Journal list menu

Volume 278, Issue 2 p. 223-235
Free Access

MNB/DYRK1A as a multiple regulator of neuronal development

Francisco J. Tejedor

Francisco J. Tejedor

Instituto de Neurociencias, CSIC and Universidad Miguel Hernandez, Alicante, Spain

Search for more papers by this author
Barbara Hämmerle

Barbara Hämmerle

Centro de Investigación Príncipe Felipe, Valencia, Spain

Search for more papers by this author
First published: 11 November 2010
Citations: 160
F. J. Tejedor, Instituto de Neurociencias, CSIC and Universidad Miguel Hernandez, Alicante, Spain
Fax: 34 965919561
Tel: 34 965919423
E-mail: [email protected]

Abstract

MNB/DYRK1A is a member of the dual-specificity tyrosine phosphorylation-regulated kinase (DYRK) family that has been strongly conserved across evolution. There are substantial data implicating MNB/DYRK1A in brain development and adult brain function, as well as in neurodegeneration and Down syndrome pathologies. Here we review our current understanding of the neurodevelopmental activity of MNB/DYRK1A. We discuss how MNB/DYRK1A fulfils several sequential roles in neuronal development and the molecular mechanisms possibly underlying these functions. We also summarize the evidence behind the hypotheses to explain how the imbalance in MNB/DYRK1A gene dosage might be implicated in the neurodevelopmental alterations associated with Down syndrome. Finally, we highlight some research directions that may help to clarify the mechanisms and functions of MNB/DYRK1A signalling in the developing brain.

Abbreviations

  • CNS
  • central nervous system
  • DS
  • Down syndrome
  • DYRK
  • dual-specificity tyrosine phosphorylation-regulated kinase
  • NRSF
  • neuron-restrictive silence factor
  • Introduction

    MNB/DYRK1A is a protein kinase that belongs to the dual-specificity tyrosine phosphorylation-regulated kinase (DYRK) family. MNB/DYRK1A is highly conserved from insects to humans [1] and it displays characteristic properties that are discussed in detail in one of the three minireviews in this series [2]. Orthologous genes have been cloned independently in various organisms and named Minibrain (Mnb) or Dyrk1A.

    The evidence from diverse experimental systems has shown various possible functions of MNB/DYRK1A in central nervous system (CNS) development, including its influence on proliferation, neurogenesis, neuronal differentiation, cell death and synaptic plasticity (see Table 1). These data, together with the localization of the human MNB/DYRK1A gene on chromosome 21 [3,4] and its overexpression in the brain of fetuses with Down syndrome (DS, trisomy 21) [5], have provided support to several hypotheses implicating MNB/DYRK1A in neurodevelopmental alterations underlying the cognitive deficits of DS (previously reviewed in [6,7]). These facts have certainly stimulated and conditioned the research into the neurobiological functions of MNB/DYRK1A. More recently, the observation that MNB/DYRK1A is overexpressed in the adult DS brain [8], together with biochemical data, also implicated MNB/DYRK1A in various neurodegenerative processes. This issue is extensively covered in the second accompanying paper of this minireview series [9].

