Molecular Plant Advance Access published online on August 5, 2008
Molecular Plant, doi:10.1093/mp/ssn022
A Genome-Wide Transcription Analysis Reveals a Close Correlation of Promoter INDEL Polymorphism and Heterotic Gene Expression in Rice Hybrids
a National Institute of Biological Sciences, Zhongguancun Life Science Park, Beijing 102206, China
b Institute of Biophysics, Chinese Academy of Sciences, 15 Datun Road, Beijing 100101, China
c Department of Botany, College of Life Sciences, Hunan Normal University, Changsha 410081, China
d Peking–Yale Joint Research Center of Plant Molecular Genetics and Agrobiotechnology, College of Life Sciences, Peking University, Beijing 100871, China
e Department of Molecular, Cellular and Developmental Biology, Yale University, New Haven, Connecticut 06520, USA
1 To whom correspondence should be addressed. E-mail xingwang.deng{at}yale.edu, fax 1-203-432-5726, tel. 1-203-432-8908.
 |
Abstract |
|---|
 TOP Abstract INTRODUCTIONRESULTSDISCUSSIONMETHODSSUPPLEMENTARY DATAFUNDING |
|---|
Heterosis, or hybrid vigor, refers to the phenomenon in which hybrid progeny of two inbred varieties exhibits enhanced growth or agronomic performance. Although a century-long history of research has generated several hypotheses regarding the genetic basis of heterosis, the molecular mechanisms underlying heterosis and heterotic gene expression remain elusive. Here, we report a genome-wide gene expression analysis of two heterotic crosses in rice, taking advantage of its fully sequenced genomes. Approximately 7–9% of the genes were differentially expressed in the seedling shoots from two sets of heterotic crosses, including many transcription factor genes, and exhibited multiple modes of gene action. Comparison of the putative promoter regions of the ortholog genes between inbred parents revealed extensive sequence variation, particularly small insertions/deletions (INDELs), many of which result in the formation/disruption of putative cis-regulatory elements. Together, these results suggest that a combinatorial interplay between expression of transcription factors and polymorphic promoter cis-regulatory elements in the hybrids is one plausible molecular mechanism underlying heterotic gene action and thus heterosis in rice.
|
INTRODUCTION |
|---|
|
TOPAbstractINTRODUCTIONRESULTSDISCUSSIONMETHODSSUPPLEMENTARY DATAFUNDING |
|---|
Heterosis, or hybrid vigor, refers to the phenomenon in which progeny of two inbred varieties exhibits enhanced agronomic performance such as biomass production, growth rate, and fertility relative to both parents (Shull, 1908). Heterosis has been widely exploited in the breeding of maize, rice, and other crops, which is important to yield growth worldwide. For example, it is estimated that hybrid rice technology for large-scale production has a yield advantage of 15–20% over the elite inbred varieties.
The classical quantitative genetic explanations for heterosis center on a few concepts, including dominance (Davenport, 1908), over-dominance (Shull, 1908; East, 1908), and epistasis (Yu et al., 1997; Li et al., 2001). The dominance model posits that each of the inbred lines contains slightly deleterious alleles that reduce their fitness. The hybrid will benefit from the complementation of these deleterious alleles and will display a superior phenotype. The over-dominance model suggests that the heterozygous combination of alleles at a given locus is phenotypically superior to either of the homozygous combinations for that locus, thereby resulting in a superior hybrid. Both the dominance and over-dominance hypotheses are based on single locus theory, and they may be inadequate to address the molecular mechanism for heterosis (Birchler et al., 2003). Epistasis is classically defined as interactions of superior alleles at different loci from two parents, and the effects may show additivity, dominance or over-dominance (Yu et al., 1997; Li et al., 2001).
Genetic analyses have implied the involvement of multiple genetic loci in bringing about heterosis (Li et al., 2001; Stuber et al., 1992; Xiao et al., 1995; Hua et al., 2002, 2003; Lu et al., 2003). However, comprehensive assessment of the expression of the loci contributing to heterosis is largely not available due to the limited resolution of classic genetic analysis in tracking multiple loci (Birchler et al., 2003; Lippman and Zamir, 2007). Advances in DNA microarray-based analysis have made it possible to examine transcription of all individual genes in the genome in inbred parents and their F1 hybrid simultaneously. It has been widely adopted for analyzing the global gene expression of a number of organisms (Stoughton, 2005). Recently, several studies have analyzed the association of gene expression and heterosis in maize, rice, and Arabidopsis using microarrays (Swanson-Wagner et al., 2006; Guo et al., 2003, 2006; Stupar and Springer, 2006; Meyer et al., 2007; Vuylsteke et al., 2005; Huang et al., 2006). Multiple modes of gene action, including additivity, high- and low-parent dominance, over-dominance, and under-dominance, were found in those studies. However, no consensus as to how those differential expression patterns arose and how they relate to heterosis in the organism has been examined.
