Phenotype parallels protein seeding capacity in neurodegenerative diseases

Abstract

With the new era of disease‐modifying therapies for neurodegenerative diseases, a novel approach for the molecular classification of neurodegenerative diseases is needed. In this research letter, there is a summary of the advances made in Alzheimer's disease, Lewy body disorders, and progressive supranuclear palsy toward this classification.

Neurodegenerative diseases are a pathologically, clinically, and genetically diverse group of disorders that at present lack any disease‐modifying therapies [1]. Pathologically, these disorders, collectively referred to as proteinopathies, are characterized by disease‐specific protein aggregates in neurons and/or glia. Interestingly, although many of these diseases involve abnormal aggregation of the same protein, the deposits can occur in anatomically distinct regions, giving rise to specific patterns of cognitive and motor clinical phenotypes among disorders of a common protein [1].

Two of the most commonly aggregated proteins in the brain of patients with neurodegenerative disease are tau and alpha‐synuclein (α‐Syn), that due to their intrinsically disordered nature, can adopt a variety of conformations, including β‐sheet‐rich structures [2]. Converging evidence implies that in neurodegenerative diseases, a conformational change (misfolding) of the tau or α‐Syn monomer generates an aggregation nucleus (seed) that can recruit and in turn induce the aggregation of endogenous monomer [2, 3]. These seeds can self‐propagate and progressively spread between inter‐connected brain regions exploiting a variety of mechanisms of cell‐to‐cell transmission [2].

A critical unknown in the basic biology of neurodegenerative diseases is an understanding of how the same protein (tau or α‐Syn) can be associated with distinct diseases. The recent generation of pure fibrillar α‐Syn and tau polymorphs with differences in structural and phenotypic traits has led to the hypothesis that different α‐Syn and tau “strains” may be in part responsible for the pathologic and clinical heterogeneity seen among the neurodegenerative proteinopathies. The “strain” hypothesis proposes that the misfolded protein conformation determines which disease a patient will develop [4]. In prion diseases, the kinetics of the prion protein misfolding into each disease‐causing conformation typically translates into strain‐specific differences in onset and course of the disease [4]. Understanding the structure and molecular nature of these pathological polymorphs in diseased brains is crucial for the validation of the strain hypothesis in neurodegenerative diseases and for the development of targeted therapeutic interventions.

A great leap in resolving the structure of different protein aggregates has recently been provided by advances in cryo‐electron microscopy [5] where distinct structural filament folds have been found for different diseases. However, within a given disease, folding appears to be essentially identical from patient‐to‐patient when evaluated under cryo‐electron microscopy [6], leading to the question of how the aggregation of the same protein into the same protein fold can be associated with the great variability in the clinical presentations and duration of disease seen within patients with the same disease.

A possible answer to that question comes from recent work in Alzheimer's disease (AD) that has specifically implicated the cytoplasmic, soluble, oligomeric species (also known as high‐molecular weight tau (HMW‐Tau), in the pathogenesis and clinical heterogeneity of AD [3]. In contrast to fibrillar tau, these smaller oligomers, which are not evident by classical staining or by electron microscopy, can be detected biochemically [3]. These soluble forms of tau are highly heterogeneous in their size distribution and post‐translational modification patterns and increasing evidence suggests that such diversity may underlie the heterogeneity seen in neurodegenerative tauopathies [3]. A recent landmark study demonstrated that in AD patient brains, cases which contained more of this bioactive, propagation‐prone material than others were those which presented with the fastest clinical course [7]. But these HMW‐Tau species are not only found in AD brains. Our group has recently described that in the brains of patients with progressive supranuclear palsy (PSP), a four repeat (4R) tauopathy, these species could also play a role in the heterogeneity seen in the disease [8].

