Plant-based edible vaccines: from transgenic food crops to post-pandemic molecular farming and genomics-informed vaccine platforms

Plant-based edible vaccines began as an unusually attractive promise: a recombinant antigen produced in a familiar crop and delivered through the logic of food rather than injection. Rooted in earlier thinking about oral immunisation, mucosal protection, and vaccine access, this idea was later transformed by recombinant plant biotechnology into an experimentally tractable platform. This Review traces the historical and translational arc of plant-based edible vaccines, showing how each phase addressed a distinct bottleneck. Early work established that plant tissues could express immunogenic antigens and linked edible delivery with low-cost, needle-free vaccination. A second phase generated decisive proof-of-concept in potato, tobacco, lettuce, maize, rice, and other hosts, including early human immunogenicity studies, but also exposed problems of low expression, variable dosage, oral tolerance, regulatory complexity, scale-up, and uncertain manufacturability. A third phase addressed these limitations through chloroplast engineering, seed and endosperm bioencapsulation, viral vectors, transient expression, virus-like particles, glycoengineering, and improved control over antigen architecture and process design. The COVID-19 era accelerated visibility, industrial urgency, and renewed interest in oral, ambient-stable, and regionally deployable vaccine formats. Although routine marketing authorisation of a human-edible plant vaccine remains unrealised, the field has successfully broadened beyond the classical edible model toward convergent plant vaccine platforms that combine programmable host tissues, advanced delivery materials, rapid manufacturing, and clearer translational pathways. The central lesson is that plant-made vaccines progressed most when plant tissues were treated not simply as edible organs, but as designable biological matrices. The next phase is likely to be genomics- and pangenomics-informed, using sequence diversity, host domestication, and chassis engineering to identify host genotypes with favourable protease backgrounds, storage behaviour, glycosylation environments, scalability, regulatory readiness, and equitable deployment.