Rice is one of the most important stable food crops and an excellent model for grass species. Further, the availability of the complete sequences of the two rice genomes (Goff et al., 2002; Yu et al., 2002, 2005) and the readily traceable phenotypes resulting from heterosis make rice an ideal system in which to study the molecular events accompanying heterosis. As a first step to understanding the molecular basis of heterosis, we used a whole genome microarray to examine gene expression patterns in two heterotic crosses of rice. Analysis of the promoter regions of the differentially expressed genes indicated that INDEL polymorphism is an important contributor to heterotic gene action.
|
RESULTS |
|---|
|
TOPAbstractINTRODUCTIONRESULTSDISCUSSIONMETHODSSUPPLEMENTARY DATAFUNDING |
|---|
Characterization of the Two Heterotic Crosses Used in this Study
We used a rice whole genome microarray to examine gene expression patterns in two heterotic crosses (see Methods). One of the crosses involves the super hybrid rice commonly planted in China—Liang-You-Pei-Jiu (LYP9), an F1 hybrid derived from crossing of the inbred paternal line 93-11 (Oryza sativa L. ssp. Indica) and maternal line Pei-Ai 64s (PA64s, with a mixed genetic background of indica, japonica, and javanica). The F1 hybrid shows taller, stronger tillering activity and higher yield than both parents (Figure 1A, top). The other cross (Figure 1A, bottom) involves the inbred paternal line 93-11 and maternal line Nipponbare (Oryza sativa L. ssp. japonica); both strains have their genome sequenced. The F1 hybrid of this cross displays a strong hybrid vigor and higher biomass production than both parents. Both F1 hybrids show significantly increased net photosynthetic rate (Pn) (Figure 1B) and GA3 and zeatin content (Figure 1C) compared with the respective inbred parents. We selected four-leaf stage seedlings as the target for microarray analyses, as heterosis at this stage is already evident for both crosses (Supplemental Table 1).
|
) and 93-11 (
), and the lower panel is the cross between Nipponbare (Identification and Characterization of Differentially Expressed Genes
The processed microarray data was used to determine differential gene expression between inbred parents or between parental lines and their F1 hybrids (Figure 2A and 2B). A total of 3488 (9.4%) genes were identified with significant differential expression (p < 0.05) in the PA64s x 93-11 cross. Among these genes, 66% (2317 of 3488) exhibited an expression pattern that was not distinguishable from additivity (average expression level of the two parental lines), while the other 34% (1171 of 3488) genes showed non-additive expression patterns. Similarly, a total of 2416 (6.5%) differentially expressed genes were detected from the Nipponbare x 93-11 cross, with 44% (1055 of 2466) and 56% (1361 of 2416) genes displaying additive and non-additive expression, respectively. The non-additive differentially expressed genes from the two crosses were further classified into four distinct modes based on their deviation from the mid-parent prediction: high-parent dominance, low-parent dominance, over-dominance, and under-dominance (Table 1). A random set of differentially expressed genes selected from both crosses was used to validate the transcriptional level by semi-quantitative RT–PCR (Supplemental Figure 1). The PCR analysis confirmed approximately 95% (41 of 43 for PA64s x 93-11 and 33 of 35 for Nipponbare x 93-11) of differentially expressed genes identified from microarray analysis. These findings are consistent with a recent study in maize (Swanson-Wagner et al., 2006) that supports the involvement of multiple modes of gene action in association with heterosis.
|
|
We further investigated the profile changes of differentially transcribed genes in individual rice genotype. Using a hierarchical clustering algorithm (Figure 2C), we found that in the PA64s x 93-11 hybrid cross, gene expression pattern of the F1 hybrid (LYP9) was more similar to the paternal line (93-11) but deviated from that of the maternal line (PA64s). On the other hand, gene expression profile in the F1 hybrid in the Nipponbare x 93-11 hybrid cross was more similar to the maternal line (Nipponbare) than the paternal line (93-11). These results indicate that the hybrid could exhibit a gene expression pattern closer to either the paternal or maternal parent, depending on parental combination of individual crosses. We also compared gene expression profiles in the F1 hybrids between the reciprocal cross, 93-11 (maternal) x Nipponbare (paternal), and found that only a small portion (less than 10%) of differentially expressed genes seemed to be caused by maternal effects (data not shown). Therefore, it appears that the differentially expressed genes identified from the two heterotic crosses were not mainly caused by maternal effects and thus are excellent candidates for studying the molecular events associated with heterosis.
