The plant secondary cell wall is the major source of lignocellulosic biomass, a renewable energy resource that can be used for bioethanol production. To comprehensively identify transcription factors (TFs), glycosyltransferase (GT) and glycosyl hydrolase (GH) involved in secondary cell wall formation in rice (Oryza sativa), co-expression network analysis was performed using 68 microarray data points for different rice tissues and stages. In addition to rice genes encoding orthologs of Arabidopsis thaliana TFs known to regulate secondary cell wall formation, the network analysis suggested many novel TF genes likely to be involved in cell wall formation. In the accompanying paper (Hirano et al.), several of these TFs are shown to be involved in rice secondary cell wall formation. Based on a comparison of the rice and Arabidopsis networks, TFs were classified as common to both species or specific to each plant species, suggesting that in addition to a common transcriptional regulatory mechanism of cell wall formation, the two plants may also use species-specific groups of TFs during secondary wall formation. Similarly, genes encoding GT and GH were also classified as genes showing species-common or species-specific expression patterns. In addition, genes for primary or secondary cell wall formation were also suggested. The list of rice TF, GT and GH genes provides an opportunity to unveil the regulation of secondary cell wall formation in grasses, leading to optimization of the cell wall for biofuel production.

Shape is one of the most important traits of agricultural products due to its relationships with the quality, quantity, and value of the products. For strawberries, the nine types of fruit shape were defined and classified by humans based on the sampler patterns of the nine types. In this study, we tested the classification of strawberry shapes by machine learning in order to increase the accuracy of the classification, and we introduce the concept of computerization into this field. Four types of descriptors were extracted from the digital images of strawberries: (1) the Measured Values (MVs) including the length of the contour line, the area, the fruit length and width, and the fruit width/length ratio; (2) the Ellipse Similarity Index (ESI); (3) Elliptic Fourier Descriptors (EFDs), and (4) Chain Code Subtraction (CCS). We used these descriptors for the classification test along with the random forest approach, and eight of the nine shape types were classified with combinations of MVs + CCS + EFDs. CCS is a descriptor that adds human knowledge to the chain codes, and it showed higher robustness in classification than the other descriptors. Our results suggest machine learning's high ability to classify fruit shapes accurately. We will attempt to increase the classification accuracy and apply the machine learning methods to other plant species.

Seedling vigor, which is controlled by many quantitative trait loci (QTLs), is one of several important agronomic traits for direct-seedling rice systems. However, isolating these QTL genes is laborious and expensive. Here, we combined QTL mapping and microarray profiling to identify QTL genes for seedling vigor. By performing QTL mapping using 82 backcross inbred lines (BILs) of the Koshihikari (japonica) and Habataki (indica) cultivars for the rice initial growth, we identified two QTLs, early-stage plant development1/2 (qEPD1 and qEPD2), whose Koshihikari alleles promote plant height and/or leaf sheath length. Phenotypic analysis of the two substituted lines carrying the Habataki qEPD1 or qEPD2 allele revealed that qEPD2 functioned more dominantly for the initial growth of rice. From the microarray experiment, 55 and 45 candidate genes were found in the qEPD1 and qEPD2 genomic regions, which are expressed differentially between each substitution line (SL) and Koshihikari. Gibberellin 20 oxidase-2 (OsGA20ox2), which is identical to Semi Dwarf1 (SD1), was included among the 55 candidate genes of qEPD1, whereas its paralog, OsGA20ox1, was included among the 45 candidate genes of qEPD2. Consistently, introduction of the Koshihikari OsGA20ox1 allele into SL(qEPD2) increaseed its plant height and leaf sheath length significantly relative to the introduction of the Habataki OsGA20ox1 allele. Therefore, microarray profiling could be useful for rapidly screening QTL candidate genes. We concluded that OsGA20ox1 and OsGA20ox2 (SD1) function during the initial growth of rice, but OsGA20ox1 plays a dominant role in increasing plant height and leaf sheath length at the initial growth stage.

SUMMARYThe morphology of rice (Oryza sativa L.) panicles is an important determinant of grain yield, and elucidation of the genetic control of panicle structure is very important for fulfilling the demand for high yield in breeding programs. In a quantitative trait locus (QTL) study using 82 backcross inbred lines (BILs) derived from Koshihikari and Habataki, 68 QTLs for 25 panicle morphological traits were identified. Gene expression profiling from inflorescence meristems of BILs was obtained. A combination of phenotypic QTL (pQTL) and expression QTL (eQTL) analysis revealed co‐localization between pQTLs and eQTLs, consistent with significant correlations between phenotypic traits and gene expression levels. By combining pQTL and eQTL data, two genes were identified as controlling panicle structure: OsMADS18 modulates the average length of the primary rachis and OsFTL1 has pleiotropic effects on the total number of secondary rachides, number of grains per panicle, plant height and the length of flag leaves. Phenotypes were confirmed in RNA interference knocked‐down plants and overexpressor lines. The combination of pQTL and eQTL analysis could facilitate identification of genes involved in rice panicle formation.

Male sterility is defined as the loss of pollen fertility, and it represents a plant reproductive isolation symptom, along with self-incompatibility. It plays an important role in the efficient production of F 1-hybrid seeds, which results in affordable seed prices for farmers. Male sterile cultivated strawberry Fragaria × ananassa Duch. plants were found in an F 1 population and reciprocal backcrossed populations derived from a cross between 'Fukuoka S6' and 'Kaorino'. Male sterile plants were clearly distinguished from male fertile plants in those populations based on the anther color. The pollen of the male sterile plants was a lighter yellow color and not maturely shaped compared with pollen of male fertile plants. Genotyping was performed using EST-SSR markers in the three populations. Quantitative trait locus analyses for pollen fertility were conducted independently using three kinds of populations, and this revealed that male sterility was controlled by three independent chromosomal regions in these populations, which corresponded to chromosome 4 in the wild strawberry (Fragaria vesca) genome. One region was derived from 'Fukuoka S6' and the other two regions from 'Kaorino'. The segregation patterns of fertile and sterile plants in each population clearly supported the three gene theory of male sterility in cultivated strawberries. The accumulation of recessive alleles at the three regions led to male sterility, and the existence of a dominant allele in at least one region resulted in fertile pollen. Male sterile plants were also found in two self-pollenated populations derived from 'Fukuoka S6' and 'Kaorino', and the effects of the three regions were validated. The adaptability levels of the three genes with different genetic backgrounds were also evaluated using core collection cultivars and selected lines derived using recurrent selection. We also detected flanking DNA markers for the three regions associated with male sterility. The use of these markers, which are in the vicinity of quantitative trait loci and responsible for malesterility, could increase the efficiency of producing seed-propagated strawberry F 1-hybrids.