Dendrites are the branching extensions of a nerve cell that receive incoming signals from other neurons, functioning as the primary listening posts of the nervous system. Each neuron typically sprouts multiple dendrites that fan out like the limbs of a tree, and that resemblance is where the name comes from: the Greek word “dendron” means tree. But calling dendrites mere antennae undersells what they do. Research over the past few decades has revealed that dendrites actively shape and compute information before it ever reaches the cell body, making them far more than passive wiring.
The Basic Shape and Why It Matters
A dendrite begins as a thick trunk emerging from the neuron’s cell body and progressively tapers as it branches into thinner and thinner segments. This tree-like architecture is not random. Branches sprout from the dendritic shaft at preprogrammed points, and a complex mix of external signals then controls how far and in which direction each branch extends, ensuring the neuron covers its intended territory without wandering into a neighbor’s space.1Developmental Biology. Branching mechanisms shaping dendrite architecture The branching pattern even follows an internal spacing rule: neurons contain periodic structural elements arranged in a repeating pattern with a wavelength of roughly 20 micrometers, and new branches tend to emerge within specific segments of that pattern.2PubMed Central. Dendrites contain a spacing pattern
Different types of neurons produce dramatically different dendritic trees. Some interneurons have compact, bushy arbors that sample a tiny patch of brain tissue. Purkinje cells in the cerebellum, by contrast, spread enormous flat fans of dendrites that are the largest dendritic arbor of any recorded neuron in the human brain.3PubMed Central. Non-allometric expansion and enhanced compartmentalization of Purkinje cell dendrites in the human cerebellum The shape of a neuron’s dendritic tree determines what information it can sample and how it processes that information, so form and function are deeply intertwined.
Dendritic Spines, the Tiny Protrusions Where Synapses Live
Scattered along most dendrites are small bumps called dendritic spines. These mushroom-shaped, thorn-like, or thin protrusions are typically a micrometer or less in length, and they serve as the receiving end for the vast majority of excitatory synapses in the mammalian brain.4PubMed. Regulation of spine morphology and synaptic function by LIMK and the actin cytoskeleton Each spine is packed with actin filaments that give it structural support, allow it to change shape, and help organize the signaling molecules needed for communication.5PubMed Central. Actin cytoskeleton in dendritic spine development and plasticity
Spines are not static. They can grow, shrink, change shape, or disappear entirely depending on how active their synapse is. This structural flexibility is central to how the brain rewires itself during learning. Research using advanced microscopy has shown that microtubules, once thought to stay inside the main dendritic shaft, actually push into spines during bursts of synaptic activity, and these invasions appear to be linked to changes in how the synapse functions.6PubMed Central. The dynamic cytoskeleton: backbone of dendritic spine plasticity So the internal skeleton of a spine is constantly being remodeled in response to what the neuron experiences.
How Dendrites Handle Electrical Signals
Textbook descriptions sometimes portray dendrites as simple cables that passively funnel electrical signals toward the cell body. The reality is more interesting. Dendrites are studded with their own voltage-sensitive ion channels, including sodium and calcium channels, that actively amplify or dampen signals as they travel.7PubMed Central. Dendritic ion channel trafficking and plasticity When a synapse on a dendrite fires, the resulting voltage change can open sodium channels near the peak of the signal and calcium channels during both the peak and the decay, adding energy to the signal locally.8PubMed. Synaptic activation of voltage-gated channels in the dendrites of hippocampal pyramidal neurons
Signals also travel in the reverse direction. When a neuron fires an action potential at the cell body, that spike can propagate backward into the dendritic tree. In hippocampal neurons, backpropagating spikes activate calcium channels deep in the dendrites, which can then trigger a prolonged inward current that switches the neuron’s firing mode from single spikes to bursts of multiple action potentials.9PubMed. Dendritic voltage-gated ion channels regulate the action potential firing mode of hippocampal CA1 pyramidal neurons This backpropagation is thought to be one way dendrites “know” that their neuron just fired, letting them update their synapses accordingly.
Because dendrites are long and thin, a signal naturally weakens as it travels toward the cell body. Early computational models of neurons worked out detailed attenuation curves showing how much a voltage signal shrinks depending on where on the dendritic tree it originates.10PubMed Central. The attenuation of passively propagating dendritic potentials in a motoneurone cable model Active ion channels partially compensate for this loss, but the location of a synapse on a dendrite still matters for how strongly it influences the neuron’s output.
Dendrites as Miniature Computers
Perhaps the most striking finding of modern neuroscience is that individual dendritic branches can perform their own computations before the cell body ever weighs in. When roughly 10 to 50 neighboring synapses on a single dendritic branch fire at the same time, they can trigger a local regenerative event called an NMDA spike. These spikes are large in amplitude locally, around 40 to 50 millivolts, and can last for hundreds of milliseconds.11PubMed Central. The decade of the dendritic NMDA spike That makes them a form of local decision-making: a single dendritic branch essentially “votes” on whether its cluster of inputs is meaningful enough to pass along.
