Peptide Degradation Pathways: Chemistry and Prevention

For research purposes only. Not for human use. This article describes the chemistry by which synthetic peptides degrade. It is educational, is not medical advice, and does not describe administration of any compound.

Why the Chemistry Matters to a Research Result

A peptide that has degraded by 20% is a mixture of the parent and several new molecules, each with its own activity, and some of those molecules bind the same receptor, block it, or provoke responses the parent never would. Aswad and colleagues note that isoaspartate formation alone "can decrease the biological activity of a protein pharmaceutical, alter its susceptibility to proteolytic degradation, and elicit autoimmunity" (Aswad et al., 2000). When a research result fails to replicate, the peptide's degradation state is a candidate explanation that sits alongside the biology, and it is the one that can be checked by HPLC.

Manning and colleagues divide the field into chemical instability, meaning covalent changes to the molecule, and physical instability, meaning changes in conformation, association or location without a change in covalent structure (Manning et al., 2010). This article follows that division: six chemical pathways, then two physical ones, then how each is detected and which common research compounds are exposed to which. The storage guide covers the handling decisions that follow.

Chemical Pathway 1: Deamidation and Aspartate Isomerisation

Asparagine and aspartate residues degrade through a shared intermediate. The backbone nitrogen of the following residue attacks the side-chain carbonyl to form a five-membered succinimide ring; the ring then hydrolyses to give a mixture of the normal α-linked aspartate and the β-linked isoaspartate, in a ratio of roughly 2:1 isoaspartate to aspartate. For asparagine the net result is deamidation (loss of ammonia and a new negative charge); for aspartate it is isomerisation to isoaspartate with no change in mass, which makes it invisible to mass spectrometry and detectable only chromatographically or enzymatically.

Geiger and Clarke measured the kinetics in the model hexapeptide Val-Tyr-Pro-Asn-Gly-Ala: deamidation to the succinimide had a half-life of only 1.4 days at 37 °C and pH 7.4, the succinimide then hydrolysed with a half-time of 2.3 hours and racemised with a half-time of 19.5 hours, and replacing the glycine after asparagine with a bulky leucine or proline slowed degradation 33- to 50-fold (Geiger & Clarke, 1987). The residue after the asparagine or aspartate controls the rate because succinimide formation needs the chain to fold back on itself; glycine, serine and (for aspartate) glycine again allow that most easily. Aswad's review lists Asn-Gly, Asn-Ser and Asp-Gly as the sequences where rearrangement "occurs most readily" (Aswad et al., 2000). Brennan and Clarke later found that a histidine following an aspartate enhanced the cleavage reaction, suggesting His-Asp sequences may be particularly labile (Brennan & Clarke, 1995).

The pathway runs in real proteins at the rates the models predict. Johnson and colleagues aged recombinant human growth hormone at pH 7.4 and 37 °C and found isoaspartate accumulating at 1.8 sites per 100 molecules per day, concentrated at two positions: Asp-130 in a Leu-Glu-Asp-Gly sequence (isomerisation) and Asn-149 (deamidation, yielding 70% to 80% isoaspartate) (Johnson et al., 1989). Deamidation is fastest at neutral to alkaline pH and slowest in the mildly acidic range, which is one reason acidic reconstitution solvents extend solution shelf life.

Among common research compounds, sermorelin and tesamorelin (GHRH sequences with Asn8-Ser9) and epitalon (Ala-Glu-Asp-Gly) carry the highest-risk motifs; BPC-157's Asp-Asp-Ala is a lower-rate site.

Chemical Pathway 2: Oxidation

Methionine, cysteine, tryptophan, histidine and tyrosine are the oxidisable residues, in roughly that order of susceptibility under ordinary storage. Methionine oxidises to methionine sulfoxide (a mass increase of 16 Da) and, under stronger stress, to the sulfone; cysteine to disulfides, sulfenic and sulfonic acids; tryptophan to kynurenine and N-formylkynurenine via ring-opening; histidine to 2-oxo-histidine.

