Mycoprotein Scale-Up: DO Management for Fusarium venenatum Continuous Culture

Mycoprotein production using Fusarium venenatum is one of the most commercially established precision fermentation applications in the food industry — continuous culture at large industrial scale, with decades of operating experience. Yet the transition from pilot to commercial scale remains a live challenge for producers working with new strain variants or modified process conditions. The reason is almost always dissolved oxygen management.

Unlike E. coli or yeast, where overflow metabolism under oxygen limitation produces easily measurable by-products (acetate, ethanol), F. venenatum under oxygen-limited conditions shifts TCA cycle flux toward fumaric and malic acid accumulation. These organic acids don't simply represent lost carbon — they affect downstream process parameters and product quality. This article covers the DO management physiology for F. venenatum and the scale-specific modeling approach for each vessel transition.

The Physiology: Why DO Determines Product Quality

F. venenatum is an obligate aerobe that produces the mycoprotein product as its biomass — the hyphal network, when processed and textured, becomes the product. Unlike recombinant protein production where the product is secreted and quantified by titer, mycoprotein quality is partially determined by the fiber morphology (hyphae length distribution, branching frequency) and partially by the organic acid content of the harvested biomass.

The relationship between DO and fiber morphology is well-established: above approximately 20–25% air saturation, F. venenatum maintains a growth mode that produces long, well-formed hyphae with the textural properties required for product quality. Below this threshold, hyphal morphology shifts — branching increases, hyphae shorten, and the texture of processed product becomes undesirable. This is not a binary transition but a continuous function of DO with a characteristic inflection near 20% air saturation.

The relationship between DO and organic acid accumulation is a consequence of TCA cycle flux redistribution. Under oxygen-replete conditions, fumarase catalyzes the conversion of fumarate to malate within the TCA cycle, and fumarate concentration in the cell is low. Under oxygen-limited conditions, the electron transport chain slows, NADH/NAD+ ratio increases, and TCA cycle flux redistribution causes fumarate and malate to accumulate intracellularly and eventually to be secreted. Published values for fumarate secretion rates in F. venenatum under moderate oxygen limitation (DO 10–20% air saturation) range from 0.8 to 2.4 mmol fumarate/g DCW·h — values that become significant at commercial cell densities.

Scale-Up Oxygen Transfer Challenges

Fusarium venenatum fermentation is typically run as continuous culture (chemostat) rather than fed-batch — a key operational distinction from most recombinant protein processes. In a continuous culture at a specific growth rate μ of 0.05–0.12 h⁻¹ (typical range for mycoprotein production), the oxygen demand is more predictable than in a fed-batch with exponentially increasing cell density. However, the absolute OUR at commercial cell density is still substantial:

OUR = μ × Yx/s^-1 × qO2_specific × DCW

At a DCW of 30 g/L (typical commercial mycoprotein titer), μ = 0.08 h⁻¹, and a specific oxygen uptake rate of 18 mmol O₂/g DCW·h (measured), the volumetric OUR is 540 mmol O₂/L·h or approximately 1,730 mg O₂/L·h. This demand is well within the capability of a well-configured 500L pilot vessel (kLa of 150–250 h⁻¹ at typical operating conditions), but becomes a challenge at commercial scale where kLa drops substantially.

The kLa reduction at commercial scale

Commercial mycoprotein vessels typically operate at 50,000–150,000L — far larger than standard precision fermentation pilots. At this scale, maintaining kLa above the threshold required for the process OUR demand requires either very high P/V (power per unit volume), which is energy-intensive, or supplemental oxygen enrichment of the sparge gas. The industry has largely adopted oxygen enrichment as the primary lever for DO management at commercial scale, but the modeling of how much enrichment is needed at each scale point requires accurate kLa estimation.

For intermediate-scale vessels (1,000–10,000L), where many growing mycoprotein producers currently operate, the kLa challenge is more about impeller selection and operating conditions than oxygen enrichment. A 5,000L vessel with a standard 3-Rushton turbine stack at 2.5 kW/m³ specific power input may achieve kLa of 180–240 h⁻¹ in water — reduced to approximately 130–170 h⁻¹ in fermentation broth with the characteristic viscosity and foaming properties of F. venenatum culture.

Continuous Culture vs Fed-Batch: Scale-Up Implications

The choice of continuous culture for mycoprotein production is operationally sensible (steady-state quality, no batch-to-batch variation, high volumetric productivity), but it changes the scale-up modeling problem compared to fed-batch.

