Published papers

Gene expression levels exhibit stochastic variations among genetically identical organisms under the same environmental conditions. In many recent transcriptome analyses based on RNA sequencing (RNA-seq), variations in gene expression levels among replicates were assumed to follow a negative binomial distribution, although the physiological basis of this assumption remains unclear. In this study, RNA-seq data were obtained from Arabidopsis thaliana under eight conditions (21-27 replicates), and the characteristics of gene-dependent empirical probability density function (ePDF) profiles of gene expression levels were analyzed. For A. thaliana and Saccharomyces cerevisiae, various types of ePDF of gene expression levels were obtained that were classified as Gaussian, power law-like containing a long tail, or intermediate. These ePDF profiles were well fitted with a Gauss-power mixing distribution function derived from a simple model of a stochastic transcriptional network containing a feedback loop. The fitting function suggested that gene expression levels with long-tailed ePDFs would be strongly influenced by feedback regulation. Furthermore, the features of gene expression levels are correlated with their functions, with the levels of essential genes tending to follow a Gaussian-like ePDF while those of genes encoding nucleic acid-binding proteins and transcription factors exhibit long-tailed ePDF.

The timing of both flowering and maturation determine crop adaptability and productivity. Soybean (Glycine max) is cultivated across a wide range of latitudes. The molecular-genetic mechanisms for flowering in soybean have been determined for photoperiodic responses to long days (LDs), but remain only partially determined for the delay of flowering under short-day conditions, an adaptive trait of cultivars grown in lower latitudes. Here, we characterized the late-flowering (LF) habit introduced from the Thai cultivar K3 into a photoperiod-insensitive genetic background under different photo-thermal conditions, and we analyzed the genetic basis using quantitative trait locus (QTL) mapping. The LF habit resulted from a basic difference in the floral induction activity and from the suppression of flowering, which was caused by red light-enriched LD lengths and higher temperatures, during which FLOWERING LOCUS T (FT) orthologs, FT2a and FT5a, were strongly down-regulated. QTL mapping using gene-specific markers for flowering genes E2, FT2a and FT5a and 829 single nucleotide polymorphisms obtained from restriction-site associated DNA sequencing detected three QTLs controlling the LF habit. Of these, a QTL harboring FT2a exhibited large and stable effects under all the conditions tested. A resequencing analysis detected a nonsynonymous substitution in exon 4 of FT2a from K3, which converted the glycine conserved in FT-like proteins to the aspartic acid conserved in TERMINAL FLOWER 1-like proteins (floral repressors), suggesting a functional depression in the FT2a protein from K3. The effects of the remaining two QTLs, likely corresponding to E2 and FT5a, were environment dependent. Thus, the LF habit from K3 may be caused by the functional depression of FT2a and the down-regulation of two FT genes by red light-enriched LD conditions and high temperatures.

Nandina domestica ‘Otafukunanten’ shrubs with narrow-leaf symptoms were diagnosed with citrus leaf blotch virus (CLBV) infection based on RNA-sequencing. The full-length sequence of the CLBV strain isolated from N. domestica was determined, and the amino acid sequence shared 83.7% and 80.2% identity with the replicase polyprotein and coat protein, respectively, of the most closely related CLBV strain. All samples with narrow-leaf symptoms were infected with CLBV and cucumber mosaic virus. To the best of our knowledge, this is the first report of CLBV infection of a plant in the order Ranunculales.

The rhizome is a plant organ that develops from a shoot apical meristem but penetrates into belowground environments. To characterize the gene expression profile of rhizomes, we compared the rhizome transcriptome with those of the leaves, shoots and roots of a rhizomatous Brassicaceae plant, Cardamine leucantha. Overall, rhizome transcriptomes were characterized by the absence of genes that show rhizome-specific expression and expression profiles intermediate between those of shoots and roots. Our results suggest that both endogenous developmental factors and external environmental factors are important for controlling the rhizome transcriptome. Genes that showed relatively high expression in the rhizome compared to shoots and roots included those related to belowground defense, control of reactive oxygen species and cell elongation under dark conditions. A comparison of transcriptomes further allowed us to identify the presence of an ER body, a defense-related belowground organelle, in epidermal cells of the C. leucantha rhizome, which is the first report of ER bodies in rhizome tissue.

