What Is DSIP?

Published August 2026 · Patriot Labs Research Team · ~12 min read

A peptide named after a brain wave

Most research peptides are named for a structure, a source tissue or a parent hormone. DSIP was named for an effect — specifically, for an increase in delta-band activity on the electroencephalogram of a rabbit. The name is a hypothesis that got fixed in the literature before the hypothesis was tested properly, and everything confusing about DSIP follows from that.

The chemistry is not in dispute. A defined nonapeptide was purified, sequenced and synthesised, and the synthetic material reproduced the EEG signal that led to its isolation. What has never been resolved is whether that nonapeptide is a real endogenous signalling molecule in the animals it was extracted from, and whether the delta-EEG association reflects a physiological function or an artefact of a very particular experimental preparation.

A 2006 mini-review in the Journal of Neurochemistry, titled without hedging “a still unresolved riddle,” put it bluntly: the link between DSIP and sleep “has never been further characterized,” and the sleep-promoting hypothesis remained “extremely poorly documented and still weak.”⁷ That assessment came thirty years after the original isolation. Nothing has changed it since.

This guide treats that ambiguity as the subject rather than as a footnote. For readers new to peptide nomenclature and chemistry generally, the introduction to what peptides are covers the vocabulary used below.

The discovery: dialysate from a stimulated rabbit

The work sits inside a much older research programme. Since the early twentieth century, physiologists had proposed that sleep is driven in part by a circulating humoral factor — something that accumulates during waking and can, in principle, be transferred between animals. Marcel Monnier's group in Basel pursued this with cross-circulation experiments in rabbits, publishing a long series under the heading “Humoral transmission of sleep.”

The preparation worked like this. A donor rabbit received low-frequency electrical stimulation of thalamic nuclei, using parameters that produced delta-dominated EEG. Blood was dialysed from the donor's cerebral venous circulation, and the dialysate was infused into a recipient rabbit whose EEG was recorded. In the 1975 report, thalamic stimulation of donors raised both the concentration of a candidate peptide and its EEG activity in the recipient in parallel, and cerebral blood behaved differently from systemic blood — the authors proposed that antagonistic “waking factors” picked up in peripheral tissue were degrading or opposing the signal.¹

Purification followed. In 1977, Schoenenberger, Maier, Tobler and Monnier reported isolating from that dialysate a naturally occurring nonapeptide of molecular weight 848.98, which enhanced delta-EEG activity in recipient rabbits by roughly 43 per cent over controls, and gave it the name delta-sleep-inducing peptide.² In 1978 the same group published the amino-acid analysis, the sequence, the chemical synthesis and confirmation that the synthetic nonapeptide reproduced the effect — a mean increase in delta activity of about 35 per cent in neocortex and limbic cortex relative to control animals, described as a “highly specific delta and spindle EEG-enhancing effect.”³

Two things are worth noticing about this chain. First, the endpoint throughout is an EEG spectral measure in an anaesthetised or instrumented rabbit, not sleep behaviour in a freely behaving animal. Second, the biological activity was defined by the assay used to purify it. A purification guided by a delta-EEG bioassay will, by construction, return whatever fraction moves that bioassay — which is a legitimate way to find a molecule and a poor way to establish what the molecule normally does.

The molecule: nine residues, no family

DSIP is a linear nonapeptide with the sequence Trp-Ala-Gly-Gly-Asp-Ala-Ser-Gly-Glu, single-letter code WAGGDASGE, reported at a molecular weight of approximately 849.^2,3 It carries a tryptophan at the N-terminus and a glutamate at the C-terminus, has no cysteines and therefore no disulfide bridge, and contains no basic residues at all.

That last row does more damage than it looks. When a new peptide belongs to a known family, its family provides a receptor to test, a precursor architecture to look for and a set of structure-activity expectations. DSIP has none of that. It is an orphan sequence, which means every mechanistic question about it has had to be answered from scratch — and mostly has not been.

The composition also predicts specific instability. Aspartate followed by a small flexible residue, and serine adjacent to glycine, are among the motifs most associated with isomerisation and backbone cleavage in the general peptide-stability literature, and a free N-terminal tryptophan is an obvious target for aminopeptidases. Our guide on how peptides degrade works through those chemistries in general terms; DSIP happens to carry several of them at once.

