Reaction norm for genomic prediction of plant growth: modeling drought stress response in soybean.

KEY MESSAGE: We proposed models to predict the effects of genomic and environmental factors on daily soybean growth and applied them to soybean growth data obtained with unmanned aerial vehicles. Advances in high-throughput phenotyping technology have made it possible to obtain time-series plant growth data in field trials, enabling genotype-by-environment interaction (G × E) modeling of plant growth. Although the reaction norm is an effective method for quantitatively evaluating G × E and has been implemented in genomic prediction models, no reaction norm models have been applied to plant growth data. Here, we propose a novel reaction norm model for plant growth using spline and random forest models, in which daily growth is explained by environmental factors one day prior. The proposed model was applied to soybean canopy area and height to evaluate the influence of drought stress levels. Changes in the canopy area and height of 198 cultivars were measured by remote sensing using unmanned aerial vehicles. Multiple drought stress levels were set as treatments, and their time-series soil moisture was measured. The models were evaluated using three cross-validation schemes. Although accuracy of the proposed models did not surpass that of single-trait genomic prediction, the results suggest that our model can capture G × E, especially the latter growth period for the random forest model. Also, significant variations in the G × E of the canopy height during the early growth period were visualized using the spline model. This result indicates the effectiveness of the proposed models on plant growth data and the possibility of revealing G × E in various growth stages in plant breeding by applying statistical or machine learning models to time-series phenotype data.

Consideration of familiarity accumulated in the confined field trials for environmental risk assessment of genetically modified soybean (Glycine max) in Japan.

To date, there have been 160 regulatory approvals for environmental safety in Japan for the major genetically modified (GM) crops, including corn, soybean, canola and cotton. Confined field trials (CFTs) have been conducted in Japan for all single events, which contain various traits. The accumulated information from these previously conducted CFTs, as well as the agronomic field study data from other countries, provides a rich source of information to establish "familiarity" with the crops. This familiarity can be defined as the knowledge gained through experience over time, and used to inform the environmental risk assessments (ERA) of new GM crops in Japan. In this paper, we compiled agronomic data from the CFTs performed in Japan for 11 GM soybean events which obtained food, feed and environmental safety approvals from regulatory agencies in Japan. These CFTs were conducted by multiple developers according to Japan regulations to support the ERA of these GM soybean, covering standard measurement endpoints evaluated across developers in Japan. With this dataset, we demonstrate how familiarity gained from the CFTs of GM soybeans in Japan can be used to inform on the ERA of new GM soybean events. By leveraging this concept of familiarity, we discuss potential enhancements to the ERA process for GM soybean events in Japan.

Construction of a high-density mutant library in soybean and development of a mutant retrieval method using amplicon sequencing.

BACKGROUND: Functions of most genes predicted in the soybean genome have not been clarified. A mutant library with a high mutation density would be helpful for functional studies and for identification of novel alleles useful for breeding. Development of cost-effective and high-throughput protocols using next generation sequencing (NGS) technologies is expected to simplify the retrieval of mutants with mutations in genes of interest. RESULTS: To increase the mutation density, seeds of the Japanese elite soybean cultivar Enrei were treated with the chemical mutagen ethyl methanesulfonate (EMS); M2 seeds produced by M1 plants were treated with EMS once again. The resultant library, which consisted of DNA and seeds from 1536 plants, revealed large morphological and physiological variations. Based on whole-genome re-sequencing analysis of 12 mutant lines, the average number of base changes was 12,796 per line. On average, 691 and 35 per line were missense and nonsense mutations, respectively. Two screening strategies for high resolution melting (HRM) analysis and indexed amplicon sequencing were designed to retrieve the mutants; the mutations were confirmed by Sanger sequencing as the final step. In comparison with HRM screening of several genes, indexed amplicon sequencing allows one to scan a longer sequence range and skip screening steps and to know the sequence information of mutation because it uses systematic DNA pooling and the index of NGS reads, which simplifies the discovery of mutants with amino acid substitutions. CONCLUSIONS: A soybean mutant library with a high mutation density was developed. A high mutation density (1 mutation/74 kb) was achieved by repeating the EMS treatment. The mutation density of our library is sufficiently high to obtain a plant in which a gene is nonsense mutated. Thus, our mutant library and the indexed amplicon sequencing will be useful for functional studies of soybean genes and have a potential to yield useful mutant alleles for soybean breeding.

