Dendritic spines are small, bulbous protrusions that emanate from the shafts of dendrites in neurons, serving as the primary recipients of excitatory synaptic input throughout the mammalian brain. These microscopic structures, typically ranging from 0.5 to 2 micrometers in length, represent the fundamental units of excitatory synapse formation and are essential for proper neural circuitry function[@bourne2008]. Each dendritic spine typically forms a single postsynaptic density (PSD) opposite an axonal presynaptic terminal, creating a specialized compartment for synaptic transmission that is biochemically and structurally distinct from the parent dendrite[@sala2014].
The significance of dendritic spines in neurodegenerative diseases cannot be overstated, as these structures serve as sensitive indicators of synaptic health and functional integrity. In healthy brains, dendritic spines exhibit remarkable plasticity—they can be formed, eliminated, enlarged, or shrunk in response to neural activity, a process that underlies learning, memory formation, and experience-dependent neural circuit refinement[@holtmaat2009]. This dynamic nature, while crucial for cognitive function, also makes spines particularly vulnerable to pathological insults that characterize neurodegenerative conditions.
Dendritic spines are small, bulbous protrusions that emanate from the shafts of dendrites in neurons, serving as the primary recipients of excitatory synaptic input throughout the mammalian brain. These microscopic structures, typically ranging from 0.5 to 2 micrometers in length, represent the fundamental units of excitatory synapse formation and are essential for proper neural circuitry function[@bourne2008]. Each dendritic spine typically forms a single postsynaptic density (PSD) opposite an axonal presynaptic terminal, creating a specialized compartment for synaptic transmission that is biochemically and structurally distinct from the parent dendrite[@sala2014].
The significance of dendritic spines in neurodegenerative diseases cannot be overstated, as these structures serve as sensitive indicators of synaptic health and functional integrity. In healthy brains, dendritic spines exhibit remarkable plasticity—they can be formed, eliminated, enlarged, or shrunk in response to neural activity, a process that underlies learning, memory formation, and experience-dependent neural circuit refinement[@holtmaat2009]. This dynamic nature, while crucial for cognitive function, also makes spines particularly vulnerable to pathological insults that characterize neurodegenerative conditions.
Neurodegenerative diseases, including Alzheimer's disease (AD), Parkinson's disease (PD), Huntington's disease (HD), and amyotrophic lateral sclerosis (ALS), share a common feature: the progressive loss of synaptic integrity that ultimately leads to cognitive and motor dysfunction[@selkoe2002]. Dendritic spine abnormalities—including reduced spine density, morphological alterations, and functional impairments—have been documented across multiple neurodegenerative conditions, making these structures important pathological hallmarks and potential therapeutic targets[@harris1994]. The observation that synaptic loss correlates better with cognitive decline than traditional metrics such as neurofibrillary tangle burden or neuron loss has highlighted the central role of spine dysfunction in disease progression[@terry1991].
Understanding the molecular and cellular mechanisms that regulate dendritic spine development, maintenance, and plasticity provides crucial insights into neurodegeneration pathogenesis. This knowledge为基础 for developing therapeutic interventions aimed at preserving synaptic function and preventing the devastating cognitive decline characteristic of these disorders[@penzes2011].
Dendritic spines exhibit remarkable morphological diversity, which correlates with their functional properties and synaptic strength. Spines are typically classified into several distinct morphotypes based on their shape, size, and the ratio of head width to neck dimensions[@peters1970]. The four principal spine types include thin spines, stubby spines, mushroom spines, and filopodia, each possessing unique structural features and functional implications.
Thin spines, characterized by a long, narrow neck and a small, indistinct head, are the most abundant type in the adult brain and are associated with learning and plasticity processes[@kasai2010]. These highly dynamic structures can rapidly change shape in response to synaptic activity, making them ideal candidates for experience-dependent modification of neural circuits. The elongated neck of thin spines creates electrical and biochemical isolation between the spine head and parent dendrite, allowing for localized signaling events that can modify synaptic strength independently of the dendritic shaft[@araya2007].
Stubby spines possess a short, wide morphology lacking a distinct neck, appearing as brief protrusions from the dendritic shaft. These spines are often considered immature forms that may mature into other types or represent a transitional state during spine development[@miller1981]. Despite their simple architecture, stubby spines contain postsynaptic densities and can form functional synapses, though they may exhibit different signaling properties compared to necked spines.
