Semaglutide

Overview

Semaglutide is an acylated peptide analogue of the incretin hormone glucagon-like peptide-1 (GLP-1), belonging to the class of long-acting GLP-1 receptor (GLP-1R) agonists. The molecule was developed and characterised at Novo Nordisk, and the foundational medicinal-chemistry report was published by Lau and colleagues in the Journal of Medicinal Chemistry in 2015 ^[1]. Semaglutide was engineered directly from the earlier once-daily analogue liraglutide, with the aim of producing a once-weekly molecule by maximising reversible binding to serum albumin while preserving GLP-1R potency and full enzymatic stability ^[1][2].

Scientific interest in semaglutide stems from a combination of pharmacological properties: a long half-life in humans (approximately one week), a glucose-dependent insulinotropic effect, and a central influence on appetite regulation. The molecule has been studied in two formulations: subcutaneous (for once-weekly administration) and oral (for once-daily administration, co-formulated with the absorption enhancer SNAC) ^[2][3]. The following sections neutrally summarise the published preclinical and clinical data on structure, mechanism of action, pharmacokinetics, and study outcomes. This document is a scientific reference synthesis and contains no recommendations for use in humans.

Pharmacology and mechanism of action

Structure and molecular design

Native human GLP-1(7-37) is a peptide of 30–31 amino-acid residues that is rapidly inactivated (plasma half-life ~1.5–2 min) by dipeptidyl peptidase-4 (DPP-4), which cleaves the N-terminal His7-Ala8 dipeptide, and that also undergoes renal clearance ^[1]. Semaglutide retains the GLP-1 peptide backbone but carries three key engineering modifications relative to the native hormone ^[1][2].

First, the alanine at position 8 (the DPP-4 cleavage site) is replaced by the non-proteinogenic, helix-forming α-aminoisobutyric acid (Aib), which sterically blocks DPP-4 recognition and confers strong resistance to N-terminal cleavage while preserving receptor affinity (substitution with glycine at the same position had reduced potency) ^[1]. Second, the native lysine at position 34 is replaced by arginine (Arg34) to direct acylation site-specifically to the single remaining lysine residue, Lys26. Third, the side chain of Lys26 is derivatised with a lipophilic albumin-binding moiety ^[1].

Acylation chemistry and albumin binding

The Lys26 side chain is conjugated, via a γ-glutamic acid (γ-Glu) linker and two hydrophilic 8-amino-3,6-dioxaoctanoic acid (AEEA/OEG) spacers, to a C18 fatty diacid (octadecanedioic acid) ^[1][6]. This distinguishes semaglutide from liraglutide, which uses a C16 monoacid (palmitoyl) with a single γ-Glu linker. The systematic structure-activity studies carried out by Lau and colleagues showed that GLP-1R potency increased as the diacid length grew from C12 to C18 and then declined for diacids longer than C18; the C18 diacid with a γ-Glu-2×OEG linker was identified as the optimal combination of high albumin affinity and preserved receptor potency ^[1]. The hydrophilic OEG spacer distances the bulky albumin molecule from the receptor pharmacophore, limiting steric interference with receptor binding. The reported affinity of semaglutide for GLP-1R was approximately 0.38 nM, about three-fold lower than that of liraglutide, but with substantially higher albumin affinity ^[1]. The crystal structure of the semaglutide peptide backbone in complex with the extracellular domain of the GLP-1 receptor has been deposited in the Protein Data Bank (PDB 4ZGM) ^[1].

Reversible binding to abundant serum albumin protects the peptide from DPP-4 degradation and slows renal filtration, lowering the free, receptor-active fraction and markedly prolonging circulation time ^[1][6]. Preclinical half-lives scaled with diacid length (for example, from ~1.2 h for the C12 diacid to ~7 h for semaglutide in rats; ~46–55 h intravenously and ~63–75 h subcutaneously in minipigs) ^[1].

GLP-1 receptor signalling

GLP-1R is a class B (secretin family) G-protein-coupled receptor. Agonist binding couples predominantly to Gαs, activating adenylate cyclase and raising intracellular cAMP, which engages two effectors: protein kinase A (PKA) and the cAMP-regulated guanine-nucleotide exchange factor Epac2 ^[2][9][31]. In pancreatic β-cells this potentiates glucose-stimulated insulin secretion (GSIS)^[5]. The insulinotropic action is glucose-dependent: it enhances secretion when glucose is high and has little effect at euglycaemia, because the downstream steps require glucose metabolism (ATP-dependent closure of K_ATP channels, depolarisation, Ca2+ influx) ^[9][10]. PKA/Epac2 close residual K_ATP and voltage-dependent K+ (Kv) channels, prolong depolarisation, mobilise intracellular Ca2+, and prime/recruit insulin granules for Ca2+-dependent exocytosis ^[9]. Endosomal (sustained) cAMP signalling following receptor internalisation also contributes to insulin output ^[9].

