Skip to content
PeptideHormone

The neuron behind the hot flash

In 1979 researchers noticed that menopausal hot flashes arrive in step with pulses of LH, and could not say why. The answer was a single type of hypothalamic neuron that makes three peptides: kisspeptin, neurokinin B and dynorphin. It is the pulse generator of the reproductive axis, estrogen is its brake, and one of its outputs runs to the brain’s thermostat. How a peptide named after a chocolate led to that cell, how the idea was tested in both directions in people, and why the drugs that came out of it are pills that contain no peptide at all.

8 min read · reviewed November 2026

A clue that sat for thirty years

In 1979 a group in San Diego wired menopausal women to skin-temperature sensors, drew blood every few minutes, and noticed something odd. Each hot flash arrived together with a pulse of luteinizing hormone. The two kept the same clock (Casper et al., Science, 1979).

The obvious reading, that LH causes the flash, was wrong, and people knew it quickly: women whose pituitary had been removed, and who therefore made no LH at all, still had hot flashes. So the flash and the pulse were not cause and effect. They were two outputs of something upstream, something in the brain that ticked, and that nobody could yet name. The clue sat there for three decades, until a peptide found in a cancer lab led back to it.

Named after a chocolate

In 1996 a team in Hershey, Pennsylvania, found a gene that stopped melanoma cells from spreading. They named it KiSS-1: SS for suppressor sequence, and the rest for the town’s most famous product (Lee et al., 1996). Five years later its peptide product was matched to an orphan receptor, GPR54 (Ohtaki et al., Nature, 2001). It looked like a cancer story.

Then, in 2003, families whose children never entered puberty were found to carry broken copies of that receptor (Seminara et al., NEJM, 2003). Kisspeptin turned out to be the signal that switches on GnRH and with it the whole reproductive axis, a story told in the pulse is the message. The same genetic approach soon turned up a second peptide. Other families with the same failure of puberty had intact kisspeptin signaling but carried mutations in neurokinin B or its receptor, NK3R (Topaloglu et al., Nat Genet, 2009). Two different peptides, one identical disease. They had to be working in the same place.

One cell, three peptides

They were working in the same cell. In the arcuate nucleus of the hypothalamus (called the infundibular nucleus in humans) sits a population of neurons that make kisspeptin, neurokinin B and a third peptide, the opioid dynorphin. In 2010 they were given a name built from the three initials: KNDy neurons, pronounced “candy” (Lehman et al., 2010).

The working model gives each peptide a job. The neurons are wired to one another, and neurokinin B is how they excite their neighbors: a little activity spreads and the whole population fires together. Dynorphin is the delayed brake that shuts the burst down. And kisspeptin is the output, released onto GnRH neurons with each burst. Start, stop, send. The result is a pulse roughly every hour or so, the rhythm that the pituitary reads as an instruction to release LH. The thing upstream that ticked had been found.

Estrogen sits on top of this as the restraint. KNDy neurons carry estrogen receptors, and circulating estrogen holds their activity down. That is negative feedback: the ovary telling the brain it has heard enough.

What happens when the brake comes off

At menopause the ovaries stop making estrogen, and the restraint is gone. A pathologist, Naomi Rance, had seen the consequence years before anyone had the vocabulary for it. In the hypothalami of postmenopausal women, a specific set of neurons was physically enlarged. They were the ones containing neurokinin B, their cross-sectional area was nearly double that seen in premenopausal women, and their neurokinin B gene expression was sharply up (Rance & Young, 1991). These were KNDy neurons, running without a brake.

The link to temperature came from anatomy. KNDy neurons do not only project to GnRH neurons. They also send fibers to the preoptic area, the brain’s thermostat, where cells carry the neurokinin B receptor. Rance proposed that the overactive neurons were driving that thermostat with every burst, triggering the body’s heat-dumping responses when there was no heat to dump (Rance et al., 2013). That fit the physiology. Women with hot flashes have an unusually narrow thermoneutral zone, so a tiny rise in core temperature that would normally go unnoticed sets off flushing and sweating (Freedman, 2014).

It also explained 1979. One burst from one population of cells goes two ways: kisspeptin toward the GnRH neuron, which becomes an LH pulse, and neurokinin B toward the thermostat, which becomes a flash. Same clock, because it is the same cell.

