A Lab-Made Alzheimer's Peptide Was Already in Human Blood

A peptide designed in a lab more than a decade ago to interrupt one of Alzheimer's disease's early steps turns out to already be circulating in human blood — and a new study finds people with Alzheimer's have roughly half as much of it as everyone else.

A synthetic peptide that wasn't synthetic after all

Researchers led by Sergey Kozin at the Engelhardt Institute of Molecular Biology in Moscow originally built the tetrapeptide HAEE (Ac-His-Ala-Glu-Glu-NH2) as a tool, not a discovery. The idea was to block zinc ions from binding to amyloid-beta, one of several proposed triggers for the sticky plaques that accumulate in Alzheimer's brains. In mice, the group had already reported that HAEE could reverse zinc-driven amyloid clumping (Mitkevich et al., Aging and Disease, 2023).

While building a more sensitive way to measure HAEE in blood plasma — the kind of assay a drug developer needs before running human trials — the team instead found HAEE already there, in people who had never been given it. According to the researchers, writing in PNAS Nexus on October 20, this is the first time a peptide engineered from scratch has turned out to be a naturally occurring human molecule.

What the blood test showed

The team measured HAEE by mass spectrometry in plasma from 350 people recruited across four Russian clinical centers: 150 with no cognitive impairment and 200 with a clinical Alzheimer's diagnosis. Average HAEE was 360.5 pg/mL in the unimpaired group versus 175 pg/mL in the Alzheimer's group — under half as much, a difference the authors report as highly statistically significant. Used as a diagnostic cutoff, HAEE level distinguished the two groups with about 84% sensitivity, though the paper gives two different figures for specificity (73.5% in the text, roughly 63.5% in the table and figure) — a discrepancy worth knowing about if the result gets cited elsewhere. Notably, HAEE didn't track with disease severity: mild, moderate and severe Alzheimer's groups had statistically indistinguishable levels, which complicates a simple "less HAEE, worse disease" story.

What's known, and what isn't

This is a correlation in blood, not a demonstrated cause of disease, and the study has real limits its own authors flag. Alzheimer's status was assigned by clinical criteria and diagnostic codes alone, with no brain imaging or pathology confirmation, so the "Alzheimer's" group may include some people with other forms of dementia. The comparison group also wasn't well age-matched to the patients — the authors say healthy older volunteers were hard to recruit — which can inflate apparent differences between groups. And this is one peptide, one disease, one research group's data; it hasn't been replicated independently.

It's also worth naming the funder: the study was paid for by Lifemission LLC, a Moscow company that holds patents on using blood HAEE as both an Alzheimer's biomarker and a diagnostic test, and two of the six authors are the company's current or former staff. That doesn't make the chemistry wrong, but it's a direct financial stake in how the result gets read, and outside experts asked about it — Ottavio Arancio at Columbia and Chuanhai Cao at the University of South Florida, quoted in News-Medical's coverage — raised similar questions: whether the measurement method captures the whole molecule, why levels don't fall further as disease worsens, and how the body would even make an "engineered" sequence in the first place.

That last question sounds stranger than it is once HAEE's sequence origin is clear. It isn't an arbitrary string of amino acids — it's a synthetic copy of residues 35 to 38 of the alpha-4 subunit of the α4β2 nicotinic acetylcholine receptor, a real human protein, with its ends chemically capped for stability (Zolotarev et al., Biomolecules, 2021). Researchers picked that fragment in the first place because it sits at the spot where the receptor is thought to interact with amyloid-beta. So a short piece of a receptor that already exists in the body turning up loose in blood — presumably released when the full-length receptor is broken down — is less mysterious than a sequence invented from nothing. What the new paper doesn't establish is whether the circulating molecule is cut from that receptor by a normal biological process, made some other way, or whether the lab's capped, modified version and whatever is actually in blood are chemically identical rather than just indistinguishable by this assay. That's the detail most likely to draw scrutiny from scientists who weren't involved.

This isn't the first time a peptide's blood or tissue level has been reported lower in Alzheimer's than in healthy people — we wrote in September about catestatin, a peptide found at reduced levels in Alzheimer's-model mouse brains, where restoring it reduced damage. The pattern of "this peptide is scarcer in disease" keeps showing up; so far, in both cases, it's much easier to measure the scarcity than to prove it causes anything.

What this means for researchers

The headline claim — that a drug candidate turned out to be an existing human molecule — is the kind of result that needs an independent lab to reproduce before it reshapes anyone's thinking about amyloid biology. The next useful step isn't a treatment trial; it's a replication using imaging-confirmed Alzheimer's cases, age-matched controls, and a measurement method validated by a group with no stake in the patent. Until that happens, "HAEE is lower in Alzheimer's" is a finding to track, not a biomarker to rely on.

For anyone following the compound itself, the more interesting long-term question may be methodological rather than clinical: how often do peptides designed purely as research tools or drug candidates turn out to already exist in the body, unnoticed, because nobody had a reason to look for them in blood until now? Readers tracking the broader vocabulary of peptide research — what distinguishes a designed sequence from a naturally occurring one — can find the basics on our compound reference.