C-peptide: the insulin co-release signal tied to diabetes nerve & vessel health
A natural peptide the pancreas releases alongside insulin; acts on small blood vessels and nerves damaged in diabetes, and is used as a lab measure of insulin production, not an approved drug.
A researcher, an agent, or an algorithm wrote down the sequence and picked a target to hit.
An AI model like OpenFold3 or AlphaFold built a 3D structure and scored how well it fits the binding site.
A second contributor repeated the computation on their own hardware and the scores matched.
A chemistry service or a researcher ordered the sequence, it was manufactured, and mass spectrometry confirmed the right molecule was produced.
A binding or activity measurement confirmed that it actually does what the computer predicted — or didn't.
What this is
C-peptide (connecting peptide) is a 31-amino-acid chain that the pancreas releases every time it makes insulin. When the pancreatic beta cell processes its insulin precursor, proinsulin, into mature insulin, C-peptide is cut out and co-secreted in equimolar amounts into the bloodstream. For decades after its discovery it was dismissed as metabolic "waste," but it is now recognized as a biologically active signaling molecule in its own right, with effects on the small blood vessels and nerves that are commonly damaged in diabetes (Yosten and colleagues, Am J Physiol Endocrinol Metab, 2014). The sequence stored here (EAEDLQVGQVELGGGPGAGSLQPLALEGSLQ) is the mature, free C-peptide after cleavage; it does not include the dibasic cleavage-site residues that flank it in proinsulin.
History
C-peptide's existence became clear in 1967, when Donald Steiner and Philip Oyer at the University of Chicago demonstrated that insulin is not assembled from two separate protein chains but is instead cleaved from a single-chain precursor they named proinsulin. The interior segment removed during that processing — linking the insulin B-chain to the A-chain — was termed the "connecting peptide," or C-peptide (Brandenburg, Exp Diabetes Res, 2008). The discovery was the first unambiguous demonstration of posttranslational processing of a polypeptide precursor, establishing a paradigm for how many other peptide hormones and neuropeptides are biosynthesized. The amino-acid sequence was elucidated from porcine proinsulin by Chance and colleagues in 1968, and Rubenstein and colleagues developed a differential immunoassay for detecting C-peptide in human serum in 1969 — the basis of its clinical use as a beta-cell function biomarker (Brandenburg 2008). Scientific literature on C-peptide expanded from minimal publications in the 1960s to over 300 papers annually by 1988. Interest in its own biological activity rekindled in the early 1990s when researchers began administering it to people with type 1 diabetes who, lacking functional beta cells, were completely C-peptide-deficient.
What it does
C-peptide has two distinct roles: as a biomarker of beta-cell health, and as an active signaling molecule that appears to protect the small blood vessels and peripheral nerves from diabetes-related damage.
As a biomarker, circulating C-peptide concentration reflects how much insulin the pancreas is making, because it is produced in a one-to-one ratio with insulin but is not extracted by the liver the way insulin is. This makes it the preferred clinical measure of residual beta-cell function in people who are already taking exogenous insulin (where an insulin assay cannot distinguish injected from endogenous hormone). Retained beta-cell function — evidenced by measurable C-peptide — correlates with better glycemic control, fewer hypoglycemic episodes, and reduced rates of retinopathy and nephropathy in type 1 diabetes (Yosten and colleagues 2014).
As a signaling molecule, physiological concentrations of C-peptide (fasting plasma levels in healthy individuals: 0.3–0.6 nM; postprandial: 1–3 nM) activate intracellular pathways that stimulate Na⁺/K⁺-ATPase activity in peripheral nerve tissue and increase endothelial nitric oxide synthase (eNOS) expression and nitric oxide (NO) availability in vascular endothelium (Yosten and colleagues 2014). Both effects are impaired in diabetic microangiopathy. C-peptide also reduces reactive oxygen species production and inhibits proinflammatory cytokine and adhesion molecule expression under high-glucose conditions (Chen and colleagues, Front Endocrinol, 2023).