Human dendrites can generate these NMDA spikes as well, though the threshold is higher than in rodents. Simulations suggest this is primarily because human dendrites are wider in diameter, meaning they need more simultaneous synaptic input to reach the voltage required.12PubMed. High synaptic threshold for dendritic NMDA spike generation in human layer 2/3 pyramidal neurons This finding hints that human neurons may process clustered inputs differently from rodent neurons, which matters for understanding how our larger, more complex brains handle information.
Dendrites also integrate inputs differently depending on where those inputs land. Synapses close to the cell body sum their signals in a relatively straightforward, linear fashion and require tight timing between inputs for effective summation. Synapses on more distant dendritic branches, by contrast, are amplified with higher gain and can be integrated over broader time windows.13PubMed Central. Synaptic integration gradients in single cortical pyramidal cell dendrites This gradient means a single neuron is not just a one-size-fits-all calculator. Different parts of its dendritic tree apply different rules to incoming signals. The result is that correlated activity at a synapse can bidirectionally change how that synapse’s input is summed with its neighbors, affecting not only the strength of the connection but also the local integration rules.14Neuron. Bidirectional Changes in Spatial Dendritic Integration Accompanying Long-Term Synaptic Modifications
Dendrites and Memory Formation
The ability of dendritic spines to change their shape and molecular makeup is the structural basis of long-term memory. During long-term potentiation, the cellular process thought to underlie memory storage, individual spines undergo a choreographed sequence of remodeling. First, the actin skeleton inside the spine is rapidly reorganized and a protein called cofilin floods in. Then cofilin locks onto the actin filaments, stabilizing the spine in its enlarged state. Finally, in a later phase that depends on new protein synthesis, the scaffolding of the postsynaptic density is itself remodeled.15PubMed Central. Structural and molecular remodeling of dendritic spine substructures during long-term potentiation This multi-step process explains why some memories take time to consolidate: the molecular machinery needs hours to complete all the structural changes that make a strengthened synapse permanent.
During this remodeling, spines can also sprout temporary finger-like extensions from their heads called spine head protrusions. These protrusions depend on matrix metalloproteinase enzymes (specifically MMP-9) and on microtubule dynamics, and spines that grow them gain additional receptors for the neurotransmitter glutamate, effectively making the synapse more sensitive.16PubMed Central. Matrix metalloproteinases regulate the formation of dendritic spine head protrusions during chemically induced long-term potentiation The capacity for dendrites to manufacture some of the proteins they need on site, rather than waiting for the cell body to ship them, gives local spines a degree of autonomy in updating themselves.17PubMed. Local translation of classes of mRNAs that are targeted to neuronal dendrites
How Dendrites Grow Without Tangling
During brain development, dendrites face a spatial puzzle: they need to spread out and cover their target area without overlapping their own branches or those of neighboring neurons of the same type. The solution involves a process called self-avoidance, where sister branches from the same neuron recognize each other and retract upon contact. In mammals, this recognition is managed by a family of cell-surface molecules called clustered protocadherins. Each neuron expresses a near-unique combination of these molecules, giving it a molecular “identity tag.” When two branches from the same neuron touch, matching protocadherins trigger a retraction signal.18PubMed. Gamma-protocadherins regulate dendrite self-recognition and dynamics to drive self-avoidance When the protocadherins do not match, the branches coexist without issue.
Live imaging of retinal neurons has captured this process in real time. Dendritic protrusions that contact a sibling branch linger longer and sometimes form transient loops before retracting. When the protocadherin genes are deleted, these self-contacting retractions fail, and branches bundle together into a tangled mess that disrupts the neuron’s overall shape.18PubMed. Gamma-protocadherins regulate dendrite self-recognition and dynamics to drive self-avoidance In fruit flies, the surrounding tissue also plays a role: the epidermis secretes signaling molecules called semaphorins that help pin dendrites flat against the body wall, ensuring they stay in a two-dimensional plane where contact-based self-avoidance can work properly.19PubMed Central. Epidermis-Derived Semaphorin Promotes Dendrite Self-Avoidance by Regulating Dendrite-Substrate Adhesion in Drosophila Sensory Neurons
When Dendrites Send Instead of Receive
Although dendrites are best known for receiving signals, some neurons break that rule and release neurotransmitters from their dendrites as well. This dendritic release is surprisingly widespread. It includes classical neurotransmitters, neuropeptides like oxytocin and vasopressin, and signaling molecules like nitric oxide, and it occurs in multiple brain regions.20PubMed Central. Dendritic Release of Neurotransmitters
One of the best-studied examples is the olfactory bulb, where mitral cells and granule cells form reciprocal synapses between their dendrites. Glutamate released from a mitral cell’s dendrite excites a granule cell’s dendrite, which then sends an inhibitory signal right back.21PubMed. Olfactory reciprocal synapses: dendritic signaling in the CNS This dendrite-to-dendrite conversation allows local inhibitory circuits to operate without the signal ever traveling to a cell body. The glutamate released at these dendrodendritic synapses can even spill over to excite neighboring mitral cells, creating a form of lateral communication.22PubMed. A dendrodendritic reciprocal synapse provides a recurrent excitatory connection in the olfactory bulb These reciprocal dendritic synapses are critical for sharpening odor signals in the nose-to-brain pathway.
