Altering cortical input unmasks synaptic phenotypes in the YAC128 cortico-striatal co-culture model of Huntington disease

Abstract

Background: Huntington disease (HD) is a fatal neurodegenerative disorder caused by a CAG expansion in the huntingtin (HTT) gene, leading to selective and progressive neuronal death predominantly in the striatum. Mutant HTT expression causes dysfunctional cortico-striatal (CS) transmission, loss of CS synapses, and striatal medium spiny neuron (MSN) dendritic spine instability prior to neuronal death. Co-culturing cortical and striatal neurons in vitro promotes the formation of functional CS synapses and is a widely used approach to elucidate pathogenic mechanisms of HD and to validate potential synapto-protective therapies. A number of relevant in vivo synaptic phenotypes from the YAC128 HD mouse model, which expresses full-length transgenic human mutant HTT, are recapitulated in CS co-culture by 21 days in vitro (DIV). However, striatal spine loss, which occurs in HD patients and in vivo animal models, has been observed in YAC128 CS co-culture in some studies but not in others, leading to difficulties in reproducing and interpreting results. Here, we investigated whether differences in the relative proportion of cortical and striatal neurons alter YAC128 synaptic phenotypes in this model.

Results: YAC128 MSNs in 1:1 CS co-culture exhibited impaired dendritic length and complexity compared to wild-type, whereas reducing cortical input using a 1:3 CS ratio revealed a dramatic loss of YAC128 MSN dendritic spines. Chimeric experiments determined that this spine instability was primarily cell autonomous, depending largely on mutant HTT expression in striatal neurons. Moreover, we found that spontaneous electrophysiological MSN activity correlated closely with overall dendritic length, with no differences observed between genotypes in 1:3 co-cultures despite significant YAC128 spine loss. Finally, limiting cortical input with a 1:3 CS ratio impaired the basal survival of YAC128 neurons at DIV21, and this was partially selective for dopamine- and cAMP-regulated phosphoprotein 32-positive MSNs.

Conclusions: Our findings reconcile previous discordant reports of spine loss in this model, and improve the utility and reliability of the CS co-culture for the development of novel therapeutic strategies for HD.

Keywords: Huntington disease; YAC128, DARPP32; corticostriatal co-culture; dendrite; huntingtin; spine; synapse.

Fig. 1

YAC128 MSNs co-cultured with cortical neurons at a 1:3 CS ratio recapitulate in vivo spine loss. WT and YAC128 (Y128) co-cultures were generated at either a 1:1 or 1:3 CS ratio and processed at DIV21 for DARPP32 and MAP2 immunocytochemistry, imaging, and spine analysis in NeuronStudio. (a) Sample images of DARPP32- and MAP2-stained dendrites in CS co-culture (scale bar = 5 μm). (b) Quantification of DARPP32 staining intensity normalized to MAP2 intensity reveals no differences between genotypes or conditions [n = 30(3); two-way ANOVA with Bonferroni post-hoc analysis]. (c) Sample images of DARPP32-stained spines on secondary or tertiary dendrites in co-cultured MSNs at higher exposure (scale bar = 5 μm). The differences in numbers of (Di) total and (Dii) mature mushroom, but not (Diii) immature spines, are exacerbated in 1:3 co-cultured YAC128 MSNs [n = 32(4); two-way ANOVA with Bonferroni post-hoc analysis; *p < 0.05, ***p < 0.001]. (e) Representative Golgi staining of striatal MSNs in vivo (scale bar = 5 μm). (f) Golgi analysis confirms that reduced MSN total spine number occurs by 12 months of age in the YAC128 striatum, to a similar degree as in 1:3 co-cultures [n = 4–5 6-month-old animals and 3 12-month-old animals per genotype; two-way ANOVA with Bonferroni post-hoc analysis; **p < 0.01]. Individual data values for graph in F are available in Additional file 1. A linear correlation exists between (Gi) total and (Gii) mushroom spines versus the proportion of striatal cells at plating. A significant interaction occurs between striatal proportion and genotype [n = 30(3); two-way ANOVA with Bonferroni post-hoc analysis; *p < 0.05, **p < 0.01, ***p < 0.001]

Fig. 2

YAC128 spine instability is predominantly MSN intrinsic. WT, YAC128, and chimeric co-cultures generated at a 1:3 CS ratio were processed at DIV21 for DARPP32 immunocytochemistry, imaging, and spine analysis. (a) Sample images of DARPP32-stained spines in pure or chimeric co-cultured MSNs (scale bar = 5 μm). (Bi) Total spine density values in chimeric cultures are similar to pure cultures of the same MSN genotype. (Bii) Mature mushroom and (Biii) immature spine numbers are affected by both striatal (STR) and cortical (CTX) mHTT expression [n = 32(4); one-way ANOVA with Bonferroni post-hoc analysis; **p < 0.01, ***p < 0.001]. Student’s t test was used to compare WT STR/WT CTX and WT STR/Y128 CTX [n = 32(4); Student’s t test; #p < 0.05]

Fig. 3

YAC128 MSNs in 1:1 CS co-culture demonstrate reduced dendritic length and complexity. WT and YAC128 co-cultures were generated at either a 1:1 or 1:3 CS ratio and processed at DIV21 for DARPP32 immunocytochemistry, imaging, and dendritic analysis. (a) Sample images of MSN dendritic traces generated in NeuronStudio (scale bar = 15 μm). (Bi) Total length of the dendritic trace and (Bii) complexity by Sholl analysis are significantly reduced in 1:1 YAC128 MSNs compared to WT. Post-hoc statistical significance for Sholl analysis is shown only for WT 1:1 vs. YAC128 1:1 (*) or WT 1:3 vs. YAC128 1:3 (#) comparisons [n = 32(4); two-way ANOVA with Bonferroni post-hoc analysis; *p < 0.05, **p < 0.01, ***p < 0.001]

Fig. 4

YAC128 MSNs co-cultured at 1:1 exhibit an impaired increase in membrane capacitance with maturation. (a) Representative recording traces from WT and YAC128 MSNs in 1:1 or 1:3 co-culture at DIV14 and 21. (Bi, Bii) mEPSC frequency and (Ci, Cii) membrane capacitance (Cm) tend to increase with maturation, but a significant genotypic difference was only observed for Cm at DIV21 in 1:1 cultures [n = 12–29(3); two-way ANOVA with Bonferroni post-hoc analysis; *p < 0.05]

Fig. 5

Neuronal survival is compromised in YAC128 1:3 CS co-cultures. DIV21 WT and YAC128 co-cultures were fixed at DIV21 and stained for MAP2 and DARPP32 (D32). (a) Sample fields of view at 20X objective (scale bar = 100 μm). The numbers of (Bi) MAP2+ and (Bii) DARPP32+ neurons per field of view are reduced in YAC128 1:3 co-cultures. (Biii) The proportion of DARPP32+ neurons (# DARPP32+ divided by # MAP2+) surviving at DIV21 is also significantly lower in YAC128 1:3 co-cultures [n = 30 fields of view from three independent cultures; two-way ANOVA with Bonferroni post-hoc analysis; *p < 0.05, ***p < 0.001]