What was actually observed, and in what model

The table below separates the claims usually attached to DSIP from the measurements those claims rest on. The right-hand column is the one that matters, and it is uncomfortable reading.

Read that table as a whole and a pattern appears. The best-documented findings are the ones furthest from the peptide's name.

The missing pieces: no gene, no precursor, no receptor

For a peptide to be an endogenous signalling molecule, a standard set of things should be findable. There should be a gene encoding a precursor protein. There should be processing enzymes that liberate the mature peptide. There should be a receptor with a gene of its own, and ideally an antagonist that blocks the effect. For DSIP, none of these have been produced.

The 2006 review enumerates the gaps directly: the structure of any DSIP precursor protein, the gene for that precursor, a DSIP-specific receptor, and the gene for that receptor are all absent from the literature.⁷ The same review raises a more corrosive point — that without an isolated natural product to compare against, there is reason to doubt that the synthetic DSIP used in decades of subsequent biological work is identical to whatever was actually present in that rabbit dialysate.

The evidence that DSIP occurs naturally rests largely on immunoassay and immunohistochemistry, which detect antibody binding rather than molecular identity. Antisera raised against a short, glycine-rich, acidic nonapeptide can plausibly cross-react with fragments of larger proteins that happen to contain a similar stretch. That is why the 2006 authors landed where they did: the results, they wrote, permit the supposition that some still-unknown DSIP-like peptide or peptides exist, structurally related to DSIP but not necessarily DSIP itself.⁷

This is a substantively different situation from a peptide whose mechanism is merely incompletely mapped. With a melanocortin ligand, for instance, the receptor is cloned and the argument is about downstream detail. With DSIP the argument is still about whether the molecule is one the body makes at all.

Beyond the delta wave: endocrine, stress and redox work

Because the sleep attribution stalled, much of the later DSIP literature moved elsewhere. Three strands are reasonably well documented as observations, whatever their interpretation.

Endocrine effects. Iyer and McCann reported in 1987 that DSIP injected into the third ventricle of ovariectomised rats produced a sustained rise in plasma growth hormone, and that dispersed cultured pituitary cells from the same animals released growth hormone in a concentration-dependent manner — but only across a narrow window, peaking near 10⁻¹⁰ M and losing significance at higher concentrations.⁸ The in vivo effect was blocked by pimozide, implicating a dopaminergic route rather than a direct DSIP receptor. The 1984 Graf and Kastin review had already noted a U-shaped activity curve for DSIP with respect to both amount and infusion timing.⁴

Stress and thermoregulation. The 2006 review describes a body of work positioning DSIP as a “stress-limiting factor,” including reduction of basal corticotropin, hypothermic properties, and a reported correlation between the diurnal rhythm of DSIP-like immunoreactivity and that of body temperature.⁷ The same review catalogues reported involvement in immune responses, cardiovascular parameters, analgesia, circadian regulation, anticonvulsant activity and antioxidant activity — and treats this breadth itself as an argument that DSIP behaves like a general regulatory peptide rather than a dedicated sleep signal.

Redox and ageing models. A Russian research line has examined DSIP in oxidative-stress paradigms. Shustanova and colleagues reported shifts in the prooxidant–antioxidant balance of rat tissues and erythrocytes under cold stress, with increased activity of superoxide dismutase, catalase, glutathione peroxidase and glutathione reductase.⁹ Kutilin and colleagues later reported increased expression of the Sod1 and Gpx1 genes in brain and nucleated blood cells of rats across a range of ages.¹⁰ These are transcript- and enzyme-activity endpoints in rodent models from a closely related group of investigators, not independently replicated mechanisms.

A very broad activity profile with no identified receptor is a familiar pattern, and it is worth naming: it is what a literature looks like when a molecule is tested in many assays and the positive results are the ones written up. That is not proof of nothing happening — it is a reason to weight breadth of claims lower rather than higher. Similar caution applies to the wider nootropic peptide literature covered in the peptides studied for cognition overview and in the Semax and Selank comparison.

