"Peptide stability" describes two entirely different problems: chemical stability in the vial, measured in months at a given temperature, and peptide half-life in a biological system, measured in minutes to days. DAC, PEGylation, acetylation, amidation and D-amino-acid substitution are all engineering answers to the second problem, and understanding which enzyme each modification blocks is the fastest way to predict how an analogue will behave in a study. This guide separates the two kinds of stability, walks through the four main clearance routes a peptide faces, tabulates reported half-lives across our range, and works the decay arithmetic. Research use only; the figures below come from published pharmacokinetic work on the reference molecules, not from our own material.
Peptide half-life and vial stability: two clocks, not one
| Vial (chemical) stability | Biological half-life | |
|---|---|---|
| What limits it | Hydrolysis, deamidation, oxidation, aggregation, adsorption | Proteolysis, renal filtration, receptor-mediated uptake, hepatic clearance |
| Timescale | Years lyophilized and frozen; weeks in solution | Minutes to days |
| Controlled by | Temperature, water, light, pH, headspace | Sequence, molecular size, plasma-protein binding |
| Fixed by | Storage practice — see how to store peptides | Chemical modification of the molecule |
Confusing them produces bad conclusions in both directions: a molecule with a one-minute plasma half-life can be perfectly stable in a freezer for three years, and a molecule engineered for a week-long circulating half-life is no more resistant to a freeze–thaw event than its unmodified parent.
The four clearance routes
- Exopeptidases. Aminopeptidases chew from the N-terminus, carboxypeptidases from the C-terminus. Both are blocked by capping the relevant end.
- Endopeptidases, above all DPP-4. Dipeptidyl peptidase-4 removes the first two residues from any peptide with alanine or proline in position 2 — which describes native GLP-1, GIP, GHRH and a long list of others. This single enzyme explains most of the incretin and GHRH analogue design space.
- Renal filtration. The glomerulus filters freely below roughly 30–50 kDa. Almost every research peptide is far under that, so an unbound peptide is cleared in the first pass through the kidney. Raising apparent size above the cutoff, or binding to a large plasma protein, is the only way around it.
- Receptor-mediated internalisation and hepatic uptake. Agonists are removed with their own receptors; this route sets a floor that no modification abolishes.
What each modification does
Position-2 substitution: defeating DPP-4
Native GLP-1 has a plasma half-life of roughly 1.5–2 minutes because DPP-4 cleaves it at position 2. Replacing the Ala at that position with a non-standard residue — Aib (α-aminoisobutyric acid) in semaglutide and tirzepatide, or D-Ala in several GHRH analogues — removes the substrate recognition. The equivalent trick appears in CJC-1295 without DAC, whose four substitutions on the GHRH(1-29) backbone confer DPP-4 resistance and raise its reported half-life from GHRH's minutes to roughly 30 minutes.
Albumin binding: DAC and fatty-acid acylation
Thirty minutes is still short. The next step is to tether the peptide to serum albumin, a 66.5 kDa protein that is too large to filter and has a half-life of about 19 days. Two chemistries dominate:
- DAC (drug affinity complex) — a maleimidoproprionyl group that forms a covalent bond with the free cysteine-34 thiol of albumin. CJC-1295 with DAC reports a half-life measured in days rather than minutes as a direct result, at the cost of a continuous rather than pulsatile exposure profile — a distinction that matters greatly in growth-hormone research.
- Fatty-acid acylation — a C16 or C18 diacid chain, usually via a linker, that binds albumin non-covalently and reversibly. This is how semaglutide reaches a reported terminal half-life near 165 hours (about one week) and liraglutide about 13 hours; the difference between the two is largely linker and chain chemistry.
PEGylation: buying size
Attaching a polyethylene glycol chain does not make a peptide protease-proof; it makes it big and slippery. A 5–40 kDa PEG raises the hydrodynamic radius above the glomerular cutoff and sterically shields cleavage sites. PEG-MGF exists for exactly this reason: unmodified MGF is reported to be cleared within minutes, and the PEG chain extends that to hours. The trade-offs are lower receptor affinity per molecule, a heterogeneous product if the PEG is polydisperse, and a molecular weight on the label that is mostly polymer.
Terminal capping and non-natural residues
- N-terminal acetylation caps the free amine against aminopeptidases and adds 42.04 Da. TB-500 is supplied as Ac-LKKTETQ for this reason.
- C-terminal amidation blocks carboxypeptidases, removes a negative charge and, for many native signalling peptides, is required for receptor activity at all.
- Both together, as in N-Acetyl Semax Amidate, close both ends of a short peptide that would otherwise be degraded from either direction.
- D-amino acids and cyclisation make the backbone geometry unrecognisable to proteases that evolved on L-peptides. The notation for all of these is covered in how to read a peptide sequence.
Reported half-lives across the range
| Molecule | Reported half-life | Mechanism responsible |
|---|---|---|
| Native GLP-1 | ~1.5–2 min | DPP-4 cleavage, unmodified |
| Sermorelin (GHRH 1-29) | ~11–12 min | Unmodified, DPP-4 substrate |
| CJC-1295 without DAC | ~30 min | Four substitutions confer DPP-4 resistance |
| Liraglutide | ~13 h | C16 acylation, reversible albumin binding |
| CJC-1295 with DAC | days | Covalent albumin conjugation |
| Semaglutide | ~165 h | Aib² plus C18 diacid albumin binding |
| IGF-1 LR3 | Extended vs native IGF-1 | Arg³ substitution plus 13-residue extension reduce IGFBP binding |
Worked example: decay arithmetic
Elimination is first-order, so the fraction remaining after time t is 0.5 raised to the power (t ÷ half-life). Two consequences worth calculating explicitly:
- Clearance. With a 30-minute half-life, after 2 hours you have had 4 half-lives: 0.5⁴ = 6.25% remaining. After 3 hours, 6 half-lives: 1.6%. A molecule at this end of the scale is effectively gone within a working day.
- With a 7-day half-life, 2 hours is 0.0119 half-lives and 99.2% remains. Sampling schedules that make sense for one molecule are meaningless for the other.
- Steady state. Repeated administration approaches a plateau after roughly 5 half-lives — about 2.5 hours for the 30-minute molecule, about 35 days for the 7-day one. Study durations have to be set from that number, not from convenience.
What none of this changes
Vial stability is governed by different chemistry. As a rule of thumb drawn from Arrhenius behaviour, reaction rates for the degradation routes that matter roughly double for each 10 °C rise, which is why a lyophilized vial that is stable for years at −20 °C may be specified in weeks at 25 °C. A PEGylated or albumin-binding analogue enjoys no advantage here — the modification protects against enzymes, not against hydrolysis, oxidation or moisture ingress.
- Keep water out. Residual moisture is the main driver of solid-state degradation; a vial opened cold pulls condensation onto the powder.
- Keep oxygen and light out. Methionine, cysteine and tryptophan residues are the usual casualties.
- Limit freeze–thaw. Solute concentration at the ice front drives aggregation; aliquot instead, as covered in aliquoting peptide solutions.
The short version: modify the molecule to change what happens in a body, and change the storage to control what happens in a vial. The two problems do not substitute for each other.