Volume 15, Issue 5, Pages 903-911 (May 2007) Neuroprotection by Hsp104 and Hsp27 in Lentiviral-based Rat Models of Huntington's Disease  Valérie Perrin, Etienne Régulier, Toufik Abbas-Terki, Raymonde Hassig, Emmanuel Brouillet, Patrick Aebischer, Ruth Luthi-Carter, Nicole Déglon  Molecular Therapy  Volume 15, Issue 5, Pages 903-911 (May 2007) DOI: 10.1038/mt.sj.6300141 Copyright © 2007 The American Society of Gene Therapy Terms and Conditions

Figure 1 In vitro expression of heat-shock (hsp) transgenes. (a) E16 rat striatal neurons infected with a lentiviral vector encoding yeast hsp104. Staining with an anti-hsp104 antibody demonstrating strong transgene expression 2 weeks after infection. (b) Higher magnification showing strong hsp104 expression in the entire cell body. (c) Non-infected control culture stained with an anti-hsp104 antibody. (d) E16 rat striatal neurons infected with a lentiviral vector encoding rat hsp27. Staining with an anti-hsp27 antibody demonstrating strong transgene expression 2 weeks after infection. (e) Higher magnification showing strong hsp27 expression in the cytoplasm. (f) Non-infected control cultures stained an anti-hsp27 antibody. Scale bars (b) and (e) 20 μm. Molecular Therapy 2007 15, 903-911DOI: (10.1038/mt.sj.6300141) Copyright © 2007 The American Society of Gene Therapy Terms and Conditions

Figure 2 Neuroprotective effects of chaperones in an in vitro model of Huntington's disease. E16 rat primary striatal cultures were co-infected on day 1 with lentiviral vectors encoding wild-type (htt171 with 19Q repeats) or mutated huntingtin (htt) fragments (htt171 with 82Q repeats) and with vectors encoding hsp104 or hsp27. (a–f) Long-term survival and polyglutamine-induced pathological effects on the striatum were assessed with the neuronal marker neuronal nuclei (NeuN). Eight weeks after infection, the percentage of NeuN-positive cells was significantly lower in cultures infected with the htt171-82Q vector (d) than in cultures infected with htt171-19Q (a). The overexpression of hsp104 (e) or hsp27 (f) prevents the loss of NeuN-positive cell bodies in htt171-82Q-infected cultures. The expression of htt171-19Q with hsp104 (b) or hsp27 (c) has no effect on the number of NeuN-positive cells. (g) Quantitative analysis of NeuN-positive cells (N= 6, mean ± SEM, ***P < 0.001). One-way analysis of variance, F(5, 66) = 66.54, P< 0.001. Post hoc comparison of 82Q with 82Q + hsp104 or 82Q + hsp27, ***P < 0.001. Post hoc comparison of 19Q with 19Q + hsp104 or 19Q + hsp27, not significant. (h) The overexpression of mutated htt171-82Q leads to the formation of nuclear (V) and non-nuclear inclusions (*) in 8-week-old rat striatal cultures. (i, j) The expression of htt171-82Q with hsp104 (i, k) leads to an increase in non-nuclear EM48-positive inclusions, whereas the overall ratio of nuclear to non-nuclear EM48-positive inclusions remain unchanged for hsp27 (j, k). (k) Quantitative analysis of average difference between nuclear and non-nuclear inclusions for each group (data are mean ± SEM, *P< 0.05, **P < 0.01, and ***P < 0.001). t-test between nuclear and non-nuclear inclusions for htt171-82Q (t stat = 3.94, **P < 0.01), hsp104 (t stat = –2.24, *P < 0.05), and hsp27 (t stat = 6.09, *** P < 0.001). Molecular Therapy 2007 15, 903-911DOI: (10.1038/mt.sj.6300141) Copyright © 2007 The American Society of Gene Therapy Terms and Conditions

Figure 3 In vivo transgene expression. (a) Experimental design. Lentiviral vectors encoding hsp104 and hsp27 were injected into both sides of the striatum of adult rats (n = 16 per group). One month later, lentiviral vectors expressing htt171-19/82Q were injected into the striatum at the same coordinates (19Q on the left and 82Q on the right). As a control, ten animals were injected with lentiviral vectors expressing htt171-19/82Q alone. As a control for heat-shock protein (hsp) expression, PBS–BSA was injected into the left hemisphere in eight animals from the chaperone groups. The animals were killed 2 months after infection with htt171-19/82Q vectors. (b–e) Huntingtin (htt) expression in htt171-19Q- and htt171-82Q-injected animals was analyzed with the 2B4 antibody 8 weeks after injection. Diffuse cytoplasmic staining was observed for wild-type htt (c), whereas mutated htt was found mostly in nuclear inclusions (e). The expression of the chaperone proteins was analyzed in hsp104- and hsp27-injected animals (f–i) 12 weeks after injection, using anti-hsp104 and anti-hsp27 antibodies, respectively. Consistent with the in vitro data, hsp104 was detected in the entire cell body (g), whereas hsp27 (i) was present in the cytoplasm. Molecular Therapy 2007 15, 903-911DOI: (10.1038/mt.sj.6300141) Copyright © 2007 The American Society of Gene Therapy Terms and Conditions

