Progressive Supranuclear Palsy (PSP) represents one of the most common atypical parkinsonian syndromes, classified under the broader category of tauopathies—a group of neurodegenerative disorders characterized by the abnormal accumulation of the microtubule-associated protein tau (MAPT) within neurons and glia ^1. PSP was first described by John Steele, János ClovÁcs, and Frederick Richardson in 1964, earning the eponym "Steele-Richardson-Olszewski syndrome" ^2. The disease typically presents in the sixth or seventh decade of life, with a progressive course that leads to severe disability and death within 5-10 years of symptom onset ^3.
The clinical phenotype of PSP extends far beyond the classic tetrad of vertical supranuclear gaze palsy, pseudobulbar palsy, axial rigidity, and postural instability with falls ^4. Contemporary clinical criteria recognize multiple PSP subtypes, including PSP with predominant parkinsonism (PSP-P), PSP with predominant cerebellar ataxia (PSP-C), and the recently defined PSP with mild motor impairment (PSP-M) ^5. This clinical heterogeneity reflects the underlying neuropathological diversity, yet all PSP subtypes share the fundamental neuropathological signature of tau inclusions in specific subcortical and cortical nuclei ^6.
Understanding connectivity changes in PSP requires first appreciating the fundamental organization of healthy brain networks. The human brain operates as a complex network of spatially distributed, functionally specialized regions that are integrated through anatomical white matter tracts and synchronized through temporal correlations in neural activity ^7. This organizational architecture can be conceptualized using graph theory, where brain regions constitute nodes and structural or functional connections constitute edges ^8. The frontal cortex, basal ganglia, thalamus, and brainstem nuclei constitute major components of this rich club architecture, regions that are precisely targeted in PSP ^9. This anatomical vulnerability stems from the network position of these structures—positioned at the top of the hierarchical processing hierarchy, they serve as convergence points for information flow throughout the brain ^10.
Functional magnetic resonance imaging (fMRI) studies employing resting-state methodology have consistently demonstrated widespread connectivity abnormalities in PSP, with particular disruption of the default mode network (DMN)—a set of brain regions including the posterior cingulate cortex, precuneus, medial prefrontal cortex, and angular gyrus that demonstrate high metabolic activity during rest ^11. The DMN, critical for self-referential processing, memory consolidation, and prospective thinking, shows marked hypoconnectivity in PSP, reflecting the characteristic frontal and parietal involvement ^12.
Whitwell et al. demonstrated that PSP patients exhibit significantly reduced functional connectivity within the DMN, particularly between posterior cingulate and medial frontal regions ^1. This disruption correlates with performance on executive function and attention tasks, suggesting that DMN dysfunction contributes to the prominent cognitive impairments observed in PSP ^13. Importantly, these functional alterations parallel the characteristic "cortical hypometabolism" patterns observed in FDG-PET studies, indicating a close structure-function relationship ^14.
The salience network, anchored by the anterior insula and dorsal anterior cingulate cortex, plays a critical role in detecting behaviorally relevant stimuli and coordinating the switching between the DMN and central executive networks ^15. PSP patients demonstrate both hyperconnectivity and hypoconnectivity within salience network structures, likely reflecting compensatory responses to pathology in connected regions ^16. This altered salience network function may underlie the apathy, emotional processing deficits, and impaired social cognition observed in PSP ^17.
The frontoparietal control network (FPCC), supporting flexible cognitive control and goal-directed behavior, shows consistent disruption in PSP. Meta-analyses have identified reduced frontoparietal connectivity as a robust finding across PSP cohorts, correlating with executive dysfunction and response inhibition deficits ^18. These changes likely reflect the characteristic tau pathology affecting frontostriatal circuits, particularly the dorsolateral prefrontal cortex and its connections to the basal ganglia ^19.
Diffusion tensor imaging (DTI) studies have revealed extensive white matter damage in PSP, extending far beyond the midbrain and basal ganglia to involve association fibers connecting frontal, parietal, and temporal cortices ^20. The characteristic "hummingbird sign" on structural MRI reflects atrophy of the midbrain and superior cerebellar peduncle, but DTI reveals more widespread microstructural damage that precedes visible atrophy ^21.
