HAEE peptide designed against Alzheimer amyloid turns up naturally in human blood

A small HAEE peptide that a Moscow team designed as a possible Alzheimer drug appears to be made by the human body as well. Writing in PNAS Nexus, the researchers report finding it in the blood plasma of all 150 healthy older volunteers they tested, and at roughly half that level in 200 people with a clinical diagnosis of Alzheimer's disease. The work is a single measurement study, funded by a company that holds patents on the test, and no other group has yet repeated it.

From drug candidate to blood molecule

Amyloid plaques, clumps of the amyloid beta peptide that build up in the brain in Alzheimer's disease, contain unusually high levels of zinc, and zinc makes amyloid beta stick together quickly in the test tube. The group led by Sergey Kozin at the Engelhardt Institute of Molecular Biology focused on a four amino acid stretch of amyloid beta, positions 11 to 14, that holds on to zinc. Using molecular docking and simulations, they picked a matching tetrapeptide, histidine, alanine and two glutamates, capped at both ends, as the best fit to block that site. They call it HAEE. In earlier preclinical work, injected HAEE lowered amyloid load in the brains of transgenic mice and appeared to cross the blood brain barrier in rats.

The new paper began as groundwork for a clinical trial. Before giving a drug to people, developers need a reliable way to measure it in blood, so the team built an assay that extracts small peptides from plasma and identifies HAEE by liquid chromatography and mass spectrometry, with a heavy carbon labelled copy as an internal standard. When they ran plasma from untreated healthy volunteers, the HAEE peptide was already there. The authors say this is, as far as they know, the first time a tetrapeptide created artificially has turned out to be a natural component of the human body. They also searched human protein sequences and list six proteins that could in principle be cut to release it.

What the blood tests showed

The study used plasma from 350 people at clinics in Moscow, Saint Petersburg, Tomsk and Nizhny Novgorod, collected in 2023 and 2024. The 150 controls had no diagnosed cognitive problems and an average age of about 71. The 200 patients had carried a clinical diagnosis of Alzheimer's disease for at least two years and were split into four stage groups of about 50 each. Mean HAEE in controls was about 360 picograms per millilitre, against about 175 in the patient group as a whole, a drop of roughly half that was highly significant. Levels did not differ meaningfully between mild and severe stages, which suggests the gap, if real, appears early and does not deepen as the disease advances.

Mean HAEE in controls was about 360 picograms per millilitre, against about 175 in the patient group as a whole, a drop of roughly half that was highly significant.

Used as a test, low HAEE separated patients from controls with an area under the curve of 0.83, where 1.0 would be perfect and 0.5 no better than chance. At the cut off the authors recommend, sensitivity was 84.1%. Specificity, the share of healthy people correctly called negative, appears as 73.5% in the results text but 63.5% in the summary table and figure, an inconsistency the paper leaves unexplained. At the lower figure, more than one healthy person in three would be flagged. The authors point out that total amyloid beta in plasma is roughly similar in patients and healthy people, and they propose that a shortage of HAEE, not of amyloid, could disturb a natural pairing between the two molecules.

The study at a glance

Several features of the design limit how far the numbers can be pushed. Patients were diagnosed on clinical grounds alone, without amyloid scans, spinal fluid tests or newer blood markers, so some may have had other dementias, as the authors acknowledge. The healthy group was hard to recruit and slightly younger, although an effect size check suggested the age gap was small. All samples came through one commercial biobank operator, were stored frozen at minus 20 degrees and travelled for up to three days, and the paper does not say whether storage time differed between groups. There was no separate validation group, so the cut off was chosen and tested on the same 350 people.

Money and patents are part of the picture. The research was organised and funded by Lifemission, a Moscow start up, which the paper says holds patents on endogenous HAEE as a blood biomarker for Alzheimer's disease and on a diagnostic test system. The senior author co founded the company and was its science director, and the first author is a former employee. Such ties are common in early biomarker work, and the paper discloses them, but they mean the people with most to gain from a positive result designed, ran and reported the study. The paper was received in May 2026, accepted in August and published open access in October, after an earlier preprint. Our discovery section tracks other peptides found in unexpected places.

What the HAEE peptide finding could mean

The authors go further than their data. They suggest HAEE may act as a natural protector of the brain whose shortage could help trigger Alzheimer's disease, and that giving synthetic HAEE could make up the deficit. Neither idea was tested here. A lower level in patients could just as well be a result of the disease, of other illnesses, of diet, of medicines or of how samples were handled. The study also does not show where the peptide comes from, whether it reaches the brain in meaningful amounts, or whether it binds amyloid beta inside the body. An independent group would need to answer those questions, ideally with blinded measurements in a cohort whose diagnosis is confirmed by brain imaging.

For now the honest framing is modest. A tetrapeptide designed to block zinc driven clumping of amyloid beta has been detected in human plasma, and in one Russian cohort it was about half as abundant in people diagnosed with Alzheimer's disease. The paper itself notes that HAEE does not fit the established classes of Alzheimer blood markers, which track amyloid, tau, nerve damage or brain inflammation. No trial of HAEE as a treatment in people has reported results, and it remains an experimental compound. Anyone who sees it promoted for memory or brain health should treat that as unsupported. The full paper is open access on PMC. Our clinic coverage follows tests that reach patients.

Conclusion

The HAEE peptide story is an intriguing twist: a molecule drawn up as a drug candidate seems to circulate naturally, and at lower levels in people with clinically diagnosed Alzheimer's disease. But it rests on one company funded cohort, a clinical diagnosis without biomarker confirmation, a cut off tested on the same samples and an unexplained specificity discrepancy. Independent replication has to come before any talk of a blood test or a therapy.

Source Kopylov AT, Yakovlev RY, Trukhin AV, Poluektov YM, Anashkina AA, Kozin SA. Beyond Alzheimer's story: How an engineered molecule gained an endogenous essence. PNAS Nexus, October 2026. DOI 10.1093/pnasnexus/pgag320

Amino Weekly reports research. It is not medical advice and never gives doses.