From dual to quintuple agonism for next-generation pharmacology to treat obesity and type 2 diabetes: synergistic incretin and nuclear receptor signaling

Abstract

Receptor-targeted polypharmacology is emerging as a promising strategy for the treatment of obesity, type 2 diabetes, and metabolic dysfunction-associated steatohepatitis. GLP-1–GIP–lanifibranor, a unimolecular conjugate combining GLP-1R/GIPR co-agonism with pan-PPAR activation, enables receptor-guided intracellular delivery of lanifibranor to incretin receptor-expressing cells while limiting systemic off-target exposure. In obese mouse models, the conjugate produced greater reductions in body weight, adiposity, food intake, and hyperglycemia than semaglutide, GLP-1–GIP co-agonism, or lanifibranor alone, while significantly improving insulin sensitivity. Mechanistic analyses demonstrated receptor-dependent delivery and identified PPARδ signaling as a principal mediator of glycemic improvement independent of weight loss. Unlike unconjugated lanifibranor, the conjugate did not induce anemia, fluid retention, renal dysfunction, or adipocyte differentiation, supporting the concept that tissue-restricted PPAR activation may mitigate classical adverse effects of systemic PPAR agonism. These findings establish peptide-directed nuclear receptor targeting as a potentially important platform for next-generation metabolic therapeutics, although substantial translational uncertainties remain regarding clinical efficacy, safety, and long-term applicability.

Introduction

The past decade has witnessed a paradigm shift in how medicine conceptualises and treats metabolic disease. Obesity and type 2 diabetes, long regarded as conditions demanding modest pharmacological ambition, have become the proving ground for some of the most sophisticated polypharmacological engineering in modern drug development. At the centre of this renaissance lies the incretin axis: a system of gut-derived peptide hormones, principally glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic polypeptide (GIP), that orchestrate postprandial insulin secretion, suppress glucagon release, slow gastric emptying, and exert far-reaching effects on energy homeostasis through central nervous system circuits. Receptor agonists that harness these pathways have progressed from single-agent GLP-1R agonists, such as semaglutide and liraglutide, to dual GLP-1R/GIPR co-agonists, exemplified by tirzepatide, which in head-to-head comparison outperformed semaglutide for weight reduction in the SURMOUNT-5 trial [1] and for glycemic control in the SURPASS-2 trial [6]. The incremental therapeutic gain offered by each successive layer of receptor engagement has emboldened researchers to ask how far multi-target polypharmacology might ultimately go.

Parallel to the incretin story, the peroxisome proliferator-activated receptor (PPAR) family, three nuclear transcription factor isoforms (PPARα, PPARγ, PPARδ) that regulate lipid and glucose metabolism, inflammation, adipogenesis and mitochondrial biogenesis, has attracted sustained pharmaceutical interest. PPARγ agonists (thiazolidinediones) improve insulin sensitivity but are encumbered by a characteristic class adverse effect of weight gain, fluid retention, and in the case of rosiglitazone, heightened cardiovascular risk [14], ultimately causing multiple agents to be withdrawn or severely restricted. Fibrates, acting primarily through PPARα, improve dyslipidaemia with an acceptable safety record but only modestly affect adiposity. The pan-PPAR agonist lanifibranor, which activates all three isoforms with moderate and balanced affinity [3, 10, 18], improved liver histology and resolution of non-alcoholic steatohepatitis (NASH) in the phase 2b NATIVE trial [5] and is currently the subject of the pivotal NATiV3 phase 3 trial for metabolic dysfunction-associated steatohepatitis (MASH, NCT04849728), with topline results anticipated in the second half of 2026. Yet even lanifibranor, at the doses required for hepatoprotection, produces body weight gain, anemia and peripheral edema, limitations that have so far confined its clinical utility to patients in whom liver disease is the dominant concern.

