The influence of huntingtin protein size on nuclear localization and cellular toxicity

Abstract

Huntington disease is an autosomal dominant neurodegenerative disorder caused by the pathological expansion of a polyglutamine tract. In this study we directly assess the influence of protein size on the formation and subcellular localization of huntingtin aggregates. We have created numerous deletion constructs expressing successively smaller fragments of huntingtin and show that these smaller proteins containing 128 glutamines form both intranuclear and perinuclear aggregates. In contrast, larger NH2-terminal fragments of huntingtin proteins with 128 glutamines form exclusively perinuclear aggregates. These aggregates can form in the absence of endogenous huntingtin. Furthermore, expression of mutant huntingtin results in increased susceptibility to apoptotic stress that is greater with decreasing protein length and increasing polyglutamine size. As both intranuclear and perinuclear aggregates are clearly associated with increased cellular toxicity, this supports an important role for toxic polyglutamine-containing fragments forming aggregates and playing a key role in the pathogenesis of Huntington disease.

Figure 1

(a) Diagrammatic representation of the cDNA expression constructs used in this study. Bars represent huntingtin protein, with amino acids numbered according to GenBank/EMBL/DDBJ accession number L27350. The locations of the polyglutamine tract (Q) are indicated. The names of the constructs, 436, 771, 989, 1597, 1955, and 10366 correspond to the nucleotide position of the translation termination codon. The amino acid length of each protein is also indicated at the right. (b) Western blot of huntingtin truncation constructs. HEK 293T cells were transfected with the 771, 989, 1597, 1955, and 10366 constructs, containing 15 glutamines. 36 h after transfection, cell lysates were prepared. Equal amounts of protein were loaded on a 5–20% SDS-PAGE gradient gel, and Western blotted using anti-huntingtin antibody BKP1, which recognizes the NH₂ terminus. Multiple bands in the lanes represent different conformations of the protein or huntingtin multimers. The amino acid length (aa) of each expressed protein is also indicated.

Figure 1

(a) Diagrammatic representation of the cDNA expression constructs used in this study. Bars represent huntingtin protein, with amino acids numbered according to GenBank/EMBL/DDBJ accession number L27350. The locations of the polyglutamine tract (Q) are indicated. The names of the constructs, 436, 771, 989, 1597, 1955, and 10366 correspond to the nucleotide position of the translation termination codon. The amino acid length of each protein is also indicated at the right. (b) Western blot of huntingtin truncation constructs. HEK 293T cells were transfected with the 771, 989, 1597, 1955, and 10366 constructs, containing 15 glutamines. 36 h after transfection, cell lysates were prepared. Equal amounts of protein were loaded on a 5–20% SDS-PAGE gradient gel, and Western blotted using anti-huntingtin antibody BKP1, which recognizes the NH₂ terminus. Multiple bands in the lanes represent different conformations of the protein or huntingtin multimers. The amino acid length (aa) of each expressed protein is also indicated.

Figure 2

(a) Immunofluorescence of 293T cells transfected with the 436, 771, 989, 1597, 1955, and 10366 constructs, each containing 128 glutamines. 36–40 h after transfection, the cells were processed for immunofluorescence using mAb 2166 (red) or BKP1 (green) to detect huntingtin. The cells were counterstained with the DNA stain DAPI (blue) to label the nucleus (the DAPI stain is not visible in the cells shown for 1597-128 and 10366-128). Huntingtin located in the nucleus appears pink as the red stain is overlaid with blue. Huntingtin from the 436, 771, and 989 constructs (top row) appears as large intranuclear aggregates. Huntingtin from the 1597, 1955, and 10366 constructs (bottom row) show perinuclear aggregates and diffuse cytoplasmic stain. (b) Size-exclusion chromatography of a lysate from 293T cells transfected with HD1955-128 and treated with tamoxifen. 1-ml fractions were collected at 4°C, and samples were immediately analyzed by reducing SDS-PAGE and Western blotting with BKP1 antibody. Large molecular weight complexes were not retained by the column and flowed through into the void volume. Huntingtin aggregates isolated in the void volume were observed as species that did not denature upon SDS-PAGE and did not exit the stacking gel (V _o = 42 ml; lanes 2 and 3). Under the same conditions, no aggregation was observed in the whole cell lysate (lane 1) nor in the monomeric huntingtin fractions (V = 72 ml; lanes 4 and 5). Bar in a: (436-128) 11.3 μm; (771-128) 5.31 μm; (989-128) 11.2 μm; (1597-128) 11.97 μm; (1955-128) 12 μm; (10366-128) 11.4 μm.

