Peptide Adsorption to Labware: Surface Loss, Recovery Failure, and Why a 99% Purity COA Cannot Guarantee Assay Concentration

Peptide adsorption to glass and plastic container walls can strip 80 to 90 percent of a dilute peptide solution out of the liquid phase within seconds of preparation. A vial certified at 99 percent purity by RP-HPLC can still deliver a small fraction of the intended concentration to a cell assay, because the loss happens after the certificate of analysis is issued, inside the researcher’s own labware.

This is one of the least discussed failure modes in peptide research. Purity, identity, and net peptide content are all measured on the lyophilized powder as it leaves the analytical laboratory. None of those measurements describe what happens once the powder is reconstituted, diluted into buffer, and pipetted between tubes. At the low micromolar and nanomolar concentrations typical of receptor binding work and cell based assays, the dominant source of concentration error is frequently not synthesis quality at all. It is the container.

What Surface Adsorption Actually Is

Peptides are amphipathic by construction. A sequence that folds into an amphipathic helix presents a hydrophobic face and a charged face simultaneously, and both faces find something to bind on ordinary labware. Borosilicate glass carries surface silanol groups that deprotonate above roughly pH 3, leaving a net negative surface that attracts cationic peptides electrostatically. Polypropylene presents the opposite problem: a hydrophobic, largely uncharged surface that binds nonpolar side chains through hydrophobic association. There is no common laboratory material that is inert to both interaction modes, which is why substituting glass for plastic solves the problem far less often than researchers assume.

The quantity that governs severity is the ratio of container surface area to solution volume. A fixed number of binding sites on the wall removes a fixed number of peptide molecules. When the solution is concentrated, that fixed loss is a trivial percentage. When the solution is dilute, the same absolute loss becomes most of the sample. This is the mechanistic reason adsorption behaves like a threshold effect rather than a smooth proportional tax, and it explains why the problem is invisible during stock preparation and severe at the working dilution.

The Concentration Threshold Where Recovery Collapses

The most systematic quantification of this effect comes from Kristensen, Henriksen, and Andresen, published in PLoS ONE in 2015 (10(5):e0122419). The group used analytical RP-HPLC to measure recovery of three cationic membrane active peptides, mastoparan X, melittin, and magainin 2, after one hour of incubation in borosilicate glass vials, standard polypropylene tubes, or Protein LoBind tubes. The peptide solutions were 220 microlitres in volume, spanning 1, 2, 5, 10, and 20 micromolar, in 10 millimolar HEPES buffer at pH 7.4.

At 1 micromolar in either borosilicate glass or polypropylene, only 10 to 20 percent of the expected peptide was recovered. All three peptides behaved the same way despite differing in sequence, length, and charge distribution. Recovery climbed steadily as the starting concentration rose, which is the signature of progressive wall saturation rather than degradation. Protein LoBind tubes, a surface engineered to reduce protein and peptide binding, retained substantially more peptide in solution across the full concentration range, though they did not eliminate loss entirely.

The practical translation is uncomfortable. A researcher preparing a nanomolar working solution for a receptor binding curve in standard tubes may be constructing that curve from an unknown and non uniform fraction of the nominal concentration. The resulting affinity estimate is not merely noisy. It is systematically shifted, and the direction of the shift depends on labware the methods section rarely records.

Adsorption Is Immediate, Not Gradual

A common assumption is that surface loss accumulates over storage time, and that working quickly protects the sample. The kinetic data contradict this. In the same PLoS ONE work, 220 microlitre solutions of 2 micromolar mastoparan X were incubated in glass vials or polypropylene tubes for 10 seconds, 1 hour, or 24 hours. Recovery depended only weakly on incubation time. Adsorption to the container wall was essentially complete within the first few seconds after the peptide entered the solution, most likely while the tube was still being vortexed.

This finding removes the most convenient mitigation. Speed does not help, because the equilibrium is reached faster than any realistic handling step. The only variables that matter are the surface itself, the concentration, and whether something else in the solution competes for the binding sites.