Differential Regulation of Metabolic Processes
An analysis of the distribution among the Gene Ontology (GO) functional categories indicated that in both hybrid crosses, those differentially expressed genes are relatively enriched in pathways for carbohydrate metabolism, metabolism of cofactors and vitamins, amino acid metabolism, and biosynthesis of secondary metabolites (Supplemental Figure 2). We further examined the expression of individual genes encoding enzymatic steps in the standardized AraCyc-defined metabolic pathways (Mueller et al., 2003) between F1 hybrid and inbred parents (Supplemental Figure 3), which confirmed the involvement of multiple biosynthetic and secondary metabolic pathways in both rice heterotic crosses. Analysis of two representative pathways, namely carbon metabolism (Calvin cycle) and gibberellin biosynthesis, are illustrated in Figure 3. We found that in the F1 hybrid, genes involved in the Calvin cycle exhibited strong non-additive gene actions of high-parent dominance and over-dominance, which may help to reduce the effects of rate-limiting steps in carbon metabolism (Figure 3A and 3C). This observation is consistent with a significant net photosynthetic rate (Pn) increase in F1 hybrid plants (Figure 1B). Furthermore, we also found that GA3 and zeatin contents increased in F1 hybrid plants compared to parental inbred lines at the four-leaf stage (Figure 1C), which is consistent with recent reports that bioactive gibberellins are responsible for elongating shoots and biomass production (Eriksson et al., 2000; Biemelt et al., 2004). This is also consistent with observed changes in up-regulated expression of genes in the gibberellins (Figure 3B) and cytokinin biosynthetic pathways (figure not shown). Interestingly, genes involved in biosynthesis of the hormones exhibited a strong over-dominance/under-dominance mode of gene action (Figure 3B and 3C). Some of the genes involved in other biosynthetic pathways, such as glycolysis and gluconeogenesis (Supplemental Figure 3), were also verified to exhibit multiple expression profiles by semi-quantitative RT–PCR (data not shown). These results suggest that the observed modes of gene action are generally in line with the functionality of the involved genes.
|
Enrichment of INDEL Polymorphism in Rice Promoter Regions
Because of the availability of the quality genomic sequences, the Nipponbare x 93-11 cross was first examined regarding promoter sequence polymorphism between the two parental lines and its relation to differential gene expression among the three genotypes. To this end, the 3-kb upstream region from the start codon (ATG) of orthologous gene pairs between the two parental lines was compared for all sequence polymorphism sites by Smith-Waterman local alignment algorithm (see Methods), after excluding repetitive sequences. As shown in Figure 4A, there are evidently higher sequence variations between the 500 –1500-bp upstream regions, coinciding with the anticipated promoter regions. This enrichment of sequence variations in the anticipated promoter regions is more dramatic for the small insertion/deletion polymophisms (INDEL) than that of the single nucleotide polymorphism (SNP) (Figure 4B). Many putative functional cis-elements were found in those promoter INDELs by searching PLACE database (see Methods) (Supplemental Table 2).
|
2 test was used to determine the statistical significance between the two group genes. * P < 0.01, 
no significant difference.Significantly Higher Promoter INDEL Polymorphism among Heterotic Expressed Genes
To further examine a possible relationship between promoter polymorphisms and differential expression of genes, we grouped the non-differentially and differentially expressed genes and calculated their respective promoter sequence polymorphism frequency (Figure 4C). The percentage of promoter regions with INDELs for non-differentially expressed genes was 40.1%, but significantly increased to 49.8% for differentially expressed genes. The percentage of SNP between the two groups of genes was similar. Figure 4E illustrates some specific examples of differentially expressed genes with promoter INDELs between inbred parental 93-11 and Nipponbare or PA64s. In each of those cases, the promoter INDEL polymorphism coincides with a putative plant promoter cis-element (shaded in pink), thus likely affecting the cis-element function and gene expression. For example, genes (1), (3), and (4) illustrated in Figure 3A and quite a number of genes involved in the biosynthetic pathways in Supplemental Figure 3 contain promoter variation between the two parental lines (data not shown). To exclude any possible sequence error in our observed promoter sequence variation, we selected 30 pairs of genes with promoter INDELs for PCR-sequencing using 93-11 and Nipponbare genomic DNA as templates. This analysis confirmed accuracy of 93% (28 of 30). The same analysis using PA64s and 93-11 genomic DNA as templates confirmed all INDELs in the promoter regions between these two genotypes. These results indicate that promoter regions of differential expressed genes in the heterotic crosses are enriched with INDEL polymorphism.
We further investigated the promoter polymorphic variation among distinct modes of differentially expressed genes in the two heterotic crosses. As shown in Figure 4D, in the Nipponbare x 93-11 cross, the promoter region of the additive and over-/under-dominant genes all exhibited significant higher frequency of promoter INDELs compared to non-differentially expressed genes. On the contrary, the high-parent dominance gene group showed a significantly lower frequency of promoter INDELs compared to non-differentially expressed genes, while the low-parent dominance group lacks sufficient gene number to be statistically significant. Analysis of the PA64s x 93-11 cross expression data and the privilegedly access to PA64s genomic sequence generated similar results (see Supplemental Figure 4). Therefore, differentially expressed genes with different modes of gene action in heterotic crosses are correlated with distinct patterns of INDEL enrichment in their promoter regions.