    Table 1. Substrates and proteins that interact with MNB/DYRK1A in relation to its neuronal functions. Because the spatiotemporal regulation of its expression appears to be critical to understanding MNB/DYRK1A’s roles in neuronal development, we have also included possible regulators of Mnb/Dyrk1A expression and of MNB/DYRK1A kinase activity. For each protein we show: its main molecular properties, the molecular relationship with MNB/DYRK1A, the phosphorylation sites (if experimentally determined), the experimental system used to define this relationship, the possible function in neuronal development (if any), and the literature showing the relationship to MNB/DYRK1A. This list has been restricted to those genes/proteins for which there is evidence in the literature of a neuronal-related activity. Additionally, we highlight (*) those cases in which there is evidence (or strong indications) that the interaction with MNB/DYRK1A is involved in neuronal functions. ActR, regulator of activity; ExpR, regulator of expression; I, interacting protein; S, substrate; ND, not determined; CultNeu, cultured neurons; ivCNS, CNS in vivo; NCL, neural cell line; nNCL, non-neural cell line; Dif, differentiation; Other, nondevelopmental neuronal function; Prol, proliferation; Syn, synapse related; UF, unknown function; $, MNB/DYRK1A kinase primes the phosphorylation of several substrates by glycogen synthase kinase 3.
    Protein or signalling pathway Molecular nature Molecular relationship with MNB/DYRK1A Phosphorylation sites Experimental system Function Reference
    Amphiphysin Protein associated with the cytoplasmic surface of synaptic vesicles S Ser293 NCL, ivCNS Syn [54]
    β-Amyloid Peptide derived from amyloid precursor protein. Main component of amyloid plaques in Alzheimer’s disease ExpR ivCNS, NCL Other [78]
    Arip4 (androgen receptor interacting protein 4) Steroid hormone receptor cofactor I nNCL, CultNeu, ivCNS UF [79]
    APP (amyloid precursor protein) Amyloid precursor protein S Thr668 nNCL Other [80]
    ASF (alternative splicing factor) Splicing factor S, I Ser227, Ser234, Ser238 NCL, nNCL Other [81]
    bFGF Growth factor ActR NCL Dif [38]
    Caspase 9* Cystein aspartyl protease S Thr125 nNCL, ivCNS Cell death [59,82,83]
    Cyclin D1* Cell cycle regulator ? ivCNS, NCL Prol [26]
    CREB (cAMP responsive element binding protein) Transcription factor S Ser133 NCL Dif [38]
    CRY2 (cryptochrome 2) Flavoprotein, involved in circadian rhythm S Ser553, Ser557 nNCL, ivCNS Other [84]
    DNM1 (dynamin 1)* Cytoplasmic protein, involved in membrane trafficking S Ser857 nNCL, ivCNS Dif [14,16,50,51]
    Endophilin 1 Cytoplasmic protein involved in membrane trafficking I ivCNS Syn [55]
    E2F1 Transcription factor, involved in cell cycle regulation ExpR NCL, nNCL Prol, Dif [22]
    FKHR/FOXO1 (forkhead in rhabdoyosarcoma) Transcription factor S, I Ser329 nNCL UF [85,86]
    GLI1 (glioma-associated oncogene 1) Transcription factor involved in SHH signalling S Multiple sites (ND) NCL, nNCL Prol/Dif [36,87]
    GSK-3 (glycogen synthase kinase 3)* Protein kinase involved in multiple cellular processes $ nNCL, CultNeu, ivCNS Dif, other [43,88,89]
    Hip1 (huntingtin interacting protein 1) Accessory protein of the clathrin-mediated endocytosis pathway S ND NCL Dif [90]
    INI1/SNF5; SNR1 Chromatin modifying proteins I NCL, CultNeu, ivCNS Prol [23,45]
    MAP1B* Microtubule-associated protein S Ser1392 nNCL, CultNeu Dif [43]
    NFAT (nuclear factor of activated T-cells*) Transcription factor S ND ivCNS, NCL Dif [46,47]
    Notch* Cell–cell signalling transmembrane receptor protein S Multiple sites (ND) NCL, nNCL, ivCNS Prol, Dif [31]
    NRSF/REST (neuron-restrictive silence factor) Transcriptional repressor ExpR nNCL, ivCNS Prol/Dif [33]
    p53* Transcription factor S Ser15 NCL, ivCNS [27]
    PAHX-AP1 Phytanoyl-CoA α-hydroxylase associated protein 1, brain-specific protein I NCL UF [91]
    Presenilin1 Catalytic subunit of γ-secretase S Thr354 NCL, nNCL, ivCNS UF [92]
    Ras/Map kinase signalling Transmembrane signalling pathway I NCL Dif [39]
    SEPT4 (septin 4)* GTPase and cytoskeletal scaffolding protein S ND nNCL, ivCNS Syn [49]
    SIRT1 NAD-dependent protein deacetylase S Thr522 nNCL Cell death [74]
    SPRY2 (sprouty2)* Negative modulator of growth factor-mediated tyrosine kinase receptor signalling S Thr75 CultNeu, ivCNS Prol, Dif [19,93]
    STAT3 Signal transducer and activator of transcription S Ser727 nNCL UF [94,95]
    SJI1 (synaptojanin 1) Phosphoinositide phosphatase S Multiple sites (ND) ivCNS Syn [53]
    α-synuclein Cytoplasmic protein, major component of Lewy bodies S Ser87 NCL, ivCNS Other [96]
    TAU* Cytoskeletal protein, microtubule associated S Thr212 nNCL, ivCNS Other [78,88,97]
    14-3-3 14-3-3 family of regulating proteins I, ActR NCL, nNCL UF [98,99]