Trace metals drive much of the oxidation seen in practice. Zang and colleagues found that low levels of metal ions, especially iron, in digestion buffers, the chromatographic column, the LC injector and other sample-contact surfaces caused significant variability in methionine oxidation measurements, and that controlling metal levels restored consistency (Zang et al., 2012). If trace iron in an analytical system can oxidise methionine during a peptide map, trace iron or copper in a reconstitution solvent can do the same during storage. Copper is a special case for the research market because GHK-Cu is sold as a copper(II) complex: Bodnár and colleagues characterised copper(II) complexes of a methionine- and histidine-containing peptide and identified singly, doubly and triply oxidised products in which the methionine and histidine side chains had been oxidised, with ascorbic acid able to either promote or suppress the reaction depending on conditions (Bodnár et al., 2021). A copper peptide co-stored with a methionine peptide is a catalyst next to its substrate.

Light adds a second oxidation route through tryptophan. Igarashi and colleagues showed that tryptophan generates superoxide under illumination at a higher rate than any other essential amino acid and photodegrades with a half-time of about 18 hours under UVA/UVB, in both solid and solution states (Igarashi et al., 2007).

Oxidation-exposed research compounds include Semax (N-terminal methionine), MOTS-c (two methionines and a tryptophan in 16 residues), sermorelin and tesamorelin (Met27), and semaglutide (Trp25). CJC-1295 was designed with leucine in place of GHRH's Met27 and is correspondingly more resistant.

Chemical Pathway 3: Hydrolysis of the Peptide Bond

The amide bond is thermodynamically unstable in water but kinetically slow to hydrolyse at neutral pH and room temperature; uncatalysed, a typical peptide bond has a half-life of years. Two sequence contexts are much faster. Aspartate-proline bonds are cleaved readily under mildly acidic conditions and heat, so readily that Fung and colleagues used the "uniquely acid-labile aspartyl-prolyl amide bond" deliberately, heating a therapeutic protein in dilute formic acid to generate a surrogate peptide for LC-MS quantification (Fung et al., 2014). The bonds on either side of aspartate in general are more labile than average because the side-chain carboxyl participates in the reaction, which is also why Geiger and Clarke observed cleavage products from Asn-Leu and Asn-Pro peptides.

Hydrolysis is the pathway that lyophilisation controls best, since it needs bulk water. In solution, it is the pathway most accelerated by heat and by pH extremes in either direction.

Chemical Pathway 4: Diketopiperazine Formation

Peptides with a free N-terminus and proline at position 2 can cyclise: the N-terminal amine attacks the carbonyl of the second residue, releasing the first two residues as a cyclic dipeptide (a diketopiperazine) and leaving a truncated peptide. Goolcharran and Borchardt measured the kinetics with Phe-Pro model peptides and found pseudo-first-order degradation whose rate depended on the ionisation state of the N-terminal amine, with the unprotonated form more reactive, so the reaction accelerates as pH rises through the amine's pKa; phosphate and glycine buffers acted as general base catalysts, and the residue preceding the proline changed the rate substantially (Goolcharran & Borchardt, 1998).

Selank (Thr-Lys-Pro-Arg-Pro-Gly-Pro) and KPV (Lys-Pro-Val) both have proline at position 2 with a free N-terminus and are exposed. N-acetylation blocks the reaction, which is one benefit of TB-500's acetylated N-terminus, and ipamorelin's N-terminal Aib is sterically protected.

A related N-terminal cyclisation affects glutamate and glutamine at position 1, which form pyroglutamate. Bersin and colleagues found the reaction's pH dependence differed markedly between solution and lyophilised solid, and that in the pH 5.5 to 6 range the solid-state rate was greater than the solution rate (Bersin et al., 2021), a reminder that lyophilisation slows most pathways but not all of them equally.