Why an unstable, unreceptored peptide is hard to study

Some of DSIP's inconsistency is probably not conceptual at all. It is pharmacokinetic.

Graf, Saegesser and Schoenenberger examined the biostability of DSIP in blood specifically to ask whether inactivation rates explain variable effects in vivo.⁵ Incubating DSIP in human or rat blood released products with the chromatographic behaviour of free tryptophan, and product formation depended on temperature, on time and — importantly — on species. Degradation proceeds chiefly by clipping the N-terminal tryptophan, and the 2006 review characterises the resulting survival time in the intact animal as no more than a few minutes.⁷

The same 1987 study compared DSIP against two analogues, including a phosphorylated variant. The analogues degraded more slowly and, unlike DSIP itself, formed complexes; excess unlabelled peptide did not displace the radiolabel, which the authors read as non-specific binding or aggregation rather than saturable receptor occupancy. Their conclusion was that the rapid disappearance of injected DSIP reflects degradation, whereas the analogues persist through a mixture of slower degradation and complex formation — and that whether this explains the stronger, more consistent effects sometimes reported for analogues “remains to be examined.”⁵

Stack the consequences up:

• Exposure is short and species-dependent. A protocol that produced measurable EEG change in one species can deliver a different effective exposure in another using the same nominal amount, purely through differing plasma degradation. That alone can generate contradictory literature.
• The barrier question is unresolved rather than answered. The 2006 review states that the amphiphilic molecule can partly cross the blood–brain barrier and cellular membranes, and that both passive diffusion and a specific transport mechanism near the floor of the fourth ventricle have been proposed.⁷ Proposed is the operative word. Without a quantified transport measurement, the brain concentration achieved in any peripheral-administration study is unknown.
• No receptor means no negative control. A cloned receptor permits binding assays, occupancy curves and abolition of an effect with an antagonist — the standard proof that a response is receptor-mediated. None of that tooling exists for DSIP, so an observed effect cannot be tied to a mechanism, and a null result cannot be attributed to a failure of engagement rather than an absence of activity.
• Non-monotonic responses break the usual inference. When higher amounts produce smaller effects — the U-shaped curve noted in review⁴ and visible in the pituitary cell data⁸ — two laboratories using different amounts can honestly report opposite results without either being wrong.
• Aggregation confounds tracer work. Non-displaceable binding of labelled analogues⁵ means radiolabel distribution studies may be tracking aggregates and non-specific adsorption rather than the free peptide.

Any one of these would complicate a research programme. Together they go a long way toward explaining why fifty years produced a large literature and a small amount of settled knowledge.

Bench handling and material identity

Two practical points follow from the chemistry rather than from the biology.

First, identity verification matters more than usual here. Because there is no receptor binding assay and no functional potency assay in routine use, the only meaningful check on DSIP material is analytical — mass and chromatographic purity. A short acidic nonapeptide is straightforward to characterise by mass spectrometry, and a certificate of analysis reporting the expected mass and an HPLC purity figure is the whole of what can reasonably be verified before an experiment.

Second, the sequence carries the specific liabilities described above — a free N-terminal tryptophan, an Asp in a flexible glycine-rich context, and a Trp side chain among the more photo- and oxidation-sensitive in the amino-acid set. None of that is exotic; it means the general handling logic in the degradation guide applies with less margin than it would for a cyclised or otherwise protected sequence. Comparable considerations apply to other short neuropeptides in the same catalogue category, such as Pinealon.

What the evidence does not establish

This section carries more weight than usual for this compound, because the gap between the name and the data is unusually large.

It does not establish that DSIP is an endogenous peptide. No gene, no precursor protein, no processing pathway. The natural-occurrence evidence is antibody-based detection of “DSIP-like” material, and the 2006 review explicitly declines to treat that as identification, proposing instead that an unknown related peptide may account for the signal.⁷

It does not establish a mechanism. There is no identified receptor, no receptor gene, no antagonist and no binding data of defined molecular identity.⁷ Where a downstream pathway has been implicated at all — the dopaminergic route in the growth-hormone work⁸ — it is inferred from pharmacological blockade in one laboratory, not from a mapped signalling chain.

