Plasmids
Alternate codon non-G4C2 pcDNA3.1(+)-GFP-DPR100 plasmids are previously described44. To generate an EGFP control plasmid, GFP was extracted from pEGFP-N1 (Clontech, 6085-1) and subcloned into the pcDNA3.1(+) vector using BamHI and NotI restriction sites. pBabe-mCherry-GFP-LaminB1 retroviral vector is previously described20, from which a pBabe-mCherry-GFP-LaminB1S391A construct was subsequently generated by site-directed mutagenesis. A LaminB1313-585 fragment plasmid for bacterial expression was made to generate recombinant purified LaminB1 for in vitro kinase assay use. Chemically synthesised DNA encoding the human LaminB1313-585 fragment was subcloned into a pETDuet-1 vector and transformed into E. coli Rosetta II cells.
Mammalian Cell Line Culture and Pharmacological Treatments
SH-SY5Y (ATCC, CRL-2266), PtK2 (ATCC, CCL-56) and MRC-5 fibroblast (ATCC, CCL-171) cells were grown in Advanced DMEM/F12 with 10% FBS, 5% penicillin-streptomycin and 5% L-glutamine (Gibco). For immunohistochemical assays, SH-SY5Y cells were grown in coverslip-bottom 96 well plates (ibidi, 89626). For biochemical assays and EV generation cells were grown in 6 well plates and 150 mm dishes, respectively. PtK2 cells were grown in 35 mm glass coverslip bottom dishes (Griener Bio-one).
SH-SY5Y and PtK2 cells were treated with either DMSO (0.1%) or BafilomycinA1 (Sigma) in DMSO 0.1% at concentrations and time durations indicated in figure legends. In p38 inhibition experiments, SH-SY5Y were treated with 20 µM SB203580 (Tocris Bioscience, 1202). In experiments measuring DNA damage, SH-SY5Y cells were treated with 10 µM etoposide (Sigma, E1383)23 as a positive control condition. In experiments probing markers of apoptosis, as a positive control a sample of cells were treated with 1 µM staurosporine (Sigma, S6942) for 4 hours prior to the fixation time point.
Rodent Primary Cortical Neuron Culture
Cells were isolated from E18 Sprague Dawley rat embryos (Charles River) and cultured in Neurobasal medium containing 0.25% penicillin-streptomycin, 0.25% Glutamax and 2% B-27 (all Gibco), on 13 mm diameter glass coverslips coated in 0.1 mg/ml poly-D-Lysine
GFP-DPR100 Transfection
All GFP-DPR100 transfections were performed in rodent primary cortical neuron cultures at 6 DIV. Cells were transfected with 0.5 µg DNA per well for either EGFP alone, GFP-PR100, GFP-GR100 or GFP-GA100 using 1 µl per well of Lipofectamine-3000 (Invitrogen) in OptiMEM media (Gibco), according to the manufacturer protocol. After 4hrs, transfection media was replaced with the original culture media. At 72hrs post-transfection, cells were fixed for immunofluorescence staining.
Retroviral Infection
To generate a stable mCherry-GFP-LaminB1S391A cell line, 10 µg pBabe-mCherry-GFP-LaminB1S391A was transfected into Phoenix retroviral packaging cells (ATCC, CRL-3213). Media was replenished the following day, collected 48hrs later and passed through a 0.45 µm filter. Filtered media was then added to MRC-5 fibroblasts with 8 µg/ml polybrene (Santa Cruz, sc-134220). Media was replaced the following day, with 1 µg/ml puromycin added for selection after 48hrs until all non-transduced control cells were dead.
In Vitro Immunofluorescence Staining
All cells were fixed with 4% paraformaldehyde for 10 minutes at room temperature and permeabilised with 0.1% Triton X-100 (Sigma). Non-specific immunoreactivity was blocked with 1% BSA or 5% NGS for 30 minutes at room temperature. Primary antibody incubation was performed for 1 hour at room temperature.