Mushroom spines feature a large, spherical head connected to the dendritic shaft by a thick neck, representing the most stable and mature spine type[@harris1992]. The large head volume provides substantial space for postsynaptic machinery, including neurotransmitter receptors, signaling molecules, and cytoskeletal elements. Mushroom spines are typically associated with strong, stable synaptic connections and are resistant to elimination compared to thinner spine types[@yang2009]. The morphological characteristics of mushroom spines make them particularly important for long-term memory storage, as their stability provides a structural basis for persistent synaptic modifications.
Filopodia are long, thin, headless protrusions that extend from dendrites and represent the most dynamic of spine-related structures. These actin-rich structures actively explore the neuropil and can initiate synaptogenic contacts with presynaptic terminals, serving as precursors to functional spines[@ziv1996]. The transformation from filopodium to mature spine involves the recruitment of postsynaptic proteins and the establishment of a proper postsynaptic density.
The structural integrity and functional plasticity of dendritic spines depend on a sophisticated molecular architecture centered on the actin cytoskeleton. Actin filaments form the core structural scaffold of spines, comprising approximately 5-10% of total cellular actin in neurons[@cingolani2008]. The spine actin cytoskeleton is highly dynamic, with continuous polymerization and depolymerization regulated by numerous actin-binding proteins, including cofilin, Arp2/3 complex, and various myosin motors[@hotulainen2010].
The postsynaptic density (PSD) is a specialized electron-dense structure located at the tip of dendritic spines that contains the molecular machinery for synaptic transmission. The PSD scaffold is composed of hundreds of proteins, including PSD-95, Homer, Shank, and GKAP, which organize neurotransmitter receptors, adhesion molecules, and signaling enzymes into functional signaling complexes[@sheng2011]. PSD-95, a major scaffold protein, clusters ionotropic glutamate receptors (particularly AMPA and NMDA receptors) at the postsynaptic membrane and links them to intracellular signaling pathways that regulate synaptic plasticity[@kennedy2000].
Excitatory synaptic transmission in spines is mediated primarily by glutamate receptors, with NMDA receptors (NMDARs) and AMPA receptors (AMPARs) serving as the principal mediators of fast synaptic transmission. The composition and properties of these receptors determine the postsynaptic response to presynaptic glutamate release and are dynamically regulated by activity-dependent mechanisms[@bredt2003]. NMDA receptors, with their unique calcium permeability and voltage-dependent magnesium block, serve as molecular coincidence detectors that trigger long-term potentiation (LTP) and long-term depression (LTD)—the cellular correlates of learning and memory[@malenka2004].
Spine morphology is also influenced by extracellular matrix components and cell adhesion molecules that mediate trans-synaptic interactions. Integrins, cadherins, and immunoglobulin superfamily proteins contribute to spine development, stability, and plasticity by linking the spine cytoskeleton to the presynaptic terminal and surrounding extracellular environment[@dalva2007]. The proper organization of these adhesion complexes ensures structural stability while allowing for activity-dependent remodeling.
Dendritic spines contain a specialized subset of cellular organelles and membrane compartments that support their unique functions. Endoplasmic reticulum (ER) cisternae, particularly smooth ER, can penetrate spine necks and heads, creating calcium stores that regulate local calcium signaling[@spacek1997]. The spine apparatus, a specialized form of smooth ER characterized by stacked cisternae connected by dense plates, is associated with synaptic plasticity and calcium release mechanisms[@fifkov1982].
Spines also contain endosomes and recycling compartments that regulate the trafficking of membrane proteins, including neurotransmitter receptors. The endocytic zone adjacent to the PSD is a specialized domain where receptor internalization and recycling occur, providing a mechanism for activity-dependent modulation of synaptic strength[@blanpied2003]. Lysosomes and autophagosomes have also been observed in larger spines, where they may contribute to local protein turnover and quality control mechanisms[@goetzl2016].
The spine membrane is enriched in specific lipid species and cholesterol that influence receptor dynamics and signaling platform organization. Lipid rafts—membrane microdomains enriched in cholesterol and sphingolipids—concentrate certain receptors and signaling molecules, facilitating specific signal transduction cascades[@allen2007]. The unique lipid composition of spines contributes to their distinctive physical properties and supports the specialized functions of the postsynaptic compartment.