Serial Transfers Compound the Loss

The same investigators tracked what happens across multiple containers. Solutions of 250 microlitres at 5 micromolar were transferred successively through one to four borosilicate glass vials, polypropylene tubes, or Protein LoBind tubes, with a one hour incubation between each transfer. After four transfers through standard glass or polypropylene, recovery approached zero. Protein LoBind tubes again preserved considerably more material through the same sequence.

Serial dilution schemes, aliquoting a reconstituted vial into single use tubes, and moving a sample between a preparation tube, a storage tube, and an assay plate are all multi container workflows. Each step is a fresh set of unsaturated binding sites. A protocol that looks rigorous on paper can quietly discard nearly all of the compound before the first measurement is taken.

Key Research Findings

• At 1 micromolar in borosilicate glass vials or polypropylene tubes, only 10 to 20 percent of expected peptide was recovered after 1 hour, consistently across mastoparan X, melittin, and magainin 2 (Kristensen et al., PLoS ONE, 2015, 10(5):e0122419).
• Recovery rose with increasing peptide concentration across 1 to 20 micromolar, consistent with progressive saturation of a finite number of wall binding sites rather than chemical degradation.
• Adsorption kinetics were fast: recovery at 10 seconds, 1 hour, and 24 hours of incubation differed only weakly, indicating equilibrium is reached within seconds.
• Four successive transfers of 5 micromolar solutions through glass vials or polypropylene tubes reduced recovery to near zero.
• Varying buffer NaCl from 2 to 149 millimolar did not strongly change recovery in polypropylene or Protein LoBind tubes, indicating the interaction is not purely electrostatic.
• Across eight radiolabelled endocrine peptides incubated 48 hours in treated and untreated glass and plastic, siliconization reduced recovery while added bovine serum albumin improved it; lyophilizing peptide with BSA in the optimal container returned more than 89 percent recovery for every peptide tested (Goebel-Stengel et al., Analytical Biochemistry, 2011, 414(1):38-46).
• Peptide identity, not just container material, determines the magnitude of loss, and no single surface was optimal for all peptides in the 2011 comparison.

Charge and Ionic Strength Predict Less Than Expected

An intuitive model would predict that raising buffer ionic strength screens electrostatic attraction and rescues cationic peptides from a negatively charged glass surface. The data only partly support this. When Kristensen and colleagues prepared 2 micromolar solutions in HEPES buffers containing 2, 100, or 149 millimolar NaCl, they observed at most a slight upward trend in recovery with salt concentration in borosilicate glass, and no consistent salt effect in polypropylene or Protein LoBind tubes. Control experiments confirmed the HPLC method itself was insensitive to the salt concentration, so the null result is real rather than an artifact of quantification.

The earlier work by Goebel-Stengel and colleagues reaches a compatible conclusion from a different direction. That group incubated eight radiolabelled endocrine peptides, including ghrelin, sulfated cholecystokinin-8, corticotropin releasing factor, GLP-1, insulin, leptin, nesfatin-1, and peptide YY, for 48 hours in glass and plastic tubes that were either untreated or coated with siliconizing fluid. The peptides were deliberately chosen to span a wide range of net charge, size, terminal group chemistry, and modification. No single physicochemical descriptor predicted which peptides adsorbed most. Binding capacity varied by surface and by peptide in a way that resisted simple rules.

The honest conclusion from both studies is that adsorption behaviour must be measured empirically for a given peptide and a given container. It cannot be reliably inferred from sequence, isoelectric point, or hydrophobicity index alone.

Carrier Proteins, Siliconization, and Low Binding Plastics

Three mitigations appear repeatedly in the literature, and they are not equally effective. Adding a carrier protein such as bovine serum albumin works by competitive occupancy: the far more abundant protein saturates the wall binding sites, leaving the peptide in solution. In the 2011 Analytical Biochemistry work this was the single most effective intervention, and combining BSA with lyophilization in an appropriately selected tube produced recovery above 89 percent for all eight peptides tested. The limitation is obvious. Carrier protein is unacceptable in any workflow feeding mass spectrometry, and it can interfere with binding assays.

