New Perspectives of Dyrk1A Role in Neurogenesis and Neuropathologic Features of Down Syndrome.

Down syndrome (DS) is one of the most common genetic disorders accompanying with mental retardation, cognitive impairment, and deficits in learning and memory. The brains with DS also display many neuropathological features including alteration in neurogenesis and synaptogenesis and early onset of Alzheimer's disease (AD)-like symptoms. Triplication of all or a part of human chromosome 21, especially the 21q22.1~21q22.3 region called 'Down syndrome critical region (DSCR)', has been considered as the main cause of DS. One gene product of DSCR, dual-specificity tyrosine-phosphorylation-regulated kinase 1A (Dyrk1A), has been highlighted as a key contributor to the neural consequences of DS. This minireview summarizes accumulating recent reports about Dyrk1A involvement in the neuritogenesis, synaptogenesis, and AD-like neurofibrillary tangle formation, which is mainly focusing on Dyrk1A-mediated regulation of cytoskeletal proteins, such as tubulin, actin, and microtubule-associated protein tau. Understanding the molecular mechanisms of these phenomena may provide us a rational for new preventive and therapeutic treatment of DS.

INTRODUCTION

Since Dr. John L. H. Down first described the patients with mental retardation and characteristic facial appearance [1], Down syndrome (DS) has been characterized as one of the most common genetic disorders with an incidence of 1 in every 700~800 live births. DS patients also display cognitive impairment, learning and memory deficit, a high risk of leukemia, congenital heart disease, and hypotonia [2-4]. The main cause of DS is trisomy of all or a part of human chromosome 21 [2, 5]. Several studies of the partial trisomy 21 patients characterized a region of human chromosome 21 [21q22.1~21q22.3; named as 'Down syndrome critical region (DSCR)'] as a key suspect of DS symptoms [6-8].

Among the 33 presumed genes in DSCR, dual-specificity tyrosine-phosphorylation-regulated kinase 1A (Dyrk1A) has been intensively studied due to its close association with various cellular and neuronal processes [9]. Dyrk1A is a proline-directed serine/threonine kinase [10] that phosphorylates more than 20 substrates involved in various cellular processes [11]. More importantly, Dyrk1A up-regulation by trisomy 21 is implicated in the neural defects observed in the patients with DS [11].

Interestingly, accumulating data for recent years have suggested that Dyrk1A is involved in the regulation of cytoskeletal proteins such as tubulin, actin, and microtubule-associated protein tau and the alteration in neurogenesis of DS. In this context, this minireview focuses and discusses Dyrk1A and its link to neuropathologic features of DS.

ALTERATION OF NEUROGENESIS IN DOWN SYNDROME

The brains from DS patients have shown alteration in neurogenesis and synaptogenesis. The cortices from DS infant patients (from birth to 14 years) showed 20~50 percent fewer neuronal densities compared to age-matched controls [12]. The reduction in neuron numbers was also observed in the middle-aged patients [13]. This phenomenon correlates with findings from the fetal brains [14] and cultured fibroblasts with DS [15] that showed altered cell proliferation. The arrest of neurogenesis is accompanied with early arrest of brain growth as well as higher frequency of Alzheimer's disease (AD)-like plaque and tangle formation [13, 16, 17]. Also, the brains with DS showed a significant reduction in dendritic spine number in the hippocampus [18, 19]. Alteration in synaptogenesis can be supported by a recent study that the DS model mice show significant changes in spine morphology [20].

DYRK1A, MICROTUBULE, AND NEURITOGENESIS

Among a number of DSCR gene products, Dyrk1A is the most attractive protein that shows close association with neuritogenesis. In nerve growth factor-induced PC12 cell neuronal differentiation (a cell line derived from a pheochromocytoma of the rat adrenal medulla), Dyrk1A overexpression prolonged mitogen-activated protein kinase cascade and promoted neurite outgrowth [21]. In contrast, stable overexpression of Dyrk1A in immortalized H19-7 hippocampal neural progenitor cells caused a failure in basic fibroblast growth factor-induced neuronal differentiation [22]. Cortical neurons from Dyrk1A transgenic adult mice also showed a reduction in the length and number of dendrites [23]. Meanwhile, knockdown of Dyrk1A by specific short hairpin RNA in cultured cortical neurons caused a reduction in neurite length and tau-1-positive axons as well as an increase in neurite branching [24]. The compromised neuritogenesis by Dyrk1A knockdown was considered as a consequence of reduced phosphorylation at serine 1392 residue of microtubule-associated protein 1B (MAP1B) and of altered microtubule dynamics [24]. In addition, a potent and specific inhibitor of Dyrk1A, harmine, showed a capacity that can reduce the number of neurites in cultured hippocampal neurons [25]. Although there are many gaps in knowledge due to different experimental systems and approaches, it is obvious that Dyrk1A protein levels may contribute to neurite formation and altered neuritogenesis seen in DS.

