Filtering peptide solutions

Filtration is the most routine step in preparing a peptide solution and one of the easiest to get quietly wrong. The same membrane that removes particles can also remove the peptide, shed extractables into it, or create a sterility claim the filter was never qualified to support. The variables are few, but they interact.

Part of the handling & laboratory practice pillar — the end-to-end chain from receipt to result.

Four different jobs

“Filter it” can mean four unrelated things, and a filter chosen for one may be wrong for another:

Purpose	Typical pore size	What it does not do
Remove visible particles and haze	0.45 µm or larger	Does not sterilise, does not remove dissolved impurities
Sterilise a solution	0.22 µm	Does not remove endotoxin, viruses or mycoplasma
Reduce bioburden before storage	0.22 µm	Does not make the solution sterile if the receiving vessel is not sterile
Remove aggregates	Depends on aggregate size	Does not remove monomers that will re-aggregate; may remove product

The distinction between sterilising and depyrogenating is the one most often lost. A 0.22 µm membrane retains bacteria; it does not retain endotoxin, which is a molecule of a few tens of kilodaltons and passes through. Removing pyrogen requires adsorption, ultrafiltration or dry-heat depyrogenation of glassware — not a syringe filter. This is covered in endotoxin and bioburden.

Membrane chemistry

Membranes differ in what they bind and what they leach. For peptide work the relevant comparison is:

Membrane	Aqueous compatibility	Organic compatibility	Binding tendency	Notes
PES (polyethersulfone)	Good	Moderate	Low	Common low-binding choice for aqueous protein and peptide solutions
PVDF	Good	Good	Moderate	Widely used for HPLC solvent and sample preparation; can bind hydrophobic peptides
Nylon	Good	Good	Higher	Binds peptides and proteins noticeably; also adsorbs some small molecules
Cellulose acetate	Good	Poor	Low to moderate	Low protein binding, limited solvent range
PTFE	Poor unless pre-wet	Excellent	Low	For organic phases; hydrophobic and needs alcohol pre-wetting for aqueous use
Regenerated cellulose	Good	Moderate	Low	Low binding, good for aqueous buffers

Housing material matters too. Some syringe filter housings are not resistant to pure acetonitrile, DMSO or dichloromethane; the membrane may survive while the housing softens or leaches. Check the manufacturer's solvent compatibility chart for the exact device, not just the membrane family.

Adsorption: the loss you do not see

A filter is a high-surface-area material, and peptides adsorb to surfaces. At milligram-per-millilitre concentrations the loss may be a fraction of a percent; at micrograms per millilitre it can be a large part of the sample. The first millilitres passing through a new filter carry most of the loss, because binding sites are still available.

Practical mitigations, in order:

Pre-rinse. Pass a small volume of the same buffer — or of the sample itself, if volume allows — and discard it. This saturates binding sites before the collected fraction.
Choose low-binding membranes. PES or regenerated cellulose for aqueous work; avoid nylon for dilute peptide solutions unless tested.
Minimise contact. Use the smallest filter and the shortest path that does the job; a large filter for a 1 mL sample is unnecessary surface area.
Recover the hold-up volume. Syringe filters retain liquid in the housing and membrane. Account for the dead volume, or push it through with air if the procedure allows.
Measure before and after. A280 or A214 before and after filtration on a dilute sample will tell you whether loss is material, rather than assuming it is not.

At very low concentrations, the right answer is often not to filter at all: centrifuge at low speed and take the supernatant, which avoids a high-surface-area contact step. The wider problem of surface loss is covered in adsorption to labware.

Pore size and the sterility claim

0.22 µm is the conventional sterilising grade for aqueous solutions, and 0.45 µm is a clarification grade. Two qualifications matter:

Mycoplasma. If a solution must be mycoplasma-free, 0.22 µm is not the accepted barrier; 0.1 µm filtration is used for that claim. A 0.22 µm filter removes bacteria, not mycoplasma.
Virus. Neither pore size removes viruses. A sterile-filtered solution is not virus-free, and no filter-based claim should suggest otherwise.

Filter ratings are also nominal. A “0.22 µm” membrane has a pore-size distribution, and retention depends on the challenge organism, the load and the pressure. That is why sterilising filtration is validated with a bacterial challenge and an integrity test rather than by the number on the package.

Integrity testing and aseptic practice

A sterilising filter is only a sterilising filter if it is intact. The bubble-point or pressure-hold test performed after filtration is the evidence; without it, the sterility claim rests on the package label. In a laboratory setting:

Work in a laminar-flow hood or a clean bench, with sterile tubing and a sterile receiving vessel.
Pre-wet and pre-rinse the filter as the procedure requires.
Filter at a pressure the device is rated for; excessive pressure can breach the membrane or force particles through.
Perform the integrity test after filtration, and record the result.
Label the filtrate with date, filter lot, pore size and volume filtered.

Filtering into a non-sterile tube defeats the purpose. So does reusing a filter across sessions, or storing a wet filter for later use.

Throughput and pressure

A filter has a capacity, and exceeding it shows up as rising back-pressure, slow flow or a burst. For peptide solutions the load is usually low, but three factors reduce throughput: viscosity (high concentration, co-solvents), particle load (undissolved cake, precipitates) and membrane fouling by aggregates.

If a solution blocks a 0.22 µm filter quickly, do not push harder. Clarify through a larger pore size first, or centrifuge, then filter the supernatant. A prefilter stack — a glass-fibre or larger-pore layer above the sterilising membrane — is the standard answer for particle-heavy solutions. Forcing a blocked filter is how membranes fail and how particles reach the filtrate.

When not to filter

The solution is already clear and the volume is small. The loss and the extractables risk may exceed the benefit.
The peptide is aggregating. Filtration removes the visible aggregate and leaves the monomer, which can re-aggregate; it treats the symptom and hides the trend. See aggregation and self-assembly.
The solvent attacks the housing. DMSO, DMF and pure acetonitrile require a compatible device; otherwise the filter becomes a contamination source.
The experiment needs the particle. Some assays deliberately include undissolved material or a suspension; filtering changes the sample.
Sterility is the goal but the receiving vessel is open. Fix the aseptic chain first; the filter cannot compensate.

What to record

Filtration is a sample-preparation step that can change the result, so it belongs in the record:

Membrane material, pore size, diameter and manufacturer/lot.
Volume filtered and volume recovered, including the discard from pre-rinse.
Pressure or vacuum used, and whether flow was normal.
Integrity test result, if a sterilising claim is made.
Any visible change — colour, turbidity, precipitate on the membrane.

If a result later looks low, the first question is often whether the peptide stayed on the filter. A record that names the membrane and the volume makes that question answerable.

For laboratory research use only. This article describes laboratory filtration practice; it makes no claim about any specific product, filter or sterile status, and does not support administration to humans or animals.

Editor's take

Filtration is sold as clarification and behaves as a loss step. At small volumes, the hold-up and the membrane adsorption can remove more material than the filter was meant to clean.

Limitations

This is laboratory practice guidance, not a validated stability study. Follow your institution SOPs and the product COA; individual compounds may behave differently.

Products referenced in this guide

Bacteriostatic water 10 mL