Chemical Pathway 5: β-Elimination and Disulfide Scrambling

Cysteine, serine and threonine can undergo β-elimination at alkaline pH, losing the side-chain group and leaving a dehydroalanine residue. In disulfide-bonded peptides the consequence is more than one lost residue: β-elimination of a disulfide releases a free thiol, which can attack another disulfide and reshuffle the bonding pattern. Costantino and colleagues traced the covalent aggregation of moist lyophilised insulin to exactly this mechanism, "intermolecular thiol-catalyzed disulfide interchange following β-elimination of an intact disulfide bridge", and found the process was suppressed by lower temperature and water content and, unexpectedly, by Cu²⁺, which presumably oxidised the free thiols before they could exchange (Costantino et al., 1994). Among research peptides the pathway is relevant to disulfide-containing sequences (oxytocin, vasopressin analogues, some conotoxins) and to any cysteine peptide at pH above 8.

Chemical Pathway 6: Racemisation

Every chiral α-carbon can racemise, converting an L-residue to its D-form with no change in mass and only a small change in retention time. Geiger and Clarke observed racemisation at the succinimide intermediate with a half-time of 19.5 hours in their model peptide (Geiger & Clarke, 1987), so any residue that passes through a succinimide (asparagine, aspartate) racemises as a side reaction of deamidation. Racemisation of other residues in storage is slow at neutral pH but is accelerated by base. The bigger source of D-residues in a research peptide is synthesis, not storage: Badgujar and colleagues review how racemisation during coupling and D-contamination in starting amino acids introduce enantiomeric impurities that ordinary LC-MS cannot detect (Badgujar et al., 2024).

Physical Pathway 1: Aggregation

Aggregation covers reversible oligomerisation, irreversible amorphous precipitation, and ordered fibril formation. It is driven by exposed hydrophobic surfaces, by high concentration, by interfaces (air-liquid, ice-liquid, container walls) and by anything that unfolds a structured peptide. Korang-Yeboah and colleagues studied teriparatide and found that formulation variables governed fibrillar aggregation and gelation, that stabilising the folded structure slowed both, and that oxidation rate was inversely related to solution concentration (Korang-Yeboah et al., 2021). Freezing is a dedicated aggregation stress: Miyahara and colleagues found that human growth hormone aggregated substantially on freezing whether or not dissolved oxygen was present and concluded that "interfacial stresses" dominated (Miyahara et al., 2026). Preservatives contribute too; Heljo and colleagues showed that m-cresol, and to a lesser degree phenol and benzyl alcohol, increased the size of reversible peptide oligomers (Heljo et al., 2015).

Meyer, Ho and Manning make the connection between the physical and chemical pathways explicit: excipients that maintain native conformation and reduce aggregation "have now been found to slow both deamidation and oxidation, whether in solution or in the solid state" (Meyer et al., 2002). A peptide that is aggregating is usually also degrading chemically faster, because the unfolded state exposes the residues involved.

Fatty-acid-conjugated peptides such as semaglutide are designed to self-associate and bind albumin; in a vial with no albumin, that same property predisposes them to aggregation, which is why the GLP-1 class needs stricter solution handling than a short unmodified peptide.

Physical Pathway 2: Adsorption

Peptides adsorb to glass and plastic, and at working-solution concentrations the loss is significant. Grohganz and colleagues measured cetrorelix adsorption following a Langmuir isotherm to a plateau of 0.4 µg/cm², decreasing in the order glass > polypropylene = polyethylene > PTFE, reduced by lipophilic solvent, acidic pH or surfactant, and varying considerably between glass vials from different suppliers (Grohganz et al., 2004). Adsorption is not degradation in the chemical sense, but a solution that has lost a third of its peptide to the container wall behaves, in an assay, exactly like one that has degraded by a third.