Alzheimer's disease, the most common cause of dementia worldwide, is characterized by the accumulation of extracellular amyloid-beta (Aβ) plaques and intracellular neurofibrillary tangles composed of hyperphosphorylated tau protein. Both pathological hallmarks exert profound effects on dendritic spine structure and function, contributing to the synaptic failure that underlies cognitive decline[@spires2005]. The relationship between Aβ and spine dysfunction has been extensively studied using animal models, human tissue, and in vitro systems, revealing multiple mechanisms by which this amyloidogenic peptide compromises synaptic integrity.
Soluble oligomeric Aβ, rather than fibrillar plaques, appears to be the primary toxic species responsible for synaptic dysfunction[@walsh2007]. These oligomers bind to synapses with high affinity, disrupting NMDA receptor signaling, impairing long-term potentiation, and promoting the internalization of AMPA receptors[@hsieh2006]. Aβ oligomers also activate dendritic spine calcium signaling pathways, leading to aberrant activation of calcium-dependent proteases, phosphatases, and kinases that modify the spine cytoskeleton[@shankar2007].
The effect of Aβ on spine morphology varies depending on the oligomeric species, concentration, and exposure duration. Acute Aβ exposure typically causes spine loss and morphological alterations, while chronic exposure leads to more subtle changes in spine dynamics, including reduced spine density and impaired plasticity-induced spine formation[@bittner2010]. Studies in APP/PS1 transgenic mice demonstrate that Aβ deposition is associated with significant reductions in spine density, particularly in the hippocampal CA1 region and cortical pyramidal neurons—brain regions critical for learning and memory[@moolman2004].
Tau pathology, the second major hallmark of AD, also profoundly affects dendritic spine integrity. Hyperphosphorylated tau mislocalizes from axons to dendrites, where it accumulates in spines and disrupts synaptic function[@hoover2010]. Tau in spines interferes with NMDA receptor signaling and promotes the internalization of AMPA receptors, similar to the effects of Aβ oligomers[@miller2014]. Furthermore, tau pathology exacerbates Aβ-induced spine dysfunction, as demonstrated by studies showing that reducing tau expression rescues Aβ-induced synaptic deficits[@roberson2007].
The postsynaptic signaling pathways regulating spine plasticity are extensively disrupted in Alzheimer's disease, affecting both the induction and expression of synaptic plasticity. NMDA receptor signaling, crucial for LTP induction, is compromised by Aβ and tau through multiple mechanisms, including receptor internalization, altered subunit composition, and disrupted downstream signaling[@snyder2005]. The calcium/calmodulin-dependent protein kinase II (CaMKII), a key effector of NMDA receptor-dependent plasticity, shows reduced activation and autophosphorylation in AD models and human tissue[@lacor2007].
AMPA receptor trafficking, which mediates the expression of synaptic plasticity, is also impaired in AD. Aβ promotes the internalization of GluA1 and GluA2 subunits through mechanisms involving clathrin-dependent endocytosis and dynamin-mediated pinching[@liu2004]. This internalization reduces synaptic AMPA receptor content, leading to synaptic depression and impaired LTP expression. The regulation of AMPA receptor cycling by scaffolding proteins like PSD-95 is disrupted by Aβ, further compromising synaptic strength[@roselli2005].
Actin cytoskeleton regulators are particularly vulnerable to AD pathology. Aβ and tau alter the activity of Rho GTPases (Rac1, Cdc42, RhoA), cofilin, and other actin-binding proteins that control spine morphology and dynamics[@zhang2019]. These alterations shift the balance toward actin depolymerization, contributing to spine shrinkage and elimination. The targeting of actin regulators by pathological species provides a direct link between protein aggregation and structural spine pathology.
Neuroinflammation, a prominent feature of AD pathophysiology, significantly contributes to dendritic spine dysfunction. Activated microglia and astrocytes release pro-inflammatory cytokines, including interleukin-1β (IL-1β), tumor necrosis factor-α (TNF-α), and interleukin-6 (IL-6), which can directly modify spine morphology and synaptic function[@huang2016]. Chronic exposure to these cytokines promotes spine loss and alters synaptic plasticity mechanisms, creating a positive feedback loop between neuroinflammation and synaptic degeneration.