Estrogenlost at menopauserestrainsKNDy neuronarcuate nucleusneurokinin B starts the burst,dynorphin stops itkisspeptinGnRH neuronone pulse per burstLH pulsefrom the pituitaryneurokinin BThermostatpreoptic area · NK3RHot flashflushing · sweatingNK3 receptor blockers act here(fezolinetant, elinzanetant)same burst,same clock
One population of neurons has two outputs. Kisspeptin drives the reproductive pulse; neurokinin B reaches the thermostat. Without estrogen to restrain the cell, the second output becomes a hot flash.

Testing it in both directions

A hypothesis like that makes two predictions, and both were tested in people. If neurokinin B triggers the flash, then giving it to someone who does not have hot flashes should produce one. In a small double-blind crossover study, ten healthy premenopausal women received a thirty-minute infusion of neurokinin B and, on another day, an inactive one. Eight of the ten flushed during neurokinin B, with the rise in heart rate and skin temperature typical of a menopausal flash. None did on the inactive infusion (Jayasena et al., 2015).

And if neurokinin B triggers the flash, blocking its receptor should stop it. The same London group gave 37 menopausal women, each having seven or more flashes a day, four weeks of an oral NK3 receptor antagonist and four weeks of placebo in random order. The drug cut the weekly number of hot flashes by 45 percentage points more than placebo (Prague et al., Lancet, 2017). Three participants had a temporary rise in liver enzymes. That detail would matter later.

Why this is strong evidence

Adding the peptide produced the symptom in people who did not have it. Blocking its receptor removed the symptom in people who did. Most mechanisms in this catalog never get tested in both directions, in humans, with a placebo.

The drugs, and their cost

The first NK3 blocker to reach the market was fezolinetant. In its phase 3 trial, women averaging seven or more moderate-to-severe flashes a day had about two and a half fewer per day on the higher dose than on placebo by week twelve, with an effect visible in the first weeks (Lederman et al., Lancet, 2023). It was approved in the United States in 2023 as the first non-hormonal drug aimed at the mechanism of hot flashes. A second compound, elinzanetant, blocks both the NK3 receptor and the related NK1 receptor. In two phase 3 trials it reduced flashes by about three more per day than placebo at twelve weeks and improved sleep (Pinkerton et al., JAMA, 2024), and it was approved in 2025 (FDA label).

The liver signal from the first small trial did not go away. In 2024 the FDA warned of rare cases of serious liver injury with fezolinetant (FDA safety communication), and the labels for the class call for liver blood tests before and during treatment. The effect size deserves a sober reading too. These drugs reduce hot flashes; in the trials they did not abolish them, and the placebo groups improved substantially as well. Estrogen remains the most effective treatment for those who can take it. What the NK3 blockers changed is that there is now a targeted option for those who cannot.

The peptide was the map, not the medicine

There is a twist worth noticing in a reference about peptide hormones. Neither of these drugs is a peptide. They are small molecules, taken by mouth, that block a peptide’s receptor. And the peptide whose discovery opened the whole field, kisspeptin, is not their target at all. Its own clinical use is elsewhere, as an experimental trigger for egg maturation in IVF, where the aim is to add a signal rather than remove one.

The pattern is the mirror of what usually happens here. Most stories in this catalog are about supplying a peptide: making it last, getting it in, engineering its half-life. This one is about silencing one, and for that job a pill that sits in the receptor is the better tool. The peptides did the scientific work. They marked the cell, the circuit and the receptor. The drug that resulted contains none of them.

  • Hot flashes and LH pulses share a clock. They coincide, but LH does not cause the flash. Both are outputs of the same hypothalamic pulse generator.
  • The generator is the KNDy neuron. One cell type makes kisspeptin, neurokinin B and dynorphin. Neurokinin B starts each synchronized burst, dynorphin stops it, and kisspeptin carries it to GnRH neurons.
  • Estrogen is the brake. When estrogen falls at menopause these neurons enlarge and overfire. Their neurokinin B output also reaches the brain's thermostat and triggers heat-loss responses.
  • Tested both ways in people. Infusing neurokinin B caused flashes in women who did not have them. Blocking its receptor reduced flashes in women who did.
  • The drugs are not peptides. Fezolinetant and elinzanetant are oral small molecules that block the neurokinin receptor. They reduce flashes by a few per day beyond placebo and carry a liver-monitoring requirement.

Educational reference on mechanism, summarized from public scientific literature and clinical-trial disclosures and simplified in places. Not individualized medical advice or a prescription. Where a dose or regimen appears it is reported from its source as research literacy — so you can read the evidence yourself — not an instruction to dose. Specific compounds and trials are named to explain the science; verify any claim against the linked primary sources.