    Here we will focus on the neurodevelopmental functions of MNB/DYRK1A. We will discuss the data revealing the main roles interpreted by MNB/DYRK1A during brain development and their possible molecular mechanisms. Additionally, and given the extensive repertoire of putative substrates and proteins with which the MNB/DYRK1A kinase may interact, we will try to highlight the genes/proteins related to its neurodevelopmental activities. We will also discuss the possible implications of MNB/DYRK1A in the neurodevelopmental alterations associated with DS. Finally, we will highlight some directions for future research that we think may help to clarify the mechanisms and functions of MNB/DYRK1A signalling in the developing brain.

    The diverse functions of MNB/DYRK1A in neuronal development

    The initial evidence for the involvement of MNB/DYRK1A in neurodevelopment was provided by the analysis of mnb mutants of Drosophila. These flies develop a smaller adult brain, particularly in the optic lobes, which appears to be caused by altered proliferation in the neuroepithelial primordia of the larval CNS. This phenotype suggests a key function of MNB/DYRK1A in the regulation of neural proliferation and neurogenesis [10]. The highly conserved structure of this kinase [1] prompted extensive studies to be carried out on its vertebrate homologues. Indeed, a smaller brain with fewer neurons in certain regions was described in haploinsufficient Dyrk1A+/− mice [11], strongly suggesting an evolutionary conserved function of MNB/DYRK1A in brain development. This idea is also supported by the fact that truncation of the human MNB/DYRK1A gene causes microcephaly [12].

    Although in mammals Mnb/Dyrk1A is expressed in most adult tissues [5,13], its expression seems to be prevalent during embryonic brain development and it gradually decreases during postnatal periods to reach low levels in the adult [13,14]. Mnb/Dyrk1A is specifically expressed in four sequential phases during the development of the mouse brain: transient expression in preneurogenic progenitors; cell cycle-regulated expression in neurogenic progenitors; transient expression in recently born neurons; and persistent expression in late differentiating neurons ([14]; summarized in Fig. 1). This rather dynamic cellular/temporal expression strongly suggests that MNB/DYRK1A plays several sequential roles in neuronal development, which we shall discuss in this section. These roles seem to be neuron specific, as the analysis of the developing chick [15,16] and mouse CNS [14] show that MNB/DYRK1A expression is restricted to neuronal lineages, although its expression in glia has been reported in primary cultures [17].

    Details are in the caption following the image

    Schematic representation of the sequential expression of Mnb/Dyrk1A during the transition from neural proliferation to neuronal differentiation. In the vertebrate neuroepithelia, Mnb/Dyrk1A mRNA is first transiently expressed in preneurogenic progenitors, before it is asymmetrically segregated during cell division and it is inherited by only one of the daughter progenitor cells, triggering the onset of neurogenic divisions. Its expression is maintained in neurogenic progenitors although at a lower level. Later, Mnb/Dyrk1A is also transiently upregulated in postmitotic precursors (newborn neurons) and downregulated as the neuron begins to migrate away from the ventricular zone (VZ). Once the migrating neuron reaches its target position, Mnb/Dyrk1A is again expressed and it translocates transiently into the nucleus preceding the onset of dendrite formation. As dendrites begin to grow, MNB/DYRK1A localizes to the apical side of the growing dendrites.

    Proliferation and neurogenesis

    There is strong evidence that Mnb/Dyrk1A is transiently expressed during the single cell cycle of preneurogenic chick and mouse embryonic neuroepithelial progenitors that precedes the onset of neurogenesis [14,15]. This expression is of particular interest as Mnb/Dyrk1A mRNA is asymmetrically segregated during cell division and it is inherited by only one of the daughter cells [15] (Fig. 1). These data, together with its co-expression in preneurogenic mouse neuroepithelia with Tis21 [15], an antiproliferative gene that is upregulated in neural progenitors that make the switch from proliferative to neuron-generating divisions [18], suggest that Mnb/Dyrk1A may act as a cell determinant of neurogenesis. Accordingly, Mnb/Dyrk1A could induce the switch from proliferative to neurogenic cell divisions in neuronal progenitors. Interestingly, it has been recently shown that MNB/DYRK1A protein is actively distributed during adult neural stem cell division. Consequently, the inherited MNB/DYRK1A kinase acts as an inhibitor of epidermal growth factor receptor degradation by phosphorylating sprouty2, a modulator of tyrosine kinase receptor signalling [19]. In accordance with this, adult neural stem cells derived from Dyrk1A+/− mice exhibit defects in self-renewal.

