Sustainable production of Cocoa with plant cell cultures

Europe's Green Deal and Bioeconomy Strategy both highlight the need to strengthen the resilience of Europe's bioeconomy, increase the strategic autonomy of bio-based supply chains, and decouple economic growth from the use of scarce land and water resources as well as biodiversity loss. Plant cell culture offers a promising solution by enabling the controlled, climate- and land-independent, pesticide-free production of high-value natural products in closed bioreactor systems, while reducing transport requirements and dependence on fragile global supply chains. Unlike microbial fermentation, plant cell culture unlocks the plant kingdom—the richest source of bioactive secondary metabolites. The growing investments by private companies in plant cell culture start-ups since 2020 further underscore the increasing commercial potential of this technology.

Plant cell culture was first described more than a century ago and is based on the cultivation of plant cells in sterile liquid media for the production of biomass or valuable natural compounds. Its industrial feasibility has already been demonstrated convincingly. For example, Taxus cell cultures grown in bioreactors exceeding 75,000 L supply a major share of the global demand for paclitaxel. These examples demonstrate the technical maturity of the technology and its suitability for manufacturing high-value biopharmaceuticals. At the same time, plant cell culture offers significant opportunities far beyond the pharmaceutical sector. Demand for sustainably produced plant-derived ingredients is rapidly growing in cosmetics, nutraceuticals, and food applications. However, transferring the technology into these high-volume markets requires substantially greater process scalability, robustness, and economic efficiency than the production of high-value pharmaceuticals.

Consequently, the widespread industrial adoption of plant cell culture in these sectors is limited less by the fundamental feasibility of the technology than by its economically viable and reproducible implementation at industrial scale. Three closely interconnected challenges are primarily responsible. First, the development of new cell lines and their translation into scalable manufacturing processes still relies largely on empirical trial-and-error approaches, resulting in long development times and high costs. Second, industrial deployment is constrained by limited access to fit-for-purpose bioreactors and contract manufacturing organisations with plant cell culture expertise. Third, the lack of standardised data and process frameworks hampers the development of robust manufacturing processes while simultaneously complicating regulatory approval.

These challenges share a common root cause: the absence of structured, high-quality datasets and standardised development and optimisation workflows capable of enabling the data-driven and AI-assisted engineering of new plant cell cultures and manufacturing processes.

COCO-AI addresses these challenges by developing an AI-enabled platform for the data-driven engineering and optimisation of plant cell cultures. To this end, multi-omics, imaging, and bioprocess data are integrated across the entire development pipeline to transform today's largely empirical workflows into standardised, accelerated, and reproducible engineering processes. Using cocoa as a lead product for establishing the pipeline and the other plant species for pipeline validation, we will (i) accelerate the development of friable callus lines through AI-assisted assessment and targeted optimisation of friability, (ii) shorten suspension cell line development by predicting optimal culture media, growth conditions, and elicitation strategies, and (iii) translate the resulting knowledge into standard operating procedures (SOPs) and a transferable AI agent applicable across multiple plant species.

The project's objective is to establish a scalable and regulatory-ready development framework for plant cell culture that accelerates process development, facilitates industrial implementation, and enables the sustainable deployment of plant cell cultures for food, nutraceutical, cosmetic, and other bio-based applications. The key innovation of COCO-AI is the integration of an AI agent across the complete plant cell development and bioproduction pipeline, transforming empirical trial-and-error workflows into predictive, data-driven engineering processes.

To achieve this ambition, COCO-AI brings together nine leading partners from Europe, North America, the Middle East, and Asia, combining complementary expertise in plant cell biology, bioprocess engineering, artificial intelligence, and industrial scale-up. Together, the consortium integrates advanced multi-omics technologies, high-throughput process development, and machine learning approaches to overcome the longstanding bottlenecks of plant cell culture, namely slow development cycles and the lack of standardised engineering workflows.

Fraunhofer IME will coordinate the project and lead its overall scientific and administrative management. In close collaboration between the departments Model-based Product and Bioprocess Engineering and Precision Fermentation, Fraunhofer IME will develop bioreactor-ready suspension cell lines by combining high-throughput process development, AI-guided experimental design, and proprietary enabling technologies, thereby establishing robust cultivation processes suitable for industrial scale-up.