What Is C-Peptide? The Molecule That Proves Insulin Is Yours

Peptide science

C-peptide is the 31-amino-acid stretch that gets cut out of proinsulin every time a beta cell manufactures insulin — and for most of the last sixty years it was treated as packaging waste. That reputation was never quite fair. C-peptide turned out to be one of the most useful molecules in endocrinology, not because of anything it does, but because of what it proves: that the insulin in a person’s bloodstream was made by their own pancreas. It also anchors one of the longest-running unresolved arguments in peptide biology.

What is C-peptide?

Insulin is not synthesised as insulin. The gene product is preproinsulin, a single chain of 110 amino acids (UniProt entry P01308). A 24-residue signal sequence at the front is stripped as the chain enters the endoplasmic reticulum, leaving proinsulin: the B chain, then a connecting segment, then the A chain, all in one piece.

That connecting segment is C-peptide — residues 57 to 87 of the precursor, or 33 to 63 if you count from the start of proinsulin. It is 31 amino acids long and weighs roughly 3,021 daltons. Its structural job is real and non-trivial: by holding the A and B chains in the correct relative position, it lets the three disulfide bonds of mature insulin form in the right places rather than at random.

Once folding is complete, the connector is surplus. Two prohormone convertases do the cutting inside maturing secretory granules — PC1/3 at the B-chain junction and PC2 at the A-chain junction — and carboxypeptidase E trims the leftover basic residues. What emerges is mature insulin plus a free C-peptide, packaged in the same granule and released into the portal vein in a strict one-to-one ratio. (A small remainder, on the order of two to three per cent, escapes as intact proinsulin and partly processed intermediates.)

Why C-peptide outlasts insulin in the blood

Equal secretion does not mean equal circulating levels, and the reason is anatomy. Both molecules enter the portal vein and reach the liver first. The liver is studded with insulin receptors and removes roughly half the insulin on that single pass. C-peptide has no receptor there and is not extracted; it sails through and is cleared instead by the kidneys, with five to ten per cent excreted in urine unchanged.

The kinetics follow from that. In a 1985 somatostatin-suppression study in healthy volunteers, the initial half-life measured over the first eight minutes was 3.9 minutes for insulin and 10.2 minutes for C-peptide; both decayed biphasically. The figures clinical laboratories quote for the practical, effective half-life are about 5 to 10 minutes for insulin and 30 to 35 minutes for C-peptide.

Net effect: peripheral C-peptide runs five to ten times higher than insulin and fluctuates far less, so the molar insulin-to-C-peptide ratio in a peripheral sample is normally at or below 1 — despite the pancreas having released them one for one. For more on how clearance route shapes what a blood test can see, see our primer on peptide pharmacokinetics and peptide half-life.

What C-peptide is actually used for

Because C-peptide is produced only by the endogenous pathway, it reports on the pancreas rather than on the pharmacy. Injected insulin cross-reacts with insulin immunoassays and anti-insulin antibodies distort them further; C-peptide sidesteps both problems. The main applications are:

• Classifying diabetes. Residual beta-cell function separates type 1 from type 2 and helps identify latent autoimmune diabetes in adults. Persistently very low values point to an insulin-dependent picture.
• Flagging monogenic diabetes. A stimulated urinary C-peptide-to-creatinine ratio at or above 0.2 nmol/mmol distinguished HNF1A and HNF4A maturity-onset diabetes of the young from long-duration type 1 diabetes with 97 per cent sensitivity and 96 per cent specificity in a 2011 study. It does not, however, separate that form of MODY from type 2 diabetes.
• Insulinoma versus injected insulin. In an insulin-secreting tumour, insulin and C-peptide are both high. In surreptitious insulin administration, insulin is high and C-peptide is suppressed — the classic dissociation.
• Transplant and post-pancreatectomy monitoring, and tracking beta-cell preservation in early type 1 diabetes trials.

Is C-peptide biologically active? A thirty-year argument

Starting in the 1990s a body of work proposed that C-peptide is not inert at all — that it binds a cell-surface receptor, raises endothelial nitric oxide, stimulates sodium-potassium ATPase, and improves microvascular flow. If true, the chronic C-peptide deficiency of type 1 diabetes would be a second, independent driver of diabetic complications, and replacing it would be a therapy.

The hypothesis got a proper test, and it did not survive it. In 2013 the orphan receptor GPR146 was proposed as the C-peptide receptor; a 2020 independent study in cells expressing human GPR146 found no measurable response at concentrations up to 33 micromolar, no internalisation, and concluded the assignment was wrong. Meanwhile a company-sponsored phase 2b trial randomised 250 people with type 1 diabetes and mild-to-moderate peripheral neuropathy to weekly injections of a long-acting PEGylated C-peptide or placebo for 52 weeks. Drug exposure was ample and sustained. The primary endpoint, sural nerve conduction velocity, improved by 1.30 and 0.64 m/s on the two doses — and by 1.19 m/s on placebo. The trial missed its primary endpoint, and no C-peptide product has ever been approved anywhere.

A 2021 review summarising the field put it plainly: the concept of C-peptide as a hormone is presently not supported, while plasma C-peptide remains a standard measure of beta-cell activity. That is a fair statement of where the science sits. Treat any claim that C-peptide is a therapy as unsupported by the trial record.

Measurement caveats worth knowing

• Assays are not interchangeable. Reference laboratories explicitly warn against comparing results across methods when following a patient over time.
• Kidney function changes the answer. Because clearance is renal, C-peptide rises in renal impairment — and urinary C-peptide testing becomes unreliable at any degree of impairment.
• Fasting versus stimulated. A mixed-meal tolerance test, with the peak around 90 minutes, is the more reproducible stimulation method; the glucagon stimulation test peaks near 6 minutes but is more often nauseating.
• Pre-analytical handling matters. C-peptide is a small linear peptide and is susceptible to proteolysis, so prompt separation of serum is part of the method — an ordinary consequence of peptide degradation chemistry.
• Biotin interferes with the widely used electrochemiluminescence platform; laboratories ask for a washout before the draw.

Frequently asked questions

Is C-peptide the same thing as insulin?

No. They are two different products of one precursor molecule, released together in equal numbers. Insulin is the hormone with a receptor and a job; C-peptide is the connector that made insulin fold correctly and was then cut loose.

Can a C-peptide test tell whether someone is taking injected insulin?

That is one of its established uses. Injected insulin does not carry C-peptide with it, so exogenous insulin raises measured insulin while suppressing C-peptide — a pattern that differs from an insulin-secreting tumour, where both rise together.

Does a high C-peptide mean a healthy pancreas?

Not by itself. Elevated values also occur in renal failure, obesity and insulin resistance, with sulfonylurea use, and in insulinoma. The number only means something alongside glucose, kidney function and the clinical context.

Is there a C-peptide drug?

No. The only large controlled trial of a long-acting C-peptide missed its primary endpoint in 2016 and the programme ended. There is no approved therapeutic C-peptide product.

References

Informational only — not medical advice. Nothing here is a treatment recommendation or a dose for any person; prescribing information for linaclotide and plecanatide is linked above and should be read in full by anyone with a clinical question. Consult a qualified healthcare professional about your own care. Intended for readers 21 and over.