It does not establish a reproducible sleep effect even in animals. The foundational EEG results came from one group using one specialised preparation. Subsequent work has variously failed to confirm slow-wave or paradoxical-sleep promotion, found only minor effects, or reported impairment.⁷ The cross-species picture is not merely noisy but self-contradictory: a 1984 review described the feline response as predominantly affecting REM sleep,⁴ while a 1987 feline study using intracerebroventricular administration reported increased deep slow-wave sleep with REM measures unchanged.⁶

It does not support any conclusion about people. Human work on DSIP was attempted in the 1980s and did not converge. The published clinical material from that period includes small uncontrolled pilot studies — one such report enrolled seven patients, with baseline periods used as the comparison rather than a randomised control group¹¹ — which is a study design that cannot separate a compound effect from regression to the mean, expectancy or natural fluctuation. DSIP was never developed into an approved medicine in any jurisdiction, and nothing in this literature indicates that DSIP does anything for a person's sleep or for any other human outcome.

Findings cluster by laboratory. The isolation and characterisation work is dominated by the Basel group; the redox and ageing work by a single Russian research line;^9,10 the endocrine work by one American group.⁸ Independent replication across laboratories — the thing that would actually settle any of this — is the scarcest commodity in the whole file.

Sample sizes and endpoints are modest by modern standards. The primary studies are small-animal experiments with spectral EEG or hormone-assay endpoints, mostly predating pre-registration, blinded scoring conventions and multiple-comparison correction as routine practice. That does not make them wrong. It does mean effect sizes reported in the 1970s and 1980s should be read as provisional.

Absence of long-term safety data. There is no body of chronic toxicology or long-term exposure data for DSIP of the kind that would exist for a developed compound. This is a research chemical, sold and handled as one.

Frequently asked questions

What does DSIP stand for? Delta sleep-inducing peptide. The name was assigned by the Basel group that isolated it and records an association with delta-band EEG activity measured in rabbits — not a demonstrated physiological role. Later reviewers have argued the name commits to more than the data supports.

What is the amino acid sequence of DSIP? Trp-Ala-Gly-Gly-Asp-Ala-Ser-Gly-Glu, single-letter WAGGDASGE, a nonapeptide with a reported molecular weight near 849. The sequence was determined by amino-acid analysis and sequencing and confirmed by synthesis of material with matching EEG activity, reported by Schoenenberger and colleagues in 1978.³

Is DSIP an established endogenous sleep factor? No, and that is the central unresolved question. The 2006 Journal of Neurochemistry mini-review concluded that there is still no strong evidence of the natural occurrence of DSIP, and that no DSIP gene, precursor protein or specific receptor has been identified. Findings across laboratories and species have been inconsistent, including reports of no effect and of impaired sleep.⁷

Why is DSIP considered difficult to study? It is degraded rapidly in blood, chiefly by removal of the N-terminal tryptophan, so exposure after administration is short and species-dependent.^5,7 Reported concentration-response relationships are non-monotonic rather than graded.^4,8 No receptor has been identified, so there is no binding assay or antagonist to anchor a mechanism. Together these make both positive and negative results hard to interpret.

Does DSIP cross the blood–brain barrier? Partly, according to review, on the basis of its small size and amphiphilic character — but the mechanism is proposed rather than demonstrated. Both passive diffusion and a specific transport route near the floor of the fourth ventricle have been suggested, and no quantified transport figure is established.⁷

Why do DSIP analogues appear in the literature? Because the parent peptide degrades quickly. Analogues including a phosphorylated variant were shown to degrade more slowly in plasma and serum, though they also formed non-displaceable complexes consistent with aggregation or non-specific binding, which complicates their interpretation in tracer studies.⁵

Is DSIP approved for human use? No. All Patriot Labs products are sold strictly for in-vitro research and laboratory use only. They are not approved for human or veterinary consumption, and nothing on this page describes use in a person.

All Patriot Labs products are sold strictly for in-vitro research and laboratory use only. Not for human or veterinary consumption. This guide is educational and describes peptide chemistry and published research in general terms; it is not medical advice, does not describe how to use any product, and the references cited do not constitute a product claim.