Genome sequencing reveals agronomically important loci in rice using MutMap

Abstract

In MutMap, a mutant is crossed directly to the original wild-type line and then selfed, allowing unequivocal segregation in second filial generation (F2) progeny of subtle phenotypic differences. This approach is particularly amenable to crop species because it minimizes the number of genetic crosses (n = 1 or 0) and mutant F2 progeny that are required. 

 have  not incorporated the findings of the genomics revolution

RESULTS

MutMap method

1.We first use a mutagen (for example, ethyl methanesulfonate) to mutagenize a rice cultivar (X) that has a reference genome sequence. Mutagenized  plants  of  this  first  mutant  generation  (M1)  are  self-pollinated and brought to the second (M2) or more advanced genera-tions to make the mutated gene homozygous. Through observation of phenotypes in the M2 lines or later generations, we identify recessive mutants  with  altered  agronomically  important  traits  such  as  plant height, tiller number and grain number per spike. 利用诱变剂诱导具有参考基因组序列的品种，第一代突变体 M1 自交产生 M2，连续自交获得纯和单株。通过观察 M2 系或后代中的表型，我们确定隐性突变体具有改变的农艺学重要特征。

2. Once the mutant is identified, it is crossed with the wild-type plant of cultivar X, the same cultivar used for mutagenesis. The resulting first filial generation (F1) plant is self-pollinated, and the second generation (F2) progeny (>100) are grown in the field for scoring the phenotype.一旦鉴定出突变体，就将其与品种 X 的野生型植物杂交，该品种用于诱变。产生的第一代孝子代（F1）植物是自花授粉的，第二代（F2）后代（> 100）在田间生长以对表型评分。

3.. Among the F2 progeny, the majority of SNPs will segregate in  a  1:1  mutant/wild  type  ratio.  However,  the  SNP  responsible  for the change of phenotype is homozygous in the progeny showing the mutant phenotype. 在 F2 子代中，大多数 SNP 将以 1：1 突变体/野生型比例分离。然而，负责表型改变的 SNP 在显示突变表型的后代中是纯合的。

4.If we collect DNA samples from recessive mutant F2 progeny and bulk sequence them with substantial genomic coverage (>10×  coverage),  we  expect  to  have  50% mutant and 50% wild-type sequence reads for SNPs that are unlinked to the SNP responsible for  the  mutant  phenotype.  However,  the causal  SNP  and  closely  linked  SNPs  should show 100% mutant and 0% wild-type reads. 无关的 SNP 具有 50％突变和 50％野生型序列读数突变表型，紧密相连的 SNP 应显示 100％突变和 0％野生型读码。

5. we expect that  this  index  would  equal  1  near  the  causal  gene  and  0.5  for  the unlinked loci.我们期望该指数在目标基因附近等于 1，而对于未连锁的基因座等于 0.5。

6. SNP indices can be scanned across the genome to find the region with a SNP index of 1, harboring the gene responsible for the mutant phenotype.在整个基因组中扫描 SNP 索引，以找到 SNP 索引为 1 的区域，其中包含负责突变表型的基因。

MutMap applied to pale green leaf mutants

we applied MutMap to two mutants showing pale-green leaf phenotypes with slightly lower chlorophyll concentrations compared with wild type (Hit1917-pl1 and Hit0813-pl2)

Red regression lines were obtained by averaging SNP indices from a moving window of five consecutive SNPs and shifting the window one SNP at a time. The x-axis value of each averaged SNP index was set at a midpoint between the first and fifth SNP.通过对五个连续 SNP 的移动窗口中的 SNP 指数求平均值并一次将窗口移位一个 SNP 来获得红色回归线。每个平均 SNP 指数的 x 轴值设置为第一个和第二个之间的中点。

We  crossed  these  mutants  to  the  Hitomebore  wild  type  in  2009, and obtained F1 progeny. F1 plants were self-pollinated, and >200 F2progeny were obtained for each cross. For both mutants we observed segregation between wild-type and mutant phenotypes in field-grown F2  progeny,  with  a  wild  type/mutant  ratio  of  3:1  (Supplementary Table  2),  suggesting  that  each  mutant  phenotype  was  caused  by  a recessive  mutation  in  a  single  locus.  For  each  cross,  we  isolated DNA of 20 F2 progeny showing the mutant phenotype, and bulked the  samples  in    an  equal    ratio.  This  bulked  DNA  was  subjected  to whole-genome sequencing using an Illumina GAIIx sequencer. 鉴定突变类型为隐形突变，F2 符合 3：1。

 We obtained  70  million  and  133  million  sequence  reads  (75  bp)  for Hit1917-pl1 and Hit0813-pl2, respectively, corresponding to >5 Gb of total read length with >12× coverage of the rice genome (370 Mb;Supplementary  Table  3).  These  reads  were  aligned  to  a  reference sequence  of  Hitomebore  using  MAQ  software  Aligned  data  were passed  through  a  filter  to  reduce  spurious  SNP  calls  caused  by sequencing and alignment errors (Online Methods). As a result, we found that mutants harbor 1,001 (Hit1917-pl1) and 1,339 (Hit0813-pl2) transition-type (G→A and C→T) SNPs with high-quality scores , presumably caused by ethyl methanesul-fonate mutagenesis. SNP筛选，滑动窗口得到拟合曲线，获得高质量snp(100%)和过度型snp。