Possibilities of direct introgression from Brassica napus to B. juncea and indirect introgression from B. napus to related Brassicaceae through B. juncea.

The impact of genetically modified canola (Brassica napus) on biodiversity has been examined since its initial stage of commercialization. Various research groups have extensively investigated crossability and introgression among species of Brassicaceae. B. rapa and B. juncea are ranked first and second as the recipients of cross-pollination and introgression from B. napus, respectively. Crossability between B. napus and B. rapa has been examined, specifically in terms of introgression from B. napus to B. rapa, which is mainly considered a weed in America and European countries. On the other hand, knowledge on introgression from B. napus to B. juncea is insufficient, although B. juncea is recognized as the main Brassicaceae weed species in Asia. It is therefore essential to gather information regarding the direct introgression of B. napus into B. juncea and indirect introgression of B. napus into other species of Brassicaceae through B. juncea to evaluate the influence of genetically modified canola on biodiversity. We review information on crossability and introgression between B. juncea and other related Brassicaseae in this report.

High-Throughput Phenotyping of Soybean Biomass: Conventional Trait Estimation and Novel Latent Feature Extraction Using UAV Remote Sensing and Deep Learning Models.

High-throughput phenotyping serves as a framework to reduce chronological costs and accelerate breeding cycles. In this study, we developed models to estimate the phenotypes of biomass-related traits in soybean (Glycine max) using unmanned aerial vehicle (UAV) remote sensing and deep learning models. In 2018, a field experiment was conducted using 198 soybean germplasm accessions with known whole-genome sequences under 2 irrigation conditions: drought and control. We used a convolutional neural network (CNN) as a model to estimate the phenotypic values of 5 conventional biomass-related traits: dry weight, main stem length, numbers of nodes and branches, and plant height. We utilized manually measured phenotypes of conventional traits along with RGB images and digital surface models from UAV remote sensing to train our CNN models. The accuracy of the developed models was assessed through 10-fold cross-validation, which demonstrated their ability to accurately estimate the phenotypes of all conventional traits simultaneously. Deep learning enabled us to extract features that exhibited strong correlations with the output (i.e., phenotypes of the target traits) and accurately estimate the values of the features from the input data. We considered the extracted low-dimensional features as phenotypes in the latent space and attempted to annotate them based on the phenotypes of conventional traits. Furthermore, we validated whether these low-dimensional latent features were genetically controlled by assessing the accuracy of genomic predictions. The results revealed the potential utility of these low-dimensional latent features in actual breeding scenarios.

Random regression for modeling soybean plant response to irrigation changes using time-series multispectral data.

Plant response to drought is an important yield-related trait under abiotic stress, but the method for measuring and modeling plant responses in a time series has not been fully established. The objective of this study was to develop a method to measure and model plant response to irrigation changes using time-series multispectral (MS) data. We evaluated 178 soybean (Glycine max (L.) Merr.) accessions under three irrigation treatments at the Arid Land Research Center, Tottori University, Japan in 2019, 2020 and 2021. The irrigation treatments included W5: watering for 5 d followed by no watering 5 d, W10: watering for 10 d followed by no watering 10 d, D10: no watering for 10 d followed by watering 10 d, and D: no watering. To capture the plant responses to irrigation changes, time-series MS data were collected by unmanned aerial vehicle during the irrigation/non-irrigation switch of each irrigation treatment. We built a random regression model (RRM) for each of combination of treatment by year using the time-series MS data. To test the accuracy of the information captured by RRM, we evaluated the coefficient of variation (CV) of fresh shoot weight of all accessions under a total of nine different drought conditions as an indicator of plant's stability under drought stresses. We built a genomic prediction model (MTRRM model) using the genetic random regression coefficients of RRM as secondary traits and evaluated the accuracy of each model for predicting CV. In 2020 and 2021,the mean prediction accuracies of MTRRM models built in the changing irrigation treatments (r = 0.44 and 0.49, respectively) were higher than that in the continuous drought treatment (r = 0.34 and 0.44, respectively) in the same year. When the CV was predicted using the MTRRM model across 2020 and 2021 in the changing irrigation treatment, the mean prediction accuracy (r = 0.46) was 42% higher than that of the simple genomic prediction model (r =0.32). The results suggest that this RRM method using the time-series MS data can effectively capture the genetic variation of plant response to drought.