The conceptual breakthrough that underlies the paper under review was the recognition that the adverse systemic effects of PPAR agonists largely reflect their indiscriminate action across all tissues expressing these ubiquitous nuclear receptors. If one could restrict nuclear receptor engagement to a defined cellular subset, specifically those cells that express GLP-1R or GIPR, the beneficial metabolic effects of PPAR agonism might be preserved while the off-target liabilities are substantially attenuated. This hypothesis was validated in a proof-of-concept study in 2022, in which covalent attachment of the PPARα/γ dual-agonist tesaglitazar to a GLP-1R agonist backbone produced receptor-dependent delivery into GLP-1R-expressing cells and superior metabolic outcomes relative to either component alone, without affecting body weight in GLP-1R-KO mice [15]. In a recent paper published in Nature [13], Liskiewicz, Novikoff, and colleagues from Timo Müller’s laboratory at Helmholtz Munich extends this platform decisively, replacing the dual PPAR agonist with the triple agonist lanifibranor, and upgrading the incretin backbone from a GLP-1R monoagonist to the DPP4-protected GLP-1R/GIPR co-agonist MAR709, thereby engaging five molecular targets simultaneously within a single unimolecular agent. This incremental but logically compelling escalation of pharmacological scope raises questions that extend beyond mouse metabolic physiology: it challenges how the field frames the relationship between targeted drug delivery and nuclear receptor biology, and it forces a candid appraisal of what preclinical efficacy data in rodent obesity models can and cannot predict about therapeutic potential in humans.

The study reports the synthesis and preclinical characterisation of GLP-1–GIP–lanifibranor (GLP-1–GIP–Lani), a unimolecular conjugate in which lanifibranor is covalently tethered to MAR709, a pharmacokinetically optimised GLP-1R/GIPR co-agonist [13]. The authors first demonstrate that the conjugate is functionally indistinguishable from its unconjugated GLP-1–GIP backbone with respect to incretin receptor signalling, cAMP production via GLP-1R and GIPR, and glucose-stimulated insulin secretion in isolated mouse islets. This preserved incretin activity is critical: it establishes that covalent attachment of a bulky small-molecule cargo does not disrupt the ligand–receptor interaction at the plasma membrane.

The targeting specificity is validated in HEK293T cell systems co-transfected with incretin receptors and individual PPAR isoforms. GLP-1–GIP–Lani induced expression of the PPAR target gene PDK4 comparably to free lanifibranor, but only in cells expressing GLP-1R or GIPR, not in double-incretin receptor-negative cells [13]. This receptor-dependency recapitulates the “address label” logic first established in the tesaglitazar conjugate.

In diet-induced obese (DIO) mice, GLP-1–GIP–Lani outperformed GLP-1–GIP co-agonism and semaglutide in terms of body weight reduction, fat mass loss, food intake suppression, and glycaemic improvement across multiple experimental paradigms. At a daily dose of 10 nmol/kg, GLP-1–GIP–Lani produced greater reductions in body weight, fat mass, and blood glucose compared with either GLP-1–GIP or lanifibranor alone. Hyperinsulinaemic-euglycaemic clamp experiments demonstrated significantly superior insulin sensitivity, and suppression of endogenous glucose production was confirmed by pyruvate tolerance testing and decreased hepatic expression of Pcx and Pepck1. Bulk RNA sequencing revealed over 5,400 differentially expressed genes in the liver after GLP-1–GIP–Lani treatment versus vehicle, compared with only 913 for GLP-1–GIP and 57 for Lani, a striking illustration of the synergistic transcriptional effect of combining incretin signalling with targeted nuclear receptor activation. Importantly, GLP-1–GIP–Lani also prevented body weight gain in leptin receptor-deficient db/db mice, a model in which tirzepatide, semaglutide, and retatrutide show only modest efficacy [13].

The safety profile demonstrated in chronic DIO experiments is notable. Unlike free lanifibranor, the conjugate did not increase adipocyte differentiation in preadipocytes, did not cause anaemia, and did not produce fluid retention or renal impairment. Cardiovascular end-points including ejection fraction, fractional shortening, stroke volume, and cardiac output were improved relative to vehicle controls, while blood pressure remained unchanged. These findings are directly attributable to the targeted delivery concept: by restricting lanifibranor exposure to incretin receptor-expressing cells, adverse on-target, off-tissue effects are substantially avoided.

