Fermentations are often split into phases such as growth and production, which are distinct in terms of substrate availability and metabolic pathway induction. Successful transition between phases can influence performance. For instance, a glucose feed that was initiated too late could result in starvation and a feed started too early could result in overfeeding; both scenarios may impact strain performance. However, timing of phase transitions within a fermentation, such as depletion of the batch carbon source, may vary between strains or occur outside of working hours, adding to the complexity of assessing strain performance. 

To tailor a fermentation process to specific strains, physiological cues known as metabolic shifts can be leveraged as signals for phase change. Metabolic shifts happen when cells exhaust or switch carbon sources. They are often observed in fermentations and are accompanied by measurable metabolic signals, such as a sharp rise in Dissolved Oxygen (DO), increase in pH, and decrease in oxygen uptake rate (OUR) or carbon dioxide evolution rate (CER). Here (Figure 1), a metabolic shift results in a spike in DO (A), which is used to initiate substrate feeds. However, DO signals, particularly early in the fermentation, have the tendency to be noisy (B), resulting in falsely triggered feeds.

‍

Instead, a more robust signal would depend on off-gas measurements (Figure 2). Beyond a threshold, these measurements are less prone to noise, and a drop in off-gas CO₂ concentration can be used to initiate feeds resulting in fewer falsely triggered feeds.

‍

Here, the fermentation script monitors average off-gas CO₂ concentration (Figure 3, orange lines) for a period of time, scans for a decrease past a certain threshold, signifying a metabolic shift, and triggers the feed using a feed trigger algorithm in the bioreactor recipe (Figure 3, blue lines). The specific hallmarks of a metabolic shift are unique to each organism and each fermentation process. As such, a number of factors, such as the length of the reference CO₂ measurement window or the absolute decrease in CO₂ concentration required for the trigger can be tuned over a number of runs in order to optimize conditions (Figure 3).

‍

Additional requirements for triggering a feed may be added to avoid false triggers, such as a “sleep timer” to prevent premature triggers or backup DO spike to prevent missed feed triggers. In the example in Figure 3, the optimized conditions were validated by successfully triggering feeds in all 16 reactors included in the trial, demonstrating a robust protocol.

Culture Biosciences offers individual off-gas sensors on each reactor providing high-resolution data that can be used to make process steps, such as feed triggers, more robust. Culture’s experienced software team can quickly create and validate sophisticated control scripts based on metabolic shifts. These tightly controlled fermentation conditions translate into high-quality data and experimental success for Culture’s clients.

For more on feed strategies for process development, including key challenges to keep in mind and best practices, check out this resource.