Reactivated endogenous retroviruses promote protein aggregate spreading

Abstract

Prion-like spreading of protein misfolding is a characteristic of neurodegenerative diseases, but the exact mechanisms of intercellular protein aggregate dissemination remain unresolved. Evidence accumulates that endogenous retroviruses, remnants of viral germline infections that are normally epigenetically silenced, become upregulated in neurodegenerative diseases such as amyotrophic lateral sclerosis and tauopathies. Here we uncover that activation of endogenous retroviruses affects prion-like spreading of proteopathic seeds. We show that upregulation of endogenous retroviruses drastically increases the dissemination of protein aggregates between cells in culture, a process that can be inhibited by targeting the viral envelope protein or viral protein processing. Human endogenous retrovirus envelopes of four different clades also elevate intercellular spreading of proteopathic seeds, including pathological Tau. Our data support a role of endogenous retroviruses in protein misfolding diseases and suggest that antiviral drugs could represent promising candidates for inhibiting protein aggregate spreading.

Fig. 1. Upregulation of murine endogenous retrovirus in donor cells increases intercellular aggregate induction.

a Experimental workflow. Donor N2a NM-HA^agg clone passage 7 (P7) and 16 (P16) or donor EVs were added to recipient N2a NM-GFP^sol cells. Analysis was performed 16 h post-exposure. b Confocal images of recipient and donor cells P7 or P16. Shown are Z-stacks. Nuclei were stained with Hoechst. Note that donors have not been stained for NM-HA. Insets: Close-ups. c Percentage of recipient cells with induced NM-GFP aggregates upon coculture with donors P7 and P16. d Confocal images of recipient cells exposed to EVs from donors P7 and P16. e Percentage of recipient cells with NM-GFP aggregates. f Particles isolated from the conditioned medium of donor cells P7 and P16 post-thawing were analyzed by nanoparticle tracking. g Primary neurons expressing soluble NM-GFP were exposed to donor EVs of early or late passage. Quantitative analysis of neurons with NM-GFP aggregates. h Volcano plot of total cell proteome of donor N2a NM-HA^agg cells at lower and higher passage numbers. Proteins were ranked according to their P-value and their relative abundance ratio (log2 fold change) in cells of P16 compared to cells of P7. i Volcano plot of proteomes of EVs derived from donor N2a NM-HA^agg cells P16 versus P7. Proteins were ranked as above. j qRT-PCR analysis of env and gag mRNA of N2a NM-HA^agg cells. Shown is the fold change in expression in donor cells P16 versus P7. k Increased MLV Env and Gag expression upon continuous cell culture of cells. Env and GAPDH were detected on the same blot, Gag, and NM-HA on a second blot. Shown are Env surface unit SU (gp70) and Gag polyprotein Pr65 and capsid (CA). All data are shown as the means ± SD from three (g, j), six (f), or nine (c, e) replicate cell cultures. Three (c, e–j) independent experiments were carried out with similar results. P-values calculated by two-tailed unpaired Student’s t-test (c, e–i) or one-way ANOVA with Bonferroni´s multiple comparisons (j). ns: non-significant. Source data are provided as a Source Data file.

Fig. 2. Donor cells of late passage produce both EVs and active retroviral particles.