Mechanistic receptor-specificity was confirmed through genetic and pharmacological receptor inhibition. Weight loss and food intake suppression were attenuated in DIO Vglut2/Glp1r-KO mice, diminished in Gipr-KO mice, and absent in double incretin receptor KO (DIR-KO) mice. The blood glucose-lowering effect was abolished by selective PPARδ antagonism using GSK3787, confirming that glycaemic benefit beyond weight loss is mediated specifically through PPARδ signalling. Intriguingly, PPARδ antagonism did not blunt weight loss, suggesting that the weight-reducing arm of the molecule’s action is either PPARδ-independent or operates through alternative PPAR isoforms or non-PPAR mechanisms.

At the central nervous system level, GLP-1–GIP–Lani induced equivalent neuronal FOS activation in the arcuate nucleus, area postrema, and nucleus tractus solitarius relative to GLP-1–GIP [13], consistent with the inability of both molecules to cross the blood-brain barrier. However, fibre-photometric and whole-cell patch recording studies showed that GLP-1–GIP–Lani more robustly activated POMC neurons in the arcuate nucleus [13], suggesting that the incretin receptor-dependent transport of lanifibranor into neurons or supporting cells near fenestrated brain vasculature augments the anorectic signal through nuclear receptor-driven transcriptional reprogramming. The hindbrain showed particularly robust proteomic remodelling by GLP-1–GIP–Lani, with 350 differentially regulated proteins compared with 94 for GLP-1–GIP, an observation that merits considerably deeper investigation.

The principal conceptual contribution of this work is the demonstration that the peptide-drug conjugate platform, pioneered through the GLP-1–tesaglitazar conjugate [15], is generalizable beyond dual nuclear receptor targeting and can be extended to a pan-PPAR agonist of current clinical relevance. The transition from a PPARα/γ dual-agonist to a PPARα/γ/δ triple agonist adds the dimension of PPARδ signaling, a receptor with neuroprotective, anti-inflammatory, and lipid-oxidative roles that has remained therapeutically underexplored, in part because germline PPARδ-KO mice are embryonically lethal, limiting mechanistic work. The observation that PPARδ antagonism selectively ablates the glycaemic but not the weight-loss benefit creates an interesting pharmacological dissection of the molecule’s multi-target actions, and suggests that PPARδ agonism within incretin receptor-expressing pancreatic, hepatic, or adipose cells improves glucose homeostasis through weight-independent mechanisms.

The dose comparison is also intellectually important. Lanifibranor at 30 mg/kg/day (approximately 68.98 µmol/kg) is required to improve liver metabolism as a standalone agent in preclinical models, whereas the conjugate achieves comparable or superior metabolic effects at 10 nmol/kg, a 6,898-fold lower dose. This extraordinary dose compression is the direct result of cellular concentration: receptor-mediated endocytosis of the conjugate accumulates lanifibranor within incretin receptor-expressing cells far beyond what circulating concentrations of free drug could achieve at pharmacologically safe doses. This principle of receptor-guided drug concentration is well-established for antibody-drug conjugates in oncology but has been less systematically explored for metabolic indications, and the present study adds compelling preclinical evidence for its feasibility.

Several important mechanistic questions remain unresolved. First, delineating which tissue compartment is primarily responsible for the enhanced efficacy of GLP-1–GIP–Lani relative to GLP-1–GIP is exceptionally challenging. The liver and skeletal muscle do not express GIPR or GLP-1R at significant levels, and the authors correctly attribute the chronic transcriptional reprogramming of these tissues to indirect effects of improved systemic metabolism, weight loss, reduced adiposity, and enhanced insulin sensitivity. Yet whether a small fraction of conjugate enters these tissues through alternative uptake mechanisms, or whether circulating lanifibranor released from the conjugate after endosomal processing contributes to effects in PPAR-expressing non-incretin tissues, cannot be definitively excluded. These uncertainties are linked to the absence of a detailed pharmacokinetic and biodistribution comparison between conjugated and unconjugated lanifibranor, which would be required to disentangle true targeting effects from differences in exposure dynamics.

