Age-Dependent Formation of TMEM106B Amyloid Filaments in Human Brain

Abstract

SummaryMany age-dependent neurodegenerative diseases, like Alzheimer’s and Parkinson’s, are characterised by abundant inclusions of amyloid filaments. Filamentous inclusions of the proteins tau, amyloid-β (Aβ), α-synuclein and TDP-43 are the most common. Here, we used electron cryo-microscopy (cryo-EM) structure determination to show that residues 120-254 of the lysosomal type II transmembrane protein 106B (TMEM106B) also form amyloid filaments in the human brain. We solved cryo-EM structures of TMEM106B filaments from the brains of 22 individuals with neurodegenerative conditions, including sporadic and inherited tauopathies, Aβ-amyloidoses, synucleinopathies and TDP-43opathies, as well as from the brains of two neurologically normal individuals. We observed three different TMEM106B folds, with no clear relationship between folds and diseases. The presence of TMEM106B filaments correlated with that of a 29 kDa sarkosyl-insoluble fragment of the protein on Western blots. The presence of TMEM106B filaments in the brains of older, but not younger, neurologically normal individuals indicates that they form in an age-dependent manner.

Homotypic fibrillization of TMEM106B across diverse neurodegenerative diseases

“…Interestingly, TMEM106B fibril structures similar to those identified in our study in patients across a wide range of proteinopathies as well as neurologically normal aged individuals was recently reported (Schweighauser et al, 2021). Although the findings are similar, our study is distinct in focusing primarily on genetic and sporadic forms of FTLD-TDP (n = 8, Table S1), whereas the work described in the recent report draws mainly from various tauopathies and synucleinopathies.…”

“…Although this suggests that TMEM106B(120-254) fibrils may not be a consistent feature of all tauopathies, our absence of evidence could simply be due to the limited n-number and varying brain regions in our cryo-EM studies of other tauopathies. It has also recently been suggested that variations in tissue fractionation protocols may significantly affect TMEM106B fibril yield (Schweighauser et al, 2021).…”

“…OPEN ACCESS (Schweighauser et al, 2021) does not preclude a role in disease etiology as a similar observation is seen for tau (Brier et al, 2016;Chatterjee et al, 2021) and a-synuclein filaments (Hatton et al, 2022). Additionally, it is unclear whether the formation of TMEM106B amyloid fibrils represents a primary pathogenic process or a non-specific secondary phenomenon (e.g., general downstream marker of lysosomal stress).…”

“…Additionally, it is unclear whether the formation of TMEM106B amyloid fibrils represents a primary pathogenic process or a non-specific secondary phenomenon (e.g., general downstream marker of lysosomal stress). The observation of TMEM106B fibrils in some, but not all, cases of PSP (Figures 2, S3D, and S3E) (Shi et al, 2021) and DLB (Figure 2) (Schweighauser et al, 2021) suggests that the degree of fibrillization is highly variable in neurodegenerative diseases. Nevertheless, the combined trend across the two studies is that in the majority of cases with known neurodegeneration, fibrillization of TMEM106B is substantially greater when compared to agematched controls (Figure S6) (Schweighauser et al, 2021).…”

“…Importantly, we and others (Jiang et al, 2022) contend that western blots do not reliably detect TMEM106B aggregation. An antibody has recently been generated that detects nonfibrillar TMEM106B by binding to residues (239-250), which are buried in the fibril core (Schweighauser et al, 2021). Such an antibody can only measure accumulation of monomeric TMEM106B in the soluble or insoluble fractions and is therefore not suitable for detecting fibrils.…”