Microglial phagocytosis, while normally important for synaptic pruning during development, becomes dysregulated in AD and may contribute to excessive spine elimination. The complement system, particularly C1q and C3, tags synapses for microglial engulfment in AD models, and blocking this pathway protects against synapse loss[@stephan2012]. The overactivation of microglial pruning mechanisms, combined with impaired synaptic maintenance, creates an environment where spines are particularly vulnerable to elimination.
Astrocyte dysfunction also contributes to spine pathology in AD. These glial cells normally provide metabolic support, regulate extracellular glutamate levels, and release trophic factors that support synaptic integrity. In AD, astrocyte function is compromised, leading to impaired glutamate clearance, altered metabolic support, and reduced secretion of synaptogenic factors[@prezcordn2018]. The disruption of astrocyte-neuron interactions compounds the direct effects of Aβ and tau on spines.
Parkinson's disease is characterized by the progressive degeneration of dopaminergic neurons in the substantia nigra pars compacta and the accumulation of Lewy bodies composed primarily of alpha-synuclein (α-syn) aggregates. While PD is traditionally considered a movement disorder, cognitive dysfunction and dementia are common in advanced stages, highlighting the involvement of cortical and hippocampal circuits[@kalia2015]. Dendritic spine abnormalities have been documented in PD models and human tissue, revealing a previously underappreciated synaptic pathology.
Alpha-synuclein, a small presynaptic protein involved in synaptic vesicle trafficking, can misfold and aggregate in both sporadic and familial forms of PD. The toxicity of α-syn extends beyond dopaminergic neurons to affect cortical and hippocampal pyramidal neurons, where it causes spine loss and functional impairment[@volpicellidaley2011]. Studies in rodent models overexpressing wild-type or mutant α-syn demonstrate significant reductions in spine density, particularly in cortical layer V pyramidal neurons and hippocampal CA1 pyramidal cells[@tanner2011].
The mechanisms by which α-syn affects spines include both cell-autonomous and non-cell-autonomous pathways. Within neurons, α-syn aggregation disrupts synaptic vesicle cycling, impairs neurotransmitter release, and alters postsynaptic signaling[@burre2018]. The accumulation of α-syn in dendritic compartments, where it is normally absent, interferes with local protein trafficking and signaling mechanisms that maintain spine integrity[@khalil2018].
The dopaminergic system profoundly influences dendritic spine morphology and plasticity, particularly in the striatum and cortex. The degeneration of dopaminergic neurons in PD disrupts this modulatory influence, contributing to spine abnormalities in affected circuits[@surmeier2017]. In the striatum, which receives dense dopaminergic input from the substantia nigra, dopamine depletion leads to significant spine loss on medium spiny neurons—a hallmark of PD pathophysiology[@villalpandoestrada2020].
Dopamine acts through D1 and D2 receptors to modulate spine plasticity through distinct mechanisms. D1 receptor activation promotes LTP and spinogenesis, while D2 receptor activation favors LTD and spine elimination[@day2006]. The loss of dopaminergic input in PD disrupts this balance, favoring pathways that promote spine loss. Additionally, levodopa treatment, the primary therapy for PD, can cause dyskinesias associated with further spine alterations, highlighting the complex relationship between dopamine and spine morphology[@zhang2006].
Cortical dopaminergic innervation, though less dense than striatal input, also modulates spines in prefrontal and other cortical regions. Dopamine deficiency in PD contributes to cognitive deficits through effects on prefrontal cortical circuits, where spine density and plasticity are reduced[@xu2018]. The restoration of dopaminergic signaling with pharmacological interventions can partially reverse these deficits, though complete recovery is often not achieved.
Mitochondrial dysfunction and oxidative stress are central pathogenic mechanisms in PD that contribute to spine pathology. Complex I deficiency, a hallmark of sporadic PD, impairs neuronal energy metabolism and increases reactive oxygen species (ROS) production[@schapira1990]. The high energy demands of spines, particularly during plasticity events, make them particularly vulnerable to mitochondrial dysfunction.