a Viral particles and EVs were precipitated from a conditioned medium of N2a NM-HA^agg cells of different passage numbers using polyethylene glycol. Reverse transcriptase (RT) activity was determined at P1, early passage (EP), and late passage (LP) using a colorimetric RT assay (Roche). b MLV-susceptible Melan-a cells were exposed to the conditioned medium of the donor clone. Western blot analysis was performed 6 days later using antibody ABIN457298 against Env/Gag. c The 100,000 × g pellet from the conditioned medium from donor clone N2a NM-HA^agg (late passage) was subjected to an Optiprep density gradient. Fractions were analyzed for particle numbers, reverse transcriptase (RT) activity, particle morphology by electron microscopy (EM), protein content by Western blot (WB), and aggregate-inducing activity. d Particle numbers of gradient fractions are determined using ZetaView. e Density gradient fractions were analyzed for Alix, endogenous Env/Gag (ABIN457298) and NM-HA. Alix and NM-HA were detected on the same blot, Env was detected on a separate blot by WB. f Reverse transcriptase (RT) activity identifies viral particles (fractions 8–11). g Transmission electron microscopy of particles in fractions 2, 3, 9, and 10. Scale bar: 500 nm. h NM-GFP aggregate-inducing activity of gradient fractions in recipient cells. N2a NM-GFP^sol cells exposed to different fractions were analyzed for induced NM-GFP aggregates 16 h post-exposure. i Recipient cells were cocultured with donor cells (LP) in the presence of different dilutions of anti-MLV Env antibody mAb83A25 for 1 h. Alternatively, recipients were cultured with donor EVs that had been pre-incubated with anti-Env antibodies for 1 h. Anti-Env antibodies were present throughout the experiment. Partly created with Biorender.com. j The percentage of recipient cells with induced NM-GFP aggregates was determined 16 h post coculture or EV addition. All data are shown as the means ± SD from three (a, d, f, h), 12 (j, coculture) or 6 (j, EVs) replicate cell cultures. Three (a, d, f, h, j), two (e), or one (g) independent experiments were carried out with similar results. P-values were calculated by one-way ANOVA with Tukey’s multiple comparisons. Source data are provided as a Source Data file.

Fig. 3. Epigenetic modulation of MLV expression affects aggregate spreading.

a Donor N2a NM-HA^agg cells of late passage (LP) were transfected with three siRNAs against endogenous env, gag, or non-silencing siRNA as control and subsequently cocultured with recipients. b Reduction of donor transcripts following siRNA treatment assessed by qRT-PCR. Shown are fold changes relative to mock control. c Env and Gag expression in N2a NM-HA^agg cells (LP) transfected with siRNA. Note that viral transcripts include a genomic transcript coding for gag/pol and env and a spliced transcript coding only for env. Antibodies were anti-MLV Env mAb83A25, anti-MLV Gag ab100970 and anti-GAPDH. Env and Gag Western blots were reprobed for GAPDH. Asterisk marks the uncleaved Env precursor, circle marks the SU domain. Modulation of Env expression is expected to change the rate of newly translated precursor to processed Env. Protein levels normalized to GAPDH are shown. d Donors cocultured with recipients 72 h post-transfection. Shown is the percentage of recipient cells with NM-GFP^agg normalized to control. e Donors (EP) were treated with 5-azacytidine (Aza), 5-aza-2´-deoxycytidine (Dec), or DMSO for 3 days and then cultivated in a normal medium for 5 days. Cells were subsequently cocultured with recipients for 16 h. f Expression of MLV env (left panel) or gag (right panel) transcripts in donor cells 5 days post-treatment assessed by qRT-PCR. g Western blot of Env and Gag after inhibitor treatment. The same blot was probed with anti-MLV Env mAb83A25, anti-MLV Gag ab100970 and anti-GAPDH. h Percentage of NM-GFP^agg recipient cells. i Donors (LP) were treated with l-methionine (l-M), betaine, choline chloride (CC), or medium control for 6 days. MLV env or gag transcripts were analyzed by qRT-PCR. j Expression of Env, Gag, and GAPDH probed on the same blot. k Percentage of NM-GFP^agg cells. All data are shown as the means ± SD from three (b, f, i), six (d, h), or 12 (k) replicate cell cultures. Three (b, d, f, h, i, k) independent experiments were carried out with similar results. P-values were calculated by one-way ANOVA with Dunnett’s post hoc test. Source data are provided as a Source Data file.