Fractional anisotropy (FA) reductions in PSP are most pronounced in white matter tracts connecting subcortical structures to frontal cortical regions, including the thalamic radiations, corpus callosum, and frontostriatal projections ^22. These structural changes show strong correlations with clinical severity measures, including the PSP Rating Scale (PSPRS), suggesting that white matter disruption directly contributes to clinical disability ^23.
The pattern of connectivity disruption in PSP exemplifies the "disconnection syndrome" concept originally proposed by Geschwind, wherein behavioral deficits in neurological disease result from disconnection between brain regions rather than focal damage alone ^24. The selective vulnerability of specific network nodes—particularly rich club hubs in the basal ganglia, thalamus, and brainstem—produces a characteristic pattern of distributed functional disruption that cannot be explained by local pathology alone ^25.
Graph-theoretic analyses have confirmed that PSP patients demonstrate reduced network efficiency, decreased clustering coefficients, and altered hub connectivity compared to healthy controls ^26. These topological changes reflect the progressive nature of the disease, with more advanced cases showing greater network fragmentation and loss of small-world properties ^27.
A fundamental understanding of network changes in PSP requires consideration of how tau pathology spreads through the brain. The observation that neurofibrillary tangle distribution in PSP follows specific anatomical pathways—rather than spreading randomly—led to the hypothesis that tau may propagate along neuronal connections in a "prion-like" manner ^28. This model proposes that pathological tau species are released from affected neurons, taken up by connected neurons, and templating the conversion of normal tau into pathological conformers ^29.
Multiple lines of evidence support this network-based propagation model. Experimental studies have demonstrated that tau pathology in mouse models spreads along anatomically connected pathways, preferentially affecting synaptically coupled neurons ^30. Human neuroimaging studies have shown that patterns of cortical atrophy in PSP correspond to the spatial distribution of tau pathology, consistent with trans-synaptic spread ^31. Furthermore, the characteristic regional vulnerability within the brain can be predicted by connectional anatomy—regions with greater connectivity to earlyaffected sites demonstrate earlier and more severe pathology ^32.
The molecular cascade underlying tau propagation involves several key steps: misfolding and oligomerization of tau protein, release into the extracellular space, neuronal uptake, and templating of endogenous tau misfolding ^33. Post-translational modifications of tau—including phosphorylation, acetylation, and truncation—influence these processes and may determine the propagation kinetics and regional vulnerability ^34.
The tau protein (encoded by the MAPT gene on chromosome 17) exists in six isoforms generated by alternative splicing of exons 2, 3, and 10. The inclusion of exon 10 generates tau isoforms with four microtubule-binding repeats (4R tau), which predominate in PSP pathology ^35. This selective accumulation of 4R tau distinguishes PSP from Alzheimer's disease, where both 3R and 4R tau accumulate, and suggests specific molecular mechanisms underlying 4R tau aggregation and propagation ^36.
Key proteins involved in tau homeostasis include tau kinases (GSK3β, CDK5, MARK4) that phosphorylate tau and promote aggregation, tau phosphatases (PP2A, PP1) that normally dephosphorylate tau, and molecular chaperones (HSP90, HSP70) that regulate tau folding and degradation ^37. Dysregulation of this balance—shifting toward increased phosphorylation and aggregation—represents a critical step in PSP pathogenesis ^38.
Tau pathology in PSP produces profound synaptic dysfunction that directly disrupts network connectivity. Pathological tau localizes to synapses in PSP brain tissue, where it may interfere with normal synaptic function and plasticity ^39. Studies have demonstrated reduced synaptic density in PSP cortex, correlating with cognitive impairment and functional connectivity disruption ^40.
At the molecular level, tau interacts with multiple synaptic proteins, including proteins involved in neurotransmitter release, receptor trafficking, and dendritic spine morphology ^41. The redistribution of tau from axons to dendrites—a phenomenon observed in both PSP and other tauopathies—may promote excitotoxicity by enhancing dendritic calcium influx ^42.
