Dialysis desalting and buffer exchange

Buffer exchange is the moment a protein leaves the solvent it was purified in and meets the solvent the next experiment requires. Dialysis, a desalting column and a centrifugal spin device all let small molecules pass a barrier that holds the protein back. They do not share a clock, a volume change, or a failure mode. The practical decision is which format survives the protein you have, and how you will prove the new buffer is actually there. Making and checking that buffer is its own craft, set out in preparing a buffer and checking pH. If the sample clouds during the swap, read that result with aggregation and cloudy protein samples. The gel that tells you whether the protein stayed behind is reading a protein gel.

Three formats, three kinds of patience

Dialysis puts the sample in a bag or a cassette whose membrane has a stated molecular-weight cutoff, then stirs it against a large external volume. Salts, imidazole, urea and other small solutes drift out. The protein, if it is larger than the cutoff by a comfortable margin, stays. Exchange is gradual. That gradual pace is why dialysis is used for sensitive refolds and why it is frustrating when you needed the sample before lunch. One change of external buffer is rarely a finished exchange. Protocols commonly replace the external bath more than once. The volume of the bath, not the label on the bag, does most of the work.

A desalting column, gravity or spun, is a short size-exclusion bed. The protein travels in the void and the salts lag behind. It is fast. The sample volume has to fit the bed. Oversizing the load merges the protein with the salt and undoes the point. You often accept some dilution. Collect fractions if the protocol says the separation is tight, and do not pool the tail blindly.

A centrifugal buffer exchange uses a spin concentrator. You concentrate, dilute with the new buffer, and concentrate again. Each cycle reduces the old solute. Several cycles are normal. The membrane can foul, the protein can precipitate on the filter, and the hold-up volume is not zero. It is a good format for small volumes and a harsh one for proteins that hate being forced against a surface.

Cutoff, leakage and dilution

Choose a cutoff well below the protein mass. A rule of thumb you will see is a membrane rated several-fold under the mass, not a membrane that matches it. A 25 kilodalton protein does not belong in a 30 kilodalton cutoff just because the numbers are neighbours. Shape matters. An elongated protein can thread a pore that a globular protein of the same mass cannot. If the yield collapses, run a gel of what passed through. Loss through the membrane looks like magic until you stain that fraction.

Dilution is part of the result. A desalting column that returns a larger volume has lowered the concentration even when recovery of mass is excellent. A dialysis bag can also take up or lose water with the osmotic balance. Measure concentration after the exchange by a method the new buffer allows. Do not copy the pre-exchange number onto the new tube.

Precipitation during the swap is the failure that hurts. Leaving high salt, high imidazole, or molar denaturant can drop a protein out of solution as the good solvent disappears. If a dialysis bag fills with cloudiness, the protein may be in the solid, not in the "exchanged" liquid you pipette off the top. Spin, run both fractions on a gel, and reconsider the destination buffer. Sometimes a stepwise dilution, or an intermediate salt, keeps the chain soluble. That is an experiment, not a guarantee.

Prove the buffer, do not recite it

Conductivity is the simple check for leftover salt. A sample whose conductivity still matches the elution buffer has not exchanged, whatever the protocol title says. pH is the partner check. Electrodes, temperature and calibration are described in the buffer page already linked. Read the sample, not only the carboy you poured from. Imidazole and urea change assays and sometimes pH behaviour. A280 after exchange still needs a blank of the new buffer. A Bradford or BCA plate needs to know whether detergent or reductant came along.

Branch points when the control fails

If conductivity stays high, you have not finished. Change the dialysis bath again, or repeat a spin cycle, or stop pooling the desalting tail. If pH is wrong, the new buffer may be mis-made, or the sample's old buffer may be overwhelming a weak new one. Fix the buffer as a buffer problem. If protein appears in the flow-through, the cutoff is too high or the membrane has failed. If the membrane retentate is cloudy, you are concentrating an aggregate. Stop and read the pellet. Running that cloudy retentate straight onto an analytical column is how a bed gets plugged.

A gel of the exchanged sample catches proteolysis that occurred during a long dialysis. Fragments mean the inhibitors, the temperature, or the time were wrong. They do not mean the cutoff was clever.

Research use, heat and enquiry notes

Membranes, concentrated salts and denaturants are ordinary laboratory chemicals with ordinary hazards. This is not a clinical fluid process and not biosafety approval. Discard spent denaturant baths as your institution asks.

Long dialysis in a warm room is a microbiological and a folding risk. A stirred bath that sits out during a hot afternoon is not the cold exchange you wrote in the notebook. Power cuts stop stir plates and warm cold rooms. If the bath stopped, do not assume equilibrium. Check conductivity, look for cloudiness, and record the interruption. Humidity does not change the cutoff, but it does change how carefully you can weigh hygroscopic buffer salts. Prepare the destination buffer with the same discipline as any other pH-critical solution.

Say, in an enquiry, which solute must leave, which buffer must arrive, the protein mass, the volume, and whether precipitation has already been seen. Reagents and chemicals covers buffer components. A quote can carry those constraints. Where exchange sits inside a larger purification question, the custom protein expression and purification page is an enquiry reference for discussing the method. It does not mean the exchange is run as a service.