Hippo Signaling Pathway Dysregulation in Human Huntington's Disease Brain and Neuronal Stem Cells

Abstract

The Hippo signaling pathway is involved in organ size regulation and tumor suppression. Although inhibition of Hippo leads to tumorigenesis, activation of Hippo may play a role in neurodegeneration. Specifically, activation of the upstream regulator, mammalian sterile 20 (STE20)-like kinase 1 (MST1), reduces activity of the transcriptional co-activator Yes-Associated Protein (YAP), thereby mediating oxidative stress-induced neuronal death. Here, we investigated the possible role of this pathway in Huntington's disease (HD) pathogenesis. Our results demonstrate a significant increase in phosphorylated MST1, the active form, in post-mortem HD cortex and in the brains of CAG knock-in Hdh^Q111/Q111 mice. YAP nuclear localization was also decreased in HD post-mortem cortex and in neuronal stem cells derived from HD patients. Moreover, there was a significant increase in phosphorylated YAP, the inactive form, in HD post-mortem cortex and in Hdh^Q111/Q111 brain. In addition, YAP was found to interact with huntingtin (Htt) and the chaperone 14-3-3, however this interaction was not altered in the presence of mutant Htt. Lastly, YAP/TEAD interactions and expression of Hippo pathway genes were altered in HD. Together, these results demonstrate that activation of MST1 together with a decrease in nuclear YAP could significantly contribute to transcriptional dysregulation in HD.

Figure 1

YAP subcellular localization in human post-mortem cortex and NSCs derived from HD patients. (a) Representative images from unaffected control (top; n = 7) and HD (bottom; n = 10) human cortical tissue stained with YAP (green), NeuN (red), and DAPI (blue) (left to right). Merged images (far right) show overlap of YAP, NeuN and DAPI. Arrows highlight representative cells; scale bar = 50 μM. (b) The percentage of NeuN+ cells that were YAP+ (YAP/NeuN%) was significantly lower (p = 0.0250, Mann-Whitney U Test) in HD (black) compared to control (grey) human tissue. (c) The percentage of DAPI+ cells that were YAP+ (YAP/DAPI%) was significantly lower (p = 0.0020, Mann-Whitney U Test) in HD compared to control tissue. (d) Representative images from unaffected control (WAD9) (top; n = 5) and HD (G018) (bottom; n = 6) human embryonic-derived NSCs stained with YAP (green; left) and DAPI (blue; middle). Merged images (right) show overlap of YAP and DAPI. Scale bar = 25 μM. (e) Overall YAP intensity was not different (p = 0.1255, Mann-Whitney U Test) between control (grey) and HD (black) NSCs. (f) Nuclear YAP staining was significantly lower (p = 0.0043, Mann-Whitney U Test) in HD compared to control tissue. Data are presented as median and min-max values. *p < 0.05, **p < 0.01.

Figure 2

YAP mRNA expression in post-mortem human cortex. (a) Quantitative real time PCR assay illustrates no change in YAP mRNA relative to GAPDH levels (p = 0.397, Mann Whitney U test) between unaffected control (n = 5; grey) and HD (n = 12; black) patients. (b) Quantitative real time PCR assay also demonstrates no change (p > 0.05, Two-way ANOVA) in full length YAP (YAP-FL; left) or YAPΔC isoforms with 13, 48, and 61 nucleotide inserts (left to right: YAP-Ins13, YAP-Ins48, and YAP-Ins61, respectively) between unaffected control and HD patients. A two-way ANOVA demonstrated that there was no significant effect of genotype [F(3, 68) = 0.775, p = 0.512], YAP isoform [F(1, 68) = 0.282, p = 0.597], or interaction [F(3, 68) = 0.383, p = 0.765] in control (n = 8; grey) and HD (n = 11; black) patients. Data are presented as median values.

Figure 3

MST1/2 is activated and YAP is inactivated in post-mortem human cortex and the Hdh mouse model of HD. (a) Representative western blots for pMST1/2, pYAP and GAPDH (top to bottom) with human cortex samples from unaffected control (left; n = 4) and HD (right; n = 4) patients. (b) There was a trend towards an increase (p = 0.065, Mann-Whitney U Test) in pMST1/2 levels (left) from HD compared to controls. pYAP protein levels (right) were significantly increased (p = 0.015, Mann-Whitney U Test) in HD compared to controls. (c) Representative western blots for MST, YAP and GAPDH (top to bottom) with human cortex samples from unaffected control (left; n = 4) and HD (right; n = 4) patients. (d) There was no change in MST1 (left) (p = 0.343, Mann-Whitney U Test) or YAP (right) (p = 0.200, Mann-Whitney U Test) between HD and controls. (e) Representative western blots for LATS2, pLATS2 and GAPDH (top to bottom) with human cortex samples from unaffected control (left; n = 8) and HD (right; n = 8) patients. (f) There was no change in LATS2 (left) (p = 0.2039, Mann-Whitney U Test) or pLATS2 (right) (p = 0.8785, Mann-Whitney U Test) in HD compared to controls. (g) Representative western blots for phosphorylated MST1/2 (pMST1/2) and GAPDH (top to bottom) from HdhQ7/Q7 (left; n = 3) and HdhQ111/Q111 (right; n = 4) striatum and for MST1, phosphorylated YAP (pYAP), YAP and GAPDH (top to bottom) from Hdh^Q7/Q7 (left; n = 3) and Hdh^Q111/Q111 (right; n = 5) striatum. (h) Quantification of westerns blots showed no change in pMST1/2 (left) (p = 0.4000, Mann-Whitney U Test) and increased pYAP (right) (p = 0.0357, Mann-Whitney U Test) protein levels in Hdh^Q111/Q111 relative to Hdh^Q7/Q7 striatum. (i) Representative western blots for pMST1/2, MST1, and GAPDH (top to bottom) from Hdh^Q7/Q7 (left; n = 3) and Hdh^Q111/Q111 (right; n = 4) cortex. (j) Quantification of western blots shows increased pMST1/2 (p = 0.0286, Mann-Whitney U Test) protein levels in Hdh^Q111/Q111 (right; n = 4) relative to Hdh^Q7/Q7 (left; n = 4) cortex. (k) Quantification of immunofluorescence staining illustrated a trend towards a decrease (p = 0.1000, Mann-Whitney U Test) in nuclear YAP (YAP/DAPI%) levels in Hdh^Q111/Q111 (n = 3) relative to Hdh^Q7/Q7 (n = 3) samples. Data are presented as median and min-max values. * p < 0.05.

