Three generations of GLP-1 research molecules. Each one engages more receptors than the last. Each one shows up in published preclinical work with a slightly different metabolic fingerprint.
Here's the short version. Semaglutide hits one receptor. Tirzepatide hits two. Retatrutide hits three. That progression took about twenty years of medicinal chemistry to engineer, and the receptor pharmacology behind it is genuinely interesting.
This article walks through what the published research literature actually says about these molecules. The chemistry that makes them tick. The receptors they engage. The preclinical models researchers have used to characterise them. And what to look for when sourcing reference material for laboratory work.
All discussion is research-use-only. Third-person framing. No human-use language.
Chemistry and structural backbone
Native GLP-1 is a 30-amino-acid peptide. It comes from the proglucagon precursor, gets cleaved in the gut by L cells, and lives for about one to two minutes in research animals before getting chewed up by an enzyme called DPP-4 (dipeptidyl peptidase-4) and cleared by the kidneys.
That short half-life is the problem every GLP-1 analogue tries to solve.
Semaglutide: three modifications
Semaglutide solves it with three structural changes characterised in the published literature.
- Position-eight alanine swapped for alpha-aminoisobutyric acid (Aib), a non-natural residue that DPP-4 cannot cleave
- A C18 fatty diacid chain attached to lysine 26 via a linker, letting the molecule bind reversibly to albumin and float around for about a week
- An arginine substitution at position 34 to make sure the fatty chain attaches at the right spot
Tirzepatide: a chimera
Tirzepatide is a 39-amino-acid synthetic chimera. The backbone is mostly GIP (glucose-dependent insulinotropic polypeptide), but it borrows tricks from GLP-1 research. Same Aib substitution at position two. A C20 fatty diacid for albumin binding.
Retatrutide: triple engagement
Retatrutide goes further. Same 39-residue framework, but engineered to engage three class-B G-protein-coupled receptors at once. Aib residues at positions two and thirteen. C20 diacid lipidation. The design pulls structural motifs from GIP, GLP-1, and glucagon.
The class-B receptors share enough sequence similarity that a single peptide can be engineered to hit all three. That insight is what made tri-agonism possible.
What synthesis looks like
Research-grade material is made by solid-phase peptide synthesis using Fmoc protection chemistry. The lipidated side chain gets attached in a separate solution-phase step. Method choice matters because the linker length, the fatty chain length, and the attachment site all change how the molecule behaves.
Mechanism at the receptor level
All three molecules act on class-B secretin-family G-protein-coupled receptors. Three receptors are relevant here.
- GLP-1R (the GLP-1 receptor)
- GIPR (the GIP receptor)
- GCGR (the glucagon receptor)
They share roughly 40 to 50 percent sequence identity in the seven-transmembrane domain. They all use the same two-domain binding model characterised in cryo-electron microscopy research: the C-terminal of the peptide grabs the extracellular domain, and the N-terminal slides into the transmembrane bundle to trigger G-alpha-s coupling.
How each agonist behaves
Semaglutide Near-full agonist at GLP-1R. Potency in the low picomolar to low nanomolar range depending on the cell system. Selectivity for GLP-1R over the other two receptors exceeds three orders of magnitude in published binding studies.
Tirzepatide Balanced dual agonist at GLP-1R and GIPR. In transfected cell lines, it is actually more potent at GIPR than at GLP-1R. It shows partial agonism at GLP-1R compared to native peptide on certain endpoints. This biased pharmacology has been examined extensively in beta-arrestin and cAMP comparisons.
Retatrutide Engages all three receptors. Balanced agonism at GIPR and GLP-1R, somewhat lower relative activity at GCGR. The exact rank order shifts depending on cell background and assay format.
Receptor residence time matters. Published surface plasmon resonance research has measured the on-rate and off-rate of each agonist at each receptor, and slow off-rate correlates with longer downstream signalling.
Pathway biology and downstream signalling
What happens after the receptor is engaged? That depends on the receptor and the cell type.
In pancreatic beta cells
GLP-1R activation in preclinical islet research couples through G-alpha-s. cAMP goes up. Protein kinase A activates. So does Epac2, a cAMP-activated guanine nucleotide exchange factor. The downstream effects include:
- Enhanced glucose-stimulated insulin secretion
- Modulation of voltage-gated potassium channels
- Changes in beta-cell gene transcription via CREB-family factors
In the brain
GLP-1R is expressed in three central nervous system regions characterised extensively in rodent neurophysiology research.
- Arcuate nucleus of the hypothalamus
- Area postrema
- Nucleus tractus solitarius
Receptor activation in these regions modulates neuronal firing in satiety circuitry. Published research has characterised the projections from these hypothalamic neurons to multiple downstream brain regions involved in autonomic and behavioural pathway biology.
GIPR and GCGR additions
GIPR signalling has been characterised in adipose tissue and in pancreatic alpha and beta cells, with effects on lipid handling and glucagon secretion that differ from GLP-1R.
GCGR activation (relevant for retatrutide) has been characterised in hepatic and adipose research models as influencing gluconeogenesis, lipolysis, and energy expenditure pathways.
The combined receptor engagement produces metabolic signatures in rodent models that single or dual agonists cannot reproduce.
Research domains and published preclinical literature
The preclinical literature on the GLP-1 family is enormous. Glucose homeostasis, body composition, hepatic lipid metabolism, cardiovascular biology, central nervous system pathways.
Diet-induced obese rodent models
In these models, semaglutide produces dose-dependent reductions in food intake and body weight, with concurrent improvements on glucose tolerance testing.
Tirzepatide has been examined in similar rodent and non-human primate models. Published comparative studies generally report greater effects on body weight and glycaemic endpoints than equimolar semaglutide. Whether that comes from GIPR co-engagement or from biased signalling at GLP-1R is still debated in the published literature.
NASH and beyond
Retatrutide research expanded into non-alcoholic steatohepatitis (NASH) rodent models, with reported reductions in hepatic triglyceride content and modulation of inflammatory gene expression.
Published research has also examined GLP-1 receptor agonist effects on:
- Cardiovascular biology (rodent infarction and atherosclerosis models)
- Renal function (diabetic nephropathy models)
- Neuroinflammation (rodent neurodegeneration models)
- Inflammatory bowel pathway biology
- Polycystic ovary research
- Skeletal muscle protein turnover
Single-cell sequencing
Recent published research has used single-cell RNA sequencing to characterise cell type-specific responses to receptor activation in hypothalamic, hepatic, and adipose research preparations. Comparative analyses across the three agonist generations have started to map how differential receptor engagement translates into distinct downstream gene expression programmes.
Other mass-handling research peptides like MOTS-c and AOD-9604 engage different pathway biology and serve as comparison compounds in some preclinical metabolic study designs.
Comparative published literature
Head-to-head preclinical comparisons of semaglutide, tirzepatide, and retatrutide are published. The typical design uses matched molar dosing and looks at body weight, food intake, fasting glucose, glucose tolerance, plasma insulin, and hepatic triglyceride content.
The general hierarchy
Published reports characterise a fairly consistent ranking in research models.
- Tirzepatide > Semaglutide on body weight and glycaemic endpoints at matched exposures