CB (TSA/CB) and Waiyin 2/CB (WY/CB) with a common male parent CB. CB is from California, USA. Tesanai 2 is from Guangdong, China. It is a high yielding variety notable for its large panicles. Waiyin 2 is from International Rice Research Institute (IRRI) with larger grains. Seventy three percent of 167 RFLP marker enzyme combinations tested revealed polymorphisms between the parents of the two crosses. One hundred and seventy one plants for each F2 population were genotyped with 93 and 101 RFLP markers in TSA/CB and WY/CB respectively. Two linkage map comprised of 89 marker loci and 93 marker loci of 12 linkage groups were constructed respectively using MAPMAKER (Lander et al. 1987).
    Each of 8 characters were subjected to QTL mapping with both ANOVA (The PROC GLM procedure in the Statistical Analysis System, SAS Institute Inc. 1988) and interval mapping procedures (MAPMAKER/QTL, Lander and Botstein 1989). In TSA/CB, 22 QTLs for the 8 characters were detected by one way ANOVA (data not shown) and the same QTLs were also detected by interval analysis (Table 1). In WY/CB, 19 QTLs for 6 traits were detected by one way ANOVA and again, all 9 QTLs were detected by interval analysis (Table 2). These results indicated that both statistical procedures produced were very similar results. The QTLs identified are named with trait abbreviation followed by chromosomal number (Table 1 and 2). QTLs controlling grain weight per plant (GWT) were located within 5 intervals (gwtl, gwt2, gwt4, gwt5 and gwt8) in TSA/CB, gwtl within RG374-RG394 on chromosome 1, gwt2 within RG157-RG171 on chromosome 2, gwt4 within RG143-RG214 on chromosome 4, gwt5 within RG13-RG573 on chromosome 5 and gwt8 within RZ66-RG598 on chromosome 8 (Table 1). In WY/CB, however, only 3 intervals containing QTLs (gwtl, gwt4a and gwt5a) controlling GWT were detected. QTL gwtl was located within RG374-RG394 on chromosome 1 as in TSA/CB, gwt4a within RG788-RG190 on chromosome 4 and gwt5 within CD082-RG360 on chromosome 5 (Table 2).
    Among the 41 QTLs various traits detected in both populations, 23 QTLs can explain phenotypic variance larger than 10%. Some of the QTLs such as np4, gwt4a, tgwt10, nft>8, explained more than 20% of variance (Table 1 and 2). If a QTL can explain larger variation and is significantly different from other QTLs (e.g. much larger LOD score), one would assume that it would be a major gene than a QTL. QTL np4 (Table 1 ) from TSA seems having such characteristics. Number of panicle (NP) is generally considered under the control of both major and minor genes. The np4 might be a major gene which joins other QTLs such as np2 with lesser effect to control the development of panicle.
 


Table 1. QTLs detected for yield components based on interval analysis
            (MAPMAKER/QTL) in the Tesanai 2/CB F2 population
 

Trait1 QTL2 Interval LOD3 %variation4 explained a-5 d-6 d/[a]7
GWT gwtl RG374-RG394 2.93 10.7 -1.96 12.36 6.32
gwt2 RG157-RG171 4.11 11.4 -9.14 7.10 0.78
gwt4 RG143-RG214 3.10 8.7 -7.91 0.28 0.04
gwt5 RG13-RG573 2.39 11.0 1.27 12.72 10.03
gwt8 RZ66-RG598 2.31 7.8 -8.04 0.51 0.06
NP np2 RG157-RG171 3.62 9.3 -1.39 2.27 1.64
np4 RG143-RG214 9.68 26.1 -2.92 0.18 0.06
NG ngl RG374-RG394 4.41 17.2 -1.39 48.18 34.57
ng2 RG25-RG157 2.79 9.0 -21.46 27.65 1.29
ng8 RG978-RZ66 2.28 13.8 -30.55 9.49 0.31
ngl2 RG341-RG235 2.48 7.4 1.77 31.53 17.79
TNS tns3 RG104-RG348 2.32 6.8 -16.56 -45.87 -2.77
tns8 RZ562-RG978 4.84 15.5 35.70 53.64 1.50
tnsl2 RG457-RG341 2.34 7.4 -37.89 -9.61 -0.25
SF sfl RG374-RG394 3.86 12.2 -5.10 25.00 4.92
TGWT tgwtl RG173-RG532 3.01 12.6 1.63 -1.50 -0.92
tgwt4 RG143-RG214 2.74 8.5 -1.39 0.55 0.39
tgwt5 RG182-RG13 2.73 14.8 -1.55 -1.33 -0.86
SD sd3 RG104-RG348 2.13 5.7 -5.43 -15.04 -2.77
sd8 RZ562-RG978 3.73 10.9 11.79 14.69 1.25
sdl2 RG457-RG341 2.21 8.0 -14.06 -4.04 -0.29
sfl nfb8 RG108-RZ562 6.49 20.9 1.91 -0.49 -0.25
1GWT=Grain weight per plant, NP=Number of panicles per plant, NG=Number of grains per panicle, TNS=Total number of spikelets per panicle, SF=Spikelet fertility, TGWT=1000-grain weight, SD=Spikelet density, NFB=Number of first branches per panicle.
2 QTLs are named by trait abbrevations plus chromosomal number.
3 Log 10-likelihood.
4Percent phenotypic variance explained.
5Additive gene effect at the putative QTL loci.
6Dominance effect at the putative QTL loci.
7Degree of dominance.


