Most discontinued compounds disappear quietly. SAR441255 did the opposite. Sanofi killed the program in the third quarter of 2019, roughly four months after the Phase 1 trial began — and then, in January 2022, published a full clinical and translational report in Cell Metabolism anyway. The paper reads like a posthumous case file: complete in vitro pharmacology, a 42-day non-human primate study, PET receptor occupancy imaging, and first-in-human results, all for a molecule that no longer had a development path.
That sequence is why SAR441255 occupies an unusual position in metabolic research. It is one of very few triple GLP-1/GIP/glucagon receptor agonists with published human target-engagement data, it is genuinely balanced across all three receptors in a way most of its competitors are not, and — because no company is protecting a commercial program around it — it has continued to appear in academic imaging work as recently as late 2025.
This guide covers what the primary literature actually establishes about SAR441255, and what a laboratory sourcing it should verify before putting it into an assay.
Chemical identity and design
SAR441255 is a 39-residue synthetic peptide with a C-terminal amide, built on the exendin-4 scaffold rather than on native GLP-1 or glucagon. That choice is the foundation of everything else about the molecule.
Exendin-4 carries a C-terminal extension (residues 30–39) that folds back to cage Trp25 — the so-called tryptophan cage. This intramolecular packing gives the peptide markedly better helicity, physicochemical behavior, and metabolic stability than the native hormones it mimics. Sanofi’s medicinal chemistry team used that stable chassis and then edited fifteen positions onto it:
- Position 2: α-aminoisobutyric acid (Aib) replaces the native residue. This blocks dipeptidyl peptidase-IV cleavage — the principal clearance route for both GLP-1 and GIP — and simultaneously strengthens activation at all three receptors.
- Positions 17 and 21: GLP-1 residues, introduced to tune GLP-1 receptor activity.
- Position 18: a glucagon residue.
- Positions 19 and 28: GIP residues, added to build GIP receptor activation into the same chain.
- Positions 3, 13, 20 and 29: changes made to balance potency across the three receptors and improve chemical stability in buffer. Leucine at position 13 was singled out as especially important for solution stability; the other three drive GIP receptor activation.
- Positions 32, 34, 35 and 39: exendin tail mutations that address aggregation at acidic pH in the presence of phenolic preservatives — a formulation problem, not a pharmacology one.
- Position 14: the defining feature. A C16 palmitic acid chain is attached to the ε-amino group of a lysine through two γ-glutamic acid spacers, creating an albumin-binding arm that extends circulating half-life.
The reported peptide mass is 4863.6 Da (calculated 4863.63, observed 4863.67 by LC/MS), with HPLC purity above 95% in the published synthesis. There is no public CAS registry number; the compound is identified by its Sanofi development code and by patent application WO2018100135. That absence matters analytically, and we return to it below.
What “balanced” actually means here
The word balanced gets used loosely across the triagonist literature. For SAR441255 it has a specific, measurable meaning.
In HEK293 cells expressing recombinant human receptors, cAMP accumulation assays returned mean EC50 values of 1.03 pM at GLP-1R, 1.01 pM at GCGR, and 0.73 pM at GIPR — within a factor of 1.4 of one another, and comparable to the endogenous ligand at each receptor. Maximal activity matched the cognate hormones as well.
Sanofi then repeated the exercise in cell systems expressing each receptor at endogenous density, which is the harder test:
| Receptor | System | SAR441255 EC50 | Endogenous ligand EC50 |
|---|---|---|---|
| GLP-1R | 1.1B4 human pancreatic β-cells (cAMP) | 27 pM | 29 pM (GLP-1) |
| GCGR | Primary human hepatocytes (cAMP) | 2.1 nM | 3.0 nM (glucagon) |
| GIPR | Human adipocytes (lipolysis) | 42 pM | 102 pM (GIP) |
The potency tracked the native hormone in all three systems. This is the profile that separates SAR441255 from most multi-agonists, which are typically skewed — tirzepatide, for instance, has roughly fivefold lower GLP-1R affinity than native GLP-1 while retaining full GIPR affinity.
