Four procedures get called “cleaning” and they are not the same thing. Regeneration removes what did not elute, after every run, and protects the step that follows. Cleaning in place removes what accumulated — denatured product, host-cell protein, DNA, lipid — with caustic. Sanitisation controls bioburden. Storage keeps the column stable while it is idle. The caustic step is the one that mainly consumes the ligand, and that is what sets the service life of a Protein A column.
The four procedures, side by side
| Procedure | Purpose | When | Typical agent | Flow |
|---|---|---|---|---|
| Regeneration / strip | Remove product and impurities the elution did not — and keep them from being fixed on the medium by the caustic step | Every cycle, after elution | 0.1 M acetic acid pH 2.9, or glycine-HCl pH 2.5–3.0, 2–3 CV | Same direction as the run |
| Cleaning in place (CIP) | Remove what elution and strip left behind: denatured product, host-cell protein, DNA, lipid | Every cycle in a platform process; less often in low-throughput lab work | Sodium hydroxide, 10–20 mM for native ligands up to 0.1–1.0 M for alkali-stabilised ones, 10–15 min contact, 2–3 CV | Commonly reverse flow on a packed bed, because the inlet fouls first; per datasheet on other formats |
| Sanitisation | Control bioburden between campaigns | Campaign start and end, or per SOP | Sodium hydroxide per the medium's specification — alcohols such as 20 % ethanol or 2 % benzyl alcohol are storage solutions, not sanitisation agents | As for CIP |
| Storage | Keep the column stable while idle | Overnight and longer | Storage solution per the medium — commonly 20 % ethanol or 2 % benzyl alcohol, 5 CV, 2–8 °C | Same direction as the run |
The distinction matters because the most common mistake is to treat a capacity problem as a regeneration problem. Stripping harder after every run does not recover capacity lost to fouling, and it does cost ligand. One note on direction that is easy to get backwards: on a packed bed, vendors commonly specify the caustic step in reverse flow, because a packed bed fouls at the inlet and up-flow lifts that material off instead of pushing it through the bed. For any other format, the datasheet decides.
What actually limits service life
Two mechanisms run in parallel and they need different responses:
- Ligand loss. Alkaline cleaning hydrolyses the protein ligand over time, and acid exposure during elution and strip contributes as well. This is permanent, cumulative and roughly proportional to the exposure you have applied — which is why service life is quoted in cleaning cycles, and why the mildest elution that still works is also a lifetime decision.
- Fouling. Feed components deposit on and in the medium. This shows up as rising back pressure and falling capacity, and it is partly reversible with the right cleaning step. The mechanisms, and why open-channel media are slower to block, are covered in why chromatography media foul.
Published lifetime work puts the usable range for Protein A media somewhere between tens and a few hundred cycles, with the cleaning regime and the feed composition — not the number of bind-and-elute cycles — doing most of the explaining. Two useful entry points are Zhang et al. on maximising functional lifetime and Pathak and Rathore on the impact of feed composition. Cycle numbers from a datasheet only transfer to your process if the cleaning regime transfers with them.
How to measure when a column is done
Do not wait for a failure. Measure the decline:
- Record a breakthrough curve on a fresh column under fixed conditions — same feed concentration, same residence time, same breakthrough level. That is your baseline.
- Repeat it at intervals, for example every 10 or 20 cycles. Keep the arithmetic and the hold-up correction identical between runs, or the drift you measure is your method rather than the medium — see how to read a capacity figure.
- Plot capacity against cumulative cleaning cycles, not against runs. A straight decline points at the ligand; a step change points at fouling or at something that happened in the feed.
- Define the end point in advance — a percentage of the starting capacity that your process can absorb — and retire the column when it is reached, rather than when a batch fails.
Pressure is the second trace to keep: a rising back pressure at constant flow is the earliest sign that the next CIP needs to be stronger or earlier.
MonoCore™ Protein A and Protein G
For our own affinity capsules the conditions come with the datasheet. What is specified publicly: the MonoCore™ capsule tolerates pH 4–13 in long-term use and pH 2–14 short term, and the limiting component there is the housing rather than the matrix — on an affinity capsule the ligand is the sensitive part. Avoid prolonged exposure to strong acids and to hot or concentrated alkali rather than fixating on a single pH value. The ion exchange and HIC capsules, where the chemistry is not a protein ligand, carry their caustic conditions in their datasheets.
Run the cleaning step at a third to a half of the loading flow rate and follow it with a full re-equilibration before the next run. How you tell whether it worked: back pressure at a fixed flow rate, capacity against cumulative cleaning cycles, and leachate over the same history. The elution buffer belongs in that picture too — the lower the elution pH, the more ligand is lost per cycle, which is one more argument for the mildest workable acid in Protein A elution buffer.
Frequently asked questions
What is the difference between regeneration and CIP?
Regeneration is the acid strip after every cycle that removes what did not elute. CIP is the periodic caustic step that removes accumulated, non-eluting material. Regeneration is routine and cheap; CIP is what consumes the ligand and therefore the column's life.
How many cycles does a Protein A column last?
Published ranges span roughly tens to a few hundred cycles, and the spread is explained by the cleaning regime and the feed, not by the number of runs. The only number that applies to your process is the one you measure on it, by tracking capacity against cumulative cleaning cycles.
Which NaOH concentration should I use?
The one specified for your medium — and the specification is what to ask for, not a general rule. Alkali-stabilised recombinant ligands tolerate considerably more than native Protein A, and exceeding the specification trades service life for cleanliness. For a MonoCore™ affinity capsule, what is public is the frame: pH 4–13 in long-term use and pH 2–14 short term as limits set by the housing, cleaning at a third to a half of the loading flow rate, and a reverse-direction clean where the device is rated for it. The exact concentration and contact time for Protein A and Protein G come with the datasheet.
Should CIP run in reverse flow?
Often, yes — on a packed bed it is the vendor default rather than the exception, because the inlet fouls first and up-flow removes that material instead of driving it deeper. Check the pressure rating in that direction, follow it with a re-equilibration, and for any format that is not a packed bed take the answer from its datasheet rather than from the analogy.
How should the column be stored?
Cold, at 2–8 °C, in the storage solution specified for your medium and after a full re-equilibration — roughly 5 column volumes are enough to displace the working buffer. MonoCore™ affinity capsules are stored in 20 % ethanol with 10 % glycerol, 50 mM Tris and 150 mM NaCl at pH 7.6; for the ion exchange and HIC capsules it is 20 % ethanol. For pauses of a few hours, equilibration buffer is fine.
Still an open question?
Tell us the molecule, the feed volume and the system you run it on, and we will say whether a MonoCore™ capsule is the right starting point — or whether it is not.