Chinese scientists have developed the first three-dimensional spatiotemporal cell atlas covering the full life cycle of rice, offering new insights into its growth and development and helping identify genes with specific functions that could support precision breeding for high and stable yields.

The study was led by the Yazhouwan National Laboratory in collaboration with 12 leading institutions, including BGI Research, Huazhong Agricultural University, Southern University of Science and Technology, Wuhan University and the Institute of Genetics and Developmental Biology of the Chinese Academy of Sciences. The findings were published in the journal Cell on Tuesday.

Rice growth and development depend on the coordination of different types of cells. However, scientists have lacked a comprehensive picture of how thousands of genes work together to regulate the plant throughout its life cycle. In the past, researchers generally focused on individual genes or specific tissues, making it difficult to understand the development process as a whole.

To address this gap, the research team used the Japonica rice variety Zhonghua 11 as a model. They combined several technologies to examine rice development, including methods that can show where genes are active within tissues, single-cell analysis, large-scale gene sequencing and artificial intelligence.

The resulting atlas covers 10 major types of organs and tissues and 61 developmental stages, from the early embryo to mature organs. It contains 851,725 high-quality single-nucleus datasets and 347,640 spatial data units. The researchers identified 119 cell types and 133 subtypes and mapped how different cell states are related as rice develops.

The atlas functions as a three-dimensional map showing where different types of cells are located in tissues and how their gene activity changes over time. It connects changes at the molecular level with the visible growth and development of the rice plant.

"With this high-resolution map, researchers gain a 'full life cycle view', enabling them to observe how different cells coordinate and interact, ultimately assembling a comprehensive life map of rice," said Jia Pengfei, co-corresponding author of the study and a researcher at the Institute of Genetics and Developmental Biology.

One notable finding concerns the endosperm, the main edible part of a rice grain. The researchers found a clear division of labor between different regions of the developing endosperm.

Genes involved in carbohydrate metabolism and starch production were more active in the dorsal region, while genes related to the production of storage proteins were concentrated in the ventral region. Experiments in which the researchers altered these genes further showed that they could directly change how starch and protein accumulated in the two regions.

The researchers also found that a single key regulatory factor can perform different functions depending on the type of cell in which it operates. The findings show that the distribution of nutrients within a rice grain is controlled at highly specific locations, revealing a more complex system of nutrient allocation than previously understood.

The findings could provide a new approach to breeding rice with better taste and nutritional value. Rather than relying mainly on average measurements taken from an entire grain, researchers could use the atlas to identify and regulate genes at specific locations and stages of development.

To make the findings available for further research, the team has established an interactive online atlas that allows users to search for genes, view where genes are active in tissues and compare different datasets.

The researchers have also developed RICE scGPT, an artificial intelligence-based model designed to analyze individual rice cells. Trained using data from the study as well as publicly available datasets, the model can help researchers identify different cell types and combine data from different sources, making the datasets easier to compare and reuse.

Jia said the ultimate goal is to apply the findings to agricultural production, helping advance the seed industry and strengthen food security through data-driven innovation.