Highlights
Short-term duodenal infusion of recombinant RORDEP1 (r-RORDEP1) was safe and well tolerated in 17 healthy, normal-weight men.
Compared with placebo, r-RORDEP1 produced an early rise in plasma glucagon-like peptide-1, insulin, and C-peptide, while lowering gastric inhibitory polypeptide and glucose after a standardized liquid mixed meal.
The intervention also increased the Matsuda index of insulin sensitivity, suggesting an acute improvement in whole-body insulin sensitivity.
These first-in-human findings are directionally consistent with preclinical work implicating RUMTOR-derived peptides as microbiome-linked modulators of host metabolism.
Background and Clinical Context
The idea that gut microbes shape human metabolism has moved from broad association studies toward mechanistic interrogation. Most translational work has focused on microbial metabolites such as short-chain fatty acids, bile acid derivatives, and tryptophan catabolites. Far fewer studies have evaluated microbiome-derived polypeptides as candidate endocrine or paracrine mediators of host metabolic regulation. The current trial addresses that gap by testing recombinant RORDEP1, one of two RUMTOR-derived peptides synthesized by specific strains of the human commensal bacterium Ruminococcus torques.
According to the investigators, preclinical studies suggested that RORDEPs lower blood glucose through effects on incretin secretion and hepatic insulin sensitivity. That mechanistic hypothesis is biologically plausible. The proximal small intestine is a major site of nutrient sensing and gut hormone release, and signals arising there can rapidly influence insulin secretion, glucagon-like peptide-1 (GLP-1) biology, gastric inhibitory polypeptide (GIP; also called glucose-dependent insulinotropic polypeptide), and hepatic glucose handling. If a microbial peptide can reproduce part of that signaling axis, it could represent a new class of host-microbiome therapeutics.
From a clinical standpoint, the unmet need is clear. Type 2 diabetes and related metabolic disorders remain highly prevalent, and despite major advances with GLP-1 receptor agonists, dual incretin agonists, and sodium-glucose cotransporter-2 inhibitors, there is continued interest in therapies that target upstream physiology, including gut endocrine signaling and insulin sensitivity. A safe microbial peptide capable of modulating incretin biology could eventually expand the therapeutic toolbox, although any translation from healthy volunteers to metabolic disease remains highly preliminary.
This trial should therefore be viewed primarily as an early human pharmacology and safety study. Its importance lies less in immediate clinical application and more in whether it provides a credible signal that the RORDEP platform has physiological activity in humans.
Study Design
Trial overview
This was a randomized, placebo-controlled, crossover clinical trial conducted at Gentofte Hospital, Denmark. Seventeen healthy men aged 18 to 35 years with normal body weight were enrolled, and all completed the trial. Participants were randomized through block randomization to receive either r-RORDEP1 or placebo as the initial intervention. Both participants and investigators were blinded to treatment assignment.
Population
The study population was deliberately narrow to minimize biological variability in this first-in-human setting. Key exclusion criteria included use of any medication, antibiotic use within the preceding 3 months, lactose intolerance, smoking, alcohol or drug abuse, and use of probiotics or creatine during the study period. The restriction to healthy, normal-weight young men improves internal validity for physiological signal detection, but it also limits generalizability to women, older adults, people with obesity, and patients with dysglycemia.
Intervention and comparator
After a standardized liquid mixed meal, participants received r-RORDEP1 through a naso-duodenal tube. The dosing regimen consisted of an initial bolus of 0.0108 mg/kg body weight followed by a continuous infusion of 0.25 micrograms/kg/min for 170 minutes. Placebo was administered in the comparator condition. Direct duodenal delivery was presumably chosen to maximize local exposure at the presumed site of action and avoid uncertainties regarding oral stability, gastric degradation, and variable intestinal delivery.
Endpoints
The primary endpoints were changes in plasma incretin concentrations and peptide YY. Secondary endpoints included safety and tolerability, along with changes in plasma insulin, C-peptide, and glucose. The investigators also assessed whole-body insulin sensitivity using the Matsuda index, a surrogate measure derived from glucose and insulin dynamics during meal or oral glucose tolerance testing.
Key Findings
Safety and tolerability
All 17 participants completed the study, which is reassuring for feasibility and short-term tolerability. Duodenal infusion of r-RORDEP1 was reported to be well tolerated, with no changes in biochemical measures of hematological, liver, or renal function. In an early-phase translational study, this is an important result. Microbiome-derived peptides may raise theoretical concerns about immunogenicity, off-target hormonal effects, or unexpected gastrointestinal intolerance. While this small trial cannot exclude uncommon adverse events, it provides an initial safety signal that supports continued development.
Equally important is what the study does not show. The exposure was acute, the sample was small, and participants were metabolically healthy. These conditions are appropriate for first-in-human exploration, but they do not address chronic dosing safety, immunogenicity over time, effects in patients with comorbidities, or interactions with background medications.
