Freeze-thaw stability
Freeze-thaw studies exist because material gets frozen and thawed far more often than anyone plans. A reference standard is thawed for every assay, a clinical lot is shipped and re-frozen, and a research aliquot is used three times before someone notices. The study answers how much damage each cycle does.
The damage is not the cold
Freezing itself is mild. The stresses come from what happens around the ice.
Cryoconcentration is the main one. Ice forms as pure water, so every solute concentrates into the shrinking liquid fraction. A protein at one milligram per milliliter can reach tens of milligrams per milliliter in that channel, alongside correspondingly concentrated buffer salts, which is exactly the condition that promotes association.
pH shift follows from the same process. Sodium phosphate is the notorious case: the dibasic form crystallizes first on cooling, which strips base from the remaining liquid and drops the pH by two units or more. A formulation buffered at pH 7.2 can present its protein with pH 4.5 during freezing. Histidine and citrate do not do this to anything like the same degree, and for a protein that fails freeze-thaw in phosphate, changing the buffer is the first thing to try.
The ice-water interface is large, cold and denaturing. Protein adsorbs to it and unfolds there, which is why slow freezing, which produces large ice crystals and less surface, is often gentler than snap freezing for a protein in a simple buffer.
Interfaces and shear during thaw add to it, particularly if the material is warmed unevenly and a concentrated layer sits at the boundary.
Designing a study that answers something
Decide the question first. A study supporting a handling recommendation needs three to five cycles under the conditions people actually use. A study supporting a formulation choice needs the same cycles applied across several formulations in parallel, and the comparison rather than the absolute number is the output.
Specify the cycle completely. Freezing rate, target temperature, hold time, thaw method and thaw temperature all change the result, and a study that says "three freeze-thaw cycles" without them cannot be repeated. Snap freezing in liquid nitrogen and a slow descent in a controlled-rate freezer are different experiments.
Include the container. Vial geometry, fill volume and headspace change the freezing front and the interfacial area, and a result from a full two-milliliter tube does not transfer to a thinly filled bag.
Run an unfrozen control held at the same temperature for the same total time. Without it, loss from the freeze-thaw arm cannot be separated from loss that would have happened anyway.
What to measure
Measure aggregation by size exclusion chromatography, and measure it as both soluble high molecular weight species and recovered monomer mass. These are different failures: soluble aggregate shows up as a new peak, while insoluble aggregate simply does not elute and appears as reduced total recovery. A study reporting only percent monomer purity will score a sample that lost half its material to the filter as unchanged.
Add a subvisible particle count where the material is a formulation candidate, because the particles that form on freezing are frequently below the visible threshold and above the size that matters for immunogenicity.
Measure activity or binding on the thawed material. Aggregation is the common failure, but a protein can also lose activity through a conformational change that size exclusion will not see.
The handling that removes most of the problem
Aliquot once, on receipt, into single-use volumes. This is the intervention that eliminates the issue entirely and it costs an hour.
Avoid sodium phosphate for anything that will be frozen. Histidine at around pH 6 is a reasonable default for antibodies and many proteins.
Include a cryoprotectant. Sucrose or trehalose in the few percent range protects by staying amorphous and remaining preferentially excluded from the protein surface. A nonionic surfactant addresses the interfacial component, and the two work on different mechanisms, so both are common in the same formulation.
Thaw quickly and uniformly, at room temperature or in a water bath rather than slowly in a refrigerator, and mix gently once fully thawed rather than while ice remains.