Phosphodiesterase 10A (PDE10A) localization in the R6/2 mouse model of Huntington's disease

Abstract

In Huntington's disease (HD) mutant huntingtin protein impairs the function of several transcription factors, in particular the cAMP response element-binding protein (CREB). CREB activation can be increased by targeting phosphodiesterases such as phospohodiesterase 4 (PDE4) and phosphodiesterase 10A (PDE10A). Indeed, both PDE4 inhibition (DeMarch et al., 2008) and PDE10A inhibition (Giampà et al., 2010) proved beneficial in the R6/2 mouse model of HD. However, Hebb et al. (2004) reported PDE10A decline in R6/2 mice. These findings raise the issue of how PDE10A inhibition is beneficial in HD if such enzyme is lost.

R6/2 mice and their wild type littermates were treated with the PDE10A inhibitor TP10 (a gift from Pfizer) or saline, sacrificed at 5, 9, and 13 weeks of age, and single and double label immunohistochemistry and western blotting were performed. PDE10A increased dramatically in the spiny neurons of R6/2 compared to the wild type mice. Conversely, in the striatal cholinergic interneurons, PDE10A was lower and it did not change significantly with disease progression. In the other subsets of striatal interneurons (namely, parvalbuminergic, somatostatinergic, and calretininergic interneurons) PDE10A immunoreactivity was higher in the R6/2 compared to the wild-type mice. In the TP10 treated R6/2, PDE10A levels were lower than in the saline treated mice in the medium spiny neurons, whereas they were higher in all subsets of striatal interneurons except for the cholinergic ones. However, in the whole striatum densitometry studies, PDE10A immunoreactivity was lower in the R6/2 compared to the wild-type mice.

Our study demonstrates that PDE10A is increased in the spiny neurons of R6/2 mice striatum. Thus, the accumulation of PDE10A in the striatal projection neurons, by hydrolyzing greater amounts of cyclic nucleotides, is likely to contribute to cell damage in HD. Consequently, the beneficial effect of TP10 in HD models (Giampà et al., 2009, 2010) is explained by the efficiency of such compound in counteracting this phenomenon and therefore increasing the availability of cyclic nucleotides.

Introduction

Huntington's disease (HD) is an autosomal-dominant inherited neurodegenerative disorder characterized by motor dysfunction, cognitive decline and emotional and psychiatric disturbance (Wilson et al., 1987). The genetic mutation involves the IT15 gene (The Huntington's Disease Collaborative Research Group, 1993) and is characterized by a CAG expansion beyond the normal 10–35 repeat range, resulting in formation of a mutant huntingtin protein with an expanded poly-glutamine repeat region (Albin and Tagle, 1995). Although mutant huntingtin is expressed throughout the brain, the medium spiny neurons of the striatum are particularly vulnerable to the toxicity of this protein (Albin et al., 1992). Indeed, striatal pathology is believed to underlie the motor symptoms that are typical of the disease.

The cause of striatal and cortical neuron loss in HD is under intense investigation. Wild-type huntingtin has a permissive role in the cortical synthesis of brain derived neurotrophic factor (BDNF), and its mutation impairs BDNF production and its transport to the striatum (Zuccato et al., 2001). Moreover, evidence suggests that mutant huntingtin interacts with and impairs the function of a number of other transcription factors (Sugars and Rubinsztein, 2003), in particular the cAMP response element-binding protein (CREB) (Steffan et al., 2000, Sugars and Rubinsztein, 2003, Sugars et al., 2004). CREB-mediated transcription is required for the survival of adult CNS neurons (Hardingham et al., 1998, Mantamadiotis et al., 2002) and induction of CREB signaling has a protective role in several animal models of neurodegeneration (Block et al., 2001, Walton and Dragunow, 2000, Walton et al., 1996). Inhibition of CREB-mediated gene transcription has been hypothesized to contribute to neuronal loss in HD (Jiang et al., 2003, Kazantsev et al., 1999, Nucifora et al., 2001, Steffan et al., 2000, Steffan et al., 2001) and a decreased transcription of CREB-regulated genes was observed in HD transgenic animals (Luthi-Carter et al., 2000, Nucifora et al., 2001, Wyttenbach et al., 2001). Therefore, drugs targeted at counteracting CREB loss of function may be considered as powerful tools to treat HD.