Glucagon suppression, gastric emptying, and central effects

Semaglutide suppresses glucagon secretion from pancreatic α-cells in a glucose-dependent manner, lowering hepatic glucose production. Two non-mutually-exclusive mechanisms have been described: a paracrine pathway, in which GLP-1R activation on δ-cells raises somatostatin that inhibits α-cells via SSTR2; and a direct action on α-cells, where a sparse population of GLP-1R on α-cells raises cAMP and, via PKA, inhibits P/Q-type voltage-dependent Ca2+ channels, reducing glucagon exocytosis ^[11]. Importantly, the counter-regulatory glucagon response to hypoglycaemia is largely preserved, because the effect is glucose-dependent ^[11].

Semaglutide slows gastric emptying, blunting postprandial glucose excursions and prolonging satiety; this is mediated largely through vagal afferent/central pathways linking intestinal GLP-1 signalling to the brainstem, with the effect on gastric emptying showing tachyphylaxis under chronic exposure ^[12]. Centrally, GLP-1R agonism reduces appetite and energy intake. Semaglutide does not cross the blood-brain barrier freely; distribution studies in rodents show access via the circumventricular organs (area postrema, median eminence, organum vasculosum, subfornical organ), with c-Fos activation in the hindbrain, hypothalamus (arcuate nucleus), and secondary relays such as the lateral parabrachial nucleus ^[13]. In the arcuate nucleus, GLP-1R agonists depolarise and activate anorexigenic POMC/CART neurons while indirectly inhibiting orexigenic NPY/AgRP neurons, shifting the melanocortin system toward satiety ^[13]. Together, these pancreatic, gastrointestinal, and neuronal actions account for the improvements in glycaemic control^[28] and reductions in body weight observed in preclinical and clinical studies.

Study outcomes

The clinical research programme for semaglutide comprises three large coordinated study series: SUSTAIN (subcutaneous formulation in type 2 diabetes), PIONEER (oral formulation in type 2 diabetes), and STEP (subcutaneous formulation for body-weight regulation), together with dedicated cardiovascular-outcomes trials. What follows is a neutral summary of published results and is not a recommendation for use in humans.

Glycaemic control

The SUSTAIN programme evaluated once-weekly subcutaneous semaglutide. The key SUSTAIN 1–5 and 7 trials tested the investigational maintenance doses against a graded set of comparators across the spectrum of type 2 diabetes: placebo as monotherapy (SUSTAIN 1), the DPP-4 inhibitor sitagliptin (SUSTAIN 2), extended-release exenatide (SUSTAIN 3), insulin glargine (SUSTAIN 4), placebo on a background of basal insulin (SUSTAIN 5), and a head-to-head comparison with the GLP-1R agonist dulaglutide (SUSTAIN 7) ^[3][15]. Treatment periods ranged from 30 to 56 weeks with a gradual dose-escalation regimen. The primary efficacy endpoint was change in glycated haemoglobin (HbA1c) from baseline, with change in body weight as a confirmatory/secondary endpoint. In these trials, the reported reductions in HbA1c were on the order of approximately 1.0–1.5 percentage points for the lower investigational dose and approximately 1.3–1.8 (up to ~2.0) percentage points for the higher investigational dose, with body-weight reductions of approximately 3.5–6.5 kg ^[3]. In SUSTAIN 7, the lower investigational dose of semaglutide reduced HbA1c by 1.5% versus 1.1% for the lower investigational dose of dulaglutide, and the higher investigational dose of semaglutide reduced HbA1c by 1.8% versus 1.4% for the higher investigational dose of dulaglutide, with correspondingly greater reductions in body weight ^[16].

The PIONEER programme evaluated the first oral GLP-1R agonist (co-formulated with the absorption enhancer SNAC) in an investigational once-daily oral regimen. In PIONEER 1 (26 weeks, monotherapy versus placebo), the reported placebo-corrected reductions in HbA1c increased with dose (by the treatment-policy estimate, approximately 0.6%, 0.9%, and 1.1% from the lower to the higher investigational level) with a dose-dependent change in body weight, most pronounced at the higher investigational level ^[17]. Other PIONEER trials compared oral semaglutide with placebo and with active comparators (for example, empagliflozin, sitagliptin, liraglutide), with randomisation stratified by background glucose-lowering therapy and end-of-treatment endpoints at 26, 52, or 78 weeks. Efficacy was analysed within the estimand framework (treatment-policy estimand versus trial-product estimand), which accounts for different handling of treatment discontinuation and rescue medication ^[3].

