Written and maintained by CASRAI Editorial Board
Last updated
Short answer
Dendritic spines are small protrusions along the dendrites of spiny neurons, and they are the site of most excitatory synapses. A typical spine is 1-3 µm long, with a rounded head on a thinner neck. In the lab, the practical question is how spines are counted and sized, because the method decides whether you get a snapshot or a time course.
Definition and scope
A dendritic spine is a protrusion from a dendrite whose head receives mainly excitatory synaptic input. Reviews describe typical lengths of 1-3 µm and volumes of roughly 0.01-0.8 µm³, with a cytoskeleton built mainly from actin (PMC5243184). The proposed roles are regulating the efficacy of a single synapse, expanding the computational capacity of the neuron, and compartmentalizing postsynaptic signaling. Cajal’s early proposal was that spines increase the dendritic surface area available for synaptic connections.
Where are dendritic spines found?
Spines sit along dendrites of spiny neurons, and in those neurons they form the site of excitatory synapses (PMC8468246). Density is not uniform: it is highest in early childhood, declines through adolescence to a stable adult level, and varies between dendritic branches of the same cell. Because density differs by branch, a count is only meaningful alongside the branch, segment length and developmental stage it came from.
Spine types: thin, stubby and mushroom
Electron-microscopy-based classification divides spines into three types (PMC8468246):
| Type | Shape as described |
|---|---|
| Mushroom | Large head, short thin neck |
| Thin | Small head, long thin neck |
| Stubby | No clear head or neck |
Head diameter and volume correlate positively with synaptic strength, and some work suggests neck width and length may be better indicators (PMC5243184). Labs that sort spines into these classes should state the measurement thresholds used, since they differ between software packages and laboratories.
Dendritic spine density and plasticity
Long-term potentiation (LTP) has been linked to increased spine density and/or head diameter, while long-term depression (LTD) has been linked to spine retraction or shrinkage. Spine turnover is far higher in development than in adulthood: roughly 10-15% of spines over 24 hours in development versus roughly 1-2% in adulthood (PMC5243184). That difference matters when choosing imaging intervals: an interval that captures change in a young animal can miss almost everything in an adult. The same review concludes that spine dynamics are likely both a cause and a consequence of behavior.
How spines are quantified: choosing a method
The methods in the literature differ mainly in whether they give a snapshot or follow the same spines over time (PMC5243184).
| Method | What you get | Main constraint in the source |
|---|---|---|
| Golgi-Cox staining | Post-mortem snapshot of morphology | Cannot follow the same spine over time |
| Fluorescent dyes | Post-mortem snapshot of morphology | Cannot follow the same spine over time |
| Two-photon imaging in vivo | Repeated imaging of spines in awake animals | With a thinned skull, image quality degrades beyond about 50 µm depth |
For the stain itself, see the Golgi stain guide; for the optics, see two-photon microscopy and epifluorescence microscopy. For activity readouts rather than structure, see calcium imaging.
Checklist before comparing spine counts across papers
- Units: published densities are reported per µm, per 10 µm or per 20 µm of dendrite. Convert to one unit before comparing; the figures are not interchangeable.
- Classification thresholds: the cut-offs separating thin, stubby and mushroom spines vary by software and laboratory. Compare class proportions only when the criteria match.
- Dendritic location: density varies by branch, so match the branch type.
- Developmental stage: density and turnover change from early childhood and development to adulthood.
- Snapshot versus longitudinal: fixed-tissue counts and repeated in vivo imaging answer different questions, so do not pool them.
Spine pathology in disease
Reported associations from the literature (PMC8468246): higher spine density in several brain regions in autism spectrum disorder; lower density in the prefrontal and auditory cortex in schizophrenia; loss of synapses and spines as a hallmark of Alzheimer’s disease; abnormal morphology or density in intellectual disability; and increased density in preclinical models of neuropathic pain. These are research findings about tissue and animal models, not diagnostic criteria.
Common misunderstandings
- “More spines means stronger synapses.” Density and size are separate measures. Head size tracks synaptic strength, and LTP has been tied to either density or head diameter increases.
- “Spines are fixed once formed.” Turnover is measurable in adults, at roughly 1-2% over 24 hours, and much higher in development.
- “A Golgi count shows change over time.” Fixed-tissue methods give a single snapshot per animal.
FAQ
What are dendritic spines?
Small protrusions on dendrites, typically 1-3 µm long, whose heads receive mainly excitatory synaptic input.
Where are dendritic spines found?
Along the dendrites of spiny neurons, where they form the site of excitatory synapses. Density varies by dendritic branch and with age.
What are the types of dendritic spines?
Mushroom, thin and stubby, distinguished by head and neck shape in electron-microscopy-based classification.
How are dendritic spines counted?
Golgi-Cox and fluorescent dyes give post-mortem snapshots; two-photon imaging allows repeated imaging in awake animals. Always normalize density to a common unit of dendrite length before comparing studies.
How much do spines turn over?
About 10-15% in 24 hours in development and about 1-2% in adulthood, according to a review of in vivo imaging work.
Are dendrites voltage-gated?
Dendritic membranes are not passive cable. A review states that dendritic plasma membranes express a plethora of voltage-gated ion channels, and that several of them are expressed in dendrites at higher densities than at the soma. These channels mediate active backpropagation of action potentials and let dendrites initiate local dendritic spikes that act as dendritic outputs (PMC6394607). That review is about dendrites generally, not spine heads specifically.
What are the four types of synapses?
There is no single fixed list of four in the sources read here. A 2021 review describes the chemical synapse as typically configured as presynaptic axon terminals onto postsynaptic dendrites or somata, giving axo-dendritic and axo-somatic synapses, and describes axo-axonic synapses, in which an axon terminates on another axon or axon terminal, as a less-studied, non-canonical type (PMC8053672). Teaching texts often add further categories or classify by other criteria, so state which scheme you are using. Spine-bearing dendrites are the postsynaptic side of the axo-dendritic case.
What is dendritic arborization?
It is the growth and branching of a neuron’s dendrites into a tree-like arbor. A review of dendrite development describes organized stages of dendrite morphogenesis, in which dendritic arbor growth and elaboration is followed by retraction and pruning (PMC3833426). Arborization concerns the branching pattern of the whole dendritic tree, whereas spines are the small protrusions along its branches; spine counts are normally reported per length of a given branch.
This page is general scientific background for researchers and lab staff, not medical advice.
Sources
- Structural and functional plasticity of dendritic spines – root or result of behavior? PMC5243184, read 2026-11-07.
- Dendritic Spine Initiation in Brain Development, Learning and Diseases and Impact of BAR-Domain Proteins. PMC8468246, read 2026-11-07.
- Stores, Channels, Glue, and Trees: Active Glial and Active Dendritic Physiology (Ashhad and Narayanan). PMC6394607, read 2026-11-07.
- Cell-intrinsic drivers of dendrite morphogenesis (Puram and Bonni, 2013). PMC3833426, abstract read 2026-11-07.
- Axo-axonic synapses: Diversity in neural circuit function (2021). PMC8053672, abstract read 2026-11-07.
