2. 1 The amylose content of core germplasm and the phenotypic variation of eight characteristic value which starch viscosity of RVA spectrum
The 9 traits phenotypic variation of 179 core germplasm materials (table 1, figure 1). The 179 core germplasm material can be divided into two classes of indica and japonica rice to analysis of phenotypic distribution frequency: In addition to a small class of glutinous rice materials which amylose content is less than 10% of the AC, the rest of the split between 15% and 30%, and AC of indica varieties mainly concentrated in between 15% and 27%, while japonica rice material appeared two peaks, one was between 19% and 21%, another was between 23% and 25%; PKV, HPV, CPV and CSV with similar frequency distribution, including indica material regardless of overall data or peak position are higher than that of japonica rice in the character of material viscosity values, it is worth noting that indica’s two peaks in the CPV traits, a wave is similar with that of japonica rice, the other one is bigger than it; Indica was less than japonica in the BDV viscosity value, while the viscosity value of the whole is bigger than that japonica rice in the SBV; As you could see that indica and japonica had the similar frequency distribution straight side frequency chart in PeT and GT.
According to the result of various traits correlation (table 2), we should know that AC, PKV, HPV, CPV and CSV had very significant positive correlation. PKV was different from other traits relatively, which had significant positive correlation with AC and CSV. There are significant positive correlation between GT and AC, CPV, CSV, SBV, PeT at the same time, BDV exist significant negative correlation only with SBV, and it existed significant positive correlation with PKV, while no correlation with other traits.
2. 2 The AC of core germplasm and QTL results of starch RVA spectrum characteristic value
According to the figure 2, AC, PKV, HPV, CPV and CSV were the high degree of correlation of characters, if one QTL loci have effect on one of the other characters, and it also had effect on the rest of the several characters at the same time usually; BDV and SBV are significantly negative correlation, the QTL they scanned out had a certain similarity, and most of them had high specificity and not had effect on other traits.
AC scanned to 23 QTL in genome-wide, including 10 QTLs which were reported about the starch synthesis related genes in the same section. There are two effect value high QTLs qAC-1 and qAC - 6-1 of LOD values were 15.5 and 10.0, predicted genes were AGPL2 and GBSSI respectively.
PKV was detected 20 QTLs, including 8 QTLs which were reported about the starch synthesis related genes in the same section.
HPV was detected 11 QTLs, including 8 QTLs were reported about the starch synthesis related genes in the same section. One of the predicted gene GBSSI for QTL qHPV - 6-1 with a 12.5 LOD values.
CPV was detected 23 QTLs, including 9 QTLs were reported about the starch synthesis related genes in the same section. We also found a same LOD QTL which value of 10.7 in the GBSSI gene segments .
CSV was detected 18 QTLs, including 9 QTLs were reported about the starch synthesis related genes in the same section.
BDV and SBV were detected 27 QTLs, including 16 QTLs in BDV and 11 QTLs in SBV; 12 QTLs were reported about the starch synthesis related genes in the same section. The two traits were detected to a QTL which effect value is quite high in GBSSI gene loci at the same time, LOD values were 11.1 and 13.3 respectively.
GT was scanned to 13 QTLs, but theirs effects are not high, the biggest QTL appeared in the ALK gene loci which LOD was 5.3.
PeT was detected 28 QTLs, including 10 QTLs were reported about the starch synthesis related genes in the same section. Two QTLs of LOD values are 13.7 and 10.5 respectively in AGPL2 and ALK gene segments.
There were a lot of QTLs region had the effect on multiple traits in 12 chromosomes of rice. The QTL on Chromosome 1 about 25Mb AGPL2 genetic regions had contribute to AC, PKV, HPV, CPV, CSV, GT and PeT; the QTL near the chromosome 2 about 12Mb had contribute to AC, PKV, HPV, CPV, CSV, and PeT; The QTL of chromosome 2 about 22Mb SUT5 genetic regions had contributed to AC, PKV, HPV, CPV, CSV, GT and PeT; The QTL of Chromosome 5 about 5.5 Mb had contribute to AC, PKV, HPV, CPV, CSV, GT and PeT. Chromosome 6 about 1.7 Mb GBSSI genetic region's QTL which was the only one had high LOD value effect to all the nine traits, The QTLs of Chromosome 6 about 6.7Mb SSIIa genetic regions of QTL, chromosome 7 about 10.6Mb, GBSSII genetic regions and near 28.6 Mb had effect all traits besides BDV and the SBV.