Detecting Degradation

Reversed-phase HPLC with UV detection sees most chemical degradation as new peaks: deamidated and isomerised species elute slightly earlier than the parent, oxidised methionine earlier still, diketopiperazine-truncated peptides much earlier, and hydrolysis fragments scattered. Mass spectrometry names the peaks: +16 Da for methionine sulfoxide, +1 Da for deamidation (Asn to Asp), −17 Da (from glutamine) or −18 Da (from glutamate) for pyroglutamate, +4 Da for tryptophan to kynurenine and +32 Da for N-formylkynurenine. Lian and colleagues' review of LC-MS for synthetic peptide therapeutics sets out the workflow and its limits, the central limit being that isomers and epimers share a mass with the parent and must be resolved chromatographically (Lian et al., 2021). Isoaspartate, which has the same mass as aspartate, is the classic blind spot; Aswad's review describes the enzymatic assay using protein L-isoaspartyl methyltransferase that detects it at low picomole levels (Aswad et al., 2000). Aggregation is detected by size-exclusion chromatography, dynamic light scattering or, for fibrils, thioflavin T fluorescence.

Forced-degradation studies deliberately apply each stress and map the products. Yuan and colleagues subjected pramlintide to acid and base hydrolysis, oxidation, photolysis and heat, found extensive degradation under hydrolysis, oxidation and heat but minimal photodegradation, and matched four impurities to predicted degradation products by LC-MS (Yuan et al., 2018). The same design applied to any research peptide would tell a laboratory which storage variable matters most for that sequence. The purity-testing guide explains the chromatographic reading in more detail.

Susceptibility of Common Research Compounds

The table is derived from the sequences and the pathway chemistry above, not from published stability studies of each compound, of which there are few for research-market peptides.

Slowing Degradation

Each pathway has a lever. Deamidation, hydrolysis and diketopiperazine formation slow at mildly acidic pH (around 4 to 6). Oxidation slows without dissolved oxygen, trace metals or light, so degassed solvent, chelator, amber glass and a freezer help, and keeping copper complexes away from methionine peptides helps more. Aggregation slows at lower concentration, without interfaces, and without repeated freezing, so aliquot once and store in low-binding tubes. Lyophilisation removes bulk water and slows almost everything, with the pyroglutamate exception noted above and the moisture-driven aggregation that Costantino documented if the powder is allowed to take up water. The storage guide turns these into a handling protocol, and the reconstitution guide covers solvent selection.

Frequently Asked Questions

Which degradation pathway is fastest at neutral pH? Deamidation at an Asn-Gly site, with a half-life of about 1.4 days at 37 °C in Geiger and Clarke's model peptide. Most other pathways run over weeks to months under the same conditions.

Can mass spectrometry detect every degradation product? No. Aspartate isomerisation and racemisation produce species with the same mass as the parent. They need chromatographic separation or, for isoaspartate, an enzymatic assay.

Why does lyophilisation not stop all degradation? It removes bulk water, which slows hydrolysis and deamidation, but oxidation continues at the cake surface, moisture uptake restarts aggregation, and pyroglutamate formation can be faster in the solid state than in solution at some pH values.

Does GHK-Cu degrade other peptides? Copper(II) catalyses oxidation of methionine and histidine side chains. A copper peptide stored in the same solution as a methionine peptide supplies the catalyst.

Which common research peptides are most stable? TB-500 and ipamorelin, on sequence grounds: no oxidisable residues, blocked or sterically protected N-termini, and no asparagine.