Primary antibodies: rabbit anti-LaminB1 1:500 (Abcam, ab16048); mouse anti-LaminB1 1:500 (Proteintech, 66095-1-IG); mouse anti-Mab414 1:500 (Abcam, ab24609); guinea-pig anti-p62 1:500 (Progen, GP62-C); rabbit anti-cleaved PARP-1 1:500 (Abcam, ab32561); rabbit anti-cleaved caspase-3 1:400 (Cell Signalling Technology, 9661S); chicken anti-MAP2 1:1000 (Abcam, ab5392); rabbit anti-p21 1:500 (Cell Signalling Technology, 2947); mouse anti-phospho-γH2AX 1:1000 (Sigma, 05-636).
Secondary antibody incubation was performed for 2 hours at room temperature. Secondary antibodies, all 1:500: goat anti-rabbit Alexa Fluor 488 (Invitrogen, A11034); anti-mouse Alexa Fluor 546 (Invitrogen, A11030); anti-guinea pig Alexa Fluor 555 (Invitrogen, A21435); anti-rabbit Alexa Fluor 546 (Invitrogen, A11035); anti chicken Alexa Fluor 633 (Invitrogen, A21103). Cells were incubated with 1ug/ml Hoechst-33342 (Invitrogen, H3570) for 10 minutes at room temperature. Primary neurons on coverslips were then mounted on slides with Dako (Agligent, S3023), all other cell preparations were imaged directly in microplates.
TUNEL Assay
TUNEL assay was carried out in fixed, permeabilised cells or Drosophila whole brains with a commercial kit (Roche, 12156792910) according to manufacturer instructions. Prior to the TUNEL assay, positive control samples were incubated with DNase I (Promega, M6101) in 50 mM Tris-HCl (pH 7.5) and 1 mg/ml BSA at 37°C.
Extracellular Vesicle Isolation
Conditioned media was harvested from SH-SY5Y cells treated with BafilomycinA1 or DMSO for 24 hrs and centrifuged at 300 ×g for 5 minutes to remove dead cells. The supernatant was then extracted into ultracentrifuge tubes (Beckman Coulter, 344058) for processing into distinct pellet fractions corresponding to apoptotic bodies (‘8K’ fraction), larger extracellular vesicles (‘20K’ fraction) or smaller extracellular vesicles (‘100K’ fraction). To obtain these fractions, sequential ultracentrifugation was carried out at 4°C (SW 32 Ti rotor, Optima XPN-80, Beckman Coulter). For each fraction a wash step was included whereby the supernatant from the first ultracentrifugation was removed and temporarily stored at 4°C, whilst 1X PBS was added to the pellet and ultracentrifugation repeated at the same speed and duration. The ultracentrifugation steps were as follows. 8K pellet fraction: 8,000 ×g for 30 minutes; 20K pellet fraction: 20,000 ×g for 60 minutes; 100K fraction: 100,000 ×g for 70 minutes. Once isolated all final clean pellets obtained were stored at -70°C prior to further biochemical processing. Total protein and double strand DNA measurements of the EV fractions were obtained by UV-Vis spectrophotometry. 2µL of each sample was loaded onto the pedestal of a Thermo Scientific™ NanoDrop™ One Microvolume UV-Vis Spectrophotometer and blanked with 0.1µM filtered PBS between samples. Measurements were taken in duplicate, and the average of those measurements are reported.
Western Blots
Whole cell lysates or snap frozen Drosophila heads were prepared for Western blotting as follows. Cells were trypsinised, resuspended in RIPA buffer and incubated on ice for 30 minutes. Drosophila were mechanically lysed in Laemmli buffer. Samples were centrifuged at 13,000 rpm in a standard benchtop centrifuge at 4°C for 5 minutes. The supernatant was then extracted into fresh tubes. Pierce BCA protein assay was performed to determine total protein concentration of samples, using a commercially available kit (ThermoFisher Scientific, 23225) according to instructions.
