On this page
Ammonia accumulates from two sources at once: cells produce it as they metabolise glutamine, and glutamine degrades spontaneously in warm medium. Read the shape of the ammonia curve against glutamine and glutamate, not the number alone: the shape tells you whether the lever is feed strategy, glutamine handling, a process event, or sample handling.
Why ammonia is the first curve to pull
When viability starts sliding mid-run, ammonia is the cheapest suspect to check and the most commonly missed. It accumulates from two sources at once: cells produce it as they metabolise glutamine, and the glutamine in your medium produces it on its own, glutamine degrades spontaneously in warm medium, fed or not. So ammonia rises even in a run where the cells are doing nothing wrong, and it rises faster exactly when you feed more glutamine to fix a different problem.
Ammonia inhibits growth and can shift product quality. The point of this note is not a magic number, the level your culture tolerates is a property of your cell line, your pH and your process, it is how to read the shape of the curve against the rest of the run, because the shape tells you which lever to pull.
Our amino acid panel measures ammonia in the same run as the 19 amino acids and glutamine/glutamate themselves (AQC derivatization, HPLC), which is what makes the reading below possible from a single spent-media sample set: ammonia alone is a symptom; ammonia next to glutamine and glutamate is a diagnosis.
Related: Glutamine, Glutamate and Ammonia Across a Fed-Batch Run (the study summary) and Spent Media Analysis: What Amino Acid Profiles Tell You About Your Culture (the broader spent-media picture).
Four curve shapes, four different problems
1. Steady climb from day one, glutamine still high. The medium is cooking itself. Glutamine is degrading in the feed or the vessel faster than the cells use it. Levers: fresher glutamine or a more stable glutamine source, feed glutamine in smaller, more frequent later doses rather than up front, check feed storage temperature and hold times. Feeding more glutamine here makes the ammonia worse, not better.
2. Ammonia tracks cell density, glutamine nearly exhausted. This is metabolic, not chemical: a dense, hungry culture stripping glutamine to glutamate and beyond. Levers: it's a feed-strategy problem, total nitrogen and the glutamine feeding schedule, not a contamination hunt. If viability holds, this shape can be normal late-run behaviour; the question is whether the peak arrives before or after your harvest window.
3. Sudden spike between two samples. Something changed: a feed event (concentrated glutamine bolus), a pH excursion (ammonia toxicity is pH-dependent, the same total ammonia hurts more as pH rises), or a lysis event releasing intracellular contents as viability fell for another reason. Read this shape backwards: the spike may be the consequence of the viability drop, not its cause. Check what else moved in the same interval before blaming ammonia.
4. High ammonia, flat everything else, from the first sample. Suspect the starting conditions: the basal medium, the inoculum carryover, or the sample itself. Spent-media samples keep changing after you draw them, glutamine keeps degrading in the tube. A sample that sat warm before freezing will read high on ammonia and low on glutamine no matter what the vessel looked like. Freeze samples promptly and note draw-to-freeze time; an ammonia curve is only as honest as its sample handling.
Setting your own action line
Do not borrow a threshold from a paper on a different cell line. Build yours from your own history:
- For your last several runs, good and bad, line up ammonia (per sample day) against viability and titer.
- Find the ammonia level above which your runs stopped recovering. That, minus a safety margin, is your alert line; the level at which good runs peaked and still finished well is your watch line.
- Sample densely enough to see the shape: a start, mid-run and pre-harvest point at minimum; daily through the window where viability historically turns. Ammonia measured once is a number; measured across the run it is a curve, and only the curve diagnoses.
What to send, and what you'll get back
Send at least 200 µL per sample (400 µL preferred), as a set, time-course samples from one run, frozen promptly, labelled with run day and draw time, not a single tube. One sample can say "ammonia was high on day 9"; the set says which of the four shapes you have. The report returns ammonia alongside glutamine, glutamate and the other amino acids on the 19-amino-acid panel for each sample, so the ammonia curve arrives with the context that interprets it: what was fed, what was consumed, what was left.
For a whole run, RUN-1 covers 24 time points for $1,990, about $83 per time point. Order a time course See an example report
Overnight turnaround available, with expedited service on request. Standard results in 3 to 5 business days from receipt.
Scope
- Ammonia is one variable. The curve can indict the feed strategy; it cannot by itself prove causation, pH, osmolality, lactate and the viability method all belong in the same review.
- Thresholds are process-specific. Any lab that hands you a universal "toxic above X" number for your line is overselling.
- Sample handling is part of the measurement. Warm, delayed samples manufacture ammonia after the draw; we will flag sample sets whose pattern suggests it, but we cannot undo it.
- This note describes interpretation of a measured panel. Results support process-development decisions, with the chromatogram and the calibration in every report.
OpenChemWorks Laboratory · Reviewed by the OpenChemWorks laboratory · Updated October 30, 2026