Many natural peptide hormones work for minutes and then disappear from the blood. That suits the body, which releases them in short bursts, but it is a problem for a medicine that has to be injected. The difference between a peptide that lasts six minutes and one that lasts a week often comes down to a few chemical changes at specific positions in the chain. This page explains what those changes are and why they work, using peptides covered elsewhere on this site.
It is a different question from how a peptide keeps in a vial. Storage is about water, pH, oxygen and temperature; in the body the main players are enzymes, excretion and binding to other molecules. For storage, see peptide stability and storage.
From minutes to a week
The six peptides on the chart span more than three orders of magnitude. Sermorelin is an unmodified fragment of a natural hormone. The others carry modifications that protect a vulnerable bond, bind the peptide to a blood protein, or both. The rest of the page explains these mechanisms.
What a half-life does and does not tell you
A half-life measured in blood is the time it takes for the concentration to fall by half. The concentration can fall for several reasons: the peptide is cut by enzymes, removed from the body, or moves out of the blood into tissues. The half-life itself does not say which.
The routes are often not fully known. For tesamorelin, an approved medicine, the label states that no formal metabolism studies have been done in humans[2]: even for a licensed peptide, the exact routes of elimination can remain uncharacterised.
Half-life is also not the same as the duration of the effect. After an intravenous dose of ipamorelin in healthy men, growth hormone rose to a single sharp peak at about 40 minutes and fell to very low levels by 6 hours at all doses, while plasma ipamorelin persisted for longer than the growth hormone response[3]. The time course of an effect and the half-life of the drug are different measurements.
Enzymes that cut peptides
Blood contains many peptidases, and different peptides are vulnerable to different ones. One enzyme, dipeptidyl peptidase-4 (DPP-4), removes the first two amino acids in a single step from susceptible peptides that have a free N-terminus and a suitable second residue. GHRH and GLP-1, both with alanine at position 2, are among its substrates.
- GHRH and sermorelin. In human plasma, DPP-4 cut growth hormone-releasing hormone and its shorter forms between residues 2 and 3; a specific DPP-4 inhibitor blocked the cut, and so did replacing the alanine at position 2 with D-alanine. The 1–29 fragment, which is sermorelin, was also susceptible to a separate trypsin-like cleavage between residues 12 and 13 (the cleavage pattern differed among GHRH fragments), and 94% of it was degraded after 60 minutes in plasma[6]. After intravenous injection its half-life was about 6 minutes[1].
- GLP-1. In human plasma at 37 °C, GLP-1 was converted to an inactive form lacking its first two residues with a half-time of about 20 minutes. Cooling to 4 °C or adding a DPP-4 inhibitor prevented it, and the shortened form was a major metabolite in the blood of healthy people[7]. The same cut happens in collected blood samples unless protease inhibitors are added[8].
DPP-4 is not the whole story. In the same blood-sample study, the hormone oxyntomodulin was also cut by other, unidentified peptidases[8], and GHRH had its second, trypsin-like cleavage site[6]. Which enzyme matters depends on the sequence.
How chemists extend half-life
1. Block the cut site. The simplest change is at the bond the enzyme attacks. The 1989 GHRH study showed that a D-amino acid at position 1 or 2 prevents DPP-4 cleavage[6]. Modern peptide drugs more often use Aib (2-aminoisobutyric acid), an amino acid with two methyl groups on its α-carbon and no L or D form. Semaglutide carries Aib at position 8 of the GLP-1 sequence, the second residue of the active hormone, specifically to protect against DPP-4[4], and retatrutide has Aib at position 2[5]. Tesamorelin takes a different route: it keeps the full GHRH sequence and adds an acyl cap to the N-terminus.
2. Hitch a ride on albumin. Albumin, the most abundant protein in blood, has a half-life of about three weeks. A fatty acid attached to a peptide makes it bind albumin reversibly, and the stronger that binding, the longer the peptide was expected to stay in circulation; this was the working hypothesis behind semaglutide[4]. Liraglutide, with a C16 fatty acid attached to a lysine through a short linker, has a half-life of 13–15 hours after injection under the skin, enough for once-daily dosing[4].
Semaglutide was designed by changing this side chain systematically. The developers compared fatty acids of different lengths and linkers of different composition and chose an 18-carbon diacid joined to lysine 26 through a linker of γ-glutamate and two short oligo(ethylene glycol) units. In mini-pigs its half-life after intravenous injection was 46 hours; in humans it is about 165 hours, allowing once-weekly injection[4]. Retatrutide uses the same principle with a C20 diacid and has a half-life of about six days[5].