Samples were boiled in 1X Laemmli buffer with 10% β-mercaptoethanol at 95°C for 5 minutes. Separation was then carried out by SDS-PAGE by loading 30 µg of protein, and then transferred onto nitrocellulose membrane (ECL, 0.2 µm pore size, Amersham GE) using a Mini-PROTEAN Tetra Cell tank (BioRad). Non-specific binding was blocked with 5% (w/v) milk powder or BSA in TBST overnight at 4°C. Primary antibodies were incubated overnight at 4°C. Primary antibodies: rabbit anti-GAPDH 1:2000 (Cell Signalling Technology, 2118); rabbit anti-LaminB1 1:1000 (Abcam, ab16048); rabbit anti-β-actin 1:1000 (Signalway, 21338); mouse anti-p62 1:1000 (Abnova, H00008878-1701); mouse anti-β-tubulin 1:1000 (DSHB, E7), mouse anti-Drosophila LaminB 1:1000 (DHSB, ADL67.10), mouse anti-p53 1:2000 (Santa Cruz, sc-126), mouse anti-phospho-γH2AX 1:1000 (Sigma, 05-636), rabbit anti-p21 1:1000 (Cell Signalling Technology, 2947). Anti- HRP- or dye-conjugated secondary antibodies in 5% (w/v) milk powder or BSA in TBST were incubated for 1 hour at room temperature. Secondary antibodies: goat anti-rabbit IgG (CalbioChem, DC03L); goat anti-mouse IgG (CalbioChem, 3239294101); goat anti-mouse IRDye 800 Cw 1:10,000 (LI-COR, D10825-15); goat anti-rabbit IRDye 680 RD (LI-COR, D11102-15), all 1:10,000. Specific protein signals were detected by chemi-illuminescence using SuperSignal West Pico Chemiluminescent Substrate (ThermoFisher Scientific, 34579), or via dye fluorescence using an Odyssey Imaging System (LI-COR).
To prepare isolated extracellular vesicle samples for Western blot, pellets were resuspended in 1X PBS and diluted 1:3 in RIPA buffer. Total protein concentration of samples was determined using a commercial micro-BCA kit (ThermoFisher Scientific, 23235) according to instructions. 2X Laemmli buffer with 10% β-mercaptoethanol was then added 1:1 to the pellet sample volume in the ultracentrifuge tube. Tubes were then placed in boiling water for 5 minutes.
LaminB1 ELISA
Quantitative measurement of LaminB1 protein concentration in extracellular vesicle pellet samples was carried out by ELISA using a commercially available kit (Abcam, ab252351-10001) according to instructions.
NanoParticle Tracking Analysis
Extracellular vesicle samples in 0.1 µM double filtered 1X PBS were loaded onto a Nanosight NS300 with 488 nm scatter laser and high sensitivity camera (Malvern Instruments Ltd., Malvern, UK). Nanoparticle tracking and quantification of size and number of imaged particles was performed with NTA2.1 software (Nanosight, Malvern, UK).
p38 and LaminB1 In Vitro Kinase Assay
Expression and purification of recombinant active p38α and a LaminB1313–385 fragment was performed as previously described45. An in vitro kinase assay was performed whereby 1mg/ml LaminB1313 − 586 was incubated with either 30 or 60 ng active p38α for either 0.5, 1 or 2 hours. The reaction was carried out in 1X kinase buffer (25 mM Tris/HCl pH 7.5, 5 mM β-glycerolphosphate, 2 mM DTT, 0.1 mM Na3VO4, and 1 mM MgCl2). ATP (550 µM) was added to the incubation mixture to start the reaction at 37°C. At the end of reaction, the samples were collected with 2X sample buffer (20% glycerol, 6% SDS in 0.12 M Tris pH 6.8, 10% β-mercaptoethanol, and 0.4% Bromophenol blue), heated to 95°C for 10 minutes, and run on 10% SDS gel under denaturing conditions. Western blot was then executed as previously outlined, staining with Pro-Q Diamond Phosphoprotein Gel Stain (ThermoFisher Scientific, P33300).
