Heat shock protein 70 chaperone overexpression ameliorates phenotypes of the spinal and bulbar muscular atrophy transgenic mouse model by reducing nuclear-localized mutant androgen receptor protein

Abstract

Spinal and bulbar muscular atrophy (SBMA) is an inherited motor neuron disease caused by the expansion of the polyglutamine (polyQ) tract within the androgen receptor (AR). The nuclear inclusions consisting of the mutant AR protein are characteristic and combine with many components of ubiquitin-proteasome and molecular chaperone pathways, raising the possibility that misfolding and altered degradation of mutant AR may be involved in the pathogenesis. We have reported that the overexpression of heat shock protein (HSP) chaperones reduces mutant AR aggregation and cell death in a neuronal cell model (Kobayashi et al., 2000). To determine whether increasing the expression level of chaperone improves the phenotype in a mouse model, we cross-bred SBMA transgenic mice with mice overexpressing the inducible form of human HSP70. We demonstrated that high expression of HSP70 markedly ameliorated the motor function of the SBMA model mice. In double-transgenic mice, the nuclear-localized mutant AR protein, particularly that of the large complex form, was significantly reduced. Monomeric mutant AR was also reduced in amount by HSP70 overexpression, suggesting the enhanced degradation of mutant AR. These findings suggest that HSP70 overexpression ameliorates SBMA phenotypes in mice by reducing nuclear-localized mutant AR, probably caused by enhanced mutant AR degradation. Our study may provide the basis for the development of an HSP70-related therapy for SBMA and other polyQ diseases.

Fig. 1.

Effects of human HSP70 overexpression on the symptomatic phenotypes of male AR-97Q mice. Rotarod task (A; n = 10), cage activity (B; n = 10), body weight (C; n = 12), and survival rate (D; n = 14) of the AR-97Q/HSP70^(−/−), AR-97Q/HSP70^(tg/−), and AR-97Q/HSP70^(tg/tg) mice. All parameters were significantly different among AR-97Q/HSP70^(−/−)mice, AR-97Q/HSP70^(tg/tg), and AR-97Q/HSP70^(tg/−) mice (p < 0.001, p < 0.05,p < 0.05, and p < 0.005, respectively). AR-97Q mice overexpressing human HSP70 lasted longer on the Rotarod and showed higher cage activity than AR-97Q/HSP70^(−/−) mice. The AR-97Q/HSP70^(−/−) mice lost weight earlier than the other two double-transgenics. Survival was prolonged in AR-97Q/HSP70^(tg/−) and AR-97Q/HSP70^(tg/tg) mice compared with AR-97Q/HSP70^(−/−) mice. E, Footprints of representative 16-week-old AR-97Q/HSP70^(−/−), AR-97Q/HSP70^(tg/−), and AR-97Q/HSP70^(tg/tg) mice. Front paws are indicated in red and hindpaws in blue. AR-97Q/HSP70^(−/−) mice exhibit motor weakness, with dragging of the legs; AR-97Q/HSP70^(tg/tg) mice walk almost normally; and AR-97Q/HSP70^(tg/−) mice walk with somewhat short steps. F, The size of steps was measured in 16-week-old AR-97Q/HSP70^(−/−), AR-97Q/HSP70^(tg/−), and AR-97Q/HSP70^(tg/tg) mice (n = 4), respectively. Each column shows an average of steps of the hindpaw. AR-97Q/HSP70^(tg/−) and AR-97Q/HSP70^(tg/tg) mice walked with significantly longer steps than AR-97Q/HSP70^(−/−) mice. *p < 0.05; **p < 0.01. Error bars indicate SD. Wt, Wild type.

Fig. 2.

Increased HSP70 expression in double-transgenic mice. A–C, Immunohistochemical study from the 16-week-old AR-97Q/HSP70^(tg/−) mice in the spinal anterior horn and skeletal muscle stained with the antibody specific for the HSP70. The immunoreactivity of HSP70 was localized to the nuclei with intense and diffuse staining, and small NIs were present in the anterior horn cell (A). A large nuclear inclusion was also present in the anterior horn cell (B). Skeletal muscle showed diffuse nuclear staining and NIs (C). D, E, Western blot analysis of total spinal cord and muscle protein lysate immunolabeled with an antibody against HSP70. AR-97Q/HSP70^(tg/−) and AR-97Q/HSP70^(tg/tg) mice express higher levels of HSP70 than wild-type (Wt) and AR-97Q/HSP70^(−/−) mice (D). The HSP70 expression level is fivefold higher in AR-97Q/HSP70^(tg/−) mice and 10-fold higher in the AR-97Q/HSP70^(tg/tg) mice than endogenous HSP70 in AR-97Q/HSP70^(−/−) mice in the spinal cord and muscle, respectively (D). The AR-97Q transgene expression did not alter HSP70 levels in the spinal cord, whereas the AR-97Q transgene expression gained the respective HSP70 levels in the muscle (D, E). Therefore, the AR-97Q transgene expression in the double-transgenics alters HSP70 levels in the muscle but not in the spinal cord. E, Western blots of nuclear and cytoplasmic extracts immunolabeled with an antibody against HSP70. HSP70 localized in the nucleus (N) as well as in the cytoplasm (CY) in the spinal cord and muscle of all lines examined. AR-97Q/HSP70^(tg/tg) mice expressed the largest amount of HSP70 in both extracts.F, Immunoprecipitation (IP) Western blots for HSP70. Soluble fractions were collected from the spinal cord and muscle, and equal protein concentrations were immunoprecipitated with an antibody to the N-terminal portion of AR and immunoblotted for HSP70. Coimmunoprecipitation of the HSP70 chaperone and the polyQ-expanded mutant AR was detected.