Siliconization, often assumed to be protective, decreased recovery for the peptides in that study. Coating a surface with a hydrophobic silicone layer removes silanol groups but creates an excellent substrate for hydrophobic association, which is the wrong trade for amphipathic sequences.

Engineered low binding polymer surfaces occupy the middle ground. They reduced loss substantially in both the 2015 concentration series and the serial transfer experiment without introducing exogenous protein. Independent proteomics work reaches the same practical conclusion. Bark and Hook, writing in the Journal of Proteome Research in 2007 (6(11):4511-4516), documented differential recovery of peptides from ordinary sample tubes and linked it directly to the reproducibility of quantitative proteomic data. Kraut and colleagues, in the same journal in 2009 (8(7):3778-3785), found that glass outperformed standard plastic but that low adsorption plastic gave the best recovery overall, with the poorly recovered peptides being predominantly hydrophobic.

Where This Sits Relative to Certificate of Analysis Data

A certificate of analysis answers a bounded question: what is in the vial as supplied. RP-HPLC area percent describes chromatographic purity, mass spectrometry confirms identity, and amino acid analysis or Karl Fischer water determination establish net peptide content against water and counterion mass. Every one of those numbers can be correct while the concentration in an assay well is off by an order of magnitude, because adsorption occurs downstream of all of them.

This is a reason to take COA data more seriously, not less. Analytical characterisation is the only fixed reference point in the chain. If the starting material is documented by an independent laboratory, a recovery discrepancy can be attributed to handling and investigated. If the starting material is undocumented, a low assay response is uninterpretable, because synthesis quality, net peptide content, and surface loss all produce the same observation. Maple Research Labs publishes independent third party analysis by Janoshik Analytical on its certificates of analysis page for exactly this reason: it converts one unknown into a known.

Practical Implications for Research Handling

Several conclusions follow directly from the data rather than from convention. Concentration is the strongest lever available, so preparing and storing concentrated stocks and diluting into the assay vessel immediately before measurement limits the time any solution spends dilute in a high surface area container. Minimising the number of transfer steps matters more than working quickly, given that adsorption equilibrates within seconds. Container material should be selected empirically for the peptide in question rather than assumed, and the choice should be recorded in the methods section alongside buffer composition.

Verification is possible without specialised equipment. Comparing measured ultraviolet absorbance at 280 nanometres against the calculated value for a sequence containing tryptophan or tyrosine reveals gross discrepancies between nominal and actual solution concentration. Where the peptide lacks aromatic residues, a recovery control run through the identical labware sequence provides the same information. Handling variables interact with solubility and aggregation behaviour, which are covered separately in our work on peptide solubility and pH optimization, on aggregation and fibrillation detection, and on reconstitution solvent selection. Surface adsorption and interfacial aggregation are mechanistically linked, since a peptide that adsorbs and unfolds at a solid or air interface can seed nucleation in the bulk solution.

Limitations of the Current Evidence

The 2015 PLoS ONE dataset rests on two independent replicates per condition for the adsorption experiments and three for the concentration standard curves, which is adequate for effects of this magnitude but leaves the finer comparisons imprecise. All three test peptides were cationic and membrane active, a class chosen for its strong surface affinity, so the 10 to 20 percent recovery figure should be read as characteristic of that class rather than as a universal constant. Anionic and neutral sequences may behave differently, and the 2011 radiolabel study makes clear that peptide specific variation is large. Neither study evaluated the full range of commercially available polymer formulations, and surface chemistry differs between manufacturers and between lots.

What survives these caveats is the qualitative conclusion, which is well replicated across four independent groups and two decades: dilute peptide solutions lose material to container walls rapidly, the loss is large enough to invalidate quantitative work, and it is invisible unless deliberately measured. Research peptides sourced with verified analytical documentation, such as the compounds listed alongside batch specific analysis including BPC-157, give researchers a defensible baseline against which handling losses can be separated from material quality.

For research purposes only. Not for human consumption. Not for diagnostic or therapeutic use.