Relationship between hybridization frequency of Brassica juncea × B. napus and distance from pollen source (B. napus) to recipient (B. juncea) under field conditions in Japan.

Several imported transgenic canola (Brassica napus) seeds have been spilled and have grown along roadsides around import ports. B. juncea, a relative of B. napus with which it has high interspecific crossability, is widely distributed throughout Japan. There is public concern about the harmful impacts of feral B. napus plants on biodiversity, but spontaneous hybridization between spilled B. napus and weedy B. juncea populations is hardly revealed. We evaluated the relationship between the hybridization frequency of B. juncea × B. napus and their planting distance in field experiments using the mutagenic herbicide-tolerant B. napus cv. Bn0861 as a pollen source for hybrid screening. The recipient B. juncea cv. Kikarashina was planted in an experimental field with Bn0861 planted in the center. No hybrids were detected under natural flowering conditions in 2009. However, the flowering period was artificially kept overlapping in 2010, leading to a hybridization frequency of 1.62% in the mixed planting area. The hybridization frequency decreased drastically with distance from the pollen source, and was lower under field conditions than estimated from the high crossability, implying that spontaneous hybridization between spilled B. napus and weedy B. juncea is unlikely in the natural environment.

Persistent C genome chromosome regions identified by SSR analysis in backcross progenies between Brassica juncea and B. napus.

Given that feral transgenic canola (Brassica napus) from spilled seeds has been found outside of farmer's fields and that B. juncea is distributed worldwide, it is possible that introgression to B. juncea from B. napus has occurred. To investigate such introgression, we characterized the persistence of B. napus C genome chromosome (C-chromosome) regions in backcross progenies by B. napus C-chromosome specific simple sequence repeat (SSR) markers. We produced backcross progenies from B. juncea and F(1) hybrid of B. juncea × B. napus to evaluate persistence of C-chromosome region, and screened 83 markers from a set of reported C-chromosome specific SSR markers. Eighty-five percent of the SSR markers were deleted in the BC(1) obtained from B. juncea × F(1) hybrid, and this BC(1) exhibited a plant type like that of B. juncea. Most markers were deleted in BC(2) and BC(3) plants, with only two markers persisting in the BC(3). These results indicate a small possibility of persistence of C-chromosome regions in our backcross progenies. Knowledge about the persistence of B. napus C-chromosome regions in backcross progenies may contribute to shed light on gene introgression.

Time-series multispectral imaging in soybean for improving biomass and genomic prediction accuracy.

Multispectral (MS) imaging enables the measurement of characteristics important for increasing the prediction accuracy of genotypic and phenotypic values for yield-related traits. In this study, we evaluated the potential application of temporal MS imaging for the prediction of aboveground biomass (AGB) in soybean [Glycine max (L.) Merr.]. Field experiments with 198 accessions of soybean were conducted with four different irrigation levels. Five vegetation indices (VIs) were calculated using MS images from soybean canopies from early vegetative to early reproductive stage. To predict the genotypic values of AGB, VIs at the different growth stages were used as secondary traits in a multitrait genomic prediction. The prediction accuracy of the genotypic values of AGB from MS and genomic data largely outperformed that of the genomic data alone before the flowering stage (90% of accessions did not flower), suggesting that it would be possible to determine cross-combinations based on the predicted genotypic values of AGB. We compared the prediction accuracy of a model using the five VIs and a model using only one VI to predict the phenotypic values of AGB and found that the difference in prediction accuracy decreased over time at all irrigation levels except for the most severe drought. The difference in the most severe drought was not as small as that in the other treatments. Only the prediction accuracy of a model using the five VIs in the most severe droughts gradually increased over time. Therefore, the optimal timing for MS imaging may depend on the irrigation levels.

Effects of irrigation on root growth and development of soybean: A 3-year sandy field experiment.

