GLP-1

Native GLP-1 occupies an unusual position: it is simultaneously the endogenous reference molecule for the most commercially successful peptide drug class in pharmaceutical history — semaglutide alone ranks among the top five best-selling drugs globally as of 2026 — and a molecule with essentially no clinical product of its own. The arc from native peptide to drug class is the central editorial story.

The clinical translation problem was the 1-to-2-minute plasma half-life. The Deacon 1995 Diabetes characterization made native GLP-1 unsuitable for any chronic dosing schedule outside continuous intravenous infusion, and the field has pursued two parallel pharmacological solutions ever since. The first is structural engineering of the peptide to resist DPP-IV cleavage and extend plasma residence — the GLP-1 receptor agonist drug class. The second is pharmacological inhibition of DPP-IV itself — the gliptin class (sitagliptin, FDA-approved 2006; saxagliptin, 2009; linagliptin, 2011; and others) — which extends the residence time of endogenous GLP-1 and GIP without administering exogenous peptide. The two occupy different positions in current diabetes practice: the receptor agonist class produces supraphysiological engagement, large weight-loss magnitudes, and the cardiovascular-outcomes benefit anchored by SUSTAIN-6, LEADER, and SELECT; the DPP-IV inhibitor class produces modest HbA1c reduction at near-physiological GLP-1 elevation, is weight-neutral, and has not produced comparable cardiovascular signal. The GLP-1 receptor pharmacology dossier develops the receptor-engagement-breadth question across the modern obesity-pharmacotherapy landscape.

The receptor agonist class divides along two engineering strategies for the half-life problem. The first uses a non-human peptide backbone — exendin-4, the Gila monster (Heloderma suspectum) salivary peptide sharing roughly 53% sequence identity with native GLP-1 but carrying a glycine at position 2 that resists DPP-IV cleavage. Exenatide (FDA-approved 2005 as Byetta) is the synthetic version; lixisenatide is an exendin-4 derivative. The second strategy uses an engineered human-GLP-1 backbone with DPP-IV-resistance modifications plus a fatty-acid acylation moiety that binds plasma albumin and slows renal clearance. Liraglutide (FDA-approved 2010 for diabetes, 2014 for weight management) was first to use that architecture. Semaglutide (FDA-approved 2017, 2021) layers an α-aminoisobutyric acid substitution at position 8 with a C18 fatty diacid on Lys26 via a γ-glutamic-acid / OEG linker — yielding an approximately week-long plasma half-life that enables once-weekly subcutaneous dosing. Tirzepatide extends the same albumin-binding architecture to a dual GIP / GLP-1 agonist on a GIP-derived backbone; retatrutide, survodutide, pemvidutide, and mazdutide extend the engineering across additional receptor combinations including glucagon. The amylin-based cagrilintide is sometimes co-formulated with semaglutide.

The oral-bioavailability problem is a related but distinct engineering challenge. The Rybelsus oral semaglutide formulation that received FDA approval in 2019 addressed it through coformulation with sodium N-(8-[2-hydroxybenzoyl]amino)caprylate (SNAC), characterized in Buckley ST, Bækdal TA, Vegge A et al., Sci Transl Med 2018, 10(467):eaar7047. SNAC produces a localized stomach microenvironment of approximately pH 5 at the tablet-mucosa interface, inactivating pepsin and enabling transcellular absorption of intact semaglutide through gastric epithelium — gastric rather than intestinal, compound-specific to semaglutide, not tight-junction-mediated. The bioavailability remains low (roughly 1%) but sufficient at therapeutic oral magnitude.

The pharmacology-of-mimetics frame is the central conceptual point. The engineered class achieves clinical outcomes — sustained appetite suppression, 15-to-25% weight loss, glycemic control, cardiovascular event reduction — that native GLP-1 cannot, because the engineered molecules deliver sustained receptor engagement that the native half-life does not permit. But the class is also constrained by what native GLP-1 can do: every clinical effect traces back to a native GLP-1 receptor present somewhere in human physiology, and the dose-response and tolerability ceiling of the class is set by underlying receptor biology rather than by molecular engineering. Class-shared GI side effects, the rodent-derived medullary-thyroid-carcinoma boxed-warning concern, the post-discontinuation weight-regain pattern, and the lean-mass-loss conversation all reflect native GLP-1 receptor biology rather than idiosyncrasies of any one engineered molecule. The fat-loss decision guide and the peptide receptor pharmacology atlas walk the comparative clinical positioning. The trajectory from exenatide's 2005 approval through retatrutide's late-stage development is the cleanest receptor-pharmacology-to-clinical-magnitude story in modern metabolic medicine, and the absence of native chronic-dosing safety data is the structural reason long-tail signals (NAION, suicidal ideation, severe gastroparesis, lean-mass-loss proportionality) require ongoing characterization at the population scale the class has now reached.