Figure 2

(a) Immunofluorescence of 293T cells transfected with the 436, 771, 989, 1597, 1955, and 10366 constructs, each containing 128 glutamines. 36–40 h after transfection, the cells were processed for immunofluorescence using mAb 2166 (red) or BKP1 (green) to detect huntingtin. The cells were counterstained with the DNA stain DAPI (blue) to label the nucleus (the DAPI stain is not visible in the cells shown for 1597-128 and 10366-128). Huntingtin located in the nucleus appears pink as the red stain is overlaid with blue. Huntingtin from the 436, 771, and 989 constructs (top row) appears as large intranuclear aggregates. Huntingtin from the 1597, 1955, and 10366 constructs (bottom row) show perinuclear aggregates and diffuse cytoplasmic stain. (b) Size-exclusion chromatography of a lysate from 293T cells transfected with HD1955-128 and treated with tamoxifen. 1-ml fractions were collected at 4°C, and samples were immediately analyzed by reducing SDS-PAGE and Western blotting with BKP1 antibody. Large molecular weight complexes were not retained by the column and flowed through into the void volume. Huntingtin aggregates isolated in the void volume were observed as species that did not denature upon SDS-PAGE and did not exit the stacking gel (V _o = 42 ml; lanes 2 and 3). Under the same conditions, no aggregation was observed in the whole cell lysate (lane 1) nor in the monomeric huntingtin fractions (V = 72 ml; lanes 4 and 5). Bar in a: (436-128) 11.3 μm; (771-128) 5.31 μm; (989-128) 11.2 μm; (1597-128) 11.97 μm; (1955-128) 12 μm; (10366-128) 11.4 μm.

Figure 3

Immunofluorescence of 293T cell transfected with the COOH-terminal huntingtin construct (amino acids 585–3,144). Transfected cells were treated with tamoxifen 36–40 h after transfection and processed for immunofluorescence using mAb 2172 to detect huntingtin (red). The protein was localized exclusively throughout the cytoplasm. Bar, 4.42 μm.

Figure 4

Immunofluorescence of 293T cells transfected with the 436, 771, 989, 1597, 1955, and 10366 constructs, each containing 15 glutamines. The transfected cells were treated with tamoxifen 36-40 h after transfection and mAb 2166 (red) or BKP1 (green) was used to detect huntingtin. The cells were counterstained with the DNA stain DAPI (blue) (the DAPI stain is not visible in the 1597-15 panel). Huntingtin containing 15 repeats does not form aggregates but can be localized in the nucleus depending on the protein size. Huntingtin is localized entirely in the nucleus in the 436 and 771 constructs in the cells shown. Bar: (436-15) 8.87 μm; (771-15) 11.98 μm; (989-15) 16.18 μm; (1597-15) 16.04 μm; (1955-15) 15.2 μm; (10366-15) 21.37 μm.

Figure 5

Immunofluorescence of wild-type and null (HD ^−/−) ES cells transfected with 1955-128. The cells were treated with tamoxifen and immunostained with mAb 2166. Perinuclear aggregates were evident in both cell lines, indicating that the presence of endogenous huntingtin is not required for the formation of aggregates by transfected 1955-128. Bar: (Wild-type) 2.82 μm; (HD ^−/−) 2.87 μm.

Figure 6

Increasing toxicity with decreasing huntingtin length. In the presence of tamoxifen, progressive truncation of huntingtin resulted in increased cellular toxicity (P < 0.01). Interestingly, this trend was seen for proteins containing both 15 and 128 polyglutamines (P < 0.01). 293T cells were transfected with each huntingtin construct, or LacZ control, and cell viability in response to apoptotic stress was measured using a modified MTT assay and presented relative to control (untransfected). All wells were transfected with 0.1 μg DNA except for those transfected with the 436-15 and 128 constructs, where 0.08 μg of DNA was used because all the cells died with 0.1 μg DNA. Results for the 436 construct are normalized, taking into account decreased DNA concentration. Asterisks indicate the level of significance between huntingtin containing 15 and 128 glutamines: * P < 0.01; ** P < 0.001.