Primary hormonal responses
The clearest physiological signal appeared early after administration of the r-RORDEP1 bolus. Compared with placebo, r-RORDEP1 induced an early rise, observed at 15 or 30 minutes, in plasma GLP-1, insulin, and C-peptide, with false discovery rate-adjusted q values of 0.001, 0.001, and 0.003, respectively. These data indicate a statistically robust early endocrine response despite the small sample size.
The GLP-1 rise is particularly notable because GLP-1 is central to postprandial glucose regulation through augmentation of glucose-dependent insulin secretion, slowing of gastric emptying, and suppression of glucagon under many conditions. In the present study design, gastric emptying is less likely to explain the result because the peptide was infused directly into the duodenum after meal ingestion. The observed increase in insulin and C-peptide is consistent with enhanced beta cell stimulation, either mediated by GLP-1, by other enteroinsular pathways, or both.
C-peptide deserves emphasis because it is co-secreted with endogenous insulin but is not extracted by the liver to the same degree. The concordant increase in insulin and C-peptide therefore supports increased endogenous insulin secretion rather than a spurious analytical or distributional effect.
At the same time, r-RORDEP1 lowered plasma GIP early after dosing, with a q value of 0.02. This is mechanistically intriguing. GLP-1 and GIP are both incretin hormones, but they arise from distinct enteroendocrine cell populations and can have different spatial triggers along the gut. An increase in GLP-1 with a concomitant decrease in GIP suggests that RORDEP1 does not simply amplify generalized gut hormone secretion. Instead, it may act through a selective nutrient-sensing or receptor-mediated pathway, potentially with regional specificity in the proximal intestine.
The abstract does not report a significant effect on peptide YY, one of the predefined primary outcome measures. Without full data, one should avoid overinterpreting this omission, but it suggests that the hormonal profile may be selective rather than broad-based.
Glucose and insulin sensitivity outcomes
r-RORDEP1 also reduced plasma glucose early after administration compared with placebo, with a q value of 0.006. In healthy participants receiving a mixed meal, postprandial glucose excursions are usually modest, so demonstrating a significant acute reduction in this setting strengthens the argument that the peptide is biologically active in humans.
The intervention increased the Matsuda index of insulin sensitivity, with a p value of 0.049. This finding aligns with the preclinical hypothesis that RORDEPs improve insulin sensitivity, including hepatic insulin sensitivity. However, caution is required in interpretation. The Matsuda index is a useful and widely used surrogate but does not isolate hepatic from peripheral insulin sensitivity and is not equivalent to gold-standard clamp methodology. In addition, a borderline p value in a 17-person crossover study should be seen as hypothesis-generating rather than definitive proof of improved insulin action.
Still, the overall metabolic pattern is coherent: increased GLP-1, increased insulin secretion, lower glucose, and a favorable shift in a surrogate insulin sensitivity measure. Coherence across related endpoints matters in early-phase physiology studies, particularly when the direction of effect matches prior animal work.
Interpreting the timing of effect
The timing is one of the most informative aspects of the study. The primary endocrine changes occurred within 15 to 30 minutes after the bolus. That rapid onset supports a direct intestinal signaling mechanism rather than a delayed secondary effect. It also suggests that dose timing and delivery route could be crucial variables in later development. If the biological effect is tightly linked to luminal exposure in the proximal intestine, oral formulations may require specialized delivery technology, or developers may need to identify the receptor and design more stable agonists.
Mechanistic and Translational Interpretation
The study strengthens the concept that specific bacterial gene products, not just broad taxonomic shifts in the microbiome, can influence host metabolism in measurable ways. This is a critical conceptual step. Much of microbiome science has struggled with reproducibility because associations between taxa and disease are highly context-dependent. A defined recombinant peptide offers a more tractable translational path than whole-microbiome manipulation because it can be dosed, manufactured, and studied with standard pharmacological tools.
The increase in GLP-1 alongside a decrease in GIP may point toward selective activation of specific enteroendocrine pathways, perhaps involving L cells more than K cells, or a broader network effect within the proximal small intestine. Whether the effect is receptor-mediated on enteroendocrine cells, neural, paracrine, or dependent on interactions with the mucosal barrier is unknown from this trial. Another unresolved issue is whether the glucose-lowering signal is driven mainly by enhanced insulin secretion, improved insulin sensitivity, suppressed endogenous glucose production, or some combination of these mechanisms.
Preclinical studies apparently suggested improved hepatic insulin sensitivity. The present trial does not directly measure hepatic glucose production, so that aspect remains inferential. Future studies incorporating tracer methodologies, hyperinsulinemic-euglycemic clamps, or mixed-meal tests with isotope labeling would be valuable to localize the site of metabolic action.
Strengths of the Trial
The study has several notable strengths for an early translational investigation. First, the randomized, placebo-controlled, crossover design is efficient and appropriate for detecting acute physiological effects, because each participant serves as his own control. This reduces inter-individual variability in hormone and glucose responses. Second, both participants and investigators were blinded, limiting expectation bias. Third, direct duodenal infusion helps isolate biological activity by ensuring reliable exposure at the intended intestinal site. Fourth, the use of adjusted q values for several hormonal endpoints suggests attention to multiplicity, an important consideration in endocrine profiling.
