A novel lncRNA FAM151B-DT regulates degradation of aggregation prone proteins

Abstract

Neurodegenerative diseases share common features of protein aggregation along with other pleiotropic traits, including shifts in transcriptional patterns, neuroinflammation, disruption in synaptic signaling, mitochondrial dysfunction, oxidative stress, and impaired clearance mechanisms like autophagy. However, key regulators of these pleiotropic traits have yet to be identified. Here, we used transcriptomics, mass spectrometry, and biochemical assays to define the role of a novel lncRNA on tau pathophysiology. We discovered a long non-coding RNA (lncRNA), FAM151B-DT, that is reduced in a stem cell model of frontotemporal lobar dementia with tau inclusions (FTLD-tau) and in brains from FTLD-tau, progressive supranuclear palsy, Alzheimer's disease, and Parkinson's disease patients. We show that silencing FAM151B-DT in vitro is sufficient to enhance tau and α-synuclein aggregation. To begin to understand the mechanism by which FAM151B-DT mediates tau aggregation and contributes to several neurodegenerative diseases, we deeply characterized this novel lncRNA and found that FAM151B-DT resides in the cytoplasm where it interacts with tau, α-synuclein, HSC70, and other proteins involved in protein homeostasis. When silenced, FAM151B-DT blocks autophagy, leading to the accumulation of tau and α-synuclein. Importantly, we discovered that increasing FAM151B-DT expression is sufficient to promote autophagic clearance of phosphorylated tau and α-synuclein, and reduce tau and α-synuclein aggregation. Overall, these findings pave the way for further exploration of FAM151B-DT as a promising molecular target for several neurodegenerative diseases.

Fig. 1. FAM151B-DT expression is significantly reduced in MAPT mutant neurons and tauopathy brains.

A Schematic representation of the discovery of lncRNAs that are differentially expressed (DE) in MAPT mutant iPSC-neurons compared with isogenic controls. FAM151B-DT was among the 15 common dysregulated lncRNAs in MAPT R406W (n = 3), P301L (n = 2), and IVS10 + 16 (n = 4) neurons [11]. B FAM151B-DT is significantly reduced in MAPT R406W, P301L, and IVS10 + 16 neurons compared with isogenic controls. Graph represents Log₂FC. C FAM151B-DT is significantly reduced in human FTD with tau inclusions (n = 3) and neuropathology free controls (n = 3); AD (n = 84), PSP (n = 82) brains compared with control brain tissues (n = 76). Graph represents Log₂FC. *, p ≤ 0.05; ***, p < 0.001; ****, p < 0.0001.

Fig. 2. FAM151B-DT expression regulates tau seeding.

A Schematic representation of the FRET-based tau seeding assay. B qPCR for FAM151B-DT reveals significant reduction of FAM151B-DT levels after siRNA treatment in tau biosensor cells. Data is representative of three independent experiments with biological triplicates in each experiment. C Silencing FAM151B-DT significantly increased integrated FRET density compared to control siRNA-treated tau biosensor cells. Quantification of the FRET signal, normalized to controls. Error bars represent SEM, n = 20,000 cells per experiment. Data is representative of three independent experiments with biological triplicates in each experiment. D qPCR for FAM151B-DT reveals significant increase in FAM151B-DT levels upon overexpression in tau biosensor cells. Data is representative of three independent experiments with biological triplicates in each experiment. E FAM151B-DT overexpression significantly decreased the integrated FRET density compared to ctrl-vector-treated tau biosensor cells. Quantification of the FRET signal, normalized to controls. Error bars represent SEM, n = 20,000 cells per experiment. Data is representative of three independent experiments with biological triplicates in each experiment. Student’s t-test, *, p ≤ 0.05; **, p < 0.001; ****, p < 0.0001.

Fig. 3. Proteins interacting with FAM151B-DT.

A–B Nuclear and cytosolic isolation of RNAs to define the localization of FAM151B-DT relative to well characterized lncRNAs (nuclear localized MALAT1 and cytosolic TUG1). A iPSC-derived neurons. B SH-SY5Y cells. C–D Tau interacts with FAM151B-DT. Data is representative of three independent experiments with biological triplicates in each experiment. C Immunoprecipitation of tau (Tau5) in SH-SY5Y cells. D RNA immunoprecipitation (RIP)-qPCR validation of the tau-bound RNA fraction shows robust FAM151B-DT expression. GAPDH included as a negative control. Data is representative of three biological replicates. Data is representative of three independent experiments with biological triplicates in each experiment. Two-way ANOVA, ***, p < 0.001. E Comprehensive identification of RNA binding proteins (ChIRP) workflow. Schematic representation includes the positions of even and odd probes along the entire FAM151B-DT RNA sequence. F ChIRP-qPCR reveals enrichment of FAM151B-DT, compared to the housekeeping transcript GAPDH. One way ANOVA, ****, p < 0.0001. Data is representative of three independent experiments with biological triplicates in each experiment. G ChIRP-mass spectrometry hits of proteins retrieved by even and odd oligos specific for FAM151B-DT compared with non-targeting LacZ control oligos. 215 total proteins were identified across the three samples using a protein probability ≥99%; ≥2 peptides mapped; and ≥95% peptide probability (coverage). Three biological replicates were used for each set of oligos (e.g. even or odd). H GO enrichment analysis of the FAM151B-DT proteome (101 proteins) in SH-SY5Y cells. I Schematic of string network analysis of FAM151B-DT proteome reveals multiple clusters (see Supplementary Fig. 5). J Visualizing cluster 1. Protein-protein interaction enrichment p-value = 1.0e^-16. K Top 10 most significant pathways in cluster 1 identified by KEGG-Pathway enrichment analysis.