One caveat the authors flagged themselves: the receptor activity profile is skewed at murine and monkey receptors even though it is balanced at human ones. Any rodent or primate result therefore has to be read with that species mismatch in mind, and cannot be extrapolated to human receptor stoichiometry directly.
The comparison that motivated the whole design is the glucagon arm. Against a dual GLP-1R/GCGR agonist from the same program, SAR441255 was roughly eighteen-fold more potent at the glucagon receptor in the comparative panel (EC50 6.6 pM versus 119.3 pM). Adding glucagon activity that aggressively normally costs you glycemic control. It didn’t here — and the authors attribute that to the GIP component counterbalancing the diabetogenic pressure of chronic GCGR agonism.
Preclinical record
Diet-induced obese mice. Twenty-five-week-old obese female C57BL/6NHsd mice received subcutaneous SAR441255 twice daily for 28 days at 0.3, 1, 3, 10 or 30 µg/kg. Twice-daily dosing was chosen because the half-life in lean mice is only about 5.9 hours. By day 26, vehicle-treated obese controls had gained 11.5% body weight; the treated groups changed by +9.7%, +6.9%, +5.8%, −4.8% and −14.1% respectively. A dual GLP-1R/GCGR agonist run at the same 30 µg/kg dose produced only −6.3%.
Liver endpoints moved in parallel. Mean serum ALT fell 73% and 81% in the 10 and 30 µg/kg groups versus obese controls, with AST down 46–58% across the top three doses, alongside dose-dependent reductions in liver weight.
Obese diabetic cynomolgus monkeys. Over 42 days at a maintenance dose of 11 µg/kg, body weight fell 12.6 ± 1.74% and HbA1c fell 1.37 ± 0.34%, reaching values below 5% — normoglycemic for the species. The dual GLP-1R/GCGR comparator achieved −8.1% body weight with a similar HbA1c reduction. The substantially greater weight loss without a glycemic penalty is the central preclinical finding.
A useful negative result also came out of this study: total ketones, 3-hydroxybutyrate and FGF21 were not reliable readouts of glucagon receptor engagement in the presence of concurrent GLP-1R and GIPR activation. Researchers designing biomarker panels around triagonists should not assume these carry over from selective glucagon pharmacology.
Receptor occupancy: the part that keeps getting extended
SAR441255 is, to date, unusually well characterized by positron emission tomography — first by Sanofi, and later by academic groups working with material the company continued to supply.
In the original 2022 work, lean cynomolgus monkeys given 11 µg/kg subcutaneously showed a 72.5% reduction in pancreatic GLP-1R tracer binding and a 40.6% reduction in hepatic GCGR tracer binding at approximately Cmax. GIP receptor occupancy could not be assessed because no radiotracer of sufficient potency existed at the time.
That gap was closed in a 2025 eBioMedicine study using a porcine model and two tracers, including a newer GIPR-specific ligand. The dose–occupancy relationship in pancreas was clean: 16% occupancy at 1 µg/kg, 56% at 2.5 µg/kg, 73% at 4 µg/kg, and a plateau of 83% at 14 µg/kg that did not increase further at 100 µg/kg. Estimated EC50 was 1.8 µg/kg in pancreas.
Two findings from that paper deserve attention:
Central engagement. GLP-1R occupancy reached 77% in the pituitary and up to 60% in the medial hypothalamic area — the region containing the arcuate and periventricular nuclei, the brain’s appetite centre. The dose–occupancy curve was shifted right relative to pancreas (EC50 3.9 µg/kg pituitary, 2.5 µg/kg medial hypothalamus), suggesting that centrally mediated effects require higher exposure than the peripheral incretin effect.
GIPR engagement is real but modest. At 100 µg/kg, pancreatic GIPR tracer uptake fell 23 ± 8.5% — statistically significant, but the authors were appropriately cautious, noting that porcine pancreatic GIPR expression is substantially lower than human and that the tracer bound weakly in vivo in this species.