Fig. 3.

Colocalization of the nuclear-localized HSP70 chaperone with mutant AR. Immunohistochemical analysis for the antibody specific to the HSP70 as well as to the expanded polyQ stretch (immunostained with a monoclonal antibody, 1C2) in the spinal cords of 16-week-old AR-97Q/HSP70^(tg/tg) mice (A–E) and SBMA patients (F–M). Double-labeling immunohistochemistry revealed diffuse nuclear staining for goat anti-HSP70 (A) and expanded polyQ (B), suggesting that HSP70 and mutant AR are colocalized in the spinal motor neurons of AR-97Q/HSP70^(tg/tg) mice. Immunofluorescence double-staining with antibodies against HSP70 and the expanded polyQ also revealed that HSP70 and mutant AR are colocalized as shown in HSP70 (C, green), expanded-polyQ (D, red), and an overlay of the two signals (E, yellow). Diffuse staining of neuronal nuclei for HSP70 is also observed in the spinal neurons (F, J) of SBMA patients. Immunofluorescence double-staining with anti-HSP70 (green) and anti-expanded polyQ (red) antibodies revealed that the HSP70 (G) and mutant AR (H) are colocalized on the NI (shown inyellow inI) in the spinal anterior horn cell. The diffuse nuclear colocalization of HSP70 (K) and mutant AR (H) was also observed in the SBMA posterior horn cell (M). This cell also has an NI (L, M).

Fig. 4.

HSP70 decreases nuclear-localized mutant AR in double-transgenic mice. Immunohistochemical study of the spinal anterior horn (A–C) and muscle (D–F) of AR-97Q/HSP70^(−/−)and AR-97Q/HSP70 double-transgenic mice stained with a monoclonal antibody (1C2) against abnormally expanded polyQ (16 weeks old). AR-97Q/HSP70^(−/−) mice have intense and frequent staining for 1C2 in the nucleus (A, D). AR-97Q/HSP70^(tg/−) (B, E), and AR-97Q/HSP70^(tg/tg) (C, F) mice exhibit low levels of 1C2 staining in the nucleus. G, H, Quantitative assessment of diffuse nuclear staining for 1C2 in the spinal ventral horn (G) and muscle (H). Positively stained nuclei were estimated by counting in the thoracic spinal ventral horn and muscle using six transgenic mice (16 weeks of age). There are significantly more 1C2-positive cells in AR-97Q/HSP70^(−/−) mice than in AR-97Q/HSP70^(tg/−) mice or AR-97Q/HSP70^(tg/tg) mice in both tissues. Results are expressed as means ± SD for six mice. The differences in 1C2-positive cell populations are not statistically significant between AR-97Q/HSP70^(tg/−) and AR-97Q/HSP70^(tg/tg) mice. *p < 0.05; **p < 0.01; ***p < 0.001.

Fig. 5.

HSP70 decreases nuclear-localized mutant AR protein complexes as well as monomeric mutant AR. A, B, Western blot analysis of total tissue homogenates from the spinal cord and muscle of AR-97Q/HSP70^(−/−), AR-97Q/HSP70^(tg/−), and AR-97Q/HSP70^(tg/tg) mice (16 weeks of age) immunolabeled by an antibody against AR (N-20). The mutant AR appearing within the stacking gel and monomeric mutant AR were diminished in AR-97Q/HSP70^(tg/−) and AR-97Q/HSP70^(tg/tg) mice compared with AR-97Q/HSP70^(−/−) mice (A, B). Values of mutant AR were normalized to endogenous α-tubulin and expressed as the ratio to those of AR-97Q/HSP70^(−/−) mice (B). Values are expressed as means ± SD for three mice.C, Western blot analysis of nuclear (N) and cytoplasmic (CY) fractions from the spinal cord and muscle of AR-97Q/HSP70^(−/−), AR-97Q/HSP70^(tg/−), and AR-97Q/HSP70^(tg/tg) mice (16 weeks of age) immunolabeled by N-20. Mutant AR protein within the stacking gel was found primarily in the nuclear fraction. The mutant AR within the stacking gel of the nuclear fraction also significantly decreased in the spinal cord and muscle of AR-97Q/HSP70^(tg/tg)mice. R.S.I., Relative signal intensity.

Fig. 6.

HSP70 decreases large aggregated mutant AR protein and soluble monomeric mutant AR protein. A–C, Filter-trap assay of total tissue homogenates from the spinal cord and muscle of AR-97Q/HSP70^(−/−), AR-97Q/HSP70^(tg/−), and AR-97Q/HSP70^(tg/tg) mice (16 weeks of age) immunolabeled by an antibody against AR (N-20). Large aggregated mutant AR complexes were trapped by the cellulose acetate membrane (A), and soluble monomeric mutant AR passing through the cellulose acetate membrane was trapped by the nitrocellulose membrane beneath the cellulose acetate membrane (B). Endogenous α-tubulin using the nitrocellulose membrane was also shown (A, B). The normalized value of large aggregated mutant AR and soluble monomeric mutant AR against endogenous α-tubulin is shown in C. Relative values against those of AR-97Q/HSP70^(−/−)mice were expressed as means ± SD for three mice or a mean of two mice (C). The trapped AR protein was reduced in the spinal cord and muscle of AR-97Q/HSP70^(tg/−)and AR-97Q/HSP70^(tg/tg) mice in both membranes (A, B). This reduction was most evident in AR-97Q/HSP70^(tg/tg) mice (A–C), suggesting that the overexpression of HSP70 resulted in a significant, dose-dependent decrease in large aggregated and soluble monomeric mutant AR protein.R.S.I., Relative signal intensity.