The tea plant (Camellia sinensis (L.) Kuntze) of Japan is now thought to have originated in China. Actual cultivation of tea plants presumably started in Kyoto, and then spread to other regions of Japan. Tea gardens used to be composed of heterogenic, seed-derived populations—landraces—selected for climatic preferences and cultivation methods. Cultivars were bred from landraces in the modern era. A number of landraces remain in Kyoto Prefecture. However, little is known about their genetic characteristics compared with other landraces. We investigated the relationships of tea landraces and cultivars from Kyoto Prefecture and other regions. A neighbor-joining phylogram was constructed from 113 lines including 68 landraces, 44 cultivars, and one wild relative, based on simple sequence repeat markers. The lines could be classified into four groups (I–IV). In group I, most of the Kyoto landraces were close to other Japanese landraces, supporting the idea of the spread of tea plants from Kyoto to other regions in the country. The remaining lines were included into groups III and IV, apart from group II, which contained Shizuoka lines. Similar results were observed by restriction site-associated DNA sequencing analysis using 44 cultivars. Our data provide valuable information for the classification of tea lines, especially for the relationships of Kyoto lines among Japanese tea varieties.

Since ancient times, humans have bred several plants that we rely on today. However, little is known about the divergence of most of these plants. In the present study, we investigated the divergence of Mibuna (Brassica rapa L. subsp. nipposinica L. H. Bailey), a traditional leafy vegetable in Kyoto (Japan), by combining genetic analysis and a survey of ancient literature. Mibuna is considered to have been bred 200 years ago from Mizuna, another traditional leafy vegetable in Kyoto. Mibuna has simple spatulate leaves, whereas Mizuna has characteristic serrated leaves. The quantitative trait loci (QTL) and gene expression analyses suggested that the downregulation of BrTCP15 expression contributed to the change in the leaf shape from serrated to simple spatulate. Interestingly, the SNP analysis indicated that the genomic region containing the BrTCP15 locus was transferred to Mibuna by introgression. Furthermore, we conducted a survey of ancient literature to reveal the divergence of Mibuna and found that hybridization between Mizuna and a simple-leaved turnip might have occurred in the past. Indeed, the genomic analysis of multiple turnip cultivars showed that one of the cultivars, Murasakihime, has almost the same sequence in the BrTCP15 region as Mibuna. These results suggest that the hybridization between Mizuna and turnip has resulted in the establishment of Mibuna.

We constructed a high-density genetic linkage map of bronze loquat (Eriobotrya deflexa) for a three-way cross of loquat (Eriobotrya japonica) × (loquat × bronze loquat) by using single nucleotide polymorphism (SNP) markers generated by RAD-Seq, with the combination of simple sequence repeat (SSR) markers, which have been used in the previous study. The map of bronze loquat (‘Taiwan loquat No. 1’) consisted of 960 loci including 793 SNPs, 166 SSRs (98 apple, 8 loquat and 60 pear SSRs), and the loquat canker resistance gene Pse-a. This map covered 17 linkage groups over a total length of 1707.4 cM with an average distance of 1.78 cM between markers. The number of linkage groups corresponded to the basic chromosome number (n = 17). The analysis of the markers that can be anchored to apple draft genome showed that almost all loquat linkage groups were aligned to the apple draft genome. The constructed map will enable us to determine the location of quantitative trait loci and genes of interest, and to analyze genome synteny in the tribe Pyreae, subfamily Spiraeoideae of the family Rosaceae.