Models for Heterotic Gene Expression Patterns
Transcription factors regulate the binding of RNA polymerase to their target genes by binding to specific cis-regulatory DNA sequences. Polymorphisms in the cis-regulatory elements and differential expression of transcription genes in the parental lines, and their unique combination in the F1 hybrids, thus could bring about the five possible modes of gene action. In both heterotic crosses, many transcriptional factor genes were found as differentially expressed (Table 2 and Supplemental Table 3). The transcription factor genes could in turn regulate other genes that display distinct modes of heterotic gene action.
|
Based on the above analysis, we propose a working hypothesis (Figure 5) to explain how promoter INDELs and differential expression of transcription factors acting upon them could affect gene expression and contribute to different modes of gene action in F1 hybrids. As shown in Figure 5A, hybrid progeny inherits one copy of the gene from each parent. In the case of additive expression, the lack (or low activity) of transcription from one allele is assumed to be caused by the lack of (or weak) binding of a transcriptional activator to the promoter due to a cis-element mutation; the F1 hybrid progeny will thus have an expression level between the two parents. In the case of high-parent dominance, where the promoters are the same in both parents but the transcription factor is missing (or with low activity) in one parent, the progeny would have an expression level similar to the high expression parent. In the case of over-dominance, it is possible that the responsible transcription factor is only present in the parent with a dysfunctional promoter due to INDEL, but missing (weak) in the other parent that has a functional promoter. The hybrid would thus have the transcription factor from one parent and the functional promoter from the other, resulting in higher expression levels than either parent. Similarly, the transcription factor can act to repress transcription and its interaction with promoter INDELs can result in other modes of gene action, such as additivity, low-parent dominance, or under-dominance (Figure 5B).
|
|
DISCUSSION |
|---|
|
TOPAbstractINTRODUCTIONRESULTSDISCUSSIONMETHODSSUPPLEMENTARY DATAFUNDING |
|---|
Based on our microarray analysis, approximately 9.4% (3488 of 36 962) and 6.5% (2416 of 36 962) of genes exhibited differential transcription in the PA64s x 93-11 and the Nipponbare x 93-11 hybrid cross, respectively. These differential genes exhibited additive and non-additive expression patterns; however, their proportion differs in the two heterotic crosses (Figure 2). In the PA64s x 93-11 cross, a large portion (66%, 2317 of 3488) exhibited a mode of gene action that could not be distinguished from additivity, which was similar to recent studies in maize (Swanson-Wagner et al., 2006; Stupar and Springer, 2006). On the contrary, several studies reported a high proportion of non-additive expression patterns in F1 hybrids (Song and Messing, 2003; Gibson et al., 2004; Auger et al., 2005), which was consistent with our observation that a majority of the differentially expressed genes exhibit non-additive gene action (56%, 1361 of 2416) in the Nipponbare x 93-11 heterotic cross. Among those non-additive expressed genes, the proportion of genes with clear over-dominant action was 6.5% (215 of 3416) and 10% (245 of 2416) in the PA64s x 93-11 and the Nipponbare x 93-11 heterotic crosses, respectively, which was similar to results from prior studies (Vuylsteke et al., 2005; Gibson et al., 2004). Together, our results support the notion that multiple molecular mechanisms contribute to heterosis.
Many factors affect plant yield and productivity, such as the capacity of fixing carbon dioxide (CO2) and converting the fixed carbon into dry matters. Functional analysis indicated that a large number of differentially expressed genes in F1 hybrids were involved in multiple biosynthesis pathways related to carbon metabolism, such as the glycolysis pathway, the tricarboxylic acid (TCA) cycle, the calvin cycle, and the glyconeogenesis pathway (Figure 3 and Supplemental Figures 2 and 3). Our microarray data also showed that many genes involving in photosynthesis were up-regulated in F1 hybrids, similarly with findings from previous studies (Bao et al., 2005). Additionally, up-regulated genes for hormone biosynthesis are consistent with a higher content of gibberellin and cytokinin in F1 hybrids (Figure 1C). Thus, novel changes of biosynthetic processes by differentially expressed genes might result in heterotic phenotypes of F1 hybrid rice.
It is well established that the proportion and composition of transcribed genes change considerably during the lifecycle of plants, and most regulation of gene expression occurs at the transcriptional level. Comparative analysis of promoter alleles between inbreds showed that sequence polymorphism preferentially occurred in those differentially transcribed genes. In this study, mid-parent (or additive) expression of F1 hybrid indicated that the alleles from respective inbred parents did not equally contribute to transcript accumulation for variance of promoter sequence, which was consistent with the previous study (Guo et al., 2004). In the hybrids, two alleles are exposed to the common trans-acting factors (such as transcription factors), but the promoter INDEL polymorphism may lead to target genes differentially interacting with the common transcription factors, thus resulting in differential transcription between alleles tending towards mid-parent or additive levels as described in Figure 5. The majority of genes displaying non-additive profiles are expressed at levels within the range of two inbred parental lines, which might also be explained by the promoter sequence variation. For example, a transactivator or transrepressor in either of the inbred parents could lead to multiple non-additive expression patterns in the F1 hybrids (Figure 5). The fact of the promoter sequence polymorphism within non-differentially expressed genes might be due to developmental stage-specific expression fashions. In conclusion, the heterotic gene expression in the two rice hybrid crosses was in part or completely due to promoter INDEL polymorphism.