Evidence
- Human: Smaller controlled trials (46 patients, 3 months; 161 patients, 6 months) found that C-peptide replacement improved sensory nerve conduction velocity and neuropathy impairment scores in type 1 diabetic patients with peripheral neuropathy (Ekberg & Johansson, Exp Diabetes Res, 2008). A subsequent 12-month Phase IIb trial by Wahren and colleagues of a pegylated long-acting C-peptide analogue in 250 patients (Diabetes Care, 2016) showed improvement in sural nerve conduction velocity and vibration perception, but the primary endpoint did not reach statistical significance over placebo; both active and placebo groups improved comparably. Observational data from the Diabetes Control and Complications Trial indicate that residual endogenous C-peptide secretion correlates with lower rates of microvascular complications in type 1 diabetes (Yosten and colleagues 2014). A large cross-sectional study (n = 4,793) found that higher fasting C-peptide in type 2 diabetes was protective against diabetic retinopathy (odds ratio 0.73; Chen and colleagues 2023). The Veterans Affairs Diabetes Trial found each 1 pmol/mL increase in baseline C-peptide associated with a 67.2% decrease in diabetic retinopathy prevalence (Chen and colleagues 2023).
- Animal: C-peptide replacement in streptozotocin-diabetic and BB/Wor rat models has prevented nerve conduction velocity deficits, reduced glomerular hyperfiltration and microalbuminuria, and improved endoneurial blood flow. In diabetic rodents, C-peptide achieved approximately 80% correction of sensory and 60% correction of motor nerve conduction velocity deficits (Yosten and colleagues 2014). Ultra-long-lasting C-peptide delivery in diabetic mice maintained physiological concentration ranges and improved retinopathy-related neovascularization outcomes in a 2023 study (Chen and colleagues 2023).
- In vitro: C-peptide binds specifically to cell membrane preparations in nanomolar concentrations; binding is not displaced by insulin, IGF-I, or IGF-II (Yosten and colleagues 2014). It reduces NAD(P)H oxidase-dependent ROS generation by preventing RAC-1 translocation, decreases expression of ICAM-1, VCAM-1, and P-selectin, and activates AMPK-α to reduce mitochondrial oxidative stress (Chen and colleagues 2023).
Known effects
- Beta-cell function biomarker — Standard clinical use; C-peptide measurement preferred over insulin assay in treated diabetics; accepted surrogate endpoint in disease-modifying therapy trials for type 1 diabetes
- Nerve conduction improvement — Preclinical evidence strong; early-phase human trials positive; Phase IIb failed to separate from placebo on primary endpoint (Wahren et al. 2016)
- Microvascular protection (nephropathy) — Preclinical evidence strong (animal models); human observational data supportive; no controlled human trial with renal primary endpoint completed
- Retinopathy protection — Cross-sectional observational data in humans; preclinical evidence in rodent models (Chen and colleagues 2023)
- Anti-inflammatory / antioxidant — Mechanistic evidence in vitro and in animal models; not yet tested as primary endpoint in human trials
- Endothelial function / NO availability — Acute intravenous C-peptide infusion increases microvascular blood flow and NO release in type 1 diabetic patients (Yosten and colleagues 2014)
Safety signals
No approved therapeutic formulation exists, so formal safety labeling is absent. In the completed human trials, C-peptide was generally well tolerated. The 12-month Wahren and colleagues (2016) trial of pegylated C-peptide in 250 patients reported no serious adverse events attributable to the study drug. A complicating factor identified by Pinger and colleagues (Mol BioSystems, 2017) is that C-peptide preparations marketed as high-purity can contain iron contamination at near-equimolar levels, which may confound both research and any future therapeutic preparations. The same authors noted a fundamental translational gap: preclinical models treat animals within days of diabetes onset, whereas human patients typically enroll after years or decades without endogenous C-peptide, during which irreversible glycation damage may have accumulated — potentially explaining why strong animal-model effects have not fully translated to human trial outcomes.
Regulatory status
- US: No FDA-approved therapeutic formulation. C-peptide measurement is an accepted clinical laboratory test and a validated surrogate endpoint in clinical trials of beta-cell preservation therapies for type 1 diabetes (recognized in regulatory guidance context; Yosten and colleagues 2014). Cebix Inc. developed Ersatta (pegylated C-peptide) and invested over $50 million in development, but halted its programme after a Phase IIb trial failed to show benefit over placebo (Pinger and colleagues 2017).
- EU: No EMA-approved therapeutic formulation.
- WADA: Insulin is listed as a prohibited substance (S2, peptide hormones). C-peptide itself is measured in anti-doping urine analysis as an endogenous reference analyte and is not itself a prohibited substance.
