Identification of phosphodiesterase 10A (PDE10A) was first reported in mice and humans at the same period from three laboratories (Soderling et al. 1999; Fujishige et al. 1999a; Loughney et al. 1999). Primary sequence of PDE10A possesses a catalytic domain (HD domain) conserved within the 3, 5′-cyclic nucleotide phosphodiesterase family. PDE10A shows substrate specificity for both cyclic AMP (cAMP) and cyclic GMP (cGMP), and hydrolyzes these molecules to 5′-AMP and 5′-GMP, respectively. PDE10A mRNA and protein are highly expressed in the brain, particularly, in the striatal medium spiny neurons (Fujishige et al. 1999b; Seeger et al. 2003). Genetic deletion of PDE10A gene in mice as well as PDE10A inhibition by papaverine, a first reported PDE10 inhibitor, showed altered behavioral responses to several schizophrenia models (Siuciak et al. 2006; Siuciak et al. 2008); therefore, therapeutic implications of PDE10A inhibitor for psychiatric diseases have been received with considerable attention. To date, several potent and selective inhibitors for PDE10A have been reported to be efficacious in rodent models for positive, cognitive, and negative symptoms of schizophrenia (Schmidt et al. 2008; Grauer et al. 2009), while it remains to be demonstrated that whether these inhibitors provide a therapeutic benefit for schizophrenia patients or not.
Kinetic analyses of human, mouse, and rat PDE10A indicated that PDE10A has higher affinity for cAMP (Km value of 0.05–0.158 μM) than cGMP (Km value of 0.26–9.3 μM) (Soderling et al. 1999; Fujishige et al. 1999a; Loughney et al. 1999). On the other hand, Vmax values for cAMP is twofold lower than that for cGMP. As expected from this observation, the catalytic activity for cAMP is significantly inhibited by the presence of cGMP (Fujishige et al. 1999a). Two major N-terminal splice variants have been reported in humans: PDE10A1 and PDE10A2 (Kotera et al. 1999). PDE10A2 is the most predominant splice form in the brain. Fujishige et al. reported various N-terminal splice variants in rats (PDE10A2-PDE10A6), but enzymatic properties and biological functions of minor splice variants have not determined yet (Fujishige et al. 1999b). Another feature in the primary structure of PDE10A is the two GAF domains (stands for cGMP binding and stimulated phosphodiesterases, Anabaena adenylyl cyclases, and Escherichia coli FhlA) that are situated in the N-terminus of the catalytic domain. Initial study using chimeric constructs with PDE10A GAF domain and bacterial adenylyl cyclase indicated that the cyclase activity was activated by cAMP binding to the GAF domain (Gross-Langenhoff et al. 2006). Subsequent study conducted by Matthiesen et al. has demonstrated that binding of cAMP does not stimulate hydrolytic activity of full-length human PDE10A2 (Matthiesen and Nielsen 2009). Crystal structure of PDE10A GAF-B domain with cAMP has been reported (Handa et al. 2008). This study has demonstrated that a cNMP-binding pocket tightly binds cAMP, and the β1 and β2 strand in the pocket contribute to the recognition of adenine base. At present, implication of the nucleotide binding to GAF domain for the biological pathway in vivo remains to be determined.