From an evolutionary point of view, the divergence of promoter sequences between inbred lines might be the result of natural selection or domestication through artificial selection for humans' desired traits. Several modes of natural selection on promoter sequences as well as on coding sequences (Wray et al., 2003), such as positive selection, over-dominant selection, and stabilizing selection, have been proposed. Many promoters are organized into functional modules, and each of them produces a discrete effect on the total transcription output, allowing selection to modify this discrete effect independently (Arnone and Davidson, 1997). Alternatively, many promoter alleles may be functionally co-dominant and thus immediately visible to selection, which increases the efficiency of fixing beneficial alleles and eliminating deleterious ones. A number of rice studies have revealed the importance of artificial selection in the establishment of cultivated rice (Cheng et al., 2003; Vaughan et al., 2003; Lu et al., 2006), and a most recent study (Konishi et al., 2006) indicated that the 5' sequence variation of regulatory region of qSH1 gene caused loss of seed shattering among japonica subspecies of rice, and the sequence variation was a result of artificial selection during rice domestication. Thus, comparison of the promoter–transcription factor combination that results in certain modes of gene action in rice and other grass species should reveal the contribution of natural selection as well as artificial selection in shaping the outcome of heterosis.
The genome scale transcription analysis reported herein revealed that about 7–9% of the genome is differentially expressed, exhibiting all modes of gene action in two heterotic crosses. The number of genes with additive gene action in hybrid F1 progeny varied over two-fold in the two crosses, while a number of genes with the non-additive gene action mode were quite similar in the two crosses. Enrichment of promoter INDEL polymorphism was highly correlated with differentially expressed genes with additive as well as over- and under-dominance gene action, but not the high- and low-parent dominance. A working model was proposed based on these results that could explain all possible modes of gene action as a result of promoter INDELs and differential expression of transcription factors. These results indicate that the interplay between promoter INDEL and regulated expression of transcription factors is one of the molecular events for heterotic gene expression contributing to heterosis in rice.
|
METHODS |
|---|
|
TOPAbstractINTRODUCTIONRESULTSDISCUSSIONMETHODSSUPPLEMENTARY DATAFUNDING |
|---|
Plant Materials and Oligonucleotide Microarray
LYP9 (F1 hybrid), 93-11 (Oryza sativa L. ssp indica cultivar), PA64s (with a mixed genetic background of indica, japonica, and javanica), Nipponbare (Oryza sativa L. ssp. japonica cultivar), and Nipponbare x 93-11 hybrids were cultivated under typical field conditions in south China for physiological characterization and seed propagation of the isogenic strains. For hybrid seed production, LYP9 was produced in cooler areas, where the PA64s strain is male-sterility and can be easily cross-pollinated by 93-11. The hybrid seeds of Nipponbare x 93-11 were produced by manual pollination of the 93-11 pollens into Nipponbare stigma. Seedling shoots grown in environmentally controlled growth chambers at 28°C were harvested at the four-leaf stage for microarray analysis. The rice 70-mer oligonucleotide set representing 36 926 unique genes was used in this study as described previously (Jiao et al., 2005; Ma et al., 2005).
RNA Isolation, Probe Labeling, and Microarray Hybridization
Rice seedling shoots were frozen in liquid nitrogen, and total RNA was isolated using RNAwiz reagent (Ambion) and purified by the RNeasy kit (Qiagen) according to the manufacturer's instructions. For each sample, 100 mg of total RNA was labeled with aminoallyl-dUTP (aa-dUTP, Sigma-Aldrich) by reverse transcription. The aminoallyl- dUTP-labeled cDNAs were purified using a Microcon YM-30 filter (Millipore) and resuspended in 0.1 M NaHCO3. The purified cDNAs were further fluorescently labeled by conjugating monofunctional Cy2, Cy3, or Cy5 dye (Amersham) to the aminoallyl functional groups using a loop-like design. Three biological replicates were performed with each hybrid cross, and, in each replicate, alternative RNA pools for three genotypes were labeled with Cy2, Cy3, and Cy5 dye, respectively. After coupling at room temperature for 1 h, the labeling reaction was stopped by ethanolamine. The labeled probes were separated from unincorporated dye using the QIAquick PCR purification kit (Qiagen) and concentrated by a Microcon YM-30 filter, respectively. The following protocols for microarray hybridization, microarray slide washing, and array scanning were the same as described previously (Ma et al., 2001). Hybridized slides were scanned with a GenePix 4000B scanner (Axon), and independent TIFF images for triple Cy2, Cy3, and Cy5 channels were used for subsequent analysis.
Microarray Data Processing
After manual removal of spots with aberrant morphology, microarray spot intensity signals were acquired using the Axon GenePix Pro 5.0 software package without correction for background, and each slide included three troops of intensity data corresponding to Cy5, Cy3, and Cy2 channels. We first remove the dye intensity difference on each slides with the ‘LOWESS’ normalization method, which was respectively applied to log2 transformed intensities from each pair of two channels (Cy5/Cy3, Cy3/Cy2, Cy2/Cy5), and three repeats of such processing guaranteed that the dye effects had been removed. Subsequently, quantile normalization was applied to the three LOWESS-normalized microarray data to remove biases among slides. After that, all intensities of three channels can be compared to each other.