Drosophila experiments
All fly stocks were routinely maintained at 18°C in an incubator with a standard 12-hour light cycle, on a standard fly food mixture of yeast, agar, and cornmeal with nipagin and propionic acid. For ageing experiments desired progeny were maintained at 29°C and 60% humidity with a 12-hour light cycle.
The following Drosophila stocks were used: Elav-Gal4, UbiGal80ts, UAS-PR8, UAS-PR64, UAS-GR8, UAS-GR64, UAS-GA8, UAS-GA64, UAS-p38α RNAi (VDRC stocks: 34238GD, 102484KK, 52277GD), UAS-p38β RNAi (VDRC stock: 108099KK), w1118; LamBT413A{PBacDsRed}/CyO, w1118; LamBT413A/CyO, w1118 (all Well Genetics).
All non-G4C2UAS-DPR8/64 stocks with a 3’ HA tag are previously described35. To obtain flies pan-neuronally expressing inducible UAS-DPR8/64 constructs, Elav-Gal4;UbiGal80ts virgins were crossed to UAS-DPR8/64 males and maintained at 18°C. F1 progeny were transferred to 29°C for ageing for all experimental procedures. To obtain adult flies pan-neuronally co-expressing inducible UAS-PR8/64 constructs and UAS-p38 RNAi, UAS-PR8/64 and UAS-p38 RNAi virgins were separately crossed to appropriate balancer chromosome stocks. Balanced virgins from F1 progeny which incorporated UAS-p38 RNAi were then crossed to balanced F1 males incorporating the UAS-PR8/64. Finally, F1 generation males from this cross, incorporating both UAS constructs, were crossed to Elav-Gal4;UbiGal80ts virgins.
Lifespan analysis was performed as previously described46. Data are presented as survival curves and statistical comparisons between 8 and 64 repeat length genotype pairs for each DPR were carried out by log-rank test.
Negative geotaxis assays were performed as previously described46. Number of flies climbing to each cm increment was scored, and a genotype average was calculated for each timepoint across 5 trials. Flies which jumped or did not carry out a vertical climb in one movement burst were excluded for that trial.
Adult fly brain immunofluorescence was performed as previously described46. For all immunofluorescence results shown, adult female fly brains were dissected at 15 days of ageing at 29°C just prior to end-stage of lifespan of the most toxic DPR genotypes. For results where w1118 L3 stage larvae brains are shown as positive apoptosis controls, the same process was applied.
Primary and secondary antibody incubations were carried out overnight at 4°C. Primary antibodies: anti-Drosophila LaminB 1:500 (DSHB, ADL67.10); anti-HA tag 1:100 (Sigma, H6908); anti-cleaved-DCP-1 1:200 (Cell Signalling Technology, 9578); anti-cleaved caspase-3 1:200 (Cell Signalling Technology, 9661S); anti-Ref(2)P 1:50047. Secondary antibodies: 1:200 goat anti-mouse AlexaFluor 488 (Invitrogen, A11001) and goat anti-rabbit AlexaFluor 546 (Invitrogen, A11035). 1 µg/ml Hoechst-33342 nuclear stain was applied for 10 minutes at room temperature. Brains were mounted on slides with Dako (Agligent, S3023).
Human Post-Mortem Tissue Immunohistochemistry
Human post-mortem paraffin-embedded 7µm thickness frontal cortex sections were obtained from the London Neurodegenerative Diseases Brain Bank under Health Research Authority REC reference 18/WA/0206. Clinical and neuropathological case details are described in Extended Data Table 1.