    In general, 1 to 3 QTLs were detected for each trait in both populations (Table 1 and 2). Comparisons of QTLs for individual traits showed that majority of QTLs present in only one of the two populations. For example, 3 QTLs were detected for total number of spikelets per panicle (TNS) but only one (tns8) is detected in both populations. As the performance of female parents is very different and is likely controlled by a different set of operating genes. All together, we found 3 QTLs (gwtl, tns8, sd8) shared by both populations. Different QTLs detected in different populations provide an opportunity to combine them to produce new breeding lines with higher yield potential. As DNA


Table 2. QTLs detected for yield components based on interval analysis
MAPMAKER/QTL) in the Waiyin 2/CB F2 population
 

Trait1 QTL2 Interval LOD3 %variation4 explained a5 a6 d/[a]7
GWT gwt1 RG374-RG394 3.13 11.5 -6.70 -6.56 -0.98
gwt4a RG788-RG190 4.67 22.3 -8.85 -11.48 -1.30
gwt5a CD082-RG360 3.14 10.1 -8.59 -1.66 -0.19
NP np2a RG324A-RG324B 2.09 5.6 -1.06 1.13 1.06
np5 RG360-RG9 2.86 9.9 -1.99 -0.24 -0.12
np6 waxy-RG213 2.86 12.1 -1.51 -1.16 -0.76
TNS tns6 RG138-RG64 5.01 13.1 -38.04 28.14 0.74
tns8a RG108-RZ562 2.27 7.3 10.79 43.50 4.03
tns11 RG1094-RG118 4.38 12.5 -54.71 -43.41 -0.79
SF sf1a RG536-RG222 3.25 9.4 -8.27 -6.80 -0.82
sf6 waxy-RG213 2.37 12.5 11.42 7.84 0.69
sf8 RZ562-RZ617 2.25 6.7 -8.72 -0.51 -0.06
TGWT tgwt1a RZ649-RG381 2.12 5.8 1.23 1.10 0.09
tgwt2 RG171-RG437 4.23 14.8 -2.42 0.57 0.24
tgwt10 RG241-RG561 3.52 22.8 -2.89 1.32 0.46
SD sd2 RG520-RG256 3.13 12.3 3.92 -21.09 -5.38
sd6 RG138-RG64 5.15 13.3 -14.62 6.30 0.43
sd8 RG108-RZ562 2.45 7.9 3.86 16.22 4.20
sdll RG103-RG2 2.31 6.3 -5.22 13.48 2.58
1,2,3,4,5,6,7 See Table 1 for definition. markers do not subject to environmental effect, marker-aided selection of the QTLs can be conducted in early generation of breeding program. QTL mapping of rice yield and related characters in the first step to apply marker technology to breeding program for the improvement of yield potential.

References

Lander, E. S., P. Green, J. Abbrahamson, A. Barlow, M. J. Daly, S. E. Lincoln, L. Newburg, 1987.
    MAPMAKER: an interactive computer package for constructing primary genetic linkage maps of
    experimental and natural populations. Genomics 1: 174-181.
Lander, E. S. and D. Botstein, 1989. Mapping Mendelian factors underlying quantitative traits using RFLP
    linkage maps. Genetics 121: 185-199.
SAS Institute. Inc. (1988) SAS users Guide: Statistics. SAS Institute, Gary, N. C.