For detection of differentially expressed spots among three genotypes, the normalized data was log2 transformed and fitted into a mixed effect ANOVA model, with the software MAANOVA under R environment. After multiple testing between pairwise comparisons, spots with FDR-corrected P values <0.05 were regarded as differentially expressed genes. The same strategy was performed in a linear-in-genotype contrast when F1 genotype was compared to the average of the two parental lines. Spots with FDR-corrected P values <0.05 were regarded as non-additivity, when the spots with FDR-corrected P values >0.05 were regarded as no statistically significant difference from additivity. To classify the genes further distinguishably, the high-parent dominant genes and low-parent dominant genes were identified from the non-addititive group based on the criterion that F1 genotype was significantly different from one parent and no significantly different from another parent. From the additive group, the genes were identified as over-dominance or under-dominance when the F1 genotype was significantly higher or lower than both inbred parents, respectively.
Cluster Analysis
Cluster analysis was applied to all genes showing differential expression among hybrids and inbred parents from both two rice hybrid crosses, respectively. Differential expression was determined as stated above. Hierarchical clustering with the average linkage was performed using the software Cluster and visualized by the Treeview program (Eisen et al., 1998).
Promoter Variation and Motif Search
To detect the promoter variations, homologous gene pairs, with same length, mapping extron position and completely identical coding nucleotide sequence, were selected between two inbred parental genomes in each hybrid cross (PA64s x 93-11 and Nipponbare x 93-11). 5084 homologous gene pairs were screened out from the inbred lines PA64s and 93-11, and 5000 from the inbred lines Nipponbare and 93-11. After extracting the 3-kilo bases upstream regions from the start codon in these homologous gene pairs, Smith-Waterman local alignment algorithm was carried out to detect the polymorphism sites. Among those sites, deleted/inserted fragments which meet the following conditions were recorded particularly: (1) the length of inserted/deleted fragments ranges from 4 to 50 bp; and (2) these fragments are not single nucleotide repetitive sequences. Then, these small deleted/inserted fragments were searched in the Plant Cis-acting Regulatory DNA Elements database (Higo et al., 1998).
Correlation between Expression Modes and Promoter Deletion Frequency
To examine the relationship between promoter variation and gene expression profiles, the frequency of promoter deletions/insertions was detected in all differential genes. Gene expression patterns include two main groups based on the F1 expression: additivity and non-additivity. The former includes those genes exhibiting non-difference and 93-11 higher or lower than Nipponbare, and the latter includes those genes exhibiting high-parent, low-parent, over-dominance, and under-dominance. For each kind of distinct gene, 3-kb promoter alleles between inbred parents from the start codon mapped to corresponding genes were compared to each other using BLAT (Kent, 2002). The promoter alleles with identical sequences of more than 2 kb were subjected to BLAT to analyze average nucleotide numbers in the deleted/inserted fragments (gaps) according to each distinct expression profile, and then the Mann-Whitney U test was performed to estimate the different distributions in these differential genes with distinct expression patterns.
Semi-Quantitative RT–PCR
Total RNA was isolated from four-leaf seedling shoots as used in microarray analysis and treated with RNase-free DNase (Promega), and cDNA was synthesized with SuperScriptTM II First-Strand cDNA Synthesis kit (Invitrogen). PCR was performed using general standard techniques, and O. sativa actin gene was used as control.
Analysis of Net Photosynthesis and Hormones Content
The net photosynthesis was measured using a portable photosynthesis system (CIRAS-1, PP Systems, UK) by an open system in the morning between 09:00 and 11:00 h to avoid potential photoinhibition. Measurements were made by attaching a light source to the leaf chamber window under saturating photosynthetic photon flux densities (1200 µmol m–2 s–1). Data were determined at least in five leaves from five different plants of each variety.
The endogenous hormone content was measured using the Agilent 1100 high-performance liquid chromatography (HPLC) system (Agilent, Palo Alto, CA). Fresh leaves at the four-leaf stage were frozen in liquid nitrogen and then cold-dried under vacuum. 0.5 g dried leaves were added to 11 mL 80% aqueous methanol and immediately homogenized on ice. The homogenate was maintained during 15 h at 4°C in darkness with continuous shaking. Afterwards, it was centrifuged for 10 min at 4500 g at 4°C, and the clear upper phase was collected and evaporated under vacuum. Dry residue was re-dissolved in 8 mL of ammonium acetic buffer (0.1 mol L–1, pH 9.0). After centrifuging for 20 min at 15 000 rpm, the upper supernatant was purified sequentially through Polyvinylpolypyrrolidone (PVPP) column and DEAE Sephadex A225 column. Before HPLC analysis, the elution with 50% aqueous methanol was concentrated by Sep-Pak C18 column (Waters Chromatography). Standard gibberellin and zeatin were purchased from Fluka Co. (Switzerland). All solvents and buffers were HPLC-quality.