 For each mutant we identified a single unique genomic region harboring a cluster of SNPs with SNP index of 1 : the Hit1917-pl1 mutant showed a cluster of seven SNPs with SNP index of 1 on chromosome 10, whereas Hit0813-pl2 had a cluster of five SNPs with SNP index of 1 on chromosome 1 . These results show that MutMap allows rapid identification of the putative position of a causal mutation responsible for a mutant phenotype.Hit1917-pl1 突变体在 10 号染色体上显示了七个 SNP 指数为 1 的 SNP，而 Hit0813-pl2 具有在染色体 1 上 SNP 指数为 1 的五个 SNP。

                    Hit1917-pl1

                                  Hit0813-pl2

Identification of the causal SNP of a pale green leaf mutant

For Hit1917-pl1, a pale green leaf mutant, we examined SNPs with SNP index of 1 in detail. Of seven SNPs with SNP index of 1, two corresponded to exons of protein-coding genes.7 个 SNP-index 为 1 的 SNP 中有 2 个位于 cds 区。

In a study of a T-DNA insertion knockout of the OsCAO1 gene, the OsCAO1 mutant has lower chlorophyll than wild type, similar to our Hit1917-pl1 mutant.基因敲除验证。

We also made a knockdown mutant of OsCAO1 by transforming wild-type plants with an RNA interference (RNAi) construct targeting the OsCAO1 gene. The Hit1917-pl1 mutant transformed with wild-type OsCAO1 expressed both mutant and wild-type alleles of OsCAO1 (Fig. 3b), and its phenotype was restored to wild type .构建表达载体在突变体中恢复表达，利用 RNA 干扰野生型表达。

Application of MutMap to agronomically important traits

we applied MutMap to four semidwarf mutants (Hit1917-sd, Hit0746-sd, Hit5500-sd and Hit5814-sd) to identify the genomic regions.我们将 MutMap 应用于四个半矮突变体（Hit1917-sd，Hit0746-sd，Hit5500-sd 和 Hit5814-sd）以鉴定基因组区域。

In all cases, only a single genomic region contained a cluster of SNPs with SNP index of 1. Sometimes multiple peaks (Hit0746-sd) or a broader peak (Hit5243-sm) of plots were observed. The former may be caused by an additional mutation affecting viability of F2 and the latter may be attributed to mutation position in recombination-deficient chromosome regions (that is, close to the centromere), but in most cases these complications did not prevent identification of putative regions harboring causal mutations. The average interval of SNPs within these regions with a SNP index ≥0.9 was 2.1 Mb, and we found at most four SNPs that could have caused nonsynonymous changes of protein-coding genes ( Supplementary Table 4 )

多数情况下是单个只有单个基因组区域包含 SNP 指数为 1 的 SNPs。有时会观察到图的多个峰（Hit0746-sd）或更宽的峰（Hit5243-sm）。前者可能是由影响 F2 活力的其他突变引起的，而后者可能是由于重组缺陷型染色体区域（即靠近着丝粒）的突变位置引起的，但是在大多数情况下，这些特殊情况并不能阻止推定的鉴定因果突变的区域。这些区域中 SNP 指数≥0.9的SNP的平均间隔为2.1 Mb，我们发现最多四个SNP可能引起蛋白质编码基因的非同义变化

Discussion

Because of these problems（F2 群体存在强烈的杂种优势，影响表型）, isolation of genes with minor effects and QTL have been carried out using recombinant inbred lines (RILs) of later generations, in which the contribution of an individual gene is separately addressed and the effect of heterosis can be minimized. 由于这些问题，已经使用后代的重组自交系（RIL）进行了影响较小的基因和 QTL 的分离，其中单独处理了单个基因的作用，并使杂种优势的影响降至最低。

 If a causal SNP cannot be identified, the SNPs flanking the regions harboring causal mutations for the desired phenotypes (those with a SNP index of 1) can be used as DNA markers for marker-assisted selection by crossing the mutant to the wild type.如果不能鉴定出因果 SNP，则可以通过将突变体与野生型杂交，将具有期望表型因果突变区域（SNP-index 为 1 的那些）侧翼的 SNP 用作 DNA 标记，以进行标记辅助选择。

Methods

Alignment of short reads to reference sequences and SNP calling.

We divided SNPs into two categories: homozygous SNPs and heterozygous SNPs. Homozygous SNPs were defined as SNPs with SNP index ≥0.9 and a minimum coverage of the sites of three reads. Heterozygous SNPs were defined as SNPs with SNP index ≥0.3 and <0.9 with a coverage of the position of more than four reads. 我们将SNP分为两类：纯合SNP和杂合SNP。纯合SNP定义为SNP指数≥0.9且三个读数位点的最小覆盖率的SNP。杂合SNP被定义为SNP指数≥0.3且<0.9且覆盖超过四个读数的位置的SNP。

We further filtered SNPs with two more steps: (i) removal of common SNPs shared by at least two mutant lines and (ii) extraction of SNPs that exhibit G→A or C→T transitions, which are the most frequent changes caused by ethyl methanesulfonate mutagenesis.。我们进一步通过两个步骤对SNP进行了过滤：（i）去除至少两个突变株系共有的常见SNP，以及（ii）提取表现出G→A或C→T转变的SNP，这是由乙基引起的最频繁的变化甲磺酸盐诱变。

SNP index plot regression lines were obtained by averaging SNP indices from a moving window of five consecutive SNPs and shifting the window one SNP at a time. The x-axis value of each averaged SNP index was set at a midpoint between the first and fifth SNP.通过对五个连续 SNP 的移动窗口中的 SNP 指数进行平均，然后一次将窗口移动一个 SNP，即可获得 SNP 指数图回归线。每个平均 SNP 指数的 x 轴值设置在第一个和第五个 SNP 之间的中点。