All immunofluorescence staining, image acquisition and analysis of sections was carried out blind. Sections were warmed and dewaxed via short sequential washes in: xylene twice, 95% IMS twice and 95% IMS, followed by rinsing in distilled water. Antigen retrieval was carried out by boiling sections in pH 9.0 Tris-EDTA buffer (10 mM Tris base, 1 mM EDTA solution, 0.05% Tween 20), once cooled sections were briefly rinsed in PBS. Blocking of non-specific antibody binding was carried out by incubating sections in a blocking solution of 5% NGS, 1% BSA and 0.5% Triton X-100 in PBS for 1 hour. Sections were incubated with primary antibodies overnight at 4°C. Primary antibodies: chicken anti-MAP2 1:200 (Abcam, ab5392), rabbit anti-LaminB1 1:500 (Abcam, ab16048). Secondary antibody incubation was carried out for 1 hour. Secondary antibodies: goat anti-rabbit AlexaFluor 488 (Invitrogen, A11034) and goat anti-chicken AlexaFluor 633 (Invitrogen, A21103) both 1:250. Autofluorescence was quenched with Sudan Black B (199664, Sigma) solution for 10 minutes, followed by wash steps in 70% ethanol, distilled water, and PBS. Sections were incubated with 1 µg/ml Hoechst-33342 for 10 minutes.
K-means cluster analysis
Clustering analysis of single-cell data from LaminB1 and MAP2 immunofluorescence parameters measured in human post-mortem frontal cortex was executed in R using the kmeans function from stats package (version 4.1.1), with parameters k = 4, iter.max = 100 and nstart = 10. Input variables used were: LaminB1 nuclear circularity, cytoplasm area, nuclear area, or all of the above plus LaminB1 cytoplasm puncta (measured per area of cytoplasm). The most appropriate number of clusters k was pre-determined via the within-cluster sum of squares method whereby within-cluster sum of squares is plotted against increasing iterations of k to generate an elbow plot.
Within each identified cluster, statistical comparison of frequency of cells from different neurodegenerative disease or control categories was carried out in R by χ2 test followed by post-hoc pairwise binomial test with Bonferroni correction, implemented by chisq.test and binom.test functions respectively within the stats package.
Atomic Force Microscopy
AFM was carried out on live PtK2 cells treated either with DMSO or 10 nM Bafilomycin for 24hrs. All AFM experiments were executed on a BioScope Resolve system (Bruker) mounted on an AxioObserver.Z1 optical microscope (Zeiss) with 37°C temperature control. Prior to AFM assay, average z depth location of the nuclear surface and the z depth of cytoplasm above the nucleus were determined by confocal microscopy visualisation of a sample of PtK2 cells where cytoskeleton and nucleus had been stained by immunofluorescence for actin and DAPI, respectively. Average distance from the cell surface to nuclear surface was found to be approximately 1 µm. From this, cytoplasmic indentation during experiments was estimated to take place for probe depth distance 0–1 µm (where 0 = point of contact), as cytoplasm is much softer than the nucleus and is therefore compressed first. Nuclear surface indentation was estimated to take place from 1 µm to maximal indentation depth (< 2 µm). This was confirmed by a significant upward change in the force-indentation plot starting at 1 µm indentation.
Colloidal probes (PT.Si02.SN2.5, Novascan) were calibrated using the thermal noise method and then used for experimental AFM measurements in force curve mode above individual nuclei whose locations were determined by bright field microscopy. The following probe parameters were used: 8 µm ramp, 16 µm/s tip velocity, 15 nN force threshold. This was repeated for 200 cells for each treatment condition, for 3 different cell cultures.
Using a bespoke Python program48, elasticity was computed by fitting force indentation data in two different indentation ranges corresponding to cytoplasmic and nuclear elasticities.
Confocal Microscopy
Image acquisition of immunofluorescence for all SH-SY5Y experiments was carried out directly in 96-well microplates using an Opera Phenix high-content imaging system (PerkinElmer). Confocal scanning was carried out using a 60x 1.15 NA water immersion objective. Maximum intensity projection images were compiled from z-stacks.