Second, the central nervous system data, while intriguing, raise more questions than they answer. The blood-brain barrier impermeability of the molecule is confirmed in vitro, yet robust proteomic changes are observed in both brainstem and hypothalamus after acute administration. The most parsimonious explanation is that the molecule acts on neurons adjacent to fenestrated vasculature in circumventricular organs, a site of known GLP-1R expression, and that lanifibranor is delivered specifically to these neurons. However, as the authors themselves acknowledge, the available antibody tools for detecting GIPR in the brain are insufficient for immunohistochemical localisation of drug uptake, and the relative contributions of brainstem versus hypothalamic mechanisms to weight loss remain unclear. In addition, it should be determined whether local CNS effects reflect receptor-mediated targeting per se or altered brain exposure driven by conjugation-dependent pharmacokinetic behavior. The enhanced POMC neuronal activation demonstrated by fibre photometry is a compelling finding, and the link between GLP-1R agonism and hypothalamic POMC circuit activation [25] provides a mechanistic framework, but whether lanifibranor augments POMC firing through direct nuclear receptor action in POMC neurons or through synaptic modifications induced by PPARδ activation in local interneurons is not resolved.

Third, the body weight-reducing effects of the molecule are partially body weight-independent, as demonstrated by the weight-matched control experiment. GLP-1–GIP–Lani improved glucose tolerance beyond what food restriction to the same degree of weight loss achieved with GLP-1–GIP alone. This is consistent with the pharmacological separation of the glycaemic and weight-loss arms of the molecule, but the mechanisms of this body weight-independent glycemic benefit, whether mediated by PPARδ in pancreatic beta cells, by hepatic PPAR activation in incretin receptor-expressing hepatocytes, or by central mechanisms, remain to be established.

Translational considerations

The central translational challenge confronting this molecule is that rodent obesity models systematically amplify efficacy signals relative to the clinical setting. Diet-induced obese C57BL/6J mice at 50–60 g body weight represent a different pathophysiological substrate from the human patient with established obesity, insulin resistance, dyslipidemia, and likely hepatic steatosis who has frequently already been treated with first- or second-generation incretin agents. From a translational perspective, rodent obesity models often tend to amplify efficacy signals compared to humans, and whether the approximately 25% body weight reduction observed in DIO mice will translate to clinically meaningful weight loss beyond that achievable with tirzepatide or next-generation incretin triple agonists remains entirely open.

The clinical development trajectory of lanifibranor itself also warrants attention. In phase 2b trials, lanifibranor at 1200 mg daily demonstrated resolution of NASH in 49% of treated patients versus 22% on placebo, supporting its advancement to the NATiV3 phase 3 trial [5]. Topline results of NATiV3 are expected in the second half of 2026. If the standalone pan-PPAR agonist achieves regulatory approval for MASH, the case for a conjugate that delivers it at doses thousands of times lower with improved tolerability becomes substantially more compelling. However, the regulatory path for a complex peptide-small molecule conjugate is far longer and considerably less defined than for a conventional small molecule, and first-in-human studies would need to address pharmacokinetics of the intact conjugate, the fate of the covalent linker, and the bioavailability profile of released lanifibranor in compartments where incretin receptor expression is low. Importantly, a key unresolved issue is whether the observed in vivo superiority of the conjugate over unconjugated components is influenced not only by receptor-targeted delivery but also by pharmacokinetic differences, including altered systemic stability, tissue distribution, and intracellular drug release kinetics, which were not comprehensively dissected.

The safety profile data from chronic DIO mouse experiments is reassuring in important respects: no anaemia, no fluid retention, no histological organ pathology, and improved rather than worsened cardiac function. However, the absence of anaemia in mice should not be uncritically extrapolated to humans: the haematopoietic effects of PPARγ activation differ substantially between species, and anaemia was a clinically meaningful adverse event in the lanifibranor NATIVE trial, occurring more frequently in treated patients. Whether receptor-targeted delivery completely eliminates haematological risk in humans is a question that can only be answered in clinical pharmacology studies.