Neuronal network function requires substantial energy resources, and mitochondrial dysfunction in PSP contributes to network failure. Tau accumulation disrupts mitochondrial transport along axons, reducing energy delivery to distant synapses and contributing to synaptic loss ^43. Additionally, pathological tau interacts with mitochondrial proteins, impairing respiratory chain function and increasing oxidative stress ^47.
The vulnerability of specific neuronal populations in PSP—including large pyramidal neurons in the frontal cortex and cholinergic neurons of the basal forebrain—may reflect their exceptionally high metabolic demands and particular mitochondrial vulnerabilities ^48.
Neuroinflammation accompanies tau pathology in PSP and contributes to network dysfunction through multiple mechanisms. Activated microglia surround tau inclusions in PSP brain tissue, and positron emission tomography studies using TSPO ligands have confirmed increased microglial activation in PSP patients. Pro-inflammatory cytokines released by activated glia can modulate synaptic transmission, alter neuronal excitability, and promote further tau phosphorylation in a feed-forward cycle.
Astrocytic dysfunction also contributes to network disruption in PSP. Astrocytes normally support synaptic function through ion buffering, neurotransmitter clearance, and metabolic support; their dysfunction in PSP may impair these essential supportive functions.
The basal ganglia—comprising the caudate nucleus, putamen, globus pallidus, subthalamic nucleus, and substantia nigra—form the core of motor and cognitive circuitry disrupted in PSP. Five parallel circuits originating in the frontal cortex and projecting through the basal ganglia to the thalamus carry motor, oculomotor, cognitive, and limbic information. PSP pathology disrupts all these circuits, explaining the motor, cognitive, and behavioral symptoms of the disorder.
The subthalamic nucleus (STN), which serves as a major excitatory driver of the basal ganglia output nuclei, shows early tau pathology in PSP. This involvement may contribute to the characteristic "axial" symptoms of PSP—postural instability and gait disturbance—through disruption of motor circuits.
The substantia nigra pars reticulata (SNr) and pars compacta (SNc) demonstrate prominent neurofibrillary tangle formation in PSP. SNc dopamine neuron loss contributes to parkinsonian features, while SNr involvement disrupts the inhibitory output that normally controls thalamic activity.
The characteristic vertical supranuclear gaze palsy of PSP reflects the involvement of specific brainstem structures controlling eye movements. The rostral interstitial nucleus of the medial longitudinal fasciculus (riMLF), located in the midbrain and controlling vertical gaze, shows early tau pathology that likely underlies this cardinal feature. Similarly, the superior colliculus and its projections to the oculomotor nuclei are affected.
The pedunculopontine nucleus (PPN), a major cholinergic input to the thalamus involved in arousal and gait initiation, demonstrates significant pathology in PSP. This involvement may contribute to both the gait disturbance and the sleep disturbances common in PSP.
The thalamus serves as a central relay between subcortical structures and the cortex, making its involvement critical to understanding network disruption in PSP. The mediodorsal and ventral anterior thalamic nuclei—projecting to frontal cortical regions—show particular vulnerability, consistent with the prominent frontal symptoms. Thalamic connectivity changes in PSP reflect both direct pathology and the effects of disrupted cortical and subcortical inputs.
The characteristic patterns of network disruption in PSP have diagnostic utility that complements clinical assessment and traditional neuroimaging. Resting-state functional connectivity measures can distinguish PSP from other parkinsonian syndromes with high accuracy, potentially enabling earlier and more accurate diagnosis ^59. Machine learning approaches applied to connectivity data have shown particular promise for differential diagnosis ^60.
Key connectivity signatures that distinguish PSP include reduced connectivity within the DMN, altered connectivity between subcortical structures and frontal cortex, and specific patterns of anti-correlation between networks ^61. These signatures show sensitivity to disease progression, suggesting potential utility as progression markers in clinical trials ^62.
Combining structural, diffusion, and functional imaging data provides complementary information about the nature and extent of network disruption in PSP. The "PSP connectome" concept integrates these modalities to characterize how structural damage leads to functional disconnection and clinical impairment ^63. This multimodal approach may improve diagnostic accuracy and provide more sensitive biomarkers for therapeutic trials.