All other fluorescence image acquisition was carried out using an A1R inverted confocal microscope (Nikon), with either a 20x 0.75 NA air objective or a 60x 1.49 NA oil immersion objective. For in vitro immunofluorescence, maximum intensity projections were generated from z-stacks of 0.125 µm depth planes, of sufficient total to capture the entire z-height of the cells imaged.
For Drosophila whole brain, up to 20 single-plane images were obtained at 5 µm z-plane increments in large image mode stitching nine 1024-pixel tiles together.
For human post-mortem frontal cortex sections, images were acquired in a single z-plane in a series of 4 consecutive imaging locations starting from the tissue edge and working in towards the boundary between grey and white matter, determined from MAP2 staining morphology. Two 4-image sequences were obtained per case, from 2 independent starting points, from 1 single tissue section. Each image was comprised of nine 1024 x 1024-pixel tiles stitched together in large image field mode.
Immunofluorescence Quantification
In all instances, regularity of LaminB1 nuclear morphology was determined using circularity, defined as: circularity = perimeter2/4π × area. Using Python (v 3.8.13) all quantification of LaminB1 nuclear circularity in vitro was carried out using the nuclei model of the Cellpose algorithm49 to segment nuclei masks based on LaminB1 immunofluorescence. Circularity of masks was then calculated using the Skimage package.
Analysis of all SH-SY5Y immunofluorescence was carried out using Harmony (v 4.9, PerkinElmer). A basic pipeline was developed to identify and segment cells utilising ‘Find Nuclei’ and ‘Find Cytoplasm’ functions with appropriate filter parameters to capture the cell population. The ‘Calculate Intensity’ function was used to quantify fluorescence intensity. The ‘Find Spots’ function was used to quantify cytoplasmic p62 or LaminB1 puncta, or nuclear γH2ax foci within appropriate respective ROIs.
Analysis of primary neuron and human post-mortem immunofluorescence was carried out in NIS Elements Advanced Research Software (v5.01.00, Nikon). For human post-mortem tissue, neuronal cytoplasm and nuclear ROIs were drawn manually on the basis of MAP2 and LaminB1 signal, respectively. Identified neurons were excluded if the LaminB1 nuclear ring cross-section was incomplete and incompatible with reliable circularity quantification. A fluorescence intensity threshold was applied to LaminB1 fluorescence to detect cytoplasmic puncta. Cytoplasm area was calculated by subtracting the nuclear ROI area from the MAP2 soma ROI area.
For quantification of cleaved caspase-3 immunofluorescence intensity in primary cortical neurons, nuclei were identified based on Hoechst intensity with object size and shape restrictions applied. A GFP-positive sub-population was identified based on GFP fluorescence intensity thresholding. Finally, GFP-positive neurons were identified based on MAP2 fluorescence intensity thresholding. In the case of GFP-GA100, GFP-intense spots were identified using GFP fluorescence intensity thresholding, and then cells were screened and selected manually based on whether an identified GFP-positive spot localised to a MAP2 positive soma area. For quantification of LaminB1 karyoptosis parameters in higher magnification images of single GFP-positive cells identified & captured during acquisition, cytoplasm ROIs were manually drawn based on MAP2 signal. Intensity thresholding was applied to LaminB1 immunofluorescence to generate a nuclear ROI for circularity measurement. A second intensity threshold with size and shape filters, was applied to LaminB1 immunofluorescence in the cytoplasm ROI to quantify puncta.
Statistical analysis
Statistical tests utilised are described in figure legends and were carried out in GraphPad Prism (version 9.4.1) or RStudio (version 4.1.1). In the latter case, all tests were carried out using their eponymous function within the stats package, except for Dunnett’s post-hoc test, which was executed using the DunnTest function within the DescTools package (version 0.99.44). Where required to ensure equivalence of sample size, random sampling of datasets was carried out using the sample function within the stats package.