DYRK1A, ACTIN, AND SYNAPTOGENESIS

Dyrk1A also contributes to regulation of actin dynamics and synaptogenesis. Yeast homologue of Dyrk kinase, Pom1, interacts with Rga4 GTPase-activating protein and regulates Rga4 localization, which is involved in Cdc42 GTPase localization in yeast [26]. RNA interference screen of Drosophila kinome based on cell morphology identified minibrain (Drosophila homologue of Dyrk1A) as a regulator of actin organization [27]. Specific knockdown of minibrain caused an increase in peripheral actin and in the number of protrusions in CNS-derived cell line [27]. The repressive effect of Dyrk1A on actin dynamics is further supported by several studies with mammalian systems. Dyrk1A regulates intramolecular interaction of neural Wiskott-Aldrich syndrome protein (N-WASP) and inhibits actin-polymerizing activity of N-WASP by phosphorylating GTPase-binding domain of N-WASP [28]. While specific knockdown of Dyrk1A in COS-7 cells promoted filopodia formation, Dyrk1A overexpression caused a reduction in dendritic spine formation of cultured hippocampal neurons [28]. In strong correlation, cortical neurons from Dyrk1A transgenic mice displayed a reduction in spine density, synapse formation, and dendritic filopodia length as well as alteration in spine morphology [23]. The association of Dyrk1A with actin filaments can be further supported by co-immunoprecipitation assays with DS tissues [29].

DYRK1A, TAU, AND ALZHEIMER'S DISEASE-LIKE FEATURES IN DOWN SYNDROME

One of the major neuropathological features of DS is a sign of early onset of AD-like symptoms, characterized by the formation of amyloid senile plaques (insoluble deposits of β-amyloid) and neurofibrillary tangles (hyperphosphorylated tau aggregates) [16, 17, 30, 31]. Dyrk1A has been intensively investigated in the context of its contribution to hyperphosphorylation of tau that stabilizes microtubules. The phosphorylating capacity of Dyrk1A to threonine 212 residue of tau was first described by in vitro kinase assay and phospho-specific antibody [32]. In addition to in vitro phosphorylation, protein-protein interaction between Dyrk1A and tau was analyzed by co-immunoprecipitation [22]. Stable overexpression of Dyrk1A caused a robust increase in intracellular inclusions of phosphorylated tau in immortalized H19-7 hippocampal neural progenitor cells [22]. Consistently, Dyrk1A transgenic mouse brains displayed an increase in tau phosphorylation not only at threonine 212 residue but also at serine 202 and 404 residues [33]. Immunoblotting with a bunch of phosphorylation site-specific antibodies revealed that Dyrk1A directly phosphorylates multiple serine and threonine residues of tau (serine 199, 202, 396, 400, 404, 422, threonine 181, 205, 212, 217, and 231) in vitro and the tau hyperphosphorylation occurs both in the brains from Ts65Dn DS model mice (a partially trisomic DS mouse model of mouse chromosome 16 which contains Dyrk1A gene) and in the temporal cortices from DS patients [34]. Dyrk1A immunoreactivity in the tau-positive neurofibrillary tangles in the DS brain strongly supports their association [35]. In the same context, specific knockdown of Dyrk1A by short hairpin RNA in cultured cortical neurons caused a significant reduction in tau phosphorylation at threonine 212 residue [24]. Recently, another possible association between Dyrk1A and tau inclusions was suggested. Dyrk1A interacts with and phosphorylates proline-rich domain of serine/arginine-rich protein 55 (SRp55) that regulates splicing of tau exon 10, an exon encoding the second microtubule-binding repeat [36]. Studies from past decade strongly suggest that one of the DSCR gene products, Dyrk1A, can contribute to the hyperphosphorylated tau and its inclusion formation, which correlates with one shown in the brains with DS.

CLOSING REMARKS

As described above, Dyrk1A is closely associated with regulation of cytoskeletal protein such as tubulin, actin, and microtubule-associated protein tau through phosphorylation of various substrates. A group of substrates that are phosphorylated by Dyrk1A further contributes to regulation of neuritogenesis, synaptogenesis, and AD-like neurofibrillary tangle formation. Although many gaps in knowledge are still remaining, those extensive studies strongly suggest that the approximately 1.5-fold increase of Dyrk1A in the brains with DS may be one of the factors that lead to the neuropathologic features shown in DS patients.

Understanding molecular mechanisms of neuropathological features can offer a rationale for new preventive and therapeutic treatment of DS. One could be that inhibition of Dyrk1A activity from the excessive protein amount in DS may prevent the symptoms or make them less severe. So far, a few potent inhibitors of Dyrk1A have been identified. One of them is epigallocatechin-3-gallate (EGCG), which is the major catechin component of green tea. Treatment of EGCG promoted long-term potentiation of Ts65Dn DS model mice [37] and rescued the defects of Dyrk1A transgenic mouse brains [38]. Another candidate inhibitor of Dyrk1A is harmine although all Dyrk family proteins can be inhibited by harmine [39]. Treatment of harmine effectively reduced tau phosphorylation in neuroglioma cell line [40]. Although the current version of Dyrk1A inhibitors should be improved to get better specificity and efficacy, understanding molecular links between a strong contributor such as Dyrk1A and neuropathological features of DS and developing potent inhibitors against the identified molecular targets will bring us more closely to new preventive and therapeutic treatment of DS.