Increasing the water use efficiency of crops is an important agricultural goal closely related to the root system -the primary plant organ for water and nutrient acquisition. In an attempt to evaluate the response of root growth and development of soybean to water supply levels, 200 genotypes were grown in a sandy field for 3 years under irrigated and non-irrigated conditions, and 14 root traits together with shoot fresh weight and plant height were investigated. Three-way ANOVA revealed a significant effect of treatments and years on growth of plants, accounting for more than 80% of the total variability. The response of roots to irrigation was consistent over the years as most root traits were improved by irrigation. However, the actual values varied between years because the growth of plants was largely affected by the field microclimatic conditions (i.e., temperature, sunshine duration, and precipitation). Therefore, the best linear unbiased prediction values for each trait were calculated using the original data. Principal component analysis showed that most traits contributed to principal component (PC) 1, whereas average diameter, the ratio of thin and medium thickness root length to total root length contributed to PC2. Subsequently, we focused on selecting genotypes that exhibited significant improvements in root traits under irrigation than under non-irrigated conditions using the increment (I-index) and relative increment (RI-index) indices calculated for all traits. Finally, we screened for genotypes with high stability and root growth over the 3 years using the multi-trait selection index (MTSI).Six genotypes namely, GmJMC130, GmWMC178, GmJMC092, GmJMC068, GmWMC075, and GmJMC081 from the top 10% of genotypes scoring MTSI less than the selection threshold of 7.04 and 4.11 under irrigated and non-irrigated conditions, respectively, were selected. The selected genotypes have great potential for breeding cultivars with improved water usage abilities, meeting the goal of water-saving agriculture.

Genomic Prediction of Green Fraction Dynamics in Soybean Using Unmanned Aerial Vehicles Observations.

With the widespread use of high-throughput phenotyping systems, growth process data are expected to become more easily available. By applying genomic prediction to growth data, it will be possible to predict the growth of untested genotypes. Predicting the growth process will be useful for crop breeding, as variability in the growth process has a significant impact on the management of plant cultivation. However, the integration of growth modeling and genomic prediction has yet to be studied in depth. In this study, we implemented new prediction models to propose a novel growth prediction scheme. Phenotype data of 198 soybean germplasm genotypes were acquired for 3 years in experimental fields in Tottori, Japan. The longitudinal changes in the green fractions were measured using UAV remote sensing. Then, a dynamic model was fitted to the green fraction to extract the dynamic characteristics of the green fraction as five parameters. Using the estimated growth parameters, we developed models for genomic prediction of the growth process and tested whether the inclusion of the dynamic model contributed to better prediction of growth. Our proposed models consist of two steps: first, predicting the parameters of the dynamics model with genomic prediction, and then substituting the predicted values for the parameters of the dynamics model. By evaluating the heritability of the growth parameters, the dynamic model was able to effectively extract genetic diversity in the growth characteristics of the green fraction. In addition, the proposed prediction model showed higher prediction accuracy than conventional genomic prediction models, especially when the future growth of the test population is a prediction target given the observed values in the first half of growth as training data. This indicates that our model was able to successfully combine information from the early growth period with phenotypic data from the training population for prediction. This prediction method could be applied to selection at an early growth stage in crop breeding, and could reduce the cost and time of field trials.

Agrobacterium-mediated transformation of kabocha squash (Cucurbita moschata Duch) induced by wounding with aluminum borate whiskers.

An efficient genetic transformation method for kabocha squash (Cucurbita moschata Duch cv. Heiankogiku) was established by wounding cotyledonary node explants with aluminum borate whiskers prior to inoculation with Agrobacterium. Adventitious shoots were induced from only the proximal regions of the cotyledonary nodes and were most efficiently induced on Murashige-Skoog agar medium with 1 mg/L benzyladenine. Vortexing with 1% (w/v) aluminum borate whiskers significantly increased Agrobacterium infection efficiency in the proximal region of the explants. Transgenic plants were screened at the T(0) generation by sGFP fluorescence, genomic PCR, and Southern blot analyses. These transgenic plants grew normally and T(1) seeds were obtained. We confirmed stable integration of the transgene and its inheritance in T(1) generation plants by sGFP fluorescence and genomic PCR analyses. The average transgenic efficiency for producing kabocha squashes with our method was about 2.7%, a value sufficient for practical use.

Estrogen-inducible GFP expression patterns in rice (Oryza sativa L.).

We investigated estrogen-inducible green fluorescent protein (GFP) expression patterns using an estrogen receptor fused chimeric transcription activator, XVE, in the monocotyledonous model plant rice (Oryza sativa L.). This system has been shown to be an effective chemical-inducible gene expression system in Arabidopsis and has been applied to other plants in order to investigate gene functions or produce marker-free transgenic plants. However, limited information is available on the correlation between inducer concentration and the expression level of the gene induced in monocots. Here, we produced a transgenic rice integrated estrogen-inducible GFP expression vector, pLex:GFP, and investigated dose-response and time-course patterns of GFP induction in rice calli and seedlings for the first time. With 17-ß-estradiol treatment at >5 µM, GFP signals were detected in the entire surface of calli within 2 days of culture. Highest GFP signals were extended for 8 days with estradiol treatment at 25 µM. In three-leaf-stage seedlings, GFP signals in the leaves of pLex:GFP-integrated transgenic lines were weaker than those in the leaves of p35S:GFP-integrated transgenic lines. However, GFP signals in the roots of pLex:GFP- and p35S:GFP-integrated transgenic lines were similar with estradiol treatment at >10 µM. With regard to controlling appropriate gene expression, these results might provide helpful indications on estradiol treatment conditions to be used for the XVE system in rice and other monocots.

