386: Development and Optimization of Chemically Defined and Undefined Serum-Free Media for Lamb Muscle Cells Using Response Surface Methodology and Machine Learning Approaches

Introduction

The growing demand for sustainable and ethical alternatives in food production has driven significant advancements in cellular agriculture, particularly in the development of cultivated meat. One of the primary challenges in scaling up cultivated meat production is the reliance on fetal bovine serum (FBS) in cell culture media, which raises ethical concerns, increases production costs, and contributes to environmental impact due to cattle farming. To address these issues, this study focuses on the development and optimization of a novel serum-free media formulation for cultivated meat production, specifically tailored for lamb muscle cell cultures.

Methods

To develop an effective serum-free media for cultivated meat production, lamb muscle cells were cultured using a basal medium supplemented with essential growth-promoting components. The formulation process involved a systematic experimental design using Central Composite Design (CCD) and optimization through Response Surface Methodology (RSM). Various chemically defined components were tested across multiple experimental runs to identify the most critical factors influencing cell proliferation and viability. We used machine learning methods to analyze images of each formulation, correlating cell morphology with proliferation and developing predictive models for cell growth. The optimized media was tested in both 2D and 3D culture conditions to assess scalability.

Results

The optimized serum-free media significantly enhanced lamb muscle cell proliferation and viability, yielding growth rates comparable to FBS-containing media. RSM analysis determined the optimal concentrations of key components, providing a robust formulation for scalable use. Machine learning models accurately correlated image analysis with cell performance, ensuring consistency across different formulations and enhancing the predictive accuracy of the optimization process. Furthermore, the optimized media exhibited promising results in 3D culture conditions, demonstrating efficient cell attachment, proliferation, and morphology, which are critical for cultivated meat production.

Significance

This study provides a scalable, cost-effective, and ethically sustainable serum-free media formulation for cultivated meat production, addressing major challenges associated with FBS use. By integrating advanced optimization techniques and machine learning-driven image analysis, the developed media supports high cell growth and viability under both 2D and 3D conditions. This work offers a significant step toward the commercialization of cultivated meat, paving the way for more sustainable and ethical food systems.

Authors: Arian Amirvaresi, Arash Shahsavari, Armaghan Amanipour, Reza Ovissipour