What Is Neurotensin? The Peptide That Was Too Fragile

Peptide science

Neurotensin is one of the best-characterised neuropeptides that never became a medicine. Isolated in 1973, it produces powerful pain relief in the brain without touching opioid receptors, drops body temperature on demand, and behaves like an antipsychotic in animal models. It also falls apart in the bloodstream in about ninety seconds and cannot cross into the brain. This guide covers what neurotensin is, what it does, why the pharmacology never translated, and what researchers are doing with it now.

What is neurotensin?

Neurotensin is a 13-amino-acid peptide (a tridecapeptide) with the sequence pGlu-Leu-Tyr-Glu-Asn-Lys-Pro-Arg-Arg-Pro-Tyr-Ile-Leu. Its molecular formula is C₇₈H₁₂₁N₂₁O₂₀ (PubChem CID 25077406), giving an average mass of roughly 1,673 Da. The N-terminal residue is pyroglutamate, a cyclised glutamate that blocks aminopeptidase attack from that end.

Robert Carraway and Susan Leeman isolated it from acid-acetone extracts of bovine hypothalami while working on substance P, and published it in 1973 as “a new hypotensive peptide.” The full sequence followed in 1975. It is encoded by the NTS gene as a 170-amino-acid precursor shared with a second peptide, neuromedin N.

Two tissues dominate its production. In the brain it acts as a neurotransmitter and neuromodulator. In the gut it is made by enteroendocrine N cells, at highest density in the ileum, and released most strongly in response to dietary fat.

Three receptors, and one of them is not a GPCR

NTS1 (NTSR1) and NTS2 (NTSR2) are seven-transmembrane G-protein-coupled receptors. NTSR3 is sortilin, a single-transmembrane VPS10-domain sorting receptor — structurally unrelated, and a common source of confusion when reading the literature.

• NTS1 binds neurotensin at 0.1–0.4 nM and is Gq-preferring; in almost every system tested, agonist binding activates phospholipase C. It is selectively expressed by central dopamine neurons and is prominent in substantia nigra and ventral tegmental area.
• NTS2 binds at 2–5 nM, is levocabastine-sensitive in rodents, and is constitutively active. Oddly, at the human receptor expressed in CHO cells neurotensin has been reported to behave as an antagonist. NTS2 is the receptor responsible for the analgesia produced by centrally administered neurotensin.
• Sortilin is a trafficking receptor rather than a signalling GPCR.

The Grisshammer laboratory solved the structure of rat NTS1 bound to NT(8-13) at 2.80 Å in 2012 (PDB 4GRV). It was the first look at how a peptide agonist binds a GPCR, and it remains a reference structure for the field.

What neurotensin does — and why it depends entirely on where you put it

This is the part that makes neurotensin unusual. Its central and peripheral pharmacology are almost disjoint.

Injected into the brain, neurotensin produces hypothermia and potent antinociception. Critically, that pain relief is not blocked by naloxone, so it does not run through endogenous opioid mechanisms — which is exactly why it attracted interest as a non-opioid analgesic target. It also produces antipsychotic-like behavioural effects and reduces food intake, and it modulates dopamine transmission in nigrostriatal and mesocorticolimbic pathways.

Given peripherally, it does something else entirely: blood pressure falls, gastric acid secretion and motility are inhibited, pancreatic and biliary secretion increase, and intestinal fat absorption is promoted. A 2016 Nature paper reported an obligatory role for neurotensin in high-fat-diet-induced obesity in mice.

Why neurotensin never became a drug

Three problems, and they compound.

Speed of destruction. A classic study of neurotensin degradation reported a half-time of disappearance of 226 minutes in human plasma in vitro but only about 1.4 minutes in vivo. That gap is instructive: soluble plasma proteases are not the main problem, tissue and endothelial peptidases are. In-vitro loss was largely blocked by EDTA but not by pepstatin A, PMSF or aprotinin, pointing at metallopeptidases.

Where the cuts land. The main inactivating cleavages sit inside the active hexapeptide. Prolyl oligopeptidase cuts between Pro7 and Arg8. Thimet oligopeptidase (EC 3.4.24.15) cuts between Arg8 and Arg9. Neurolysin (EC 3.4.24.16) and neprilysin (EC 3.4.24.11) both cut between Pro10 and Tyr11, and neprilysin has a second site between Tyr11 and Ile12. Angiotensin-converting enzyme also contributes to breakdown. Neurolysin is co-localised with neurotensin receptors on synaptic membranes, which is why it is sometimes called the physiological “neurotensinase.” If you want the wider context on how enzymatic clearance shapes peptide pharmacology, see our guides to neprilysin and peptide half-life.

The barrier. Neurotensin does not meaningfully cross the blood-brain barrier. So the interesting effects — analgesia, hypothermia, antipsychotic activity — are unreachable by any ordinary route of administration.