Tissue Expression and Subcellular Localization
PDE10A transcripts are highly expressed in basal ganglia, particularly in striatum and caudate nucleus in the central nervous system (Fujishige et al. 1999a, b). In the peripheral tissues in humans, moderate expression was detected in the thyroid, testis, heart, and kidney (Fujishige et al. 1999a; Loughney et al. 1999). Immunohistochemical analysis has revealed that PDE10A protein is expressed exclusively in the cell body of the striatal GABArgic medium spiny neurons and their projecting axons in substantia nigra (Seeger et al. 2003). Striatal expression of PDE10A transcripts and proteins is conserved among humans and rodents, but there are slight species differences in other region in the brain. For example, in rats, marked expression was observed in hippocampus and cerebellum, whereas only minimum expression was detected in these tissues in mice (Coskran et al. 2006). PDE10A proteins are clearly detectable in cerebral cortex in humans. On the other hand, the expression in this area is relatively low in mice and rats. Growing evidence has suggested the idea that subcellular localization of PDEs contributes to the compartmentation of cyclic nucleotide signals. Human and rat PDE10A2 proteins were demonstrated to be localized in membrane fraction in PC12 cells, while human PDE10A1, and rat PDE10A3 were mainly expressed in cytosolic fraction (Kotera et al. 2004). Rat PDE10A2, a major splice variant in the striatum, has been mainly detected in the membrane fraction prepared from striatum by using specific antibody for PDE10A2 (Kotera et al. 2004). Xie et al. reported detailed subcellular localization of PDE10A in striatum by immunoelectron microscopy study (Xie et al. 2006). The report has indicated that the PDE10A protein was confined in vesicle-filled presynaptic terminals of striatal neurons in rats. Two modes of posttranscriptional regulation are known to alter subcellular localization of PDE10A. Kotera et al. has reported that phosphorylation at Thr-16 of PDE10A2 by PKA-mediated pathway causes alteration of subcellular localization from membrane to cytosol in PC12 cells (Kotera et al. 2004). Charych et al. has demonstrated that palmitoylation of Cys-11 of PDE10A2 is required for membrane association in vivo and distal dendritic trafficking of the protein in cultured striatal neurons (Charych et al. 2010). Phosphorylation of PDE10A2 at Thr-16 interfered with palmitoylation at Cys-11, resulting in the retention of PDE10A2 in cytosolic fraction. Zinc finger domain-containing DHHC domain-containing protein (ZDHHC)-7 and −19 were identified as candidates for the enzyme-mediating palmitoylation of PDE10A2.
Implication for Cellular Signaling
Striatal medium spiny neurons can be categorized into two distinct neurons: dopamine D1 receptor-containing (direct pathway) neurons and dopamine D2 receptor-containing (indirect pathway) neurons, which are involved in the regulation of cortico-striatal-thalamic loop. Involvement of PDE10A on corticostriatal signaling was investigated by Threlfell et al. using electrophysiological technique in vivo (Threlfell et al. 2009). Single-cell recording from direct or indirect pathway neurons, which were identified based on the response to antidromic stimulation, has revealed that administration of PDE10A inhibitors, papaverine and TP-10, increase the responsiveness of striatal neurons to cortical stimulation in vivo. Moreover, robust increase of the activity in response to cortical input was observed in striatopallidal neurons but not striatonigral neurons. As well as PDE10A, PDE4 is expressed in stiratum. Nishi et al. have reported distinct role of PDE4 and PDE10A in striatum (Nishi et al. 2008). Expression of PDE10A was detected in all DARPP-32 positive neurons, while PDE4B was mainly expressed in striatopallidal neurons rather than striatonigral neurons. PDE10A inhibition by papaverine, but not PDE4 inhibition by rolipram, caused DARPP-32 and GluR1 phosphorylation in stiatum. On the other hand, phosphorylation of tyrosine hydroxylase was increased by rolipram at presynaptic dopaminergic terminals. In striatonigral neurons, PDE10A inhibition activated cAMP-pathway (measured by DARPP-32 phosphorylation) resulting in the potentiation of dopamine D1 receptor pathway. In striatopallidal neurons, papaverine activated cAMP-pathway by potentiating adenosine A2a receptor pathway and inhibiting dopamine D2 receptor pathway simultaneously.
Knockout (KO) Mouse Phenotype
Behavioral characteristics of PDE10A KO mice (C57BL/6 background) have been reported by Siuciak et al. (2008). PDE10A KO mice showed slight decrease in spontaneous locomotor activity and delayed acquisition in conditioned avoidance response (CAR). MK-801-induced increase of locomotor activity was blunted in KO mice compared with wild-type mice, while there was no difference in PCP-induced locomotor response. Unexpectedly, basal cyclic nucleotide concentration and CREB phosphorylation were unchanged in striatum from KO mice, whereas dopamine turnover was increased in KO mouse striatum. KO mice demonstrated increased response to amphetamine and methamphetamine administration. These observations suggested the hypothesis that PDE10A may influence cortical glutamatergic neuronal activity via both D1 receptor-mediated direct pathway and D2 receptor-mediated indirect pathway. Sano et al. has reported increased social interaction behavior in KO mice of PDE10A2, a major splice form in the brain (Sano et al. 2008). Although elevation of cAMP and CREB phosphorylation in striatum was observed, PDE10A2 KO mice did not exhibit abnormalities in spontaneous locomotor activity. Behavioral test battery has revealed that responses in elevated plus maze, rotor rod, conditioned fear, and forced-swim were unchanged between PDE10A2 KO and wild-type mice.
