Somewhere in a lab at Tel Aviv University, a mouse is glowing. Not the whole mouse, just a faint bloom of fluorescence creeping up from its nose toward its skull, captured frame by frame by an imaging system that can see a peptide travel through living tissue. Five animals at a time, photographed across two and a half hours. And the glow, it turns out, is not the same in every animal. In the females it depends on something the researchers very nearly didn’t bother to check: where each mouse was in her estrous cycle.
That single decision, to track the hormonal week of each female rather than averaging them all together, is the quiet hinge of a new study in Genomic Psychiatry. And it may rescue a drug that the field had more or less given up on.
The drug is davunetide, also called NAP, a short fragment of a protein the brain manufactures to protect its own wiring. For years it carried real hope among researchers working on tauopathies, the family of disorders, Alzheimer’s among them, where a protein called tau misbehaves and neurons start to fail. Davunetide is supposed to shore up the microscopic scaffolding inside nerve cells, the microtubules along which cellular cargo gets shuttled about. In theory, it should help. In the largest trial that tested it, against a savage condition called progressive supranuclear palsy, it flatly did not. That result read like a full stop.
The average that was lying
But Professor Illana Gozes, who runs the Elton Laboratory for Molecular Neuroendocrinology at Tel Aviv University, had already spotted something odd in the rubble of those earlier trials. Pull the data apart by sex and the women seemed to respond where the men did not. That is the kind of observation that usually evaporates into statistical noise. Occasionally, though, it points at something real underneath. To find out which, her team did something almost stubbornly simple. They watched the drug move.
Tagging the peptide with a fluorescent marker and running it through a live imaging system, they followed intranasal davunetide as it travelled into the heads and bodies of mice. The females weren’t in one uniform state, of course. They were cycling. So the researchers read each animal’s place in the cycle the old-fashioned way, vaginal smears under a microscope, matched by eye against a published template. Best judgement, basically. They say as much.
The pattern was clean enough to be slightly unnerving. During proestrus and estrus, the phases when estrogen runs highest, the females took up considerably more drug in the head region than the males did. Sharpest in proestrus, where the head measurement split the sexes with a p-value of 0.00029, and the head-to-body ratio with a p-value of roughly 0.000004. As the cycle slid into metestrus, estrogen falling toward its floor, the difference between the sexes just… faded. The hormone wasn’t a bystander. It tracked the drug, step for step.
What reaches the head, importantly, is not the same as what circulates in the body. In a larger mixed group of five males and five females imaged without sorting by cycle, the females still showed higher head uptake at every single time point, with a head-to-body ratio carrying a p-value of about 0.000009. The body told one story; the head told another. That gap is itself the interesting bit.
A signal in people, and a warning in the dying
Mice are not women, and the authors are the first to say so. So they reached for a human pharmacokinetic dataset from an older study of intranasal davunetide in healthy adults: two men, six women. Tiny. Nobody is pretending otherwise. Yet the direction held. The women trended toward higher peak concentrations, the highest female peak more than double the highest male one. The men, meanwhile, hung onto the drug longer. Group the first two days together and that longer half-life in men reaches significance (p = 0.0057), while the roughly twofold higher peak in women stays a trend (p = 0.1081). A signal, not a verdict. But a signal pointing the same way as the mice.
One detail from the animal work refuses to sit politely in the methods section, even though that’s where the authors put it. In the experiments on elderly mice, the males kept dying mid-procedure. The researchers note it plainly, an increased male vulnerability, and fold it into the protocol rather than dressing it up. It is the kind of asymmetry that makes an abstract claim about sex differences suddenly, awkwardly physical. Whatever separates these bodies, it is different enough to matter right at the edge of life.
The mechanistic threads, when you gather them, all seem to pull in one direction. Estrogen shapes the integrity of the blood-brain barrier. The microtubules davunetide targets help build that barrier, and estrogen reins in their excess growth. ADNP, the parent protein, is regulated by the estrous cycle and in turn helps regulate sex hormones. “These sex-specific differences likely reflect a combination of hormonal regulation, tissue distribution, nasal physiology, and blood-brain barrier function,” the authors write, a sentence that politely refuses to name a single culprit. The drug crosses the delicate vessels of the nose and rides the bloodstream toward the brain, and that passage leans on vascular tone, which leans, in the end, on estrogen.
The limitations are real and the paper doesn’t bury them. Davunetide is still investigational. The mouse experiments often pitted two or three females against a lone male. The human cohort was minuscule. The cycle staging came down to judgement calls made with the naked eye. The authors list each one, and that candour is part of what makes the bigger argument land. They aren’t claiming a cure. They’re claiming the variable everyone averaged away was quietly carrying information.
If they’re right, it runs well past this one molecule. Alzheimer’s, the major tauopathy, strikes women at roughly twice the rate of men. A field that designs trials and sets doses without accounting for sex and hormonal state may keep churning out flat averages that hide real, living effects, and may keep shelving drugs that work for someone, just not for everyone at once. “Optimizing neuroprotective strategies will require purposeful accounting for biological sex as a core variable,” the authors conclude. We have spent a long time treating the body as one body. This modest, careful piece of work hints at something the clinic has been slow to take on board: a drug can be right for a person and wrong for the week. The woman in that failed trial and the man beside her were never, it seems, taking quite the same medicine. They only thought they were.
Source: Blatt J, Guz LS, Shabat D, Gozes I. Intranasal bioavailability is estrous-cycle regulated: Davunetide as a case study. Genomic Psychiatry, 2026. https://doi.org/10.61373/gp026r.0039
Frequently Asked Questions
Why would the same drug behave so differently in men and women?
The study points to a tangle of factors rather than one switch: estrogen’s effect on blood vessel tone and the blood-brain barrier, differences in nasal anatomy, and how tissue distributes the peptide. Because estrogen rises and falls across the cycle, a woman’s uptake of an intranasal drug may shift week to week, which is exactly the variation a mixed-sex trial tends to erase. It’s a reminder that “one dose for everyone” can quietly mean “the right dose for no one.”
Does this mean davunetide actually works after all?
Not proven, no. The work shows that more of the drug reaches the head in females when estrogen peaks, and that women in an old human dataset hit higher peak concentrations, which could explain why earlier sex-split analyses hinted at a female benefit. But bioavailability is not the same as a cure, the human sample was tiny, and the drug remains investigational. What it really argues for is a fresh trial designed around sex and hormonal state.
Why does estrogen keep coming up in brain-drug research?
Estrogen influences cerebral blood flow and the permeability of the blood-brain barrier, the gatekeeper deciding what gets into brain tissue. It also restrains the runaway growth of microtubules, the very structures this drug is meant to stabilise. That makes hormonal state a plausible lever on whether a brain-targeted compound reaches its destination at all.
Could this change how trials are run more broadly?
That’s the larger claim. If averaging men and women together can mask a real effect in one group, then any condition with a strong sex bias, and Alzheimer’s hits women at about twice the male rate, may be poorly served by sex-blind trial design. The authors are essentially asking the field to treat biological sex as a core variable rather than a footnote.
