A core tenet of food security is ensuring people have access to sufficient nutritious food. Vitamin B₁ (thiamine) is an essential micronutrient for humans, deficiency in which causes numerous diseases of the nervous and cardiovascular systems (Dhir et al., 2019). Such ailments are particularly associated with populations that have a high carbohydrate intake, and especially those that have sustenance diets comprised largely of cereals such as rice. Thiamine insufficiency is a major public health concern in Asian countries, for example, in Cambodia 27%–100% of infants and women are deficient and account for up to 45% of deaths in under 5-year-olds (Johnson et al., 2019). Rice is a staple crop for half of the global population, but seeds are low in thiamine content, and polishing (i.e. removal of the embryo and bran layers) further aggravates chronic deficiencies as up to 90% of the thiamine content is in the removed tissues (Strobbe et al., 2021). Previous biofortification attempts increased thiamine content in leaves and unpolished seeds, but the trait failed to be retained in polished grains, sometimes negatively impacting yield (Dong et al., 2016; Strobbe et al., 2021), and no field trials have been reported that are necessary to confirm achievements. Natural variation studies revealed no substantial differences in thiamine content in rice seeds and minimal correlation with the expression of known biosynthesis genes (Mangel et al., 2022), discouraging exploitation via this route. Furthermore, the molecular form of vitamin B₁ is an important consideration because the biosynthesis of the coenzyme thiamine diphosphate (TDP) is tightly regulated and necessary for cellular homeostasis, whereas supplementing with thiamine suggests that this form is innocuous (Pourcel et al., 2013). Indeed, mature cereal seeds are a natural sink for thiamine (rather than TDP, Figure S1), which is associated with certain members of the large diverse class of albumin proteins (Watanabe et al., 2001). Significantly, a sequence has been reported from this class encoding a thiamine binding protein (TBP) from Sesamum indicum (Si, sesame) (Watanabe et al., 2001).

Here, we address previous bottlenecks related to enhancing thiamine content in rice endosperm and field performance in a proof-of-concept study. We expressed SiTBP under the control of the rice endosperm-specific glutelin D-1 promoter in the rice japonica model variety TP309 (wild type) (Figure 1a), after confirmation of thiamine binding to a codon-optimized SiTBP expressed in Escherichia coli (Figure S2a–c). Among 45 transformants, 3 with independent single T-DNA insertion events were selected to homozygosity and were morphologically similar to wild type, but had increased thiamine content in the polished grain (L14, 20, 26). The lines were sown in an experimental field in Taiwan (GPS coordinates 24°04′41.3″N 120°42′53.8″E) initially for bulking and then monitored for their agricultural performance during season 1 in years 2022 and 2023. Each line was grown on three randomized plots of 6 × 8 plants each. The transgenic plants expressing SiTBP could not be distinguished morphologically from the wild type (Figure 1b). Various agronomic traits were recorded in each year such as plant height, tiller and panicle number, single-plant yield, 100-grain weight and fertility. No statistically significant differences were observed between the lines (Figure 1c–i; Figure S3a–f). The vitamin B₁ contents of leaf and unpolished seed material indicated no differences in vitamer levels of thiamine (Figure 1j,k; Figure S3g). In contrast, thiamine levels in the polished grains of the transgenic lines were increased 3 to 4-fold in each line compared to wild-type polished grains (Figure 1l; Figure S3h). Thus, SiTBP boosts the accumulation of thiamine in the endosperm without impacting yield. A proportion of this increase appears to be derived from the bran and/or embryo (Figure 1m; Figure S3i). Expression of the transgene was stable in the field, and expression of the biosynthesis genes was not changed with the exception of a small but significant decrease in the kinase OsTPKc in some of the transgenic lines (Figure S4). This represents a key step forward in vitamin B₁ biofortification by potentially achieving 35%–27% of the RDI (1.2–1.4 mg/day) for adults and lactating women with a bowl of rice (ca. 315 g), respectively.

The next step in realizing the goal of vitamin B₁-biofortified plants will be to pursue this approach in customer-preferred elite varieties. This could be combined with the appropriate expression of the dual thiamine/POLYAMINE UPTAKE TRANSPORTER identified in Arabidopsis (Martinis et al., 2016) to further facilitate enhanced endosperm thiamine content. Notably, exogenous thiamine application has been discussed as a strategy to improve content (Li et al., 2022), but this may switch off thiamine biosynthesis de novo through extant TDP riboswitches and moreover, affect plant microbiome communities. Thus, we consider it more advantageous to purposefully engineer endogenous thiamine increases in target plant tissues by expanding the strategy for increasing thiamine content in grain endosperm as we present here to enhance the nutritional quality of rice for human health.