These designs involve trade-offs. Stronger albumin binding came with lower receptor affinity: semaglutide binds the GLP-1 receptor about three times less tightly than liraglutide, a cost the developers accepted in exchange for longer exposure[4].
What modifications cannot promise
A protected peptide is not automatically a long-lived one. Ipamorelin contains two D-amino acids and an Aib residue, yet its half-life after intravenous infusion was about two hours[9][3]. A review of peptide stabilisation strategies notes that fast clearance remains one of the main obstacles for most therapeutic peptides, and that replacing natural amino acids with unnatural ones does not always give the protection hoped for[10].
The longest half-lives shown here are achieved through designs that combine resistance to enzymatic degradation with reduced clearance, notably through reversible albumin binding, and they came out of testing many variants. Lau and colleagues described a whole series of analogues before selecting semaglutide[4].
In short
- Half-lives of peptides in blood range from minutes (sermorelin, 6 minutes) to about a week (semaglutide, retatrutide).
- A falling blood level can reflect enzymatic breakdown, removal from the body or distribution into tissues; the half-life alone does not say which, and it is not the same as how long the effect lasts.
- DPP-4 can remove the first two residues from susceptible peptides with a free N-terminus and a suitable second residue, including GHRH and GLP-1; other peptidases matter for other sequences.
- Appropriate substitutions with D-amino acids or Aib can prevent DPP-4 cleavage at susceptible sites, depending on the peptide and the position; fatty acids that bind albumin can extend half-life to days.
- Modifications do not guarantee a long half-life: ipamorelin, with two D-amino acids and Aib, lasts about two hours.
References
- Prakash A, Goa KL (1999). Sermorelin: a review of its use in the diagnosis and treatment of children with idiopathic growth hormone deficiency. BioDrugs 12(2):139–157. DOI: 10.2165/00063030-199912020-00007
- U.S. Food and Drug Administration (2025). EGRIFTA WR (tesamorelin) for injection, for subcutaneous use: full prescribing information. Revised 3/2025. accessdata.fda.gov (PDF)
- Gobburu JVS, Agersø H, Jusko WJ, Ynddal L (1999). Pharmacokinetic-pharmacodynamic modeling of ipamorelin, a growth hormone releasing peptide, in human volunteers. Pharmaceutical Research 16(9):1412–1416. DOI: 10.1023/A:1018955126402
- Lau J, Bloch P, Schäffer L, Pettersson I, Spetzler J, Kofoed J, Madsen K, Knudsen LB, McGuire J, Steensgaard DB, Strauss HM, Gram DX, Knudsen SM, Nielsen FS, Thygesen P, Reedtz-Runge S, Kruse T (2015). Discovery of the once-weekly glucagon-like peptide-1 (GLP-1) analogue semaglutide. Journal of Medicinal Chemistry 58(18):7370–7380. DOI: 10.1021/acs.jmedchem.5b00726
- Coskun T, Urva S, Roell WC, Qu H, Loghin C, Moyers JS, O'Farrell LS, Briere DA, Sloop KW, Thomas MK, Pirro V, Wainscott DB, Willard FS, Abernathy M, Morford L, Du Y, Benson C, Gimeno RE, Haupt A, Milicevic Z (2022). LY3437943, a novel triple glucagon, GIP, and GLP-1 receptor agonist for glycemic control and weight loss: from discovery to clinical proof of concept. Cell Metabolism 34:1234–1247. DOI: 10.1016/j.cmet.2022.07.013
- Frohman LA, Downs TR, Heimer EP, Felix AM (1989). Dipeptidylpeptidase IV and trypsin-like enzymatic degradation of human growth hormone-releasing hormone in plasma. Journal of Clinical Investigation 83(5):1533–1540. DOI: 10.1172/JCI114049
- Deacon CF, Johnsen AH, Holst JJ (1995). Degradation of glucagon-like peptide-1 by human plasma in vitro yields an N-terminally truncated peptide that is a major endogenous metabolite in vivo. Journal of Clinical Endocrinology & Metabolism 80(3):952–957. DOI: 10.1210/jcem.80.3.7883856
- Yi J, Warunek D, Craft D (2015). Degradation and stabilization of peptide hormones in human blood specimens. PLoS ONE 10(7):e0134427. DOI: 10.1371/journal.pone.0134427
- Raun K, Hansen BS, Johansen NL, Thøgersen H, Madsen K, Ankersen M, Andersen PH (1998). Ipamorelin, the first selective growth hormone secretagogue. European Journal of Endocrinology 139(5):552–561. DOI: 10.1530/eje.0.1390552
- Al Musaimi O, Lombardi L, Williams DR, Albericio F (2022). Strategies for improving peptide stability and delivery. Pharmaceuticals 15(10):1283. DOI: 10.3390/ph15101283