Finally, the coherence between human findings and prior rodent data adds credibility. Concordance across species does not prove clinical utility, but it increases confidence that the observed human signals are not random findings.
Limitations and Caveats
The main limitation is sample size. Seventeen participants are adequate for detecting large within-subject physiological signals but insufficient for robust safety characterization or nuanced subgroup analysis. The confidence one can place in smaller effect sizes is limited.
The second major limitation is external validity. All participants were healthy, young, normal-weight men. The metabolic milieu in obesity, insulin resistance, or type 2 diabetes is substantially different, particularly with respect to incretin responsiveness, beta cell reserve, and hepatic glucose output. It cannot be assumed that the same magnitude or even direction of effect would occur in those populations.
Third, the intervention was acute and delivered by naso-duodenal tube, which is a research tool rather than a practical therapeutic route. A positive signal here does not guarantee that oral, subcutaneous, or other clinically scalable formulations will reproduce the same biology.
Fourth, the abstract does not provide effect sizes, confidence intervals, or full time-course curves. Statistical significance is informative, but translational decision-making also depends on magnitude, duration, and inter-individual variability of response. Without those details, the clinical relevance of the glucose lowering and insulin sensitivity signal remains uncertain.
Fifth, the Matsuda index, while useful, is an indirect measure. Claims regarding insulin sensitivity should therefore remain restrained until confirmed with more definitive metabolic phenotyping.
Finally, because the study focuses on short-term safety labs, it does not address immunogenicity, anti-drug antibodies, tachyphylaxis, effects on appetite or gastrointestinal motility with repeated exposure, or potential off-target effects on other endocrine pathways.
Clinical and Research Implications
This trial does not change practice, but it does establish a credible early signal for a novel microbiome-derived peptide platform. For clinicians and translational researchers, the study is important because it shows that a defined bacterial peptide can alter human postprandial hormone responses and glucose handling under controlled conditions. That is a meaningful advance in the effort to move microbiome science from descriptive correlation to testable intervention.
The next steps are clear. Dose-ranging studies are needed to define the exposure-response relationship. Longer studies must examine repeated dosing safety and durability of effect. Trials in people with overweight, obesity, prediabetes, and type 2 diabetes will be essential to determine whether the observed physiology translates into clinically useful glycemic benefit. Mechanistic studies should dissect whether the dominant site of action is enteroendocrine, neural, hepatic, or a combination. And from a drug-development perspective, formulation science will be pivotal if the mechanism requires localized delivery to the proximal small intestine.
It will also be important to determine whether RORDEP1 has additive, synergistic, or redundant effects relative to established incretin-based therapies. If its primary action is upstream stimulation of endogenous GLP-1, its value may differ from that of direct GLP-1 receptor agonism. Conversely, if it also improves hepatic insulin sensitivity through a distinct pathway, that could make it especially interesting as a complementary agent.
Conclusion
In this randomized, placebo-controlled crossover trial, short-term duodenal infusion of recombinant RORDEP1 in healthy men was safe, well tolerated, and associated with rapid metabolic effects after a liquid mixed meal. The peptide increased GLP-1, insulin, and C-peptide; lowered GIP and glucose; and modestly improved a surrogate measure of whole-body insulin sensitivity. These findings are biologically coherent and consistent with prior rodent data, supporting the concept that RUMTOR-derived peptides may influence host metabolism.
At the same time, this remains an early proof-of-physiology study rather than a therapeutic efficacy trial. The small sample, male-only healthy cohort, acute exposure, and nonclinical delivery route all limit immediate extrapolation. Even so, the study provides a strong rationale for deeper mechanistic work and for carefully staged clinical development in metabolically impaired populations.
Funding and Trial Registration
Trial registration: ClinicalTrials.gov NCT06923839.
Funding: EFSD/Lilly European Diabetes Research Programme 2021, RUCILP F-19235-01-64 – NNF21SA0070428 grant and NNF23SA0084103 grant, the latter two from the Novo Nordisk Foundation.
References
Gæde J, Fan Y, Lyu L, Gasbjerg LS, Rossing P, Hartmann B, Holst JJ, Lund AB, Knop FK, Pedersen O. A randomised clinical trial testing the safety of and metabolic responses to short-term duodenal infusion of recombinant RORDEP1 in healthy men. Diabetologia. 2026-05-23. PMID: 42177350.
Holst JJ. The physiology of glucagon-like peptide 1. Physiol Rev. 2007;87(4):1409-1439.
Nauck MA, Meier JJ. Incretin hormones: Their role in health and disease. Diabetes Obes Metab. 2018;20 Suppl 1:5-21.
Matsuda M, DeFronzo RA. Insulin sensitivity indices obtained from oral glucose tolerance testing: comparison with the euglycemic insulin clamp. Diabetes Care. 1999;22(9):1462-1470.
Muller TD, Finan B, Bloom SR, D’Alessio D, Drucker DJ, Flatt PR, et al. Glucagon-like peptide 1 (GLP-1). Mol Metab. 2019;30:72-130.
