2026 Kavli
Prize in
Neuroscience

Citation from The Kavli Prize Committee in Neuroscience

“Local Translation in Neurons – For the discovery of local protein translation in neurons and establishing its importance for brain development and plasticity”

Neurons are remarkable cells with processes extending far from the cell body. Along these processes, individual neurons form, maintain and modify thousands of synaptic connections, each of which might have different dynamic properties. How such precise and localized control is achieved was a fundamental question that Christine Holt, Kelsey Martin, Erin Schuman and Oswald Steward resolved.

Through pioneering work spanning several decades, they discovered that mRNAs can be translated locally by ribosomal machinery present in axonal growth cones, synapses and dendrites. This finding substantially transformed our understanding of how neurons work. Local translation provides a mechanism for spatial compartmentalization in which proteomes are remodelled independently from the cell body. This allows precise responses to 2 guidance of cues during axonal growth in development and enables rapid and persistent synapsespecific changes. Their work has far-reaching implications not only for neuroscience but also for cell biology as a whole.

Oswald Steward's electron microscope studies of mammalian hippocampal neurons established the structural basis for local protein translation. He demonstrated that polyribosomes are present at the base of dendritic spines and more generally at postsynaptic sites. His observations raised the possibility of a direct link between synaptic activity and local protein synthesis. This work built a foundation for subsequent functional studies that uncovered molecular mechanisms of mRNA localization and activity-dependent regulation of translation at synapses. Together with Schuman, Martin and Holt, Steward established local protein synthesis as a key mechanism for neuronal guidance and plasticity.

Christine Holt showed that local protein synthesis is required for developing tips of axons (i.e., growth cones) in the vertebrate visual system to respond to guidance cues. She went on to demonstrate that growth cones possess a rich repertoire of mRNAs, and that specific guidance cues elicit the translation of subsets of mRNAs at these locations. Her studies further determined the translatome of axonal growth cones in vivo and provided evidence that local translation is required for terminal branching and survival of central nervous system axons. Her findings demonstrated that local mRNA-based mechanisms far from the cell body enable rapid cueinduced responses that both establish and maintain neural circuitry.

Kelsey Martin uncovered mechanisms of translational control at synapses undergoing selective synaptic strengthening. Studying branch-specific synaptic plasticity in Aplysia neurons, she showed that long-term synaptic facilitation can occur independently of the cell body and requires local protein synthesis. By focusing on a peptide signalling molecule, Martin visualized synapsespecific translation in live neurons. Her work demonstrated that separate branches of the same neuron can regulate synaptic strength independently through local translation.

Erin Schuman showed that mRNAs are translated locally in dendrites both at baseline and in response to synaptic activity. She showed that local protein synthesis is required for forms of long-lasting synaptic plasticity in the mammalian hippocampus, thus explaining how individual synapses acquire the newly synthesized proteins needed to support persistent functional changes. Schuman also created powerful molecular tools to identify and visualize newlysynthesized proteins at high resolution, thereby elucidating how mRNA levels and protein synthesis and degradation are regulated within distinct subcellular compartments. Schuman's work has a broad impact on our understanding of mRNA biology and the local control of protein synthesis, extending far beyond the nervous system.

Laureates solve longstanding mystery of our impossibly efficient brains

By Miriam Frankel, science writer

From figuring out how to ride a bike to memorising words in a foreign language – human beings can learn new things in just minutes. That’s because our brains rewire easily – connections between cells can strengthen or weaken in response to what we’re doing. But this cellular process was long considered too slow and inefficient to explain our speedy learning – with scientists stumped for an explanation.

Research by this year’s winners of the Kavli Prize in Neuroscience, however, finally solved the conundrum. Oswald Steward, Erin Schuman, Kelsey Martin and Christine Holt have shown that brain cells have a distributed function and intelligence that makes them far more powerful that previously thought. The discovery overhauls our understanding of how the brain works – and raises hopes for better treatments of a wide range of brain disorders.

Brain cells, or neurons, are a bit like trees, with appendages called axons and dendrites representing roots and branches. The axon is very long – extending up to a metre in the human body. The communication between neurons that enables learning happens at the branches of these appendages in so-called synapses – tiny junctions between neurons.

Proteins are essential for such communication. They help keep synapses connected, trigger the release of signals between them and help dial up or down these connections in response to learning.

But this leads to a logistical mystery. Scientists long believed that the proteins were only produced inside the cell body, as in most other cells. This was in fact “a strong dogma” until the late 1990s, explains Nobel laureate and chair of the 2026 Kavli Prize Committee in Neuroscience Edvard Moser. But it can take many minutes or even hours for the cell to transport a protein all the way to the synapse, making it hard to explain how we can learn things in minutes. “This would require some sort of super-intelligence to make sure every little protein got to the right place,” says Moser.

Answers began emerging in 1982 thanks to the work by Oswald Steward. Using an electron microscope, he discovered that polyribosomes (machinery that enables protein synthesis) actually existed in dendrites. This suggested that proteins could be produced locally near the synapses rather than in the cell body. However, the dogma was hard to break – many scientists assumed that this machinery wasn’t crucial or functional.

This finally changed in 1990s through work by the three other laureates. In 1996, Erin Schuman was experimenting with brain slices, adding a brain-derived neurotrophic factor which can strengthen synapses. But when she added a chemical that blocked protein synthesis, the synapse strengthening stopped almost immediately – way too quickly to involve proteins that had travelled all the way there from the nucleus. She went on to physically cut dendrites off from the cell bodies and showed they could still strengthen the synapses through this chemical – without contact with the cell body.

The idea that dendrites could produce their own proteins was referred to as “crazy” by some of her colleagues. But a year later, Kelsey Martin, Christine Holt and others backed her up. She isolated a single neuron and looked at two separate branches of its axon, applying serotonin, which boosts learning, to one of them. The result? The serontonin-treated branch developed stronger synapses as a result – no instructions from the cell body needed.

The evidence now built rapidly. A few years later, Christine Holt experimented with severed axons in the developing brain and discovered they could still navigate toward chemical cues without contact with the cell body. And this was impossible when inhibitors that prevented local protein production were added – again suggesting axons have their own machinery for supplying new proteins.

The process the laureates discovered, called local translation, involves the cell body sending mRNA molecules (essentially recipes for making proteins) to these distant parts of the cells before they are even needed – enabling the proteins to be produced rapidly on site. It’s a bit like an airport: rather than the passengers having to to go to an over-crowded central office miles away if they’ve lost a boarding pass, they can just have a new one printed directly at the gate.

These discoveries ultimately broke the dogma. Today, researchers have developed ingenious ways of watching proteins live in neurons using fluorescent techniques and other methods.

Local translation has radically changed how we view the brain. For example, it explains why it’s so hard for computer programmes to mimic it through neural networks. “A neural network consists of thousands to millions of cells,” explains Moser. “But now we realise that each of its cells may have 10,000 different variations. This means you get astronomical numbers of possibilities. So it really just adds to the power of the brain; the amount of information that it can process.”

Local translation goes wrong in many diseases and disorders of the brain. In neurodevelopment disorders such as Fragile X Syndrome, which is linked to intellectual disability and autism, for example, it can become uncontrolled. The process is also affected in diseases such as dementia or ALS, where the axons and dendrites begin to deteriorate. Clinical trials are under way to tackle these problems via local translation. “It’s very likely that this research will lead to much better treatments,” says Moser.