The head-to-head against tirzepatide in the same study is worth reading carefully rather than taking at face value. Tirzepatide showed under 60% GLP-1R occupancy even at 450 µg/kg and no dose dependence — but its Tmax is 8–72 hours, and the scans were performed 2.5 hours post-dose, which is near-optimal for SAR441255 (Tmax ~3 h) and well before tirzepatide reaches peak concentration. The comparison reflects study design as much as pharmacology.
First-in-human results
The Phase 1 study (NCT04521738) was a randomized, double-blind, single-centre, placebo-controlled single-ascending-dose trial in 48 lean-to-overweight healthy adults, randomized 3:1 to SAR441255 (3, 9, 20, 40, 80 or 150 µg) or placebo. It ran in Knoxville, Tennessee between April and September 2019. All participants completed.
Pharmacokinetics. Absorption was steady, with median Tmax at 3.0–3.5 hours and monophasic elimination thereafter. Mean terminal half-life was 3.5–6.1 hours across the 20–150 µg range, with approximately dose-proportional exposure and stable apparent clearance — consistent with linear pharmacokinetics. The 3 and 9 µg cohorts largely fell below quantification. Notably, this is a short half-life for a lipidated peptide; the two-γGlu palmitoyl arm extends circulation but does not produce the multi-day exposure seen with C18-diacid designs.
Glycemic effects. Following a mixed-meal tolerance test, postprandial glucose, insulin and C-peptide all fell dose-dependently at 80 and 150 µg. The authors interpreted the simultaneous drop in insulin and C-peptide as evidence that the postprandial glucose effect was driven mainly by delayed gastric emptying — a classic GLP-1R mechanism — rather than by insulinotropic action, and cautioned that this specific effect is subject to tachyphylaxis with chronic exposure.
An unexplained observation: in the fasting state at one hour post-dose, three of six subjects at 80 µg and all six at 150 µg recorded glucose below 3.9 mmol/L. None reported symptoms, and values returned to baseline within one to two hours. The timing preceded Cmax, and fasting insulin and C-peptide did not track the glucose change. The authors offered glucagon- and GIP-mediated counter-regulation as a hypothesis but explicitly left the mechanism open.
Target engagement biomarkers. Because the three receptors have overlapping downstream effects, Sanofi used receptor-specific markers. C-telopeptide of type I collagen (CTX), a bone-resorption marker and validated GIPR readout, fell more than 50% from baseline in both high-dose cohorts — substantially larger than the modest declines reported for GLP-1R agonism alone. Plasma amino acids, a validated GCGR biomarker, dropped substantially over three hours post-dose, indicating strong hepatic glucagon receptor activation. Together these confirmed simultaneous engagement at all three targets in humans.
Tolerability. Treatment-emergent adverse events occurred in 27.8% of SAR441255 recipients versus 25.0% on placebo. Gastrointestinal events dominated, concentrated at the top two doses; nausea was reported by three of six subjects at 150 µg, all mild. There were no serious or severe events, no discontinuations, and no anti-drug antibodies. Heart rate trended upward at 150 µg on dosing day and normalized by the following day — consistent with the dose-related heart-rate increase seen in telemetered monkeys, which resolved by day 4–5 of repeat dosing.
Dose escalation stopped at 150 µg. The publication is explicit that this was unrelated to safety or tolerability and was instead “linked to changes in company strategy.”
Why the program ended
The timeline is worth laying out, because secondary sources routinely mangle it.
Sanofi’s Q2 2019 results announced that Phase 1 had been initiated. The Q3 2019 results, published three months later, listed SAR441255 as discontinued. In December 2019, incoming CEO Paul Hudson announced that Sanofi would exit diabetes and cardiovascular research entirely. The Cell Metabolism paper appeared in January 2022 — more than two years after the molecule had already been removed from the pipeline.