Figure 4

YAP-Htt-14-3-3 interactions. (a) Representative blot from a co-immunoprecipitation assay using human cortex with anti-14-3-3 antibody followed by immunoblotting with an anti-huntingtin (2166) antibody. (b) There was no change in 14-3-3 levels in the input samples (left) (p = 0.7209, Mann-Whitney U Test) or in 14-3-3/Htt interactions (right) in HD cortex (black; n = 8 input and IP) compared to control (grey; n = 8 input, n = 7 IP) as demonstrated in the IP samples (p = 0.6126, Mann-Whitney U Test). (c) Representative blot from a co-immunoprecipitation assay using human cortex with anti-YAP antibody followed by immunoblotting with an anti-huntingtin (2166) antibody. (d) There was no change in YAP levels in the input samples (left) (p = 0.1031, Mann-Whitney U Test) in HD cortex (black; n = 7) compared to control (grey; n = 7) or in YAP/Htt interactions (right) (p = 0.6667, Mann-Whitney U Test) in HD cortex (black; n = 7) compared to control (grey; n = 7) as demonstrated in the IP samples in HD cortex (black; n = 2) compared to control (grey; n = 2) (p = 0.6667, Mann-Whitney U Test). Data are presented as median and min-max values.

Figure 5

Expression patterns of Hippo pathway components in human cortex. (a) Heat map illustrating expression levels in Hippo pathway proteins from a gene expression array performed on control (n = 8) and HD (n = 14) human cortex samples. Colors indicate expression levels (blue = low to red = high). Overall there were more downregulated genes in human HD compared to control cortex. (b) Two-way ANOVA revealed a significant effect of genotype [F(2, 60) = 6.633, p = 0.0025] and treatment [F(1, 60) = 17.03, p = 0.0001] but no effect of interaction [F(2, 60) = 1.826, p = 0.1698] on gene expression from qPCR analysis. Sidak’s multiple comparison’s test demonstrated that there was a significant decrease in Meis1 (p < 0.001) and Sav1 (p < 0.05) expression in HD cortex compared to control but not a decrease in Lats2 expression (p > 0.05). (c) qPCR findings also demonstrated a significant decrease in Cyr61, another known YAP target gene, expression in HD cortex (black; n = 13) compared to control (grey; n = 8) (p = 0.0132 Mann-Whitney U Test). Data are presented as medians and min to max. *p < 0.05, **p < 0.01.

Figure 6

YAP/TEAD interactions are decreased in HD. (a) Representative blot from a co-immunoprecipitation assay using human cortex with anti-YAP antibody followed by immunoblotting with an anti-pan-TEAD antibody. (b) There was no change in TEAD levels in the input samples (p = 0.6857, Mann-Whitney U Test) (left) but there was a significant decrease in YAP/TEAD interactions (right) in HD cortex (black; n = 6) compared to control (grey; n = 7) as demonstrated in the IP samples (p = 0.0043, Mann-Whitney U Test). Data are presented as median and min-max values. *p < 0.05.

Figure 7

Disruption of YAP/TEAD interaction alters transcription. Two-way ANOVA demonstrated that there was a significant effect of gene [F(2, 12) = 23.87, p < 0.0001] but no effect of treatment [F(2, 36) = 1.144, p = 0.3298] or interaction [F(4, 36) = 2.019, p = 0.1124] on expression data from qPCR analysis. Tukey’s multiple comparison’s test a significant decrease in expression of the NADPH dehydrogenase (Dhrs4) (p = 0.0025), transcription factor 7 (Tcf7) (p = 0.0053), and vitamin D receptor (Vdr) (p = 0.0001) genes in untreated STHdh^Q7/Q7 cells (light grey) compared to untreated STHdh^Q111/Q111 cells (black; n = 5). In addition, verteporfin treatment (10 mM, 24 hours) in STHdh^Q7/Q7 cells (red; n = 5) also significantly decreased Vdr expression (p < 0.0001, Tukey’s post-hoc test) and produced a trend towards a decrease in Tcf7 expression (p = 0.1537, Tukey’s post-hoc test) compared to untreated STHdh^Q7/Q7 cells. However, verteporfin treatment did not affect Dhrs4 (p = 0.3535, Tukey’s post-hoc test) expression in STHdh^Q7/Q7 cells. Data are presented as median and min-max values. **p < 0.01, **p < 0.001, ****p < 0.0001.