Advanced neuroimaging techniques continue to reveal new aspects of network dysfunction in PSP. Graph signal processing methods can detect subtle network changes not apparent with traditional connectivity analyses ^64. Dynamic functional connectivity approaches, examining time-varying properties of brain networks, may reveal state-dependent changes relevant to symptom variability ^65.
Molecular imaging with tau PET ligands provides a unique window into the pathological substrate underlying network changes. The tau PET tracer ^18F-AV-1451 (Flortaucipir) demonstrates binding patterns in PSP that correlate with clinical severity and may predict subsequent atrophy ^66. Integrating tau PET with functional connectivity data may clarify the relationship between pathology location and network dysfunction ^67.
Understanding network propagation mechanisms has informed therapeutic strategies targeting tau pathology. Immunotherapy approaches aiming to enhance clearance of pathological tau—either by active vaccination or passive antibody administration—have advanced to clinical trials ^68. The network-based model of tau spread predicts that early intervention, before extensive network disruption, may be most effective ^69.
Small molecule tau aggregation inhibitors represent another therapeutic approach, aiming to prevent the conversion of normal tau to pathological forms ^70. Similarly, compounds targeting tau phosphorylation—either by inhibiting kinases or activating phosphatases—seek to restore tau homeostasis ^71.
Beyond directly targeting tau pathology, interventions that enhance network function may provide symptomatic benefit. Transcranial magnetic stimulation (TMS) and transcranial direct current stimulation (tDCS) can modulate cortical excitability and enhance functional connectivity, potentially compensating for network dysfunction in PSP ^72. Preliminary studies suggest that repetitive TMS to frontal regions may improve executive function in PSP patients ^73.
Cognitive rehabilitation approaches may similarly harness network plasticity, engaging affected networks to maintain function and potentially slow degeneration. The concept of "cognitive reserve"—the capacity of networks to compensate for pathology through enhanced connectivity or alternative pathways—may predict responsiveness to such interventions ^74.
The dopaminergic, cholinergic, and serotonergic systems are all affected in PSP, contributing to network dysfunction through distinct mechanisms. Dopamine replacement therapy provides modest benefit for some PSP patients, particularly those with PSP-P phenotype, but does not address the underlying pathology ^75. Cholinesterase inhibitors have shown limited efficacy for cognitive symptoms, reflecting the more widespread network disruption compared to Alzheimer's disease ^76.
The emerging field of personalized network medicine aims to integrate individual patient connectivity profiles into therapeutic decision-making. By characterizing each patient's specific pattern of network disruption, clinicians may better predict symptom progression, select optimal interventions, and monitor treatment response ^77. This approach requires validated connectivity biomarkers and a deeper understanding of the relationship between network topology and clinical features.
The MAPT gene, located in a region of chromosome 17 subject to extensive linkage disequilibrium, demonstrates strong association with PSP susceptibility. The H1 haplotype of MAPT is overrepresented in PSP cases, and specific H1 subhaplotypes may confer additional risk ^78. Understanding how MAPT genotype influences network vulnerability and response to therapy represents an important avenue for future research.
Integrating network-level analyses with molecular profiling promises to clarify the mechanisms linking tau pathology to network dysfunction. Single-cell transcriptomic studies in PSP brain tissue are beginning to reveal cell-type-specific expression changes that may underlie selective neuronal vulnerability ^79. Combining these molecular data with network connectivity maps may identify key hub genes and pathways whose disruption produces network-level effects ^80.
The study of brain network connectivity in Progressive Supranuclear Palsy has transformed our understanding of this devastating disorder, revealing how focal tau pathology produces widespread functional disruption through disconnection of vulnerable circuits. The network perspective integrates clinical, imaging, and pathological observations into a coherent framework, explaining the characteristic symptom pattern and guiding therapeutic development.
The recognition that PSP represents a "disconnection syndrome" affecting rich club hubs and their distributed connections has profound implications for diagnosis and treatment. Connectivity biomarkers show promise for earlier diagnosis and more sensitive monitoring of disease progression. Understanding the mechanisms of tau propagation along network pathways informs disease-modifying therapeutic strategies. As our understanding of PSP network biology deepens, the prospect of effective interventions—targeting both the underlying pathology and its network consequences—becomes increasingly tangible.