So SAR441255 was not abandoned for a toxicity signal, a failed endpoint, or a manufacturing problem. It was abandoned because its sponsor left the therapeutic area. Meanwhile the mechanism it validated went on to succeed elsewhere: Lilly’s retatrutide (LY3437943) and Hanmi’s efocipegtrutide pursued the same triple-receptor concept with longer-acting formats, and by the mid-2020s triagonism was among the most competitive spaces in metabolic drug development.
For laboratories, that history is a practical advantage. There is no commercial program restricting access, the primary pharmacology is fully published rather than held as proprietary data, and the compound continues to serve as a well-characterized balanced-triagonist reference point in imaging and receptor pharmacology work. If you are comparing against skewed dual agonists, our Survodutide and Mazdutide listings cover the GLP-1/glucagon side of that design space.
Analytical considerations for sourcing
This is where SAR441255 gets harder than a typical catalog peptide, and where a certificate of analysis has to do real work.
There is no CAS number to check against. Identity cannot be confirmed by registry lookup. It has to rest on intact-mass measurement against the published 4863.6 Da figure, plus sequence-level confirmation. A supplier that lists a CAS number for this compound is either citing something unverifiable or has confused it with a different molecule.
Intact mass alone cannot confirm the acylation site. The palmitoyl-γGlu-γGlu arm must sit on the ε-amine of Lys14 specifically. In the published synthesis this is controlled by orthogonal monomethoxytrityl protection on that single lysine, deprotected selectively before the lipid is coupled. A positional isomer — acylation at a different amine, or on the N-terminus — carries exactly the same molecular weight. Distinguishing the correct regiochemistry requires peptide mapping with MS/MS fragmentation, not a single deconvoluted mass.
The C-terminal amide is a 0.98 Da question. Synthesis on Rink amide resin yields the C-terminal amide; a des-amido (free acid) impurity differs by less than one dalton on a ~4.9 kDa peptide. Low-resolution MS will not resolve it.
Aib2 is a sterically hindered coupling. Incomplete incorporation of α-aminoisobutyric acid produces deletion sequences. These are visible by HPLC if the gradient is adequate, but a broad method can hide them under the main peak.
Gross mass is not peptide content. Purified peptide from preparative reversed-phase HPLC is typically isolated as the trifluoroacetate salt. On a peptide with multiple basic residues, TFA counterions plus residual water can account for a meaningful fraction of the vial’s gross mass. Quantifying by weighed powder therefore overstates molar concentration unless net peptide content is measured directly. CHN elemental analysis is the appropriate orthogonal anchor for this, and it is the method we use to establish net content rather than inferring it from HPLC area percent, which reports purity among peptide-related species and says nothing about counterion loading.
Aggregation behavior is documented. Four of the fifteen engineered positions exist specifically to prevent aggregation at acidic pH in the presence of phenolic preservatives. Buffer selection for solution work is not a neutral choice with this molecule.
Every lot we list carries third-party COA verification, with orthogonal methods rather than a single vendor chromatogram. Current specifications for SAR441255 are published on the product page.
What the literature still does not answer
Honest gaps are worth stating plainly:
- No multiple-dose human data exist. Everything in humans comes from a single-dose study in 48 healthy volunteers. The published glycemic effects may be confounded by gastric-emptying delay that would attenuate with repeat exposure.
- No data in the target population. No participants had type 2 diabetes or obesity.
- Effects on food intake and energy expenditure were never measured in the human study.
- Chronic amino acid depletion is an open question. The GCGR component drove essential and non-essential amino acids down acutely; the authors flagged long-term monitoring as necessary and unresolved.
- Human GIPR occupancy has never been imaged. The pig data are suggestive but limited by low porcine pancreatic GIPR expression.
More discussion of multi-receptor peptide design and analytical verification is available in our peptides research category.