Fig. 4. Protease inhibitors blocking MLV maturation impair intercellular protein aggregate spreading.

a Workflow of compound tests in cocultures. Donor clone N2a NM-HA^agg (LP) and recipient N2a NM-GFP^sol cells were co-seeded and viral protease inhibitors were added at three concentrations 1 h later. 12 h post-drug treatment, donor and recipient cells were analyzed for the percentage of donor cells with NM-HA^agg or recipient cells with induced NM-GFP^agg. b Quantitative analysis of the effect of protease inhibitors on cells with NM-GFP aggregates. Recipients with NM-GFP aggregates that were solvent-treated (DMSO) were set to 100%. c Workflow of compound test in cocultures. Donor N2a NM-HA^agg (LP) and recipient N2a NM-GFP^sol were co-seeded and exposed to different concentrations of Amprenavir 1 h later. Alternatively, recipient N2a NM-GFP^sol cells were pretreated with different concentrations of Amprenavir for 1 h, and cells were subsequently exposed to donor-derived EVs for 12 h. Note that donor cells (LP) from which EVs were isolated remained untreated. d Amprenavir effects on the percentage of recipient cells with induced NM-GFP aggregates in the two assays. Results were normalized to solvent control. e Donor N2a NM-HA^agg cells (LP) were treated with different concentrations of Amprenavir for 3 days. Cells were subsequently cocultured with recipient N2a NM-GFP^sol. Aggregate formation in recipients was assessed 16 h later. f Dose-dependent inhibition of NM-GFP aggregation in cocultured recipient cells upon Amprenavir treatment. Data were normalized to DMSO-treated control set to 100%. g EVs were harvested from treated donors (LP) and added to recipient cells. h Induction of NM-GFP aggregates in recipient cells by EVs derived from Amprenavir-treated donor cells (LP). i Effect of Amprenavir on particle release by donor cells (LP). All data are shown as the means ± SD from two (d), three (b, h, i), or six (f) replicate cell cultures. Three (b, d, f, h, i) independent experiments were carried out with similar results. P-values calculated by two-tailed unpaired Student’s t-test (h, i). ns: non-significant. Source data are provided as a Source Data file.

Fig. 5. Receptor polymorphisms modulate intercellular proteopathic seed spreading.

a Experimental workflow. Recipient N2a NM-GFP^sol cells were transfected with two siRNAs against XPR1, one mCat-1 siRNA, or a non-silencing siRNA control. 48 h later, recipient cells were cocultured with donor cells (LP). Alternatively, recipients were exposed to EVs isolated from conditioned medium of N2a NM-HA^agg cells (LP). NM-GFP aggregate induction was determined 16 h post-exposure or coculture. b Knock-down of XPR1 mRNA by two independent siRNAs was assessed 48 h post-siRNA transfection by qRT-PCR. Shown is the fold change of mRNA expression normalized to control (ctrl.). c Knock-down of mCat-1 mRNA. d, e Recipients with NM-GFP aggregates following receptor knock-down. Shown are the results of coculture (d) and of recipients exposed to donor-derived EVs (LP) (e). NM-GFP aggregate induction was measured 16 h post EV addition or coculture. f Transmembrane structure of XPR1. The receptor contains four extracellular loops (ECL1–4). g Polymorphic variants of xenotropic and polytropic X/P-MLV receptor XPR1 in mouse N2a and human HEK cells. Shown are mismatches in the surface-exposed loops ECL 3 and 4. ECL 3 and 4 are required for the binding of X/P-MLV. h Ectopic expression of the N2a XPR1 receptor variant in poorly permissive HEK NM-GFP^sol cells. Ectopically expressed XPR1-HA was detected using anti-HA antibodies. GAPDH served as a loading control on the same blot. i HEK NM-GFP^sol cells were transfected with mouse XPR1-HA or mock-transfected and subsequently cocultured with donor N2a NM-HA^agg cells (LP). N2a NM-GFP^sol cells served as recipient controls. j Alternatively, transfected HEK NM-GFP^sol cells were exposed to EVs from N2a NM-HA^agg cells (LP). N2a NM-GFP^sol cells served as recipient controls. All data are shown as the means ± SD from three (b–e) or six (i, j) replicate cell cultures. Three (b–e, i, j) independent experiments were carried out with similar results. P-values calculated by two-tailed unpaired Student’s t-test (c), one-way ANOVA with Dunnett’s post hoc test (b, d, e), or one-way ANOVA with Bonferroni multiple comparisons (i, j). Source data are provided as a Source Data file.

