Cultivated meat

Cultivated meat starts from a biological fact that is easy to state and hard to exploit: skeletal muscle carries its own repair population. Satellite cells sit against the muscle fibre under the basal lamina, stay quiescent until injury, then proliferate as myoblasts and fuse into new fibre. Taken out of the animal and given the right signals, they will do the same thing in a vessel. Everything difficult about the field follows from the gap between will do this and will do this at tonnage.

The proliferation ceiling

A primary satellite cell is mortal. It divides a limited number of times before replicative senescence — the Hayflick limit — and it also drifts: with each passage the population loses myogenic potency and enriches for fibroblast-like cells that grow faster but will never become muscle. Both problems have to be solved before scale is even a question, which is why the field’s real starting material is not a biopsy but an immortalised, characterised cell line, obtained either by selecting spontaneously immortalised clones or by deliberate intervention in the telomerase and cell-cycle checkpoints.

The second structural problem is anchorage. Muscle progenitors are adherent cells; they need a surface. A tank does not have one, so the surface must be supplied as microcarriers, or the line must be adapted to grow in suspension — an adaptation that changes the cell’s behaviour and has to be re-validated for differentiation afterwards.

Media is chemistry, not shopping

The culture medium is where the field’s difficulty concentrates, and the reason is chemical rather than commercial. Recombinant growth factors — FGF2 above all, with insulin, transferrin and albumin — are proteins, and proteins denature. FGF2 in particular loses activity over hours at 37 °C, so a bioreactor does not consume growth factor at the rate the cells use it; it consumes it at the rate the molecule falls apart. Thermostabilised variants attack exactly this, and they are a protein-engineering result rather than a procurement one.

Metabolism sets the other boundary. Proliferating animal cells in high glucose run aerobic glycolysis and excrete lactate; glutamine consumption releases ammonia. Both inhibit growth well before nutrients run out, which is what perfusion is actually for — not feeding, but removing.

Why thickness is a physics problem

Tissue without a vasculature is fed by diffusion alone, and oxygen runs out on the order of a hundred to a few hundred micrometres from the medium. That single number explains the shape of the industry’s products. Unstructured mince is a cell-density problem and can be solved by process engineering. A thick whole cut is a transport problem, and no amount of bioreactor volume touches it: it requires either a perfusable channel network in the scaffold or acceptance of a thin, layered geometry.

What is genuinely unsettled

Life-cycle results for cultivated meat differ by an order of magnitude between studies, because they depend on assumptions — energy mix, media inputs, whether pharmaceutical-grade purification is assumed — that no one can yet fix from operating data at scale. Published estimates of climate benefit should be read as scenario-dependent, not measured. The same caution applies to nutrition: composition is set by the differentiation protocol and the fat-to-muscle ratio, so a general claim about “cultivated meat” nutrition is not well formed.