Research use statement
SAR441255 is supplied strictly for laboratory and in vitro research use. It is not a dietary supplement, not approved by the FDA for any use, and not intended for human or veterinary administration, diagnostic use, or any household application. Nothing in this article constitutes dosing guidance. Purchasers confirm that all material will be used within a qualified research setting under the applicable terms and conditions.
References
- Bossart M, Wagner M, Elvert R, et al. Effects on weight loss and glycemic control with SAR441255, a potent unimolecular peptide GLP-1/GIP/GCG receptor triagonist. Cell Metabolism. 2022;34(1):59–74.e10. doi:10.1016/j.cmet.2021.12.005
- Khalil A, Velikyan I, Xiong M, Bossart M, Wagner M, Eriksson O. Positron emission tomography to assess drug occupancy at peripheral and central incretin receptors. eBioMedicine. 2025;122:106033. doi:10.1016/j.ebiom.2025.106033
- Finan B, Douros JD. GLP-1/GIP/glucagon receptor triagonism gets its try in humans. Cell Metabolism. 2022;34(1):3–4. doi:10.1016/j.cmet.2021.12.010
- Starling S. A new GLP1, GIP and glucagon receptor triagonist. Nature Reviews Endocrinology. 2022;18(3):135. doi:10.1038/s41574-021-00631-w
- Araki E, Sakaguchi M, Fukuda K, Kondo T. Potential of a GLP-1R/GIPR/GCGR triagonist for the treatment of obesity and type 2 diabetes. Journal of Diabetes Investigation. 2022;13(12):1958–1960. doi:10.1111/jdi.13896
- Finan B, Yang B, Ottaway N, et al. A rationally designed monomeric peptide triagonist corrects obesity and diabetes in rodents. Nature Medicine. 2015;21(1):27–36. doi:10.1038/nm.3761
- Evers A, Pfeiffer-Marek S, Bossart M, et al. Multiparameter peptide optimization toward stable triple agonists for the treatment of diabetes and obesity. Advanced Therapeutics. 2020;3(10):2000052. doi:10.1002/adtp.202000052
- Elvert R, Herling AW, Bossart M, et al. Running on mixed fuel — dual agonistic approach of GLP-1 and GCG receptors leads to beneficial impact on body weight and blood glucose control. Diabetes, Obesity and Metabolism. 2018;20(8):1836–1851. doi:10.1111/dom.13212
- Evers A, Pfeiffer-Marek S, Bossart M, et al. Peptide optimization at the drug discovery–development interface: tailoring of physicochemical properties toward specific formulation requirements. Journal of Pharmaceutical Sciences. 2019;108(4):1404–1414.
- Eriksson O, Velikyan I, Haack T, et al. Drug occupancy assessment at the glucose-dependent insulinotropic polypeptide receptor by positron emission tomography. Diabetes. 2021;70(4):842–853. doi:10.2337/db20-1096
- Bech EM, Pedersen SL, Jensen KJ. Chemical strategies for half-life extension of biopharmaceuticals: lipidation and its alternatives. ACS Medicinal Chemistry Letters. 2018;9(7):577–580.
- Chabenne JR, DiMarchi MA, Gelfanov VM, DiMarchi RD. Optimization of the native glucagon sequence for medicinal purposes. Journal of Diabetes Science and Technology. 2010;4(6):1322–1331.
- Coskun T, Sloop KW, Loghin C, et al. LY3298176, a novel dual GIP and GLP-1 receptor agonist for the treatment of type 2 diabetes mellitus: from discovery to clinical proof of concept. Molecular Metabolism. 2018;18:3–14. doi:10.1016/j.molmet.2018.09.009
- Sanofi. Q2 2019 results — R&D pipeline update (Phase 1 initiation of SAR441255). 29 July 2019. news.sanofi.us
- Sanofi. Q3 2019 results — R&D pipeline update (discontinuation of SAR441255). 31 October 2019. prnewswire.com
- ClinicalTrials.gov. Single ascending dose study of SAR441255 in healthy subjects. NCT04521738. clinicaltrials.gov/study/NCT04521738