References

1. Aswad DW, Paranandi MV, Schurter BT. Isoaspartate in peptides and proteins: formation, significance, and analysis. J Pharm Biomed Anal. 2000. PMID 10708396

2. Manning MC, Chou DK, Murphy BM, et al. Stability of protein pharmaceuticals: an update. Pharm Res. 2010. PMID 20143256

3. Geiger T, Clarke S. Deamidation, isomerization, and racemization at asparaginyl and aspartyl residues in peptides. J Biol Chem. 1987. PMID 3805008

4. Brennan TV, Clarke S. Effect of adjacent histidine and cysteine residues on the spontaneous degradation of asparaginyl- and aspartyl-containing peptides. Int J Pept Protein Res. 1995. PMID 7558585

5. Johnson BA, Shirokawa JM, Hancock WS, et al. Formation of isoaspartate at two distinct sites during in vitro aging of human growth hormone. J Biol Chem. 1989. PMID 2760065

6. Zang L, Carlage T, Murphy D, et al. Residual metals cause variability in methionine oxidation measurements in protein pharmaceuticals using LC-UV/MS peptide mapping. J Chromatogr B. 2012. PMID 22483985

7. Bodnár N, Várnagy K, Nagy L, et al. Ambivalent role of ascorbic acid in the metal-catalyzed oxidation of oligopeptides. J Inorg Biochem. 2021. PMID 34126320

8. Igarashi N, Onoue S, Tsuda Y. Photoreactivity of amino acids: tryptophan-induced photochemical events via reactive oxygen species generation. Anal Sci. 2007. PMID 17690425

9. Fung EN, Zambito F, Haulenbeek J, et al. Targeting an acid labile aspartyl-prolyl amide bond as a viable alternative to trypsin digestion to generate a surrogate peptide for LC-MS/MS analysis. Bioanalysis. 2014. PMID 25496253

10. Goolcharran C, Borchardt RT. Kinetics of diketopiperazine formation using model peptides. J Pharm Sci. 1998. PMID 9523979

11. Bersin LM, Patel SM, Topp EM. Effect of 'pH' on the Rate of Pyroglutamate Formation in Solution and Lyophilized Solids. Mol Pharm. 2021. PMID 34232660

12. Costantino HR, Langer R, Klibanov AM. Moisture-induced aggregation of lyophilized insulin. Pharm Res. 1994. PMID 8140052

13. Badgujar D, et al. Enantiomeric purity of synthetic therapeutic peptides: A review. Chirality. 2024. PMID 38448043

14. Korang-Yeboah M, Ketcham S, Shih M, et al. Effect of formulation and peptide folding on the fibrillar aggregation, gelation, and oxidation of a therapeutic peptide. Int J Pharm. 2021. PMID 33961953

15. Miyahara Y, Nagel R, Frieß W. Dissolved oxygen effects on human growth hormone stability during freeze-thaw and metal-catalyzed oxidation. J Pharm Sci. 2026. PMID 42097404

16. Heljo P, Ross A, Zarraga IE, et al. Interactions Between Peptide and Preservatives: Effects on Peptide Self-Interactions and Antimicrobial Efficiency In Aqueous Multi-Dose Formulations. Pharm Res. 2015. PMID 25893328

17. Meyer JD, Ho B, Manning MC. Effects of conformation on the chemical stability of pharmaceutically relevant polypeptides. Pharm Biotechnol. 2002. PMID 11987755

18. Grohganz H, Rischer M, Brandl M. Adsorption of the decapeptide Cetrorelix depends both on the composition of dissolution medium and the type of solid surface. Eur J Pharm Sci. 2004. PMID 14757490

19. Lian Z, Wang N, Tian Y, et al. Characterization of Synthetic Peptide Therapeutics Using Liquid Chromatography-Mass Spectrometry: Challenges, Solutions, Pitfalls, and Future Perspectives. J Am Soc Mass Spectrom. 2021. PMID 34110145

20. Yuan Y, Li YB, Tai ZF, et al. Study of forced degradation behavior of pramlintide acetate by HPLC and LC-MS. J Food Drug Anal. 2018. PMID 29389581

Research Disclaimer

The compounds named here are supplied to laboratories for research use only and are not approved for human or veterinary use. This article describes chemistry and analytical practice; no part of it describes or endorses administration of any compound to people or animals.