Fig. 6. MLV reconstitution increases intercellular NM and Tau aggregate induction.

a HEK NM-HA^agg donor cells transfected with combinations of amphotropic MLV env 10A1 vector, gag/pol plasmid, retroviral transfer vector, and/or empty vector pHCMV were cocultured with recipient Vero NM-GFP^sol cells. b Western blot of Vero NM-GFP^sol cells expressing Pit-2. GAPDH was detected on the same membrane. c Recipient Vero cells with NM-GFP cocultured with donors transfected with single plasmids. d Recipients cocultured with donors transfected with combinations of plasmids. e Recipient cells with induced NM-GFP^agg cocultured with donors that were additionally transfected with transfer vector (TV) for virus production. f Donors transfected with viral plasmids with TV coding for Luciferase-V5. Virus production was confirmed by transducing wild-type Vero cells with a conditioned donor medium. g Donors transfected with/without Luciferase-V5 coding retroviral TV and plasmids coding for gag/pol and env produce a virus that is infectious to wild-type cells, as demonstrated by immunofluorescence with anti-V5 antibodies. h Vesicle secretion upon transfection of viral plasmids. i Induction of NM-GFP aggregates in recipients exposed to conditioned medium (CM). j HEK Tau-GFP^AD donor cell population. k HEK Tau-GFP^AD donor cells transfected with combinations of plasmids coding for env 10A1, gag/pol, retroviral TV, and/or non-viral empty vector. Donors were cocultured with recipient Vero Tau-FR^sol cells. Alternatively, donor medium was added to recipient cells. l Recipients with Tau-FR^agg upon coculture with donors transfected with individual plasmids. m Recipients with induced Tau-FR^agg upon coculture with donors transfected with plasmid combinations. n Recipients with induced Tau-FR^agg upon coculture with donors transfected with plasmid combinations and TV. o Particle secretion upon transfection. p Recipients with Tau-FR^agg exposed to conditioned medium adjusted for comparable particle numbers. All data are shown as the means ± SD from two (p), three (h, o) or six (c–e, i, l, m, n) replicate cultures. Three (c–e, h, i, l–p) independent experiments were carried out with similar results. P-values calculated by two-tailed unpaired Student’s t-test (e, n) or one-way ANOVA with Tukey’s multiple comparisons (c, d, h, i, l, m, o, p). ns: non-significant. Source data are provided as a Source Data file.

Fig. 7. Different HERV glycoproteins increase intercellular protein aggregate spreading.

a Experimental workflow. Donor HEK cells stably propagating aggregated NM-HA were transfected with a plasmid coding for V5-epitope tagged HERV-W Syncytin-1 or a plasmid coding for V5-epitope tagged HERV-K Env. Cells were subsequently cocultured with recipient HEK NM-GFP^sol. b Western blot analysis of donor clone transfected with plasmids coding for V5-tagged Syncytin-1 or HERV-K Env. Samples were loaded twice for Actin detection on different blots. c Coculture of donor and recipient HEK cells. Note that we have not stained the donors in this experiment. d Quantitative analysis of the percentage of recipient cells with induced aggregates upon coculture. e Experimental workflow. Tau-GFP^AD cells were transfected with plasmids coding for V5 epitope-tagged HERV Envs and cells were subsequently cocultured with recipient HEK cells expressing Tau-FR^sol. f Western blot analysis of donor clone transfected with plasmids coding for HERV Envs. g Coculture of donor and recipient cells. h Quantitative analysis of the percentage of recipient cells with induced aggregates. i Donor HEK Tau-GFP^AD cells transfected or not with plasmids coding for HERV-K Env and -HERV-W Syncytin-1 were subsequently cocultured with human primary astrocytes expressing Tau-FR^sol. Please note that due to technical challenges, the high percentages of HERV Env expressing donor astrocytes with Tau aggregates required for cocultures cannot be achieved. j Primary astrocytes with soluble or aggregated Tau-FR following coculture. k Quantitative analysis of primary astrocytes with Tau-FR^agg following coculture with HEK Tau-GFP^AD. All data are shown as the means ± SD from six (d, h) or three (k) replicate cell cultures. Three (d, h) and two (k) independent experiments were carried out with similar results. P-values were calculated by one-way ANOVA with Dunnett´s post hoc test (d, h, k). Source data are provided as a Source Data file.