What happens to a peptide added to cell culture medium
Three processes begin at once, and none of them is accounted for by the number written on the tube. Some fraction of the peptide adsorbs to the walls of whatever container holds it. Some fraction is cleaved by proteases — those already present in serum-supplemented medium, and those the cultured cells themselves secrete over the course of the experiment. A further fraction undergoes chemical modification driven by the buffer rather than by any enzyme. What remains intact and in solution is the only material still available to interact with anything in the well, and it is a smaller quantity than the nominal concentration implies. How much smaller depends on the sequence, the vessel surface, the medium composition and the duration of the incubation, and in most published work it is not measured at all. The nominal concentration is an input to the experiment. The concentration in the well is a result, and treating the first as though it were the second is one of the more common sources of irreproducibility in peptide work in vitro.
How much peptide is lost to the container?
Enough to change a result, and unpredictably enough that it has to be checked per peptide rather than assumed.
Goebel-Stengel and colleagues measured this directly. They incubated eight ¹²⁵I-labelled endocrine peptides — ghrelin, sulfated cholecystokinin-8, corticotropin-releasing factor, glucagon-like peptide-1, insulin, leptin, nesfatin-1 and peptide YY, chosen to span a range of net charge, size and end-group chemistry — for 48 hours in glass and plastic tubes, untreated or coated with siliconising fluid, and recovered them by gamma counting. Their opening premise is the finding: "the unpredictable nature of peptide binding to surfaces requires optimization of experimental containers to be used." Important differences in binding capacity existed between surface types. Siliconisation decreased recovery. Adding 1% bovine serum albumin raised it, and lyophilising peptide-plus-BSA solutions in the container best suited to each individual peptide returned more than 89% recovery for all eight (Anal Biochem, 2011;414(1):38–46; PMID 21315060, DOI 10.1016/j.ab.2011.02.009).
Two details in that paper deserve more attention than they usually get. The first is that siliconisation, widely applied on the assumption that it reduces binding, went the wrong way. The second is that unlabelled ghrelin displaced ¹²⁵I-ghrelin from borosilicate glass, while GLP-1 and Fmoc-arginine did not — meaning the binding is specific and saturable, not a uniform surface film. A loss fraction characterised at one concentration therefore does not necessarily hold at another. The authors' conclusion is blunt: choosing the appropriate container "avoids unpredictable peptide loss that results in inaccurate measurements and false conclusions."
What degrades the peptide once it is in the medium?
Proteases from two sources, and the balance between them shifts as the culture proceeds.
The first source is the serum supplement. Böttger and colleagues compared degradation of peptides from three structural families in fresh mouse blood, in serum drawn from the same animals, and in commercial serum and plasma. Peptides were generally degraded faster in serum than in plasma, all were more stable in fresh blood than in either derived matrix, and — the finding that matters most for experimental design — the rank order of degradation rates among the peptides varied across the six incubation conditions tested. Their reading is that proteolysis measured in these assays can be misleading, because it reflects cleavage sites relevant to the in-vitro matrix rather than the intact system (PLoS One, 2017;12(6):e0178943; PMID 28575099, DOI 10.1371/journal.pone.0178943). A half-life measured in one matrix is a property of that matrix as much as of the peptide.
The second source is the cells. Rozans and colleagues built a high-throughput LC-MS assay specifically to quantify cell-induced peptide degradation in culture, on the basis that peptides are natural substrates for cell-secreted enzymes and that those enzymes can drive unwanted degradation of the peptide being studied. Their method injects unpurified culture samples directly, and they report quantifying the degradation of dozens of peptides simultaneously (J Biomed Mater Res A, 2025;113(1):e37864; PMID 39806927, DOI 10.1002/jbm.a.37864). The existence of that assay is itself the argument: cell-secreted proteolysis was worth a dedicated analytical method because the alternative was leaving it uncharacterised.
Does the buffer modify the peptide on its own?
Yes, and phosphate-buffered saline is one of the conditions under which it happens.
Erckes and colleagues monitored deamidation and isoaspartate formation by tandem mass spectrometry and confirmed isoaspartate formation under mildly basic conditions such as PBS, while amidated peptides remained stable in neutral aqueous-organic mixtures or at lower temperatures. Degradation was site-dependent, with C-terminal amides markedly more susceptible in their hands. They also note that conventional chromatography and standard mass spectrometry often fail to distinguish these products, since aspartate and isoaspartate are isomers of identical mass (RSC Med Chem, 2026;17(2):1144–1154; PMID 41541711, DOI 10.1039/d5md01025j).
This route is enzyme-independent. A sterile, cell-free buffer at physiological pH is sufficient, which means the modification proceeds during preparation and pre-incubation as readily as during the experiment itself.
Is the vehicle inert?
Not at concentrations in routine use. Verheijen and colleagues exposed 3D cardiac and hepatic microtissues to medium with or without 0.1% dimethyl sulfoxide and profiled transcriptome, proteome and DNA methylation. Both tissue types showed more than 2,000 differentially expressed genes affecting similar biological processes; microRNA analysis showed large-scale deregulation in cardiac microtissues and smaller but still substantial effects in hepatic ones; and methylation changes were tissue-specific. Their conclusion is that DMSO "is not inert" (Sci Rep, 2019;9(1):4641; PMID 30874586, DOI 10.1038/s41598-019-40660-0).
A vehicle control does not resolve this so much as define it. It establishes what the solvent does on its own, which is necessary, and it cannot establish that the solvent and the compound do not interact.
What can a certificate of analysis establish here?
The composition of a batch on a test date. Nothing downstream of that.
Our published certificates carry a measured HPLC purity, a batch number, a test date and a named laboratory — BPC-157 batch 2026-03, tested 22 January 2026, and four others in the certificate library, all analysed by Janoshik Analytical. That is a real measurement and it is the strongest evidence a supplier can offer, because it is an observation rather than a claim. It is also an observation about a sealed vial, made before the material was weighed, dissolved, filtered, diluted into medium and left at 37 °C for a working day. Every process described above happens after the last point the certificate speaks to.
Where compound-specific data on any of this exists, it should be cited. For most compounds sold for laboratory research, no published study establishes an adsorption loss on a given plate surface, a proteolytic half-life in a given medium, or a buffer-driven modification rate. That absence is the finding, and it is better stated than papered over.
Veridian Research supplies these materials strictly for in-vitro laboratory research. They are not drugs and are not approved for human or veterinary use.
The short version
A peptide added to culture medium is partly adsorbed to the vessel, partly cleaved by serum and cell-secreted proteases, and partly modified by the buffer without any enzyme involved. Each of those fractions is sequence-, surface- and matrix-specific, and none is captured by the nominal concentration. The published methods for measuring them exist. What is usually missing is the measurement.