Occurrence of metaxenia and false hybrids in Brassica juncea L. cv. Kikarashina × B. napus.

Imported genetically modified (GM) canola (Brassica napus) is approved by Japanese law. Some GM canola varieties have been found around importation sites, and there is public concern that these may have any harmful effects on related species such as reduction of wild relatives. Because B. juncea is distributed throughout Japan and is known to be high crossability with B. napus, it is assumed to be a recipient of B. napus. However, there are few reports for introgression of cross-combination in B. juncea × B. napus. To assess crossability, we artificially pollinated B. juncea with B. napus. After harvesting a large number of progeny seeds, we observed false hybrids and metaxenia of seed coats. Seed coat color was classified into four categories and false hybrids were confirmed by morphological characteristics and random amplified polymorphic DNA (RAPD) markers. Furthermore, the occurrence of false hybrids was affected by varietal differences in B. napus, whereas that of metaxenia was related to hybridity. Therefore, we suggest that metaxenia can be used as a marker for hybrid identification in B. juncea L. cv. Kikarashina × B. napus. Our results suggest that hybrid productivity in B. juncea × B. napus should not be evaluated by only seed productivity, crossability ought to be assessed the detection of true hybrids.

Improving Quantitative Traits in Self-Pollinated Crops Using Simulation-Based Selection With Minimal Crossing.

Genomic selection and marker-assisted recurrent selection have been applied to improve quantitative traits in many cross-pollinated crops. However, such selection is not feasible in self-pollinated crops owing to laborious crossing procedures. In this study, we developed a simulation-based selection strategy that makes use of a trait prediction model based on genomic information to predict the phenotype of the progeny for all possible crossing combinations. These predictions are then used to select the best cross combinations for the selection of the given trait. In our simulated experiment, using a biparental initial population with a heritability set to 0.3, 0.6, or 1.0 and the number of quantitative trait loci set to 30 or 100, the genetic gain of the proposed strategy was higher or equal to that of conventional recurrent selection method in the early selection cycles, although the number of cross combinations of the proposed strategy was considerably reduced in each cycle. Moreover, this strategy was demonstrated to increase or decrease seed protein content in soybean recombinant inbred lines using SNP markers. Information on 29 genomic regions associated with seed protein content was used to construct the prediction model and conduct simulation. After two selection cycles, the selected progeny had significantly higher or lower seed protein contents than those from the initial population. These results suggest that our strategy is effective in obtaining superior progeny over a short period with minimal crossing and has the potential to efficiently improve the target quantitative traits in self-pollinated crops.

Increased awareness and decreased acceptance of genome-editing technology: The impact of the Chinese twin babies.

Genome-editing technology has become increasingly known in recent years, and the 2018 news of genome-edited twins in China had a particularly significant impact on public awareness. In the present study we investigate the effect of Japanese media coverage on public opinions of this technology. To identify the effects we employ a questionnaire survey method on a pre-registered sample from online research company Macromill. Our repeated survey from 2016 through 2019 reveal a generally supportive attitude toward the medical application of genome-editing methods. To see this we employed a multinomial logit analysis examining the determinants of negative and positive impressions of the technology. Results show that although editing for medical purposes remained mostly acceptable, its use in fertilizing human eggs was increasingly rejected, especially in 2019, the most recent sample year. The suggestion is that while genome-editing applications in general medical fields are publicly accepted, its use in human functionality enhancement is heavily increasingly resisted. News of the twin babies in China did raise public awareness of the methods but also damaged their reputation. It therefore is important for genome researchers to hold such concerns in mind, keeping the public informed of changing technology fundamentals. As a related question we inquire into the public acceptability of genome editing for animal and plant breeding, such as in agriculture and fisheries, as well. We find the Japanese public views the medical and breeding applications of this technology to be unconnected with each other, despite that awareness of both has risen significantly in recent years.