Body-weight regulation

The STEP (Semaglutide Treatment Effect in People with obesity) programme is a coordinated phase 3a series of randomised, double-blind, placebo-controlled trials that evaluated once-weekly subcutaneous semaglutide for body-weight regulation in adults with overweight or obesity. Across the trials, the published reports used a gradual dose-escalation regimen to the maintenance level over approximately 16 weeks to mitigate gastrointestinal effects; all groups received a lifestyle or behavioural intervention ^[18].

STEP 1 randomised 1961 adults with overweight or obesity without diabetes 2:1 to semaglutide at the investigational maintenance dose or placebo over 68 weeks ^[18]. The reported mean change in body weight was −14.9% with semaglutide versus −2.4% with placebo (estimated treatment difference about −12.4 percentage points). The proportions achieving at least 5%, 10%, and 15% weight reduction were 86.4%, 69.1%, and 50.5% (semaglutide) versus 31.5%, 12.0%, and 4.9% (placebo). In STEP 2 (1210 adults with type 2 diabetes), the reported mean change in body weight was −9.6%, −7.0%, and −3.4% for the higher investigational dose, the lower investigational dose, and placebo, respectively ^[19]. STEP 3 (combination with intensive behavioural therapy) reported a change of −16.0% versus −5.7% ^[20], and STEP 5 (treatment continued to 104 weeks) reported −15.2% versus −2.6% ^[22].

STEP 4 used a withdrawal design: after a 20-week semaglutide run-in, participants who reached the maintenance dose were randomised 2:1 to continue semaglutide or switch to placebo through week 68. From week 20 to week 68, the semaglutide-continuation group changed by −7.9% versus +6.9% with the switch to placebo, illustrating weight regain on discontinuation ^[21]. An extension of STEP 1 reported that one year after treatment stopped at week 68, participants had regained approximately two-thirds of the weight lost (mean regain about 11.6 percentage points), with most cardiometabolic improvements returning toward baseline, indicating that the effects depend on continued exposure ^[23].

Cardiovascular outcomes and tolerability

The SUSTAIN-6 trial (Marso et al., 2016) was a pre-registration cardiovascular-safety trial of subcutaneous semaglutide at the investigational maintenance doses versus placebo in 3297 patients with type 2 diabetes, ~83% of whom had established cardiovascular disease or chronic kidney disease, over 104 weeks ^[14]. The primary composite endpoint (cardiovascular death, non-fatal myocardial infarction, or non-fatal stroke) occurred in 6.6% on semaglutide versus 8.9% on placebo (hazard ratio [HR] 0.74; 95% CI 0.58–0.95), achieving both non-inferiority and superiority. Non-fatal stroke was reduced (1.6% versus 2.7%; HR 0.61), and non-fatal myocardial infarction was numerically reduced (2.9% versus 3.9%; HR 0.74). A notable signal was a significant increase in diabetic-retinopathy complications (HR 1.76), partly attributed to rapid glycaemic lowering in patients with pre-existing retinopathy ^[14].

The SELECT trial (Lincoff et al., 2023) extended the cardiovascular-outcomes evidence to people with overweight/obesity but WITHOUT diabetes: 17,604 adults with prior cardiovascular disease and a BMI ≥27 were randomised to once-weekly semaglutide at the investigational maintenance dose or placebo, with a mean follow-up of ~40 months ^[24]. The primary composite endpoint occurred in 6.5% versus 8.0% (HR 0.80; 95% CI 0.72–0.90; P<0.001), a relative risk reduction of approximately 20%, alongside a mean body-weight reduction of ~9–10%. Serious adverse events were not increased overall; however, adverse events leading to permanent discontinuation were higher with semaglutide (16.6% versus 8.2%), driven by gastrointestinal disorders (discontinuation due to gastrointestinal events ~10.0% versus 2.0%) ^[24][25]. For supporting class context, PIONEER 6 (Husain et al., 2019) showed cardiovascular non-inferiority of oral semaglutide versus placebo in high-risk patients with type 2 diabetes (HR for MACE 0.79), with a nominal reduction in all-cause mortality ^[4][26].

