A. Transmission electron micrograph of K₂Q₄₄K₂ aggregates stained with uranyl acetate. Bar = 200 nm.
B. “Deep-etch” transmission electron micrograph of K₂Q₄₄K₂ aggregates. Bar = 100 nm.
C. Juxtanuclear localization of K₂Q₄₄K₂ aggregates (red). Plasma membrane was labeled with wheat germ agglutinin (green) and nuclei with bisbenzimide (blue). Bar = 2 μm.
D. Pericentriolar distribution of K₂Q₄₄K₂ aggregates. COS7 cells were incubated with FITC-K₂Q₄₄K₂ aggregates (green). Fresh medium was replenished after 1 hour and cells chased overnight. Following fixation cells were stained with antibody to γ-tubulin (red) and bisbenzimide (blue). Scale bars = 2 μm.
E. Comparison of internalized K₂Q₄₄K₂ aggregates with intracellular markers. Confocal micrographs of COS7 cells exposed to FITC-K₂Q₄₄K₂ for 10 hours. Cells were fixed and stained with antibodies against the indicated endolysosomal markers (d-f) or cytoplasmic quality control components (m-p). Cell nuclei were stained with bisbenzimide. Scale bar for panels a-i = 10 μm; for panels j-s = 2 μm.

A. Internalized K₂Q₄₄K₂ aggregates are visible attached to the actin cortex of unroofed HEK293 cells visualized by deep-etch TEM. Inset higher magnification view of region indicated by box.
B. K₂Q₄₄K₂ aggregates induce focal redistribution of intracellular HttQ₂₅. Diffuse distribution of CFP-HttQ₂₅ fluorescence in cells incubated without aggregates (top) is converted to a punctate pattern following overnight exposure to aggregated TMR-K₂Q₄₄K₂ (bottom). Scale bar = 2 μm.
C. Colocalization of CFP-HttQ₂₅ puncta and TMR-K₂Q₄₄K₂. Scale bars = 2 μm.
D. Coalescence of CFP-HttQ₂₅ fluorescence into puncta following infection with K₂Q₄₄K₂ aggregates. Cells (n ≥ 250) exposed to TMR-K₂Q₄₄K₂ aggregates (white bars) or control (black bars) were scored for punctate CFP-HttQ₂₅ appearance. Asterisk, P- ≤ 0.001. Data representative of three independent experiments.
E. Fibrillar aggregates composed of bacterially expressed, purified HttQ₅₁ are able to induce conversion of cytosolic chFP-Q₂₅ reporter. Scale bar = 2 μm. Insets, appearance of protein assemblies of HttQ₁₈ (left) or HttQ₅₁ (right) by negative stain EM. Scale bars = 20 nm.
F. Quantification of data from experiment in Fig. 2D. Number of cells counted in each experiment ≥300 cells.
G. K₂Q₄₄K₂ aggregates induce aggregation of intracellular Htt. HEK293 cells stably expressing GFP-HttQ₂₅ or Q₇₁ were infected with K₂Q₄₄K₂ aggregates for 24 hours prior to filter retardation assay. Post-lysis control is explained in Methods. Data representative of 3 independent experiments.

A. Sup35-NM and Aβ aggregates are internalized by mammalian cells. COS7 cells were infected with aggregated Sup35-NM (red; a) or Aβ (1-40) (red; b). Cells are labeled as in Fig. 1D. Scale bars = 2 μm.
B. K₂Q₄₄K₂ aggregates induce selective redistribution of CFP-HttQ₂₅. HEK293 cells transfected with CFP-HttQ₂₅, were incubated with aggregated K₂Q₄₄K₂ (c and d), Sup35-NM (e and f), or Aβ(1-40) (g and h). Scale bar = 2 μm.
C. Internalized K₂Q₄₄K₂ aggregates induce selective aggregation of cytoplasmic polyQ reporters. Filter retardation assay of HEK293 cells stably expressing the indicated reporters following incubation with the indicated aggregates (1 μM) for ~ 24 hr or K₂Q₄₄K₂ (0.2 μM) as post-lysis control. Data representative of three independent experiments.
D. Infection with Sup35-NM induces selective aggregation of cytoplasmic GFPSup35-NM. Upper panel: immunoblot of Sup35-NM sedimentation analysis from HEK293 cells expressing GFP-Sup35-NM exposed to protein aggregates as noted (1μM) prior to lysis and sedimentation into pellet (P) and supernatant (S) fractions. Asterisk, probable proteolysis product. Bottom panel: quantification (mean ± SEM) of data from 4 immunoblot experiments. Asterisk, p≤ 0.0022.

A. HEK293 cells stably expressing CFP-HttQ₂₅ were exposed to TMR-K₂Q₄₄K₂ aggregates (1 μM, solid circles; 0.1 μM, solid squares) or controls: mock infection (open circles), dextran (open triangles), or Aβ aggregates (open squares). Following the indicated number of generations in culture, cells were scored for CFP fluorescence distribution and for the presence of TMR (open diamonds) fluorescence. Each point represents counts of ≥400 cells. The solid line represents the theoretical decay of TMR fluorescence.
B. Blow-up of boxed region from Fig. 4a. Symbols as in Fig. 4a. Data representative of three independent experiments. Inset, P-values determined from binomial test of proportions for generations 24-87.
C. Intercellular transfer of polyQ aggregates in mammalian cell co-culture. Puro^r HEK293 cells stably expressing chFP-HttQ₂₅ were cultured together with puromycin-sensitive cells stably expressing GFP-HttQ₇₁. Cells were imaged after 72 hr of co-culture in DMEM (top row) or DMEM containing 4μg/ml puromycin (bottom row). Scale bars = 2 μm.
D. Quantification of chFP-HttQ₂₅ puncta formation in puro^r reporter cells cultured alone or co-cultured with cells stably expressing GFP-HttQ₂₅ or GFP-HttQ₇₁ in the presence or absence of puromycin (4μg/ml). Number of cells counted, n = 699-1276.