ATP depletion compromises the actin cytoskeleton dynamics that underlie spine morphology and plasticity. The polymerization and depolymerization of actin filaments require ATP, and energy failure leads to cytoskeletal instability and spine loss[@lin2006]. Additionally, calcium homeostasis, critical for spine function, is disrupted by mitochondrial dysfunction, leading to calcium overload and activation of degenerative pathways[@gandhi2012].
Oxidative stress damages cellular components, including proteins, lipids, and DNA, that are essential for spine structure and function. ROS can directly modify actin and actin-binding proteins, impairing cytoskeletal dynamics[@ischiropoulos2003]. The oxidative modification of synaptic proteins disrupts their function and promotes the accumulation of damaged proteins that compromise synaptic integrity.
The recognition of dendritic spine dysfunction as a central feature of neurodegenerative diseases has prompted efforts to develop therapies that preserve or restore synaptic integrity. Disease-modifying approaches targeting the underlying pathological proteins—Aβ, tau, and α-syn—are expected to indirectly benefit spines by reducing the toxic stimuli that trigger synaptic degeneration[@cummings2014]. Immunotherapies directed against Aβ (e.g., aducanumab, lecanemab) and tau (e.g., anti-tau antibodies) have shown promise in clinical trials, though their effects on spine pathology remain to be fully characterized[@sevigny2016].
Small molecules that inhibit protein aggregation or promote clearance represent another therapeutic strategy. Compounds that prevent Aβ oligomerization, enhance autophagy, or modulate proteasome activity may reduce the burden of toxic protein species and protect synapses[@menzies2015]. The blood-brain barrier permeability of such compounds remains a significant challenge, driving the development of novel delivery strategies.
Genetic approaches, including antisense oligonucleotides and CRISPR-based gene editing, offer the potential to reduce the expression of disease-causing proteins or enhance the expression of protective factors. Allele-specific silencing of mutant tau or α-syn alleles, where applicable, could provide significant clinical benefit[@miller2004]. The delivery of neurotrophic factors or synaptic proteins via viral vectors represents another experimental approach with promise for synaptic protection.
Direct targeting of synaptic mechanisms represents a complementary strategy that may provide benefits even in the presence of ongoing pathology. Compounds that enhance synaptic plasticity, promote spinogenesis, or stabilize existing spines are under active investigation for neurodegenerative diseases[@lu1997]. NMDA receptor modulators, including partial agonists and Gly site modulators, aim to enhance plasticity without causing excitotoxicity.
The modulation of actin cytoskeleton dynamics offers a direct approach to stabilize spines. Targeting Rho GTPase signaling, cofilin activity, or formin proteins can promote spinogenesis and spine stability[@tashiro2004]. However, the complexity of actin regulation and the potential for off-target effects require careful drug development.
AMPA receptor positive allosteric modulators enhance synaptic transmission and may improve cognitive function in neurodegenerative conditions. These compounds increase channel open time or favor desensitization states that promote synaptic strengthening[@oneill2004]. The challenge lies in achieving functional benefits without causing seizures or other adverse effects associated with excessive excitation.
Novel therapeutic approaches are emerging that target previously underappreciated aspects of spine biology in neurodegeneration. Microglial modulation represents a promising strategy, as reducing neuroinflammation or normalizing microglial pruning may protect synapses from excessive elimination[@hansen2018]. Colony-stimulating factor 1 receptor (CSF1R) antagonists that reduce microglial numbers have shown protective effects in animal models of AD and PD.
Astrocyte-targeted therapies aim to restore the supportive functions of these glial cells. Enhancing astrocyte glutamate uptake, promoting trophic factor release, or modulating astrocyte-neuron signaling may improve synaptic health[@keyser2008]. The development of astrocyte-specific drug delivery systems is facilitating this approach.
Gene therapy approaches delivering neurotrophic factors, such as brain-derived neurotrophic factor (BDNF) or nerve growth factor (NGF), have shown promise in preclinical models for protecting synapses and preventing degeneration[@nagahara2011]. The delivery of BDNF to hippocampus and cortex via adeno-associated virus vectors protected against Aβ-induced spine loss in mouse models, demonstrating the potential of this approach.
Finally, computational and systems biology approaches are identifying novel therapeutic targets by mapping the molecular networks that regulate spine integrity in health and disease[@winchester2014]. These integrative approaches may reveal key nodes that can be modulated to achieve broad protective effects on synaptic function.