Medaka (Oryzias sp.) is an important fish species in ecotoxicology and considered as a model species due to its biological features including small body size and short generation time. Since Japanese medaka Oryzias latipes is a freshwater species with access to an excellent genome resource, the marine medaka Oryzias melastigma is also applicable for the marine ecotoxicology. In genome era, a high-density genetic linkage map is a very useful resource in genomic research, providing a means for comparative genomic analysis and verification of de novo genome assembly. In this study, we developed a high-density genetic linkage map for O. melastigma using restriction-site associated DNA sequencing (RAD-seq). The genetic map consisted of 24 linkage groups with 2,481 single nucleotide polymorphism (SNP) markers. The total map length was 1,784 cM with an average marker space of 0.72 cM. The genetic map was integrated with the reference-assisted chromosome assembly (RACA) of O. melastigma, which anchored 90.7% of the assembled sequence onto the linkage map. The values of complete Benchmarking Universal Single-Copy Orthologs were similar to RACA assembly but N50 (23.74 Mb; total genome length 779.4 Mb; gap 5.29%) increased to 29.99 Mb (total genome length 778.7 Mb; gap 5.2%). Using MapQTL analysis with SNP markers, we identified a major quantitative trait locus for sex traits on the Om10. The integration of the genetic map with the reference genome of marine medaka will serve as a good resource for studies in molecular toxicology, genomics, CRISPR/Cas9, and epigenetics.

Using genome-wide SNP data obtained from high-throughput techniques based on double digest restriction site-associated DNA sequencing (ddRAD-seq), we elucidated the migration history and genetic diversity of the Japanese population of the ecologically important brown seaweed Sargassum thunbergii (Mertens ex Roth) Kuntze. STRUCTURE and NeighborNet analyses showed a clear genetic differentiation among populations of four geographic regions: Kyushu (POP1); Sea of Japan (POP2); Hokkaido and Tohoku (POP3); and Pacific coast from Kyushu to Kanto (POP4). Approximate Bayesian Computation (ABC) analysis indicated that POP4 diverged first, followed by the separation between POP2 (the largest effective population size) and POP3; POP1 was the last to form, shaped by the mixture of POP2 (73%) and POP4 (27%). High genetic diversity was detected in POP1 and POP2, whereas low genetic diversity was detected in POP3 and POP4. These results indicated that S. thunbergii populations of Kyushu and the Sea of Japan might have been maintained as large and stable populations gathered different lineages from China, Korea and/or Japan.

Isoflavones and soyasaponins are major specialized metabolites accumulated in soybean roots and secreted into the rhizosphere. Unlike the biosynthetic pathway, the transporters involved in metabolite secretion remain unknown. The developmental regulation of isoflavone and soyasaponin secretions has been recently reported, but the diurnal regulation of their biosynthesis and secretion still needs to be further studied. To address these challenges, we conducted transcriptome and metabolite analysis using hydroponically grown soybean plants at 6-hr intervals for 48 hr in a 12-hr-light/12-hr-dark condition. Isoflavone and soyasaponin biosynthetic genes showed opposite patterns in the root tissues; that is, the former genes are highly expressed in the daytime, while the latter ones are strongly induced at nighttime. GmMYB176 encoding a transcription factor of isoflavone biosynthesis was upregulated from ZT0 (6:00 a.m.) to ZT6 (12:00 a.m.), followed by the induction of isoflavone biosynthetic genes at ZT6. The isoflavone aglycone content in the roots accordingly increased from ZT6 to ZT18 (0:00 a.m.). The isoflavone aglycone content in root exudates was kept consistent throughout the day, whereas that of glucosides increased at ZT6, which reflected the decreased expression of the gene encoding beta-glucosidase involved in the hydrolysis of apoplast-localized isoflavone conjugates. Co-expression analysis revealed that those isoflavone and soyasaponin biosynthetic genes formed separate clusters, which exhibited a correlation to ABC and MATE transporter genes. In summary, the results in this study indicated the diurnal regulation of isoflavone biosynthesis in soybean roots and the putative transporter genes responsible for isoflavone and soyasaponin transport.