DSIP: The Sleep Peptide Isolated in 1977, Still Missing a Gene

DSIP is a nine-residue peptide, Trp-Ala-Gly-Gly-Asp-Ala-Ser-Gly-Glu, isolated from rabbit cerebral venous blood and named for the delta EEG it produced after brain infusion. The 1977 synthesis matched that sequence. Fifty years later there is still no cloned gene, no characterized precursor, and no receptor, which is why a 2006 review called it an unresolved riddle rather than a finished sleep drug.

TL;DR

• DSIP is a nonapeptide of molecular weight 849. Schoenenberger and Monnier sequenced it as Trp-Ala-Gly-Gly-Asp-Ala-Ser-Gly-Glu and showed that only the full synthetic chain, not fragments, raised delta and spindle EEG in rabbits (PNAS 1977).
• The first human work used slow intravenous infusions, 25 nmol/kg. In six healthy adults, median total sleep time in a 130-minute window rose 59% versus placebo (Schneider-Helmert et al., 1981).
• Six chronic insomniacs looked similar on an acute night. Longer sleep, fewer interruptions, a little more REM, no daytime sedation (Schneider-Helmert and Schoenenberger, 1981).
• A later double-blind insomnia trial did not hold. Monti and colleagues found no significant difference versus baseline or placebo and called the sleep change of little clinical significance (Monti et al., 1987).
• No DSIP gene, protein, or receptor has been isolated. Kovalzon and Strekalova judged the sleep-factor hypothesis poorly documented and still weak (J Neurochem 2006).

What is DSIP, chemically?

It is a short, linear nonapeptide. The sequence is tryptophan, alanine, glycine, glycine, aspartate, alanine, serine, glycine, glutamate. Graf and Kastin put the molecular weight at 849 and noted that the structure does not sit inside any of the usual neuropeptide families (Graf and Kastin, 1984).

That size is why catalogs can make it by solid-phase synthesis. See Peptides 101 for the broader map.

It is also a protease target. Nyberg, Thornwall, and Hetta purified an aminopeptidase from human cerebrospinal fluid that cleaves DSIP at neutral pH (Nyberg et al., 1990).

A catalog vial does not recreate a cerebrospinal infusion. Mixing powder is a chemistry step, covered in the reconstitution guide and the calculator, not a sleep protocol.

Why did the 1977 rabbit work look so clean?

The Basel group was chasing a humoral sleep factor. Monnier had been dialyzing sleep-promoting material from rabbit cerebral blood for years. In 1977 Schoenenberger and Monnier reported the isolated peptide, synthesized it, and tested it against fragments and swapped-residue analogs (PNAS 1977).

They infused nine synthetic peptides into 58 rabbits under double-blind conditions and scored fast-Fourier EEG. Only the full DSIP sequence raised delta and spindle activity. Fragments and swapped-residue analogs did not.

That is a real structure-activity result. It is not proof that DSIP is the body’s native sleep switch. Graf and Kastin later described a U-shaped curve for both dose and infusion time, which is what enthusiasts quote (1984 review).

What did the human infusion studies actually show?

The early human work is small, intravenous, and mostly from the same circle that isolated the peptide.

The 1981 healthy-volunteer study is the one forums treat as proof. Six people, morning infusion, double-blind crossover: sleep pressure in the lab, cleaner night sleep, no classic sedation (Schneider-Helmert et al., 1981).

The insomnia night study used the same dose, same lab, and n = 6 again. Sleep promotion showed in hour two, not hour one, and lasted into about six hours of night sleep (Schneider-Helmert and Schoenenberger, 1981). Neither paper is a modern phase 3 insomnia program.

Did DSIP beat placebo in chronic insomnia?

Monti’s 1987 trial is the paper that usually gets skipped. DSIP or placebo, 25 nmol/kg intravenously, four nights, double-blind crossover, scored by polysomnography.

Awakenings, NREM latency, and waking time fell under DSIP, but not significantly versus baseline or placebo. Total sleep and NREM rose mostly as stage 2. Slow-wave sleep and REM did not change.

Where NREM and stage 2 beat placebo, those gaps were already present at baseline. Monti called the improvement of little clinical significance (Monti et al., 1987).

That is the honest enthusiast takeaway. Early IV work was interesting. The controlled insomnia replication did not deliver a drug.

Why is there still no DSIP gene?

Kovalzon and Strekalova said the quiet part in 2006. Nobody isolated the DSIP gene, the protein, or a candidate receptor. They called the sleep-factor hypothesis extremely poorly documented and still weak (J Neurochem 2006).

DSIP-like immunoreactivity shows up in hypothalamic nuclei that are not the classic sleep network. Radioimmunoassays can light up related material without proving the nonapeptide is what the body makes.

Some artificial analogs promoted slow-wave sleep in their rabbit and rat work. DSIP itself did not, in those hands. That is analog chemistry, not a catalog-vial sleep stack.

For a peptide with a cloned receptor and human genetics, see kisspeptin-10. Sleep-adjacent Russian peptide work with a longer paper trail sits in the Selank article.

What do research catalogs sell that the papers did not use?

The human studies were slow intravenous infusions of synthetic peptide in a lab. They were not subcutaneous research vials, oral capsules, or nighttime sleep stacks. Nyberg’s CSF aminopeptidase is a reminder that the chain can be clipped before it does anything interesting (Nyberg et al., 1990).

A licensed hypnotic has a label and an approval file. A research nonapeptide has a COA if the vendor published one. Start with how to read a COA, then compare listings on the price index and clinics in the provider directory.

People who teach sleep biology and peptide pharmacology, rather than forum folklore, are on the experts page.

Bottom line DSIP is a real synthetic nonapeptide with a 1977 sequence and small 1980s infusion studies. It is not a cloned human sleep hormone, and the best controlled insomnia trial did not show a clinically meaningful effect.

Educational only. Not medical advice.