Mechanism
C-peptide is released from pancreatic beta cells when prohormone convertases PC1/3 and PC2 cleave proinsulin at two dibasic-residue junctions, liberating the insulin B-chain–A-chain heterodimer and the 31-residue C-peptide (Yosten and colleagues 2014). In circulation, C-peptide binds in nanomolar concentrations to a cell-surface receptor that pharmacological evidence places in the G-protein-coupled receptor (GPCR) family, coupled to Gαi/Gαo (pertussis-toxin sensitive). The orphan GPCR GPR146 was identified as essential for C-peptide signaling in KATOIII cells — knockdown of GPR146 blocked C-peptide-induced cFos expression, and C-peptide stimulation induced GPR146 internalization — though the evidence points to a signaling complex (signalosome) rather than a single dedicated receptor (Yosten and colleagues, J Endocrinol, 2013). Downstream of receptor engagement, C-peptide activates PI3K, ERK, PKC, and AMPK-α pathways; increases intracellular calcium; stimulates Na⁺/K⁺-ATPase in peripheral nerve and renal tubular cells; and up-regulates eNOS transcription in vascular endothelium, raising NO availability and improving microvascular tone. It also suppresses NF-κB-driven transcription of ICAM-1, VCAM-1, and proinflammatory cytokines (IL-1, IL-6, TNF-α), and can localize to cell nuclei where it interacts with histones (Chen and colleagues 2023). The combination of C-peptide with insulin produces a biological output distinct from either peptide alone, suggesting that co-secretion serves a physiological tuning function beyond simple co-release (Yosten and colleagues 2014).
Open questions
- GPR146 has been proposed as a C-peptide receptor but the full signaling complex remains uncharacterized; a dedicated high-affinity receptor has not been definitively cloned
- The translational gap between rapid-onset rodent models and long-standing human diabetes has not been bridged; no human trial has targeted early-stage type 1 diabetes with C-peptide replacement
- Whether a modified (more stable, receptor-selective) C-peptide analogue could achieve separation from placebo on clinically meaningful endpoints remains unresolved after the Ersatta failure
- Optimal therapeutic concentration window unclear: physiological levels appear protective but supraphysiological C-peptide (as seen in insulin-resistant type 2 diabetes) has been associated with increased cardiovascular risk in some cohort studies (Chen and colleagues 2023)
- The role of C-peptide in type 2 diabetes complications, where endogenous levels are elevated rather than absent, is not well characterized
Research directions for this peptide, selected from the current sources — hypotheses you can explore and model. None of it is proven yet; tap any one to see the full thinking.
Is there a separate, unidentified receptor that C-peptide actually binds to, rather than the insulin receptor it is commonly associated with?
Finding C-peptide's true receptor would unlock a fresh drug target for the nerve and kidney damage that affects tens of millions of people with diabetes, a complication that insulin alone cannot prevent.
Could replacing C-peptide in type 1 diabetes patients help prevent the retinal blood vessel damage that leads to blindness?
If C-peptide protects retinal vessels as it appears to protect nerve vessels, it could offer a new way to prevent diabetic macular edema, a common cause of vision loss in diabetes patients worldwide, particularly those with type 1 disease who have no C-peptide at all.
Would making the flexible middle section of C-peptide more rigid protect it from being broken down too quickly?
If this works, diabetic patients might need fewer injections of C-peptide replacement, making a potential therapy for nerve and kidney complications of type 1 diabetes more practical and acceptable to patients.
Does C-peptide mainly help diabetic nerves by repairing the tiny blood vessels that feed them, rather than by directly stimulating the nerve cells?
If true, this would reshape how C-peptide replacement therapy is designed for type 1 diabetes, suggesting it should be combined with blood-flow improving drugs rather than nerve-growth drugs, potentially making treatments for diabetic neuropathy more effective.
Does C-peptide only improve health outcomes in people whose bodies have completely stopped making it?
If true, it would help doctors identify exactly which diabetes patients would benefit from C-peptide therapy, avoiding costly and potentially ineffective treatment in the millions of type 2 diabetes patients who still make their own C-peptide.
▸full evidence table2 metrics
| metric | value | tool |
|---|---|---|
| ipTM | 0.2062866985797882 | boltz-2 |
| ranking score | 0.4850253760814667 | boltz-2 |
▸3-letter notation
▸recipeboltz-2 2.2.1
| parameter | value |
|---|---|
| model | boltz-2 2.2.1 |
| weights | — |
| hardware | vast_v100_32gb |
| mlx version | — |
| python | — |
| random seed | 1 |
| msa strategy | colabfold_local |
| runtime | — |
| predicted by | — |
| predicted at | 2026-05-22 |
▸citationbibtex
@peptide{pep10787,
sequence = {EAEDLQVGQVELGGGPGAGSLQPLALEGSLQ},
target = {insr},
author = {peptidemodel},
year = {2026},
status = {bioassayed}
}