    Noteworthy, the activity of Pom1p, an MNB/DYRK1A-related kinase from Schizosaccharomyces pombe, is cell cycle regulated in relation to symmetric growth and division [20]. However, Pom1p activity is high during symmetric cell division and when lost cells undergo asymmetric growth and division, the opposite to what appears to occur with MNB/DYRK1A in neural progenitors [14,15]. Moreover, mutants of mbk-1, the closest Mnb/Dyrk1A-related gene in Caenorhabditis elegans, do not show neurodevelopmental alterations [21]. Thus, new functions have probably been acquired by DYRK kinases during evolution to adapt to the new morphogenetic requirements of complex nervous systems.

    MNB/DYRK1A is also expressed in neurogenic progenitors in the Drosophila larval optic lobe [10] and in the embryonic mouse brain [14]. Although this expression seems to occur throughout the cell cycle, it is possible that the intensity of Mnb/Dyrk1A expression might vary at different cell cycle stages. Indeed, the expression of Mnb/Dyrk1A can be regulated by E2F1 [22], a transcription factor that plays a key role in the control of cell proliferation. Conversely, there is also evidence that MNB/DYRK1A may participate in the regulation of the cell cycle. For instance, it has been reported that MNB/DYRK1A interacts with SNR1 in Drosophila [23], a chromatin remodelling factor with a relevant role in cell cycle regulation [24]. Interestingly, increased levels of cyclin B1 have been detected in transgenic mice overexpressing Mnb/Dyrk1A [25] and it has recently been proposed that MNB/DYRK1A regulates the nuclear export and degradation of cyclin D1 in neurogenic mouse neuroepithelia [26]. Another very recent report has shown that the overexpression of MNB/DYRK1A induced impaired G1/G0–S phase transition in immortalized rat hippocampal progenitor cells [27]. The proposed mechanism is mediated by the phosphorylation of p53, which led to the induction of p21CIP1. There are also indications that MNB/DYRK1A is involved in the mitosis of non-neural cell lines [28]. These data establish a rather complex scenario with MNB/DYRK1A potentially fulfilling multiple actions in cell cycle regulation for which we have very little understanding of the molecular details.

    Interestingly, important evidence has recently emerged regarding the role of MNB/DYRK1A in terminating proliferation. Thus, based on the transient co-expression of MNB/DYRK1A with p27KIP1, the main cyclin-dependent kinase inhibitor in the mammalian forebrain [29], we proposed that MNB/DYRK1A is involved in the developmental signals that control cell cycle exit and early events of neuronal differentiation [14]. Indeed, it was recently reported that the overexpression of MNB/DYRK1A in the embryonic mouse telencephalon inhibits proliferation and induces premature neuronal differentiation of neural progenitors [26]. This gain of function was proposed to be driven through cyclin D1 nuclear export and degradation. Nevertheless, it has still to be proven whether the effect on cyclin D1 is a direct effect of MNB/DYRK1A or an indirect consequence of cell cycle withdrawal. Thus, confirmation of this mechanism by loss of function experiments would be important, especially as MIRK/DYRK1B, the closest homologue of MNB/DYRK1A, enhances cyclin D1 turnover [30].

    Neuronal differentiation

    In terms of the possible role of MNB/DYRK1A in early stages of neuronal differentiation, a recent report shows that the interaction and phosphorylation of the intracellular domain of NOTCH by MNB/DYRK1A attenuates NOTCH signalling in transfected neural cell lines [31]. NOTCH-mediated lateral inhibition is a key mechanism to regulate neuronal differentiation in the vertebrate CNS (reviewed in [32]). During neurogenesis, the cells in which NOTCH signalling is activated remain as progenitors, whereas those in which NOTCH activity diminishes differentiate into neurons. Thus, although the possible effects of MNB/DYRK1A kinase, as well as the underlying molecular mechanisms, need to be assessed in adequate models of the developing CNS, it is tempting to hypothesize that the MNB/DYRK1A kinase may regulate the onset of neuronal differentiation by inhibiting NOTCH signalling.