A novel strategy for promoting homoplasmic plastid transformant production using the barnase-barstar system.

Plastid transformants form biofactories that are able to produce extra proteins in plastids when they are in a homoplasmic state. To date, plastid transformation has been reported in about twenty plant species; however, the production of homoplasmic plastid transformants is not always successful or easy. Heteroplasmic plants that contain wild-type plastids produce fewer target proteins and do not always successfully transfer transgenes to progeny. In order to promote the generation of homoplasmic plants, we developed a novel system using barnase-barster to eliminate wild-type plastids from heteroplasmic cells systematically. In this system, a chemically inducible cytotoxic barnase under a plastid transit signal was introduced into nuclear DNA and barster, which inhibits barnase, was integrated into plastid DNA with the primary selection markers aminoglycoside 3'-adenylyltransferase (aadA) and green fluorescence protein (GFP) gene. As expected, the expression of the plastid barnase was lethal to cells as seen in leaf segments, but barster expression in plastids rescued them. We then investigated the regeneration frequency of homoplasmic shoots from heteroplasmic leaf segments with or without barnase expression. The regeneration frequency of homoplasmic-like shoots expressing barnase-barster system was higher than that of shoots not expressing this. We expect that the application of this novel strategy for transformation of plastids will be supportive to generate homoplasmic plastid transformants in other plant species.

Foreign DNA detection by high-throughput sequencing to regulate genome-edited agricultural products.

Although the advent of several new breeding techniques (NBTs) is revolutionizing agricultural production processes, technical information necessary for their regulation is yet to be provided. Here, we show that high-throughput DNA sequencing is effective for the detection of unintended remaining foreign DNA segments in genome-edited rice. A simple k-mer detection method is presented and validated through a series of computer simulations and real data analyses. The data show that a short foreign DNA segment of 20 nucleotides can be detected and the probability that the segment is overlooked is 10-3 or less if the average sequencing depth is 30 or more, while the number of false hits is less than 1 on average. This method was applied to real sequencing data, and the presence and absence of an external DNA segment were successfully proven. Additionally, our in-depth analyses also identified some weaknesses in current DNA sequencing technologies. Hence, for a rigorous safety assessment, the combination of k-mer detection and another method, such as Southern blot assay, is recommended. The results presented in this study will lay the foundation for the regulation of NBT products, where foreign DNA is utilized during their generation.

Regulatory Status of Genome-Edited Organisms Under the Japanese Cartagena Act.

The Japanese government recognizes the substantial values of genome-edited agricultural organisms and has defined in which cases these are covered by the existing regulatory framework to handle this technology. Genome-editing technologies could revolutionize and accelerate plant breeding owing to the simplicity of the methods and precision of genome modifications. These technologies have spread rapidly and widely, and various genome-edited crops have been developed recently. The regulatory status of genome-edited end products is a subject of controversy worldwide. In February 2019, the Japanese government defined genome-edited end products derived by modifications of SDN-1 type (directed mutation without using a DNA sequence template) as not representing "living modified organisms" according to the Japanese Cartagena Act. Here, we describe the classification and regulatory status of genome-edited end products in this decision. We hope that reporting the progress in Japan toward the implementation of this regulatory approach will provide insight for scientific and regulatory communities worldwide.

Agroinfiltration-based efficient transient protein expression in leguminous plants.

Transient protein expression is an effective tool to rapidly unravel novel gene functions, such as transcriptional activity of promoters and sub-cellular localization of proteins. However, transient expression is not applicable to some species and varieties because of insufficient expression levels. We recently developed one of the strongest agroinfiltration-based transient protein expression systems for plant cells, termed 'Tsukuba system.' About 4 mg/g fresh weight of protein expression in Nicotiana benthamiana was obtained using this system. The vector pBYR2HS, which contains a geminiviral replication system and a double terminator, can be used in various plant species and varieties, including lettuces, eggplants, tomatoes, hot peppers, and orchids. In this study, we assessed the applicability of the Tsukuba system to several species of legumes, including Lotus japonicus, soybean Glycine max, and common bean Phaseolus vulgaris. The GFP protein was transiently expressed in the seedpods of all examined legume species, however, protein expression in leaves was observed only in P. vulgaris. Taken together, our system is an effective tool to examine gene function rapidly in several legume species based on transient protein expression.