What Happens When Dendritic Spines Go Wrong
Because so much of brain function depends on spine structure and plasticity, it is not surprising that spine abnormalities show up in a range of neurological and psychiatric conditions. In autism spectrum disorders, schizophrenia, and Alzheimer’s disease, changes in spine density, shape, or turnover have been identified as important substrates of the disease process.23PubMed Central. Dendritic spine pathology in neuropsychiatric disorders The specifics differ: some conditions feature too many immature-looking spines, others show a loss of spines altogether. In schizophrenia, for instance, researchers have traced a causal chain from susceptibility genes to decreases in dendritic spines and synaptic terminals and, ultimately, to the loss of gray matter visible on brain scans.24PubMed. Schizophrenia: susceptibility genes, dendritic-spine pathology and gray matter loss
Neuroinflammation adds another layer. Microglia, the brain’s resident immune cells, normally help prune and maintain synapses. But when they become overactivated, microglia release inflammatory molecules that can damage synaptic structure and connectivity, disrupting the plasticity that dendrites depend on.25PubMed Central. Microglia-orchestrated neuroinflammation and synaptic remodeling: roles of pro-inflammatory cytokines and receptors in neurodegeneration Understanding these dendritic and spine-level changes is becoming increasingly important for developing treatments that target the structural roots of brain disorders rather than just managing symptoms.
The Support Crew Around Dendrites
Dendrites do not operate in isolation. Astrocytes, star-shaped glial cells, wrap their fine processes around dendritic spines and synaptic terminals, forming what is sometimes called the tripartite synapse. Astrocytes respond to neurotransmitters released at the synapse and can release their own signaling molecules that fine-tune synaptic function.26PubMed Central. Structural and functional plasticity of astrocyte processes and dendritic spine interactions When synaptic activity triggers long-term potentiation, the astrocytic processes wrapped around the activated spines become more motile and undergo their own structural remodeling. These changes affect how the astrocyte regulates transmission at that synapse going forward, creating a form of metaplasticity: the astrocyte’s new position and activity change the rules for future synaptic strengthening or weakening at that site.26PubMed Central. Structural and functional plasticity of astrocyte processes and dendritic spine interactions
Both astrocytes and microglia participate in the development and ongoing plasticity of dendritic spines.27PubMed. Glial Cell Modulation of Dendritic Spine Structure and Synaptic Function Glial involvement means that the health and behavior of non-neuronal cells directly affect how well dendrites can do their job, which may partly explain why diseases with prominent neuroinflammation, such as Alzheimer’s, also feature heavy dendritic pathology.
Powering All That Work
The electrical signaling and molecular remodeling that dendrites perform require substantial energy. Mitochondria, the cell’s power generators, are distributed throughout dendritic branches, and during brain development they settle into stable positions along the dendritic shaft. Most of the energy they supply goes to synapses, which are also the primary sites of calcium influx in dendrites.28eLife. Dendritic mitochondria reach stable positions during circuit development When mitochondria are sparse or dysfunctional in a particular dendritic segment, that region can struggle to maintain synaptic transmission. This energy angle adds yet another dimension to dendritic health: it is not only the signaling molecules and structural proteins that matter, but also the local fuel supply.
Do All Animals Build Dendrites the Same Way
Dendrites are found across the animal kingdom, from insects to humans. The question of whether the dendrites of a fruit fly are truly equivalent to those of a mammal is more than academic. Comparative analyses have proposed that dendrites in vertebrates and higher invertebrates share a common evolutionary origin, with the apparent differences in layout largely explainable by how cell bodies shifted position as arthropods and vertebrates diverged from a shared ancestor.29Developmental Biology. Are dendrites in Drosophila homologous to vertebrate dendrites? This matters because much of what we know about dendritic self-avoidance, branching rules, and molecular signaling was discovered first in fruit flies or worms, and its relevance to human neurons depends on this evolutionary continuity. The conservation of core mechanisms, like the protocadherin-mediated self-avoidance system in mammals and the DSCAM-based system in flies, supports the idea that the fundamental strategies for building and organizing dendrites are ancient.
Dendrites as Inspiration for Artificial Intelligence
The computational power of biological dendrites has caught the attention of researchers building artificial neural networks. Most artificial neurons are modeled as simple point units that sum their inputs and apply a threshold, ignoring the rich, nonlinear processing that real dendrites perform. Recent work on spiking neural networks has tried to close that gap by explicitly modeling dendritic morphology and nonlinear integration. These biologically inspired “dendritic spiking neurons” show substantially higher expressiveness than standard point neurons and can be integrated into modern deep network architectures.30arXiv. Scalable Dendritic Modeling Advances Expressive and Robust Deep Spiking Neural Networks The implication is that the computational tricks dendrites evolved over hundreds of millions of years might offer real advantages for machine learning systems that current architectures leave on the table.