This Week in Peptide Research: GLP-1 Mechanism Studies, LL-37 Engineering Advances, and Kisspeptin's Reach Beyond Reproduction

Peptide Science Daily reviewed new PubMed listings for its tracked compounds from September 6 through September 13, 2026. It was a quiet week with no new phase 3 trial readout and no single dominant story. The literature clustered into four threads: mechanistic and clinical questions about approved GLP-1 drugs, two protein-engineering advances for the antimicrobial peptide LL-37, three papers examining kisspeptin's effects on cardiac development, gonadotropin release, and brain receptor selectivity, and a large real-world safety analysis of desmopressin use around childbirth.

Three items this week concerned approved GLP-1 receptor agonists. In the American Journal of Physiology-Endocrinology and Metabolism, Canadian researchers reanalyzed a clinical trial in which 90 adults with type 2 diabetes (mean A1c 6.5%) were randomized to eight weeks of basal insulin glargine alone, glargine plus mealtime insulin lispro, or glargine plus exenatide, an approved GLP-1 receptor agonist marketed as Byetta and Bydureon. A1c fell progressively across the three groups (6.2%, 5.9%, and 5.8% respectively; p=0.007), and the ratio of fasting proinsulin to C-peptide, a marker of beta-cell stress, fell in parallel (p=0.01), while none of four separate measures of beta-cell compensation differed between groups. Statistical mediation analysis identified improved proinsulin processing, not better beta-cell compensation, as the factor explaining the extra glucose-lowering from lispro and exenatide compared with glargine alone. The authors said this suggests glucose control in type 2 diabetes can improve through reduced beta-cell stress without the cells actually compensating better for insulin resistance. Separately, in the European Journal of Pharmacology, Chinese researchers treated db/db mice, a model of metabolic dysfunction-associated steatotic liver disease, with dulaglutide, another approved GLP-1 receptor agonist marketed as Trulicity, for 10 weeks. Treated mice had lower liver enzymes, reduced markers of ferroptosis (an iron-dependent form of cell death), less oxidative stress and iron accumulation, and reduced inflammation and fibrosis markers, alongside broad changes to the liver's lipid profile. The authors described the findings as supporting further study of dulaglutide for this liver disease, based on the mouse model rather than clinical data. A review in the American Journal of Nephrology examined orforglipron, the first oral non-peptide GLP-1 receptor agonist, approved by the FDA in April 2026 as Foundayo for chronic weight management but not yet authorized by the TGA or EMA. The authors noted that orforglipron's phase 3 trials showed stable kidney function and no signal of kidney toxicity, but said those trials were not designed to measure kidney-specific outcomes such as albuminuria or progression to kidney failure, and often excluded patients with advanced chronic kidney disease. They said a dedicated trial examining kidney and cardiovascular outcomes in patients with both atherosclerosis and chronic kidney disease is underway, and called for further CKD-specific trials before drawing conclusions about renal benefit.

Two papers advanced the biochemistry of LL-37, an investigational human antimicrobial peptide with no FDA, EMA, or TGA approval. A team at Hokkaido University, writing in the International Journal of Biological Macromolecules, reported a new method for producing LL-37 in the lab: fusing it to calmodulin, a protein that binds sticky, amphiphilic peptide surfaces, then cleaving the tag away with an enzyme. The calmodulin-tagged version was far more soluble than LL-37 fused to conventional tags, avoided the aggregation that has long hampered efficient recombinant production of the peptide, and required no detergents. NMR spectroscopy and molecular dynamics simulations showed the calmodulin tag physically covers LL-37's aggregation-prone regions, and the same strategy also worked for the mouse version of the peptide, mCRAMP. Separately, a Polish and Italian research team writing in Dalton Transactions resolved, by their account for the first time, how LL-37 binds copper and zinc ions, comparing the full peptide with a shorter C-terminal fragment and an all-mirror-image (all-D) version of that fragment. LL-37 retained its helical shape and showed the strongest antimicrobial and antibiofilm activity of the three, an effect generally enhanced by copper or zinc binding, while the shorter fragments lost potency against bacteria but kept antifungal and antibiofilm activity. The authors highlighted the all-D fragment, expected to resist enzymatic breakdown in the body, as a promising scaffold for future antibiofilm peptide development. A separate, computational-only study in the Journal of Molecular Modeling designed a new hybrid peptide, PA-Hyb1, by combining fragments of LL-37 with two other antimicrobial peptides, cecropin A and magainin 2, to target a drug-resistance efflux pump in Pseudomonas aeruginosa. Molecular dynamics simulations estimated favorable binding compared with the antibiotic levofloxacin, but the single author stressed that PA-Hyb1 has not been synthesized or tested against live bacteria, and flagged the simplified membrane model and a preliminary host-safety-only screen as limitations on any translational claim.