Gastrointestinal events are the dominant tolerability concern for incretin agents as a class. In randomised trials and pharmacovigilance/meta-analysis data, the most frequent events are nausea, vomiting, diarrhoea, and constipation; these are typically dose-dependent, most pronounced during dose escalation, and tend to attenuate over time ^[25][27]. Network meta-analyses report that GLP-1 receptor agonists carry a higher risk of gastrointestinal events compared with placebo, with semaglutide associated with a comparatively higher risk of diarrhoea among the agents; the risk of nausea increases with dose and then tends to plateau ^[27]. Other class-level signals noted in the literature include gallbladder-related events and (rarely) pancreatitis, generally without an excess of serious events in the large cardiovascular-outcomes trials.

Pharmacokinetics and stability (research context)

Semaglutide is engineered for weekly exposure. In the initial characterisation, the affinity for the GLP-1 receptor was ~0.38 nM, and the intravenous plasma half-life in minipigs was ~46 h (mean residence time ~63.6 h after subcutaneous administration) ^[1][6]. In humans, clinical plasma-protein binding exceeds 99%, predominantly to albumin; the apparent volume of distribution is small (~12.5 L), reflecting confinement largely to the vascular/interstitial compartment ^[6]. The terminal half-life is approximately one week (~165 h), permitting once-weekly subcutaneous administration with steady state reached after approximately 4–5 weeks ^[2][6]. The long half-life results from high albumin binding (which reduces renal filtration and proteolytic exposure) and metabolic stabilisation, with FcRn-mediated albumin recycling itself contributing to persistence ^[6]. Clearance is low, and after subcutaneous administration tmax is ~1–3 days ^[6].

The metabolism of semaglutide is not confined to any single organ: it is degraded through proteolytic cleavage of the peptide backbone and sequential β-oxidation of the fatty-acid side chain ^[7]. Intact semaglutide is the principal circulating species (~69–83% of drug-related material). Both urine and faeces are important routes of excretion, with urine the major route; minimal unchanged drug is excreted. Renal impairment has limited influence on exposure, consistent with metabolism-dominated elimination ^[8].

There are two formulations. The subcutaneous formulation is an aqueous solution. The oral formulation co-formulates semaglutide with the absorption enhancer SNAC (sodium N-[8-(2-hydroxybenzoyl)amino]caprylate; sodium salcaprozate) ^[3][29]. SNAC promotes localised gastric absorption: as the tablet erodes, it creates a transiently buffered, higher-pH microenvironment that protects the peptide from pepsin and promotes monomerisation, while transiently and reversibly increasing the transcellular permeability of the gastric epithelium ^[29]. Oral absorption is rapid (tmax ~0.8–1.75 h), but absolute bioavailability is low (~0.8%) with high within-subject variability (~137%); exposure depends strongly on the fasting interval after dosing, the volume of water consumed, and body weight ^[30]. The terminal half-life is ~1 week for both the oral and subcutaneous routes, indicating route-independent elimination kinetics ^[30].

Regarding laboratory handling and stability (physico-chemical information only, for research use; not directions for use in humans): as with other therapeutic peptides, lyophilised semaglutide is markedly more stable than solutions. Recognised peptide degradation routes relevant to such molecules include backbone hydrolysis, methionine/tryptophan oxidation, asparagine/glutamine deamidation, racemisation, and physical aggregation; reaction rates increase roughly two-fold for every 10 °C. Reconstituted aqueous peptide solutions are chemically less stable than the dry solid state, and repeated freeze-thaw can promote aggregation. Quantitative statements about storage windows derive from non-peer-reviewed supplier sources and should be regarded only as indicative, not as validated specifications.

Research status and handling

Semaglutide remains the subject of active preclinical and clinical research as a model molecule of the long-acting GLP-1 receptor agonist class. The body of published data describes a cardiovascular-risk-reduction signal in high-cardiovascular-risk populations (both with and without diabetes) against an adverse-event profile that is predominantly gastrointestinal and largely dose/titration-related ^[14][24][27]. The molecule's research value derives from its well-characterised mechanism of action, predictable pharmacokinetics, and the substantial body of randomised controlled data generated in the SUSTAIN, PIONEER, and STEP programmes.

From a scientific-handling standpoint, the stability information presented in this overview is strictly physico-chemical and concerns general principles of peptide behaviour under laboratory conditions. It does not constitute instructions for the preparation, dosing, or administration to any person. All dosing and clinical-use parameters mentioned in this document are stated neutrally from the published preclinical/clinical literature in order to represent study designs accurately and are not recommendations.

This overview is intended solely for scientific and reference purposes and is not medical advice, a treatment recommendation, or guidance for use in humans.