QTL locus of multiple effect in association analysisresults were listed in table 3, From which we can see the QTLs on chromosome 2 about 12Mb, chromosome 5 about 5.5 Mb, chromosome 7 about 28 Mb and chromosome 12 about 12Mb area were scanned to contribute more than six traits with nearby which found no known starch genes, and LOD values of this were very high above 4.5.
2. 3 Single segment materials verify correlation analysis results
As shown in figure 3 and table 4, ACC10 / ZS97 single segment material NQ139 had 921 cP of BDV while the parent ZS97 had 715 cP of BDV. Both of them had significant difference. Core germplasm correlation analysis results detected that only one QTL control BDV on the locus (36Mb of chromosome 1), other RVA traits had not detected QTL in this location, it was well supported by the phenotypic data. In the import section of the single segment material has a gene OsHXK6 control rice hexokinase, we speculated that it may be a gene change in the phenotype of rice's BDV. Correlation analysis detected two QTLs control BDV and SBV about 3 Mb of chromosome 8, only value of BDV and SBV of the NIP/ZS97 single segment material NQ082 of import that area really had the most differences from that of parent ZS97, we speculated that the locus of QTL negative regulation of BDV and positive regulation of SBV. UDP - galactose transfer protein gene OsUGT2 in the segment, so we speculated that it may be a gene which change in the phenotype BDV and SBV of rice. Correlation analysis detected a control BDV QTL which LOD value of 3.3 about 35 Mb of chromosome 3, NIP/ZS97 single segment material NQ041's BDV had differences with compared with parent ZS97, but it seems that PKV, HPV and CPV also had differences from ZS97, and this point was not detected in the correlation analysis. Correlation analysis detected two QTLs control PKV and GT respectively about 24 Mb of chromosome 7 , but the NIP/ZS97 single segment material NQ079 which contains the segment only detected PKV phenotype differences, without the difference of GT, we speculated that GT was not detected is due to the effect of haploid type gene exists in other materials, but in the NIP and ZS97, and rice monosaccharide transporter gene OsMST6 close to this QTL, the gene maybe control PKV. The NIP / 9311 single segment material 13NY32 had significant difference with ZS97 on PKV, BDV and SBV, it was the sanme to the detected QTL of correlation analysis in this segment. Due to the single segment material had large import section, there are three genes related to starch synthesis pathway, they are rice beta glycosidase enzyme gene Os3BGlu6, rice monosaccharide transporter gene OsMST4 and sucrose transporters OsSUT1, they could be the main effect genes control these phenotypic variation.
2.4 The QTL with starch RVA eigenvalue known genes and starch synthesis comparative analysis of the path
Starch synthesis pathway had three ways roughly (Manish, 2012). In the early time prophaseglucose - 6 - phosphoric acid regulation of synthesis of glucose (Glc - 6 p) - 1 - phosphate (p) Glc - 1 way in the chloroplast , which step was regulated together by two categorie of genes GPT and PGM gene , then Glc - 1 p into the amyloplast became ADPGlc by mid AGPase eneyme. At the later stage ADPGlc synthesis of Amylose and amylopectin, Amylose path generated by the ADPGlc via granule combined with starch synthase Amylose (Amylose) directly; Branched chain starch synthesis approach is relatively complex, ADPGlc in turn by the starch synthase、starch branching enzyme and Amyloglucosidase action generated Amylopectin.
According to (table 5) we could see that 6 kinds of starch synthesi genes had performance on amylose content and starch RVA profile eigenvalue, starch synthase (including GBSS and SS) genes had effect in all nine characters, Wx (GBSSI) gene is the only one which contribute to all traits, and the LOD values was very high. See from this table, GPT and PGM genes which starch and sucrose synthesis previous had obvious specificity effect for BDV and SBV, while little effect on other traits or not. The results QTL IciMapping of NIP/ZS97 material scanned to a QTL location on chromosome 8 4200710 bp to 5313154 bp, LOD value and contribution rate were 3.4 and 12.4% respectively, while GPT1 genes also happened to be in here. This point also verify that the way of glucose - 6 - phosphate (Glc6p) - > glucose - 1 - phosphate(Glc1p) effect on BDV and SBV, while BDV and SBV were less susceptible to all genes except the early stage three kinds of genes of the starch and sucrose synthesis of starch synthesis enzymes. Speculation GPT, PGM and GBSS, SS are the relationship integrated upstream and downstream, there are close emergency control mechanism, and commom decided the proportion of amylose and amylopectin starch. So GBSS and SS for ADPGlc have selection bias, and the difference of ADPGlc depends on the early stage of the starch synthesis GPT and PGM genes.