The above hypothesis was first directly addressed in our studies utilizing the cyclic nucleotide phosphodiesterase type IV (PDE4) inhibitor, rolipram. PDE4 (Houslay and Adams, 2003) is one of the eleven families of phosphodiesterases that regulates through metabolic inactivation cyclic nucleotide signaling throughout the body (Conti and Beavo, 2007), including in brain (Menniti et al., 2006). PDE4 inhibition results in activation of cAMP signaling pathways and, consequently, increased CREB phosphorylation and activation (Hosoi et al., 2003, Jacob et al., 2004, Lee et al., 2004). We have shown that rolipram reduces striatal degeneration in our rat quinolinic acid (QA) model of HD pathology (DeMarch et al., 2007) as well as in a transgenic model of HD, the R6/2 mice (DeMarch et al., 2008). These beneficial effects are hypothesized to result from the increased CREB phosphorylation, and expression of the downstream neurotrophic factor BDNF, that occurred with rolipram treatment in both disease models (DeMarch et al., 2007, DeMarch et al., 2008).

The results of our studies with rolipram provide strong theoretical support for the strategy of targeting CREB signaling to promote striatal and cortical neuron survival in HD. However, these studies do not provide a direct path to a therapy since there are no PDE4 inhibitors currently available for clinical use, as these agents cause severe side effects.

Thus, we investigated another phosphodiesterase, PDE10A, which may be a particularly promising target as a potential treatment for HD (Chappie et al., 2009, Giampà et al., 2009). PDE10A is a cAMP and cAMP-inhibited cGMP PDE that is highly expressed in regions of the brain that are innervated by dopaminergic neurons such as the striatum, nucleus accumbens and olfactory tubercle (Fujishige et al., 1999a, Fujishige et al., 1999b, Loughney et al., 1999; Soderling et al., 1999).

In particular, PDE10A is highly expressed in the striatal medium spiny neurons (Coskran et al., 2006, Seeger et al., 2003; Xie et al., 2011), which are vulnerable in HD. In the striatal neurons, PDE10A is associated primarily with membranes (Kotera et al., 2004; Xie et al., 2011).

In the striatum and accumbens, PDE10A regulates both cAMP and cGMP signaling cascades (Conti and Beavo, 2007, Siuciak et al., 2006a, Siuciak et al., 2006b). Interestingly, PDE10A is the only mammalian PDE in which the GAF domain ligand is cAMP (Gross-Langenhoff et al., 2006). GAF (cGMP-stimulated PDEs, Anabaena adenylyl cyclases and Escherichia coli transcription factor Fhla) domains are regulatory domains found in many PDEs (Beavo, 1995, Conti and Jin, 1999, Soderling and Beavo, 2000) that are thought to bind small molecules (Gross-Langenhoff et al., 2006).

Notably, inhibition of PDE10A with the highly specific inhibitor TP-10 results in a significant increase in CREB phosphorylation in striatum (Schmidt et al., 2008). A recent study published by our group showed that TP-10 treatment in our rat QA model resulted in a significant sparing of striatal neurons, on one hand, and a parallel increase in activated CREB, on the other (Giampà et al., 2009). Indeed, more recently, we have demonstrated that TP-10 treatment resulted in amelioration of HD-like neuropathology and clinical signs also in the R6/2 mouse model of HD (Giampà et al., 2010).

Thus, a decrease in PDE10A levels was proven beneficial in HD models. However, changes in PDE10A during the course of the disease are not fully understood. Hebb and coworkers reported that PDE10A mRNA and protein levels decline very early in the course of disease progression in the R6/2 mice (Hebb et al., 2004, Hu et al., 2004). These findings raise the issue of how the inhibition of PDE10A would be beneficial in a disease where such enzyme is described to be lost.

We proposed to shed light on this issue by investigating, by means of immunohistochemistry, the localization of PDE10A in the striatum of the R6/2 mouse model of HD, with particular attention at the different striatal neuronal subsets. The R6/2 mouse model (Mangiarini et al., 1996) displays a progressive neurological phenotype that mimics many of the features of HD. The age of onset in R6/2 is observed as early as four weeks but more frequently occurs between nine and eleven weeks. Age at death is between 12 and 13 weeks. This line was chosen for our studies because it recapitulates striatal pathology of HD very accurately.

Moreover, we studied changes in PDE10A levels after administration of the PDE10A inhibitor TP-10, in order to shed light on the effects of enzyme activity inhibition on its expression in striatal neurons.