Plants rely on starch as their main source of energy and primary component. In photosynthetic tissues like leaves and stems, starch undergoes temporary synthesis and breakdown, whereas in storage organs such as grains, tubers, roots, and bulbs, it is stored in a more stable form [[1], [2], [3]]. Starch biosynthesis occurs via the coordinated actions of multiple enzymes including ADP-glucose pyrophosphorylases (AGPase; EC 2.7.7.27), starch synthases (EC 2.4.1.21), starch branching (EC 2.4.1.18) and debranching enzymes (EC 3.2.1.41 and EC 3.2.1.68). AGPases catalyze the first rate-limiting step in starch biosynthesis where Glc-1-P and ATP are utilized for the synthesis of ADP-glucose (and inorganic pyrophosphate; PPi) [4]. The plant AGPases are regulated by a diverse range of metabolites, with 3-PGA serving as a main activator and Pi (inorganic phosphate) acting as an inhibitor of the enzyme [5]. The AGPase activity is additionally controlled by redox potential [6], light intensity and sugar levels [7], as well as concentrations of nitrate [8] and phosphate [9].

In higher plants, the AGPases function in a heterotetrameric scaffold form, comprised of two large subunits (LS) and two small subunits (SS) [[10], [11], [12], [13]]. It was previously believed that the LS and SS AGPase subunits served as regulatory and catalytic components, respectively [14]. However, recent studies suggest that both subunits contribute to both catalytic and regulatory functions, albeit to varying extents. The SSs were found to actively participate in both catalytic and regulatory functions, whereas LSs primarily contributing to control the regulatory properties of SSs, displaying lower catalytic properties [[14], [15], [16], [17]]. The LS subunits play a crucial role in binding to the substrate (glucose-1 phosphate) and ATP, facilitating their subsequent interaction with the SSs [16,[18], [19], [20]]. This collaborative interplay promotes and enhances the catalytic activity of the SS. In potato tuber AGPases, heterologous expression of the SS alone in bacteria forms a catalytically active homotetrameric scaffold (StSS4). However, achieving maximal catalytic activity requires a considerably higher level of 3-PGA than required by the heterotetrameric enzyme [21]. In the Arabidopsis TL46 mutant line, which lacks LS, a homotetrameric SS AGPase assembly is also formed. In contrast, the LS is unable to form a homotetramer due to its insolubility. After introducing a mutation (S302N), the solubility of potato tuber LSs was greatly improved, allowing the formation of a homotetramer (StLS4) with marginal catalytic activity [20].

Plants contain two types of AGPases, plastidial and cytosolic. In rice, the genome encodes four large AGPase subunits (OsL1-OsL4) and two small subunits (OsS1 and OsS2; where OsS2 further encodes two isoforms via alternative mRNA splicing; OsS2a and OsS2b). The OsL2 and OsS2b isoforms are cytosolic, while OsL1, OsL3, OsL4, OsS1 and OsS2a are plastidial in location [[22], [23], [24], [25]]. The plastidial AGPase subunits, OsS1 and OsL1, are expressed during the early stages of endosperm development and play a role in regulating phosphorus homeostasis [23]. On the other hand, the cytosolic isoform, OsS2b, associates with OsL2 during endosperm maturation and are primarily responsible for the majority of starch synthesis during seed development [26].

In a previous study, we documented the interaction between cytosolic isoforms OsL2 and OsS2b, and proposed the catalytic activities of OsS2b variants under heat stress [27]. However, no investigation has yet conducted to explore the atomic-scale association of the OsL1:OsS1 heterotetrameric scaffold. Therefore, we employed structural bioinformatics approaches to explore the possible interfaces and critical residues responsible for the heterotetrameric association of OsL1:OsS1. Through this analysis, we noticed the mode of OsL1:OsS1 association is similar to that of the potato AGPase heterotetrameric complex (StLS:StSS), which was corroborated by protein-protein interaction analysis. These findings not only contribute to a better understanding of AGPase mechanisms in rice but also present an opportunity to manipulate the genes to enhance their activity, ultimately leading to the advancement of desirable agricultural traits.