|
SUPPLEMENTARY DATA |
|---|
|
TOPAbstractINTRODUCTIONRESULTSDISCUSSIONMETHODSSUPPLEMENTARY DATAFUNDING |
|---|
Supplementary Data are available at Molecular Plant Online.
|
FUNDING |
|---|
|
TOPAbstractINTRODUCTIONRESULTSDISCUSSIONMETHODSSUPPLEMENTARY DATAFUNDING |
|---|
This study was supported by the 863 Rice Functional Genomics Program and the National ‘863’ High-Tech Project 2003AA 210070 from the Ministry of Science and Technology of China.
|
Acknowledgements |
|---|
We thank C.M. Lu and X.Y. Chen for suggestions and technical assistance with measurement of the rice photosynthetic characteristics. No conflict of interest declared.
|
Notes |
|---|
2 These authors contributed equally to this work.
-
Arnone MI, Davidson EH. The hardwiring of development: organization and function of genomic regulatory systems. Development. (1997) 124:1851–1864.[Abstract]
Auger DL, Gray AD, Ream TS, Kato A, Coe EH Jr, Birchler JA. Nonadditive gene expression in diploid and triploid hybrids of maize. Genetics. (2005) 169:389–397.[ISI][Medline]
Bao JY, et al. Sage analysis of gene expression study of a hybrid rice strain (LYP9) and its parental cultivars. Plant Physiol. (2005) 138:1216–1231.
Biemelt S, Tschiersch H, Sonnewald U. Impact of altered gibberellin metabolism on biomass accumulation, lignin biosynthesis, and photosynthesis in transgenic tobacco plants. Plant Physiol. (2004) 135:254–265.
Birchler JA, Auger DL, Riddle NC. In search of the molecular basis of heterosis. Plant Cell. (2003) 15:2236–2239.
Cheng C, Motohashi R, Tsuchimoto S, Fukuta Y, Ohtsubo H, Ohtsubo E. Polyphyletic origin of cultivated rice: based on the interspersion pattern of SINEs. Mol. Biol. Evol. (2003) 20:67–75.
Davenport CB. Degeneration, albinism and inbreeding. Science. (1908) 28:454–455.
East EM. Inbreeding in corn. Connecticut Agric. Exp. Stn. (1908) 1907:419–428.
Eisen MB, Spellman PT, Brown PO, Botstein D. Cluster analysis and display of genome-wide expression patterns. Proc. Natl Acad. Sci. U S A. (1998) 95:14863–14868.
Eriksson ME, Israelsson M, Olsson O, Moritz T. Increased gibberellin biosynthesis in transgenic trees promotes growth, biomass production and xylem fiber length. Nat. Biotechnol. (2000) 18:784–788.[CrossRef][ISI][Medline]
Gibson G, Riley-Berger R, Harshman L, Kopp A, Vacha S, Nuzhdin S, Wayne M. Extensive sex-specific nonadditivity of gene expression in Drosophila melanogaster. Genetics. (2004) 167:1791–1799.[CrossRef][ISI][Medline]
Goff SA, et al. A draft sequence of the rice genome (Oryza sativa L. ssp. japonica). Science. (2002) 296:92–100.
Guo M, Rupe MA, Danilevskaya ON, Yang X, Hu Z. Genome-wide mRNA profiling reveals heterochronic allelic variation and a new imprinted gene in hybrid maize endosperm. Plant J. (2003) 36:30–44.[CrossRef][ISI][Medline]
Guo M, Rupe MA, Yang XF, Crasta O, Zinselmeier C, Smith OS, Bowen B. Genome-wide transcript analysis of maize hybrids: allelic additive gene expression and yield heterosis. Theor. Appl. Genet. (2006) 113:831–845.[CrossRef][ISI][Medline]
Guo M, Rupe MA, Zinselmeier C, Habben J, Bowen BA, Smith OS. Allelic variation of gene expression in maize hybrids. Plant Cell. (2004) 16:1707–1716.
Higo K, Ugawa Y, Iwamoto M, Higo H. PLACE: a database of plant cis-acting regulatory DNA elements. Nucleic Acids Res. (1998) 26:358–359.
Hua J, Xing Y, Wu W, Xu C, Sun X, Yu S, Zhang Q. Single-locus heterotic effects and dominance by dominance interactions can adequately explain the genetic basis of heterosis in an elite rice hybrid. Proc. Natl Acad. Sci. U S A. (2003) 100:2574–2579.
Hua J, Xing Y, Xu C, Sun X, Yu S, Zhang Q. Genetic dissection of an elite rice hybrid revealed that heterozygotes are not always advantageous for performance. Genetics. (2002) 162:1885–1895.[ISI][Medline]
Huang Y, Zhang L, Zhang J, Yuan D, Xu C, Li X, Zhou D, Wang S, Zhang Q. Heterosis and polymorphisms of gene expression in an elite rice hybrid as revealed by a microarray analysis of 9198 unique ESTs. Plant Mol. Biol. (2006) 62:579–591.[CrossRef][ISI][Medline]
Jiao Y, Ma L, Strickland E, Deng XW. Conservation and divergence of light-regulated genome expression patterns during seedling development in rice and Arabidopsis. Plant Cell. (2005) 17:3239–3256.