Using western blotting, we performed the first comprehensive mapping of HMW‐Tau in the PBS‐soluble fraction of 20 different brain regions from 25 neuropathologically confirmed PSP cases. We found that these soluble species were present in all regions examined, although with a high degree of variability between different brain regions. The temporal cortex and the hippocampus were the regions where most of these species were found, probably due to the presence of concomitant AD pathology, followed by the primary motor cortex. Interestingly, our analysis revealed that early affected regions in PSP [9], such as the subthalamic nucleus, the substantia nigra or the globus pallidus do not have a high presence of HMW‐Tau. An explanation for this interesting finding could be that over the course of the disease, these soluble tau species, present at the earlier stages of the disease, would become sequestered in larger insoluble aggregates at later stages of the disease [10]. Besides the high diversity of HMW‐Tau levels between regions, our brain mapping also showed great heterogeneity among PSP patients. To untangle the relevance of the diversity of HMW‐tau between PSP patients, we used 4R‐Tau seeding amplification assays (SAA) to analyze the tau seeding capacity in the primary motor cortex of all patients included in the study. Following these assays, we subclassified the PSP cases into three groups: high, intermediate, and low seeders, according to their ability to misfold recombinant 4R‐tau fragments and following the same classification that has previously been applied in AD [7] and multiple system atrophy  [11] patients. Interestingly, we found that the tau from high seeder PSP patients was more resistant to proteinase K and thermolysin digestion than the tau found in the low seeders, supporting the hypothesis that different biologically active conformers of tau exist across PSP brains and raising the intriguing possibility that different underlying pathophysiological mechanisms might occur in different subsets of patients, representing a biological basis for differences in the clinical course.

To test this hypothesis, we performed mass‐spectrometry‐based quantitative proteomics and analyzed the primary motor cortex proteome of PSP patients classified as high and low seeders. Differential expression analysis and pathway enrichment analysis revealed many enriched pathways up‐ and down‐regulated in high seeders relative to low seeders at the proteomic level, including a dysregulation of the adaptive immune system. Interestingly, a recent study using positron emission tomography  tracers that measure microglial activation and tau pathology in PSP patients, showed that neuroinflammation is key for progression of PSP [12] and our data would support this concept.

This heterogeneity in the seeding capacity between different patients with the same disease is not specific to diseases caused by tau deposits. Another recent study from our group [13] has also shown that in Lewy body disorders (LBD) despite similar amounts of α‐Syn pathology in the substantia nigra of patients with distinct dominant clinical symptoms (motor vs. cognitive), the seeding capacity of α‐Syn was dissimilar between LBD subtypes. Thus, hippocampal‐derived α‐Syn from patients with a cognitive‐predominant phenotype had significantly higher seeding capacity than that derived from patients with a motor‐predominant phenotype. Conversely, nigral‐derived α‐Syn had higher seeding capacity than that from cognitive‐predominant patients. Interestingly, nigral α‐Syn from patients with rapid disease progression (<3 years) had a higher seeding capacity than the rest of the patients included. We also performed mass‐spectrometry‐based quantitative proteomics and analyzed the substantia nigra proteome of LBD patients classified as high and low seeders. In LBD, the differential activation of the adaptive immune system did not play a major role between high and low seeders, however, we found a significant disruption in mitochondrial function and lipid metabolism in high seeder compared to low seeder cases.

These studies highlight the importance of bringing a multi‐disciplinary approach in neuropathology research. These novel approaches are vitally needed to extend our understanding of the contribution of the molecular diversity of tau, α‐Syn, and other amyloidogenic proteins to the pathobiology of neurodegenerative diseases, which has been previously limited to that demonstrated using conventional biochemistry and immunohistochemistry.

Using a multi‐pronged strategy, several groups have discovered unexpected differences in the seeding capacity of neurodegenerative disease associated proteins across a cohort of neuropathologically comparable brains. Understanding the relationship between the seeding differences and biological activity of α‐Syn and tau will be critical for future subclassification of neurodegenerative diseases, which, as proposed in other diseases, such as cancer [14] or cystic fibrosis [15], should go beyond the conventional clinical and neuropathological phenotyping and consider the structural and biochemical heterogeneity of α‐Syn and tau present in these patients.

Such an approach will facilitate the development of new personalized therapeutic strategies that might need to be directed at differently misfolded seeding competent species and might also have to be combined with other immunomodulatory strategies. Furthermore, a deeper understanding of how physicochemical factors influence the aggregation of different polymorphs or strains will provide support for the development of vitally needed, rapid, and structure‐based assays for the future diagnosis of neurodegenerative diseases.

AUTHOR CONTRIBUTIONS

Ivan Martinez‐Valbuena drafted the manuscript.

Martinez‐Valbuena I. Phenotype parallels protein seeding capacity in neurodegenerative diseases. Brain Pathology. 2024;34(2):e13238. 10.1111/bpa.13238

DATA AVAILABILITY STATEMENT

Data sharing is not applicable to this article as no new data were created or analyzed in this study.