Sterile Filtration of Peptide Solutions: Membranes, Peptide Losses and Integrity

What a sterilizing membrane removes, and what it does not

Sterile filtration passes a liquid through a membrane with pores rated at 0.2 or 0.22 µm, which retains bacteria mainly by size exclusion. It is used for solutions that cannot be heated, including almost all peptide solutions, before cell culture and similar work.

A membrane does not remove dissolved contaminants. Endotoxin passes through, so a filtered solution can still activate immune cells in culture; see endotoxin and LAL testing. Handling after the filter matters as much as the membrane itself.

Filtration does not fix poor dissolution. A cloudy solution contains undissolved or aggregated peptide that the membrane will hold back, lowering the concentration of the filtrate. The guide on solubility and aggregation covers getting the peptide fully into solution first.

Membrane materials and peptide adsorption

Common membrane materials include polyethersulfone (PES), polyvinylidene fluoride (PVDF), nylon (polyamide), cellulose acetate and regenerated cellulose, and polytetrafluoroethylene (PTFE). Many are offered in hydrophilic, chemically modified forms marketed as low binding. Hydrophobic PTFE is mainly used for organic solvents and air, and an aqueous solution will not pass through it unless it has been prewetted.

Peptides can stick to membranes in the same way they stick to tubes and vials (see peptide adsorption to labware). In a study of a 31-amino-acid peptide passed through four 0.2 µm membranes, adsorption ranked from lowest to highest as acrylate-grafted hydrophilic PVDF, a hydroxyl-modified polyamide, nylon-66 and unmodified PVDF. The authors noted that the measurements allow prediction of how much solution must pass through a given filter area to saturate its binding sites.^[1]

Work with model proteins across clarification, virus filtration, ultrafiltration and final sterile filtration found that loss depended on the protein, its concentration, the volume passed, membrane morphology and the removal mechanism.^[2] The same factors apply to peptides, and hydrophobic or fatty-acylated sequences such as Cagrilintide are the ones most worth checking. Binding is specific to each peptide and membrane, so a recovery measurement is the real answer.

A worked recovery example

The numbers here are illustrative only, to show how binding capacity, membrane area and volume interact. Suppose a membrane binds about 5 µg of a particular peptide per cm² and the filter has 4 cm² of effective area, so it can hold about 20 µg before its binding sites are full.

Filtering 1 mL of a 0.1 mg/mL solution (100 µg in total) could lose up to 20 µg, about 20% of the peptide. Filtering 10 mL of the same solution (1,000 µg) loses the same 20 µg, about 2%. Filtering 1 mL of a 10 µg/mL solution (10 µg in total) could lose nearly all of it. Dilute solutions are therefore usually filtered as a concentrated stock and diluted afterwards with sterile buffer, which the dilution calculator can plan.

Recovery is measured, not assumed. A reading taken before filtration is compared with one taken on the filtrate by the same method: if A280 falls from 0.440 to 0.405, recovery is 0.405 / 0.440 = 92%. Methods are compared in measuring peptide concentration, and the concentration calculator gives the nominal value to compare against.

Pre-flushing, rinsables and extractables

Two things happen in the first part of any filtration. Binding sites on a fresh membrane take up solute, and the device releases organic compounds from its construction materials, including compounds formed during irradiation. A study of sterilizing-grade capsules followed these "rinsables" by total organic carbon and mass spectrometry and showed that rinsing reduced them significantly; it introduced volume-per-membrane-area parameters for deciding how much to rinse.^[3]

A common laboratory practice is to pass and discard a small volume of buffer before the peptide solution, or to discard the first fraction of the filtrate itself. Surface-active additives behave like solutes here: membranes can adsorb surfactants, and a flush study of polysorbate adsorption on PES membranes found that pleated capsules behaved differently from flat discs of the same membrane, so a small flat filter is not always a good model of a larger device.^[4]

Compounds that can be released from a material under aggressive conditions are called extractables, and those that actually appear in a filtered solution are leachables (extractables and leachables). In a comparison of five membrane types, polyamide, polycarbonate and PES released more of these compounds than PTFE and regenerated cellulose; those from regenerated cellulose and PES were surface-active, and those from polyamide and regenerated cellulose increased aggregation of a model protein.^[5] Filtered buffer is a useful control in sensitive assays.

Solvent compatibility also matters. A DMSO stock or other organic solution is filtered only through a membrane listed as compatible with that solvent on the chemical-compatibility chart for the device and its housing.

Small organisms and 0.1 µm membranes

Sterilizing-grade membranes are qualified with a standard small bacterium, but organisms in a real solution can be smaller. In one study, a bacterium held in a drug solution shrank from an average length of 1.25 to 0.84 µm over 24 hours and penetrated some 0.2 and 0.22 µm nylon and modified PVDF discs, while 0.1 µm membranes qualified against both bacteria and mycoplasma gave sterile filtrate under the same conditions.^[6]

Mycoplasma have no rigid cell wall and are a known problem in cell culture. Their retention by membranes depended on the medium they were grown in and on filtration temperature: lower temperatures stiffened the cell membrane and increased retention, while bacteria with rigid cell walls showed no such effect between 2 and 35 °C.^[7] For culture media and supplements where mycoplasma are a concern, laboratories often choose a 0.1 µm membrane.

Integrity testing

A filter that is damaged, wrongly assembled or bypassed by a leak at its seal will not retain bacteria however good its pores are. The bubble point test finds the gas pressure at which liquid is pushed out of the largest wetted pores and confirms the stated pore size.

The bubble point is not sensitive to every defect. In tests on cartridge filters with laser-drilled holes, the bubble point reported by automated testers did not change with defect size until a gross leak was detected; the smallest holes that caused a failure were about 30 to 60 µm, associated with bacterial log reduction values of about 4.0 to 4.5. The authors supported pairing the bubble point with a gas diffusion test for a fuller check.^[8]

Small single-use laboratory filters are rarely tested by the user. A practical substitute is to inspect the device, stay within its stated pressure limit, avoid forcing a clogged filter (a sudden drop in resistance can mean a ruptured membrane) and, where sterility matters, incubate a sample of the filtrate in growth medium as a check.

A record checklist

For results from a filtered peptide solution to be interpretable, record:

• Filter. Membrane material, pore size, membrane area, device type and the solvent compatibility checked.
• Procedure. Volume and concentration filtered, any pre-flush or discarded first fraction, temperature and whether the filter clogged.
• Recovery. Concentration before and after filtration by the same method, and the calculated recovery.
• Controls. Filtered buffer or vehicle control, an endotoxin result if relevant, and any sterility check on the filtrate.

This guide describes general laboratory methods and published literature only. It is not a statement about the content or behaviour of any particular material. Short peptides used in cell-culture work, such as KPV and Thymosin Alpha-1, are handled with the same steps. Terms used here are defined in the glossary, and questions can be sent to support.