Three papers this week concerned kisspeptin, an investigational peptide with no FDA, EMA, or TGA approval that WADA added to its Prohibited List in 2024 as a testosterone-stimulating substance, banned in male athletes only. Researchers in Saint Petersburg, writing in In Silico Pharmacology, used AlphaFold2 structure prediction and molecular dynamics simulations to test which of five receptors linked to impulsive and compulsive behavior bind kisspeptin-10, the peptide's active core fragment, most strongly. Kisspeptin-10 bound its own receptor, KISS1R, most tightly, followed closely by neuropeptide FF receptor 1, with substantially weaker binding to dopamine and serotonin receptors. The authors said this computational result supports the idea that kisspeptin's previously reported effects on impulsive and compulsive behavior in rodents work through its own receptor rather than by directly acting on monoamine systems, though they said experimental confirmation is still needed. In Cardiovascular Toxicology, Brazilian researchers induced hypothyroidism in pregnant rats using propylthiouracil and treated some with kisspeptin-10 starting on gestational day 8. Maternal hypothyroidism reduced fetal and postnatal heart mass, impaired heart-muscle cell proliferation, and disrupted markers of cell death and blood vessel growth in offspring hearts. Kisspeptin-10 partially reversed these changes, restoring postnatal weight gain and cardiomyocyte proliferation and improving several disrupted markers, though it also raised oxidative and cell-stress markers in fetal hearts even as it lowered one such marker after birth. The authors described the effect as a partial rescue, with kisspeptin improving some outcomes while leaving a mixed safety signal in the fetal period. Separately, researchers at Texas A&M University and the University of Sao Paulo, writing in Biology of Reproduction, compared kisspeptin with gonadorelin, an approved synthetic GnRH historically marketed as Factrel and Lutrepulse, in ovariectomized cattle. Hourly injections of either peptide for eight hours induced frequent luteinizing hormone (LH) pulses in most animals (75% of kisspeptin injections and 87.5% of gonadorelin injections produced a detectable pulse), but this high-frequency dosing did not amplify the LH surge triggered by a subsequent gonadorelin challenge. The surge was smaller in Brahman cattle than in Hereford cattle regardless of treatment, which the authors attributed to breed differences in pituitary responsiveness rather than to the peptides themselves.

In Haemophilia, researchers analyzed a US claims database covering nearly 1.5 million delivery encounters between 2011 and 2021, of which 1,514 (0.10%) involved a recorded diagnosis of von Willebrand disease, the most common inherited bleeding disorder. Women with von Willebrand disease had higher odds of postpartum hemorrhage (odds ratio 1.70), a longer-than-median hospital stay (OR 1.58), and a 30-day emergency visit or readmission (OR 1.20) than women without the diagnosis, though not higher mortality. In an exploratory analysis, claims for clotting-factor concentrate were associated with higher odds of hemorrhage than no recorded treatment (OR 3.17), and claims for desmopressin, an approved drug used to raise clotting-factor levels in mild von Willebrand disease and hemophilia A, showed a similar but statistically non-significant pattern (OR 3.32, 95% CI 0.93 to 11.82). The authors cautioned that women who received treatment were likely already judged to be at higher bleeding risk before delivery, so the finding reflects confounding by indication rather than evidence that treatment caused bleeding, and said prospective studies with laboratory and dosing data are needed to clarify the picture.

Finally, a materials-science paper in Advanced Healthcare Materials, from a Georgia Institute of Technology team, tested a biodegradable microneedle-patch design intended to replace older ceramic or metal versions. Laser-fabricated "STAR particles," made from water-soluble, enzyme-degradable, or acid-hydrolyzable polymers, punctured pig skin ex vivo and increased intradermal delivery of three model drugs, including copper tripeptide-1, also known as GHK-Cu, a peptide with unclear regulatory status used topically as a cosmetic ingredient, by up to 37-fold depending on the particle material and drug. The study tested only whether the new delivery particles could get more of each drug into skin, not GHK-Cu's biological effects once delivered.

This week's items span one clinical-trial reanalysis, three rodent or cattle studies, two biochemical laboratory papers, one purely computational peptide-design study, one large real-world claims analysis, and one materials-science drug-delivery study. None is a new randomized controlled trial in humans testing a peptide for a novel indication, and none changes the approved use of any drug discussed. Dulaglutide, exenatide, orforglipron, gonadorelin, and desmopressin are approved drugs, though orforglipron's approval is limited to the FDA and to weight management, and several findings here, including dulaglutide's liver effects and kisspeptin's cardiac and behavioral effects, come from animal or computational models rather than patients. LL-37 remains investigational with no FDA, EMA, or TGA approval, and kisspeptin is investigational and, under WADA's 2024 Prohibited List, banned in male athletes. As with all research covered here, these are findings a given study reported, not established clinical fact or treatment guidance.