Importantly, the issue of sex-specificity also deserves mention. All in vivo experiments were conducted in male mice. There are well-documented sex differences in incretin receptor expression patterns, PPAR signalling, adipose tissue distribution, and the response to GLP-1R agonists in both rodent models and humans [2, 7, 24]. Therefore, the generalizability of these findings to female subjects, who constitute approximately half of the target clinical population for anti-obesity pharmacotherapy, cannot be assumed.

Positioning within the broader pharmacological landscape

The competitive landscape for advanced obesity pharmacotherapy is extraordinarily dynamic. Tirzepatide leads the commercial market, achieving body weight reductions of approximately 6.2 to 12.9 kg in phase 3 diabetes trials, with substantially greater weight loss in obese non-diabetic populations. Triple agonists extending co-agonism to glucagon receptors, most notably retatrutide (GLP-1R/GIPR/GCGR), are in late clinical development [4, 8, 9, 16, 17, 19]. Oral GLP-1R agonists including orforglipron [11, 21–23] and danuglipron [12, 20] are expanding the route-of-administration landscape. Against this backdrop, the question is not whether GLP-1–GIP–Lani is superior to existing incretin drugs in mice, it is, convincingly, but whether a complex peptide-PPAR conjugate offers a sufficiently differentiated clinical benefit profile, particularly for hepatic and cardiometabolic co-morbidities where PPAR agonism adds mechanistic dimensions beyond appetite suppression and caloric restriction.

The db/db mouse data may be the most clinically significant observation in the paper. GLP-1–GIP–Lani fully prevented body weight gain in obesity-prone leptin receptor-deficient db/db mice, a model in which tirzepatide, semaglutide, and retatrutide showed only modest ability to prevent the establishment of obesity. If this efficacy in a leptin-resistant model translates to human populations with severe obesity or those who respond suboptimally to incretin monotherapy, it would represent a genuine unmet clinical need that the current generation of agents cannot adequately address.

The role of PPARδ in the therapeutic profile of the molecule is particularly interesting. The selective PPARδ antagonism experiment demonstrating loss of glycaemic benefit without attenuation of weight loss constitutes a clean pharmacological dissection rarely achievable in complex in vivo systems. PPARδ’s role in brain metabolism and energy homeostasis is poorly understood, and the observation that GLP-1–GIP–Lani modulates the hindbrain proteome more profoundly than GLP-1–GIP opens a research avenue that may prove relevant not only for metabolic disease but potentially for neurodegenerative conditions in which PPARδ neuroprotection has been implicated.

Conclusion

The path from these preclinical observations to a clinical therapeutic is long, and the translational uncertainties are substantial, as they are for all first-in-class agents at this stage of development. What the study clearly establishes is that the receptor-targeted conjugate approach is a viable and increasingly sophisticated pharmacological strategy. Whether GLP-1–GIP–Lani itself, or a structurally refined successor, ultimately enters clinical development, the intellectual framework it embodies, using peptide receptor biology as a precision delivery mechanism for nuclear-acting small molecules, will influence the design of next-generation metabolic drugs for years to come.

Author contributions

GS wrote the manuscript.

Funding

Prof. Gaetano Santulli, MD, PhD, FAHA was supported in part by the National Institutes of Health (NIH): National Heart, Lung, and Blood Institute (NHLBI: R01-HL146691, R01-HL164772, R01-HL159062), National Institute of Diabetes and Digestive and Kidney Diseases (NIDDK: R01-DK123259, R01-DK033823), and by the American Heart Association (AHA, 24IPA1268813).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics and consent to participate

Not applicable.

Competing interests

Prof. Gaetano Santulli is the Editor-in-Chief of Cardiovascular Diabetology - Endocrinology Reports. The Editorial Office of the Journal assigned this manuscript to another Editor, who independently supervised the peer-review process.