Kent WJ. BLAT: the BLAST-like alignment tool. Genome Res. (2002) 12:656–664.
Konishi S, Izawa T, Lin SY, Ebana K, Fukuta Y, Sasaki T, Yano M. An SNP caused loss of seed shattering during rice domestication. Science. (2006) 312:1392–1396.
Li ZK, et al. Overdominant epistatic loci are the primary genetic basis of inbreeding depression and heterosis in rice. I. Biomass and grain yield. Genetics. (2001) 158:1737–1753.[ISI][Medline]
Lippman ZB, Zamir D. Heterosis: revisiting the magic. Trends Genet. (2007) 23:60–66.[CrossRef][ISI][Medline]
Lu H, Romero-Severson J, Bernardo R. Genetic basis of heterosis explored by simple sequence repeat markers in a random-mated maize population. Theor. Appl. Genet. (2003) 107:494–502.[CrossRef][ISI][Medline]
Lu J, Tang T, Tang H, Huang J, Shi S, Wu CI. The accumulation of deleterious mutations in rice genomes: a hypothesis on the cost of domestication. Trends Genet. (2006) 22:126–131.[CrossRef][ISI][Medline]
Ma L, et al. A microarray analysis of the rice transcriptome and its comparison to Arabidopsis. Genome Res. (2005) 15:1274–1283.
Ma L, Li J, Qu L, Hager J, Chen Z, Zhao H, Deng XW. Light control of Arabidopsis development entails coordinated regulation of genome expression and cellular pathways. Plant Cell. (2001) 13:2589–2607.
Meyer S, Pospisil H, Scholten S. Heterosis associated gene expression in maize embryos 6 days after fertilization exhibits additive, dominant and overdominant pattern. Plant Mol. Biol. (2007) 63:381–391.[CrossRef][ISI][Medline]
Mueller LA, Zhang P, Rhee SY. AraCyc: a biochemical pathway database for Arabidopsis. Plant Physiol. (2003) 132:453–460.
Shull GH. The composition of a field of maize. Am. Breed Assn. (1908) 4:296–301.
Song R, Messing J. Gene expression of a gene family in maize based on noncollinear haplotypes. Proc. Natl Acad. Sci. U S A. (2003) 100:9055–9060.
Stoughton RB. Applications of DNA microarray in biology. Annu. Rev. Biochem. (2005) 74:53–82.[CrossRef][ISI][Medline]
Stuber CW, Lincoln SE, Wolff DW, Helentjaris T, Lander ES. Identification of genetic factors contributing to heterosis in a hybrid from two elite maize inbred lines using molecular markers. Genetics. (1992) 132:823–839.[ISI][Medline]
Stupar RM, Springer NM. Cis-transcriptional variation in maize inbred lines B73 and Mo17 leads to additive expression patterns in the F1 hybrid. Genetics. (2006) 173:2199–2210.[CrossRef][ISI][Medline]
Swanson-Wagner RA, Jia Y, DeCook R, Borsuk LA, Nettleton D, Schnable PS. All possible modes of gene action are observed in a global comparison of gene expression in a maize F1 hybrid and its inbred parents. Proc. Natl Acad. Sci. U S A. (2006) 103:6805–6810.
Vaughan DA, Morishima H, Kadowaki K. Diversity in the Oryza genus. Curr. Opin. Plant Biol. (2003) 6:139–146.[CrossRef][ISI][Medline]
Vuylsteke M, Van Eeuwijk F, Van Hummelen P, Kuiper M, Zabeau M. Genetic analysis of variation in gene expression in Arabidopsis thaliana. Genetics. (2005) 171:1267–1275.[CrossRef][ISI][Medline]
Wray GA, Hahn MW, Abouheif E, Balhoff JP, Pizer M, Rockman MV, Romano LA. The evolution of transcriptional regulation in eukaryotes. Mol. Biol. Evol. (2003) 20:1377–1419.
Xiao J, Li J, Yuan L, Tanksley SD. Dominance is the major genetic basis of heterosis in rice as revealed by QTL analysis using molecular markers. Genetics. (1995) 140:745–754.[ISI][Medline]
Yu J, et al. A draft sequence of the rice genome (Oryza sativa L. ssp. indica). Science. (2002) 296:79–92.
Yu J, et al. The genomes of Oryza sativa: a history of duplications. PLoS Biol. (2005) 3:e38.[CrossRef][Medline]
Yu SB, Li JX, Xu CG, Tan YF, Gao YJ, Li XH, Zhang Q, Maroof MA. Importance of epistasis as the genetic basis of heterosis in an elite rice hybrid. Proc. Natl Acad. Sci. U S A. (1997) 94:9226–9231.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||