Figure 4 Neuroprotective effect of chaperones in a rat model of Huntington's disease. Analyses of DARPP-32 staining and quantification of the htt171-82Q-induced lesion. (a, b) The expression of htt171-82Q in the rat striatum led to a loss of DARPP-32 expression, which was partially rescued in (c, d) hsp104- and (e, f) hsp27-treated animals. (d–f) High-magnification images showing the presence of DARPP-32-positive neurons at the center of the zone infected with htt171-82Q in hsp104- and hsp27-treated rats. (g) Quantitative analysis of the striatal lesion (mean ± SEM, **P < 0.01). One-way analysis of variance, F(2, 34) = 20.89, P< 0.001. Post hoc comparison of 82Q with 82Q + hsp104 or 82Q + hsp27, **P < 0.01. Molecular Therapy 2007 15, 903-911DOI: (10.1038/mt.sj.6300141) Copyright © 2007 The American Society of Gene Therapy Terms and Conditions

Figure 5 Changes in the distribution of huntingtin (htt) inclusions upon chaperone gene expression. The expression of hsp104 and, to a lesser extent, hsp27 affects the distribution of htt inclusions. Two antibodies for htt, (a–c) EM48 and (d–f) N-18, and (g–i) one antibody for ubiquitin were used to analyze the formation of htt inclusions. (j–l) Confocal microscopy analysis of 4′,6-diamidino-2-phenylindole (DAPI) and EM48 staining shows the presence of nuclear (arrows; DAPI- and EM48-positive staining) and non-nuclear inclusions (arrowheads; EM48-positve and DAPI-negative staining). Molecular Therapy 2007 15, 903-911DOI: (10.1038/mt.sj.6300141) Copyright © 2007 The American Society of Gene Therapy Terms and Conditions

Figure 6 Number and size of EM48-positive huntingtin (htt) inclusions. (a) Semi-quantitative analysis of the distribution by size/localization of EM48-positive inclusions. Data are mean ± SEM of the chaperone versus the 82Q control for each subgroup of inclusions. One-way analysis of variance (ANOVA) for nuclear inclusions >10 μm2, F(2, 69) = 15, 58. Post hoc comparison of 82Q with 82Q + hsp104 or 82Q + hsp27, ***P < 0.001. One-way ANOVA for non-nuclear inclusions >10 μm2 is not significant, F(2, 69) = 2.19. One-way ANOVA for nuclear inclusions <10 μm2 is not significant, F(2, 69) = 1.39. One-way ANOVA for non-nuclear inclusions <10 μm2, F(2, 69) = 76.56. Post hoc comparison of 82Q with 82Q + hsp104 or 82Q + hsp27, ***P < 0.001. (b, c) A shift in distribution toward small objects corresponding to non-nuclear inclusions was observed when htt171-82Q was expressed with hsp104 or hsp27. Data are mean ± SD. ***P < 0.0001, **P < 0.005, and *P < 0.05 versus htt171-82Q. Molecular Therapy 2007 15, 903-911DOI: (10.1038/mt.sj.6300141) Copyright © 2007 The American Society of Gene Therapy Terms and Conditions

Figure 7 Induction of endogenous hsp70 expression by htt171-82Q. The levels of endogenous hsp70 were lower in the control htt171–19Q group (a, b) compared to the increase in endogenous hsp70 expression induced by the expression of mutated htt (c, d). (d) In htt171-82Q-expressing neurons, the hsp70 staining was found mostly in nuclear inclusions. The co-expression of htt171-82Q with hsp104 (e, f) or hsp27 (g, h) altered the subcellular distribution of hsp70, which became more diffuse and cytoplasmic. Molecular Therapy 2007 15, 903-911DOI: (10.1038/mt.sj.6300141) Copyright © 2007 The American Society of Gene Therapy Terms and Conditions

Figure 8 Induction of endogenous small heat-shock protein (shsp) expression by htt171-82Q and hsp104. In comparison with neurons expressing (a, b) the wild-type fragment of htt, the endogenous hsp27 was up-regulated in striatal neurons expressing (c, d) htt171-82Q. The overexpression of (e, f) hsp104 alone or (g, h) together with htt171-19Q led to the more marked up-regulation of endogenous hsp27 and strong neuronal and fibrillar staining. (i, j) The expression of hsp104 with htt171-82Q further increases the up-regulation of hsp27 expression, which extended over a wider area of the striatum than htt171-82Q-expressing neurons. Molecular Therapy 2007 15, 903-911DOI: (10.1038/mt.sj.6300141) Copyright © 2007 The American Society of Gene Therapy Terms and Conditions