Papaverine, an opium alkaloid used for the treatment of visceral spasm and smooth muscle relaxation, was a first-identified inhibitor for PDE10A (IC50 value = 36 nM) (Siuciak et al. 2006). Papaverine has shown to increase cGMP and CREB phosphorylation in mouse striatum; in addition, these observations were completely abolished in PDE10A KO mice. In 2008, a research group from Pfizer has reported the discovery of a novel class of inhibitor, TP-10 (Schmidt et al. 2008). TP-10 potently inhibited rat PDE10A (Ki value = 0.3 nM) and showed great selectivity (> 3000-fold) against other PDEs. TP-10 caused dose-dependent increase of cAMP, cGMP, and phosphorylated CREB levels in mouse striatum in vivo. Administration of TP-10 inhibited CAR and reversed amphetamine-induced auditory gating deficits in rodents. Clinically used D2-antagonists and atypical antipsychotics have been known to induce catalepsy. Although TP-10 induced weak catalepsy, the response was not increased by the dose elevation in contrast to the D2-antagonists and atypical antipsychotics. In 2009, detailed evaluation of MP-10, a novel compound that is closely related analogue of TP-10, in several animal models for schizophrenia was reported (Grauer et al. 2009). MP-10 demonstrated IC50 value of 1.3 nM for human PDE10A and at least 1300-fold selectivity for other PDEs. As expected from the results in TP-10 treatment, MP-10 caused significant increase of cAMP and cGMP in vivo. In addition, phosphorylation of CREB, DARPP-32, and GluR1 was increased by MP-10 treatment. MP-10 administration demonstrated weak catalepsy and disruption of CAR in rodents. Increased phosphorylation of GluR1 at Ser-845 and upregulation of enkephalin and substance-P mRNA accompanying with MP-10 treatment further supports the idea that PDE10A modulate dopaminergic and glutamatergic neurotransmission. The fact that MP-10 improved social odor memory, and increased time spent in social side in social approach/social avoidance model in mice suggest that MP-10 has a potential for the improvement of cognitive function and negative symptoms in schizophrenia patients. To date, many structurally different classes of PDE10A inhibitors have been reported from various laboratories. Comprehensive review of these inhibitors was published by Kehler and Kilburn (2009). Currently, phase 1 clinical trial using MP-10 has been conducted.
PDE10A is a dual-specific and cGMP-inhibited phosphodiesterase that is highly expressed in the striatal medium spiny neurons. This unique expression property has suggested association of this enzyme for basal ganglia function. Recent advances in the research of PDE10A using genetic deletion and selective inhibitors have suggested that PDE10A is involved in the control of dopaminergic circuit as well as the modulation of gultamatergic neuronal transmission. D2 antagonist and atypical antipsychotic drugs are widely used for the treatment of schizophrenia, while these medications exhibit several side effects such as extrapyramidal disorder and obesity, and ineffectiveness for negative symptom in schizophrenia. Distinct roles of PDE10A in the regulation of neuronal circuit may provide novel hypothesis in the regulation of brain function by PDE10A and therapeutic opportunity for schizophrenia by PDE10A inhibitor.
- Grauer SM, Pulito VL, Navarra RL, Kelly MP, Kelley C, Graf R, Langen B, Logue S, Brennan J, Jiang L, Charych E, Egerland U, Liu F, Marquis KL, Malamas M, Hage T, Comery TA, Brandon NJ. Phosphodiesterase 10A inhibitor activity in preclinical models of the positive, cognitive, and negative symptoms of schizophrenia. J Pharmacol Exp Ther. 2009;331:574–90.PubMedCrossRefGoogle Scholar
- Kotera J, Sasaki T, Kobayashi T, Fujishige K, Yamashita Y, Omori K. Subcellular localization of cyclic nucleotide phosphodiesterase type 10A variants, and alteration of the localization by cAMP-dependent protein kinase-dependent phosphorylation. J Biol Chem. 2004;279:4366–75.PubMedCrossRefGoogle Scholar
- Schmidt CJ, Chapin DS, Cianfrogna J, Corman ML, Hajos M, Harms JF, Hoffman WE, Lebel LA, McCarthy SA, Nelson FR, Proulx-LaFrance C, Majchrzak MJ, Ramirez AD, Schmidt K, Seymour PA, Siuciak JA, Tingley 3rd FD, Williams RD, Verhoest PR, Menniti FS. Preclinical characterization of selective phosphodiesterase 10A inhibitors: a new therapeutic approach to the treatment of schizophrenia. J Pharmacol Exp Ther. 2008;325:681–90.PubMedCrossRefGoogle Scholar