For decades, smell stood apart from the other senses.

Scientists could point to orderly maps in the eye, the ear, and the skin, showing how sensory cells are arranged to capture information and how those patterns connect to the brain. Smell never fit that picture. The usual view held that odor receptors in the nose were only loosely sorted into broad zones, with a lot of randomness inside them.

A new study in mice now argues that this picture was badly incomplete. Researchers led by Sandeep (Robert) Datta at Harvard Medical School report that the nose contains a detailed receptor map, one that places more than 1,000 smell receptor types in overlapping but distinct horizontal bands running from the top of the nose to the bottom. The work, published in Cell, suggests that smell may be organized more like vision, hearing, and touch than scientists had realized.

“Olfaction is super-mysterious,” Datta, a professor of neurobiology in the Blavatnik Institute at Harvard Medical School, said in background material describing the findings.

That mystery has lingered for a long time. Mice have around 20 million olfactory neurons and more than 1,000 kinds of smell receptors, each tuned to its own subset of odor molecules. By comparison, color vision relies on just three main receptor types. Scientists first identified smell receptor types in 1991, but after years of work, the field still lacked a clear map of where those receptors sat in the nose.

The new analysis tackled that problem at unusual scale. Datta and colleagues combined single-cell sequencing with spatial transcriptomics, letting them identify which receptors individual neurons expressed and where those neurons sat in the olfactory tissue. In all, they examined about 5.5 million cells from more than 300 mice, including roughly 2.3 million olfactory sensory neurons.

“This is now arguably the most sequenced neural tissue ever, but we needed that scale of data in order to understand the system,” Datta said.

What emerged was not the old picture of broad zones with random receptor placement inside them. Instead, each olfactory sensory neuron subtype, defined by the receptor it expresses, carried a distinct positional identity along the dorsoventral axis, meaning from the dorsal, or upper, region of the nose toward the ventral, or lower, region. The team describes this pattern as a receptor map made of about 1,100 distinguishable peaks of receptor expression.

These receptor populations overlapped, but not in a chaotic way. Each receptor type had its own preferred location, and those preferred locations stayed strikingly consistent from mouse to mouse. The rank order of receptor positions across samples was nearly unchanged, a level of stereotypy that cuts against the long-standing idea that receptor choice in the epithelium is mostly random within a zone.

The spatial pattern also became visible when the tissue was examined directly. Using MERFISH, a multiplexed imaging method, the group found that receptor expression varied smoothly across the epithelium rather than clustering into a handful of discrete territories. The result looked less like islands and more like layered stripes.

Datta summed up the shift plainly: “Our results bring order to a system that was previously thought to lack order, which changes conceptually how we think this works.”

The work did not stop at describing the map. It also asked how such a map gets built.

The answer appears to involve retinoic acid, a signaling molecule known to help control gene activity during development. In the tissue beneath the smell-sensing epithelium, the team found a gradient of retinoic acid-related activity that lined up with the positions of receptor types above it. Olfactory stem cells and precursor cells carried the machinery needed to respond to that signal.

When the researchers pushed retinoic acid signaling upward during regeneration, the receptor map shifted one way. When they reduced it, the map shifted the other way. Adding or removing the signal changed which receptor identities appeared where, effectively moving the map up or down along the nose.

“We show that development can achieve this feat of organizing a thousand different smell receptors into an incredibly precise map that’s consistent across animals,” Datta said.

The study argues that this positional information appears early, before a mature olfactory neuron settles on the single receptor it will ultimately express. In other words, the precursor cells seem to “know” where they are before receptor choice is finalized. That matters because each mature olfactory neuron expresses only one receptor out of the mouse genome’s 1,172 functional olfactory receptor genes.

The researchers found that precursor cells already carried graded transcriptional signatures linked to position. Those signatures influenced which receptors were weakly co-expressed early on, which receptors were later silenced, and which one was eventually chosen for strong expression. The process was not perfectly fixed, but it was clearly biased by location.

That helps explain a central feature of the new map. Each receptor occupies a tight distribution of positions, not a single point. Cells do not always choose the most spatially favored receptor, but they usually choose one close to it.

The most intriguing part may be where the map leads next.

Information from olfactory sensory neurons flows into the olfactory bulb, the first major processing station in the brain for smell. There, neurons expressing the same receptor send their axons to structures called glomeruli, creating a patterned map of odor identity. That brain map was already known to be fairly precise and stable across animals. What had remained unclear was how a seemingly fuzzy peripheral system in the nose could produce such exact targeting in the brain.

This work helps bridge that gap.

The same dorsoventral transcriptional signature that marked receptor position in the epithelium also predicted where corresponding glomeruli would sit in the olfactory bulb. When the team colored glomeruli by the positional scores of the receptors feeding them, neighboring glomeruli tended to share similar scores, and glomerular position could predict receptor identity with high accuracy. A second gene expression axis, tied to apical-basal position in the epithelium, also improved those predictions.

Taken together, the results suggest that the nose already contains the spatial instructions needed to organize the first stages of smell processing. The system is not waiting for odor receptors alone to dictate where axons should go. Instead, spatial identity in the tissue appears to coordinate both receptor choice and axon targeting.

That is a major conceptual shift. The usual way of thinking placed heavy weight on the receptor itself as a late-stage driver of neuronal identity. This study instead points to epithelial position as an organizing principle that acts earlier and more broadly.

A separate study led by Catherine Dulac’s lab at Harvard University, published in the same issue of Cell, reached consistent conclusions, adding support to the broader picture.

The findings answer one old question, but they open others.

The team is now investigating why the receptor stripes appear in this particular order. The study also explored whether related receptors tend to sit near one another or whether receptors that respond to similar chemicals cluster in space. The links turned out to be weak overall. Some closely related receptors that likely arose through gene duplication did sit near one another, and acids mainly activated dorsal class I receptors, but most odors still activated receptor populations spread widely across the epithelium.

That weak chemotopy may be useful. The analysis suggests that distributing odor tuning across the tissue could make the system more resilient to local damage. If receptors with similar roles are not all packed together, a small injury may not wipe out an entire type of odor response.

The work also points to limits. The study was done in mice, not people. The authors are now examining human tissue to see how well the same kind of receptor map holds across species. They also note that their data do not rule out the possibility that other odor features, including ones tied to meaning or valence, could show their own spatial organization.

There is also biological variability inside the apparent order. Ventral receptor types showed broader positional distributions than dorsal ones, and receptor choice retained a degree of stochasticity. The map is precise, but it is not rigid.

Still, the broader message is clear. Smell is no longer the odd sense out.

This receptor map could matter well beyond basic biology.

Datta said the findings offer foundational knowledge for efforts to treat smell loss, an area where medicine still has few options. The consequences of losing smell can reach far beyond missing pleasant odors. Smell helps warn people about danger, shapes taste, and affects emotional life. The source material notes that smell loss is linked to an increased risk of depression.

“We cannot fix smell without understanding how it works on a basic level,” Datta said.

That need became especially visible during the COVID-19 pandemic, when smell loss drew broad public attention. Datta’s group has previously studied what causes that symptom. In the new work, the researchers suggest that understanding the map of receptor identities in the nose could guide future attempts to rebuild or bypass damaged smell circuits, including possible stem cell therapies or brain-computer interfaces.

“Smell has a really profound and pervasive effect on human health, so restoring it is not just for pleasure and safety but also for psychological well-being,” Datta said. “Without understanding this map, we’re doomed to fail in developing new treatments.”

Research findings are available online in the journal Cell.