The workarounds

Stabilised analogues. NT69L and PD149163 are NT(8-13)-based compounds engineered for peptidase resistance and brain penetration. NT69L blocks apomorphine-induced climbing and cocaine-driven hyperactivity without catalepsy, and produces antinociception, hypothermia and hypotension in rhesus monkeys. PD149163 shows an atypical-antipsychotic-like profile in prepulse-inhibition testing.

A venom peptide that reached humans. Contulakin-G (CGX-1160) is a glycosylated 16-residue peptide from the venom of the cone snail Conus geographus, an agonist at all three neurotensin receptors with submicromolar potency. It carries FDA orphan drug designation and was tested in a phase 1A intrathecal trial in central neuropathic pain after spinal cord injury; four subjects were dosed and the maximum tolerated dose was set by the development of diarrhoea.

Brain shuttles. VH-N412 conjugates neurotensin to a peptide that binds the LDL receptor on brain endothelium. In mice after status epilepticus it produced rapid hypothermia along with anticonvulsant, neuroprotective and anti-inflammatory effects.

Pharmacological hypothermia. ABS-201, later renamed HPI-201, is a synthetic NTSR1 agonist built for induced cooling. It lowers body and brain temperature by 2–5 °C within 15–30 minutes without shivering, and reduced 24-hour infarct volume by roughly 30–40% in rodent stroke models, including when started 30–60 minutes after the insult. The appeal is that physical cooling is slow and impractical in most clinical settings.

Turning the receptor into an address. Because many tumours over-express NTSR1, the receptor has been repurposed as a delivery target. 177Lu-3BP-227 (also called 177Lu-IPN01087) is a DOTA-conjugated NTSR1 antagonist carrying lutetium-177. First clinical results in metastatic pancreatic adenocarcinoma imaged six patients and treated three, with kidneys dose-limiting and one partial response. The Ipsen phase 1/2 study (NCT03525392) was terminated at 14 patients, reportedly not for safety reasons but because the asset was transferred.

Neurotensin as a biomarker

The peptide itself is too unstable to measure reliably, but its precursor is not. Proneurotensin 1-117 is a stable circulating fragment, and in 2012 a JAMA study measured it in 4,632 fasting participants of the Malmö Diet and Cancer Study with a median follow-up over 13 years. Higher proneurotensin predicted incident diabetes (HR 1.28 per standard deviation), cardiovascular disease (HR 1.17) and cardiovascular mortality (HR 1.29). There was a strong sex interaction: the associations were essentially confined to women, where the hazard ratios were 1.41 for diabetes, 1.33 for cardiovascular disease and 1.44 for breast cancer. The breast cancer finding was independently replicated in 2014.

The dopamine connection, and why it stalled

Reduced cerebrospinal-fluid neurotensin was found in a subset of drug-free patients with schizophrenia, concentrations correlated negatively with illness severity, and effective antipsychotic treatment normalised them. That produced the “endogenous antipsychotic” hypothesis, and NTS1’s selective expression on central dopamine neurons gave it an anatomical basis.

What stopped it was the side-effect profile of balanced NTSR1 agonism: hypotension, hypothermia and motor impairment travel with the wanted effect. The most interesting recent attempt to separate them is SBI-553, a beta-arrestin-biased allosteric modulator that antagonises Gq signalling while still recruiting arrestin. In rodents it attenuated psychostimulant-associated behaviour without the hypothermia, hypotension or motor impairment. It remains preclinical — there is no registered human trial.

Frequently asked questions

Is neurotensin an opioid?

No. It is structurally unrelated to opioid peptides, and its analgesic effect in the brain is not blocked by naloxone. That non-opioid mechanism is the main reason it has been pursued as a pain target, though no neurotensin-based analgesic has been approved.

Why is NT(8-13) used instead of the full peptide?

Because the C-terminal hexapeptide carries essentially all the receptor activity, while the first seven residues contribute little. Working with a six-residue fragment is cheaper to synthesise and gives medicinal chemists a smaller scaffold to stabilise against the peptidases that attack that region.

What is the difference between NTSR1, NTSR2 and NTSR3?

NTSR1 and NTSR2 are G-protein-coupled receptors with different affinities and different roles — NTSR1 is tied to dopamine circuits and tumour biology, NTSR2 to central analgesia. NTSR3 is sortilin, a single-transmembrane sorting receptor that is not a GPCR at all.

Can neurotensin be given by injection to get its brain effects?

No. It does not cross the blood-brain barrier in meaningful amounts and is cleared within minutes, so peripheral administration produces only the peripheral effects — hypotension and gut effects. Every attempt to access the central pharmacology has required a chemically modified analogue or a delivery vector.

Informational only — not medical advice · 21+. VialHelp does not sell peptides and does not recommend any product, brand or protocol. Research findings described here are reported for context and are not recommendations. Consult a qualified healthcare professional for personal medical questions.