Fig. 4. FAM151B-DT and tau share interacting partners that are disrupted in MAPT mutant neurons and human brains.

A Schematic illustrating workflow. The 101 FAM151B-DT interactors were compared with tau interactors defined in iPSC-neurons, human brains, and neurofibrillary tangles (NFTs). The common set of interactors was then evaluated for differential expression at the RNA level in MAPT mutant iPSC-neurons and FTLD-tau brains. B Upset plot for FAM151B-DT interactors that are also reported to interact with tau in human iPSC-derived neurons from MAPT WT, P301L, and V337M [20], soluble, total tau from human post-mortem brain tissues from AD patients [–23], and NFT from AD patient brains [24]. C GO enrichment analysis of the FAM151B-DT interactome overlapped with tau and NFT interactomes (n = 19). D Heatmap representing differential expression of FAM151B-DT/tau shared interactome in MAPT mutant neurons and human brains from FTD with tau inclusions, PSP, and AD patients. *, p ≤ 0.05. E ChIRP-qPCR validation of FAM151B-DT pull-down using biotinylated DNA oligos. GAPDH was included as a negative control. Data are represented as mean ± SEM from three replicates. Data is representative of three independent experiments with biological triplicates in each experiment. One-way ANOVA, ****, p < 0.0001. Right panel, ChIRP-immunoblot of FAM151B-DT pull-down fractions probed with HSC70 and tau (Tau5) antibodies. F Immunoprecipitation of HSC70 in SH-SY5Y cells. Right panel, RIP-qPCR validation of HSC70-bound RNA fractions illustrates robust FAM151B-DT expression. GAPDH included as a negative control. Data are representative of three biological replicates. Data is representative of three independent experiments with biological triplicates in each experiment. Two-way ANOVA, ** p < 0.01. G Crystal structure HSC70 (PDB: 4FL9) with FAM151B-DT interaction sites annotated in red. H Schematic of the FAM151B-DT-HSC70-tau complex.

Fig. 5. FAM151B-DT modulates autophagy and tau.

SH-SY5Y cells were transiently transfected with scrambled (scr) or siFAM151B-DT siRNAs and control vector or FAM151B-DT overexpression plasmid. A Representative immunoblots of SH-SY5Y cell lines detecting HSC70, lysosome (LAMP1, LAMP2A) and autophagy (LC3 and p62) markers, and tau (AT8[ptau] and Tau5 [total tau]). B–C Quantification of western blot data from (A). B Quantification of FAM151B-DT silencing. White bar, scr. Gray bar, siFAM151B-DT. C Quantification of FAM151B-DT overexpression. White bar, vector. Gray bar, FAM151B-DT. D–E RNA expression of TFEB, LAMP2A, CTSD (Cathepsin D), and MAPT. D Quantification of FAM151B-DT silencing. White bar, scr. Gray bar, siFAM151B-DT. E Quantification of FAM151B-DT overexpression. White bar, vector. Gray bar, FAM151B-DT. Data are presented as mean ± SEM from four independent experiments. Data is representative of three independent experiments with biological triplicates in each experiment. Student’s t-test, *, p < 0.05; **, p < 0.01; ****, p < 0.0001.

Fig. 6. FAM151B-DT interacts with autophagy substrates implicated in Parkinson’s disease.

A The presence of KFERQ-like motifs were evaluated among the FAM151B-DT interactome (n = 101). Motif types were plotted. B Crystal structure of α-synuclein (PDB: 6a6b [53],) illustrating the FAM151B-DT binding sites in red based on ChIRP-MS. C ChIRP-immunoblot of FAM151B-DT pull-down fractions (see Fig. 4E) probed with pα-Syn antibody in SH-SY5Y cells. Data is representative of three independent experiments. D FAM151B-DT expression is significantly reduced in PD brains (n = 99) compared with neuropathology free controls (n = 38). *, p = 0.04. E SH-SY5Y cells were transiently transfected with scrambled (scr) or siFAM151B-DT siRNAs and control vector or FAM151B-DT overexpression plasmid. Representative western blots of phosphorylated (Ser199, pα-Syn) and total α-synuclein (α-Syn). F Quantification of western blot data from (E). FAM151B-DT silencing: white circle, scrambled (scr); gray circle, siFAM151B-DT. FAM151B-DT overexpression: white arrow, vector; gray arrow, FAM151B-DT. Data are presented as mean ± SEM from four independent experiments. G–J α-synuclein biosensor assay. G Schematic representation of the FRET-based α-synuclein seeding assay. H qPCR for FAM151B-DT reveals significant reduction of FAM151B-DT levels after siRNA treatment in α-synuclein biosensor cells. Data is representative of three independent experiments. I Silencing FAM151B-DT significantly increased integrated FRET density compared to control siRNA-treated α-synuclein biosensor cells. Quantification of the FRET signal, normalized to controls. Error bars represent SEM, n = 10,000 cells per experiment. Data is representative of three independent experiments with biological triplicates. J qPCR for FAM151B-DT reveals significant increase in FAM151B-DT levels upon overexpression in α-synuclein biosensor cells. Data is representative of three independent experiments. K FAM151B-DT overexpression significantly decreased the integrated FRET density compared to ctrl-vector-treated α-synuclein biosensor cells. Quantification of the FRET signal, normalized to controls. Error bars represent SEM, n = 10,000 cells per experiment. Data is representative of three independent experiments with biological triplicates. Student’s t-test: **, p < 0.01; ***, p < 0.001; ****, p < 0.0001.