In continuous culture, the oxygen demand at steady state is fixed by the dilution rate (D = μ) and the steady-state cell density. If D = 0.08 h⁻¹ and steady-state DCW = 30 g/L, the OUR at steady state is determined — it doesn't change over time as in a fed-batch. The scale-up question is whether the vessel can sustain this steady-state OUR at the target DO setpoint, given the vessel's kLa.

The FBA model for F. venenatum continuous culture at steady state is solved at a single set of constraints: D-determined growth rate, oxygen exchange flux constrained by kLa, and the fumarate/malate secretion fluxes allowed to vary based on the TCA cycle oxygen limitation. The model output is not a time series but a flux distribution at steady state — which tells you the predicted fumarate secretion rate and biomass-specific productivity at the target operating point.

The washout risk at scale-up

One scale-up failure mode unique to continuous culture is washout risk: if the dilution rate is set above μmax (the maximum specific growth rate under the actual conditions at scale), cells are washed out faster than they grow and the culture fails. μmax for F. venenatum is oxygen-dependent — at DO below 15% air saturation, μmax decreases, and a dilution rate that was safe at pilot scale (where DO was maintained at 30%) may exceed the effective μmax at commercial scale (where DO is 18% due to reduced kLa). The continuous culture approaches washout, the steady-state cell density drops, and productivity collapses.

This failure mode is particularly insidious because it develops slowly — cell density declines gradually over 2–4 residence times (24–48 hours at typical dilution rates) rather than catastrophically. By the time the DO and OD sensors signal the problem clearly, the culture is already significantly depleted.

DO Setpoint Selection by Scale

Based on the physiology described above, the DO setpoint for F. venenatum continuous culture should be maintained above 20–25% air saturation to ensure both fiber morphology quality and TCA cycle function. In practice, a setpoint of 25–30% at the process control level (what the DO controller targets) provides a 5–10 percentage point safety margin above the morphology transition threshold, accounting for probe position variability and mixing heterogeneity.

The challenge at commercial scale is maintaining this setpoint as kLa decreases. The practical interventions are:

1. Agitation optimization: At intermediate scale (1,000–10,000L), increasing P/V from 1.0 to 2.5 kW/m³ typically increases kLa by 50–80%. The limiting factor is shear sensitivity — F. venenatum hyphae are susceptible to shear damage at high tip speeds (>4 m/s), which itself affects product texture.
2. Oxygen enrichment: Supplementing 21% O₂ air with up to 40–60% O₂ enrichment increases the driving force for oxygen transfer (higher DO* at saturation) without requiring increased agitation. This is the standard approach at commercial scale but adds operating cost.
3. Reduced dilution rate at scale-up: If kLa is insufficient at the bench-scale dilution rate, operating at D = 0.06 h⁻¹ rather than 0.08 h⁻¹ reduces OUR demand proportionally. The trade-off is reduced volumetric productivity (g mycoprotein/L·h), which must be weighed against the capital cost of running a larger vessel at lower productivity.

Modeling the Scale Transition

A complete scale-up model for F. venenatum continuous culture involves:

1. Steady-state FBA at pilot-scale operating conditions, constrained by measured exchange fluxes (D, DCW, OUR, fumarate secretion if measurable)
2. kLa estimation at target commercial vessel geometry (impeller type, D/T ratio, operating P/V, aeration rate)
3. FBA at commercial-scale oxygen constraints: what fumarate secretion flux and cell density does the model predict when oxygen uptake is limited to kLa × (DO* − DO setpoint)?
4. Sensitivity analysis: at what kLa does fumarate secretion become significant, and what DO setpoint can the vessel maintain at the target OUR demand?

This workflow does not require knowledge of every TCA cycle enzyme kinetics in F. venenatum. It requires the stoichiometry (well-characterized from published metabolic reconstructions and closely related filamentous fungi), the measured exchange fluxes at pilot scale, and the kLa estimate at commercial scale. The output is a prediction of the fumarate accumulation risk and the minimum kLa required to maintain acceptable product quality — both calculable before you commission the commercial vessel run.

References

• Wiebe MG. Myco-protein from Fusarium venenatum: a well-established product for human consumption. Appl Microbiol Biotechnol. 2002;58(4):421–427.
• Trinci AP. The mycoprotein story. Microbiology Today. 2004;31:102–103.
• Garcia-Ochoa F, Gomez E. Bioreactor scale-up and oxygen transfer rate in microbial processes. Biotechnol Adv. 2009;27(2):153–176.
• Doran PM. Bioprocess Engineering Principles. 2nd ed. Academic Press; 2013.