    Another rather interesting possibility is that MNB/DYRK1A influences neuronal differentiation through the transcriptional regulator neuron-restrictive silence factor (REST/NRSF). Using genetic approaches, transchromosomic models of DS, embryonic stem cells with partial trisomy 21 and transgenic Mnb/Dyrk1A mice, it has been shown that an imbalance in Mnb/Dyrk1A dosage perturbs Rest/Nrsf levels, altering gene transcription programmes of early embryonic development [33]. REST/NRSF is expressed strongly during early brain development in non-neuronal tissues and in neural progenitors, cells in which it represses fundamental neuronal genes [34]. Furthermore, activation of REST/NRSF target genes is both necessary and sufficient for the transition from pluripotent embryonic stem cells to neural progenitor cells, and from these to mature neurons [35]. In addition, phosphorylation by MNB/DYRK1A also regulates the transcriptional activity of glioma-associated oncogene 1 [36], a major effector of SHH signalling, which is a key pathway in the regulation of proliferation/differentiation during vertebrate CNS development [37].

    Given the roles played by MNB/DYRK1A in sequential steps of neurogenesis and its capacity to interact with and/or modulate different signalling pathways (EGF, FGF, NGF, SHH, NFAT, etc), it is tempting to hypothesize that MNB/DYRK1A plays a key role in co-ordinating neural proliferation and neuronal differentiation. Such co-ordination is crucial for proper brain development, as premature differentiation or overproliferation can alter the balance between neuronal populations, leading to mental disorders and neuropathologies.

    MNB/DYRK1A has also been implicated in various aspects of late neuronal differentiation. Thus, MNB/DYRK1A kinase activity was upregulated in response to bFGF during the differentiation of immortalized hippocampal progenitor cells. Blockade of this upregulation inhibited neurite formation. The mechanism proposed implicates phosphorylation of the transcription factor cAMP responsive element binding protein [38]. MNB/DYRK1A overexpression also potentiates nerve growth factor-mediated neuronal differentiation of PC12 cells by facilitating the formation of a Ras/B-Raf/MEK1 multiprotein complex in a manner independent of MNB/DYRK1A kinase activity [39]. Furthermore, the upregulation of MNB/DYRK1A expression and its translocation to the nucleus precedes the onset of dendrite formation in several differentiating neuronal populations ([14,16]; see also Fig. 1). Indeed, the number of neurites developed by newborn mouse hippocampal pyramidal neurons in culture is diminished when MNB/DYRK1A kinase activity is inhibited [40], indicating that MNB/DYRK1A kinase activity is required for neurite formation. So far, the mechanisms underlying this role of MNB/DYRK1A remain unclear. In addition, we observed that MNB/DYRK1A concentrates on the apical side of dendrites in differentiating neurons [14,16], suggesting a possible role in dendrite growth. The fact that cortical pyramidal cells from haploinsuffcient Dyrk1A+/− mice were considerably smaller and less branched than those of control littermates further supports this idea [41].

    Although the mechanisms underlying the effects of MNB/DYRK1A in dendritogenesis remain unknown, several possibilities might be considered in future studies. First, a kinome RNAi screen implicated MNB/DYRK1A in the regulation of actin-based protrusions in CNS-derived Drosophila cell lines [42]. Thus, MNB/DYRK1A could be involved in regulating actin dynamics, an important process in the regulation of neuronal morphology. Second, it has been shown that MNB/DYRK1A primes specific sites of MAP1B for glycogen synthase kinase 3β phosphorylation, an event that seems to be associated with alterations in microtubule stability [43]. It has also been shown that Drosophila MNB interacts with SNR1 [23], a member of the SWI/SNF complex, which is involved in the morphogenesis of dendritic arbors in Drosophila sensory neurons [44]. Moreover, MNB/DYRK1A interacts with INI1 (the SNR1 mammalian orthologue) in transfected neural cell lines [45]. In addition, the MNB/DYRK1A kinase has been shown to be a negative regulator of nuclear factor of activated T-cell signalling [46,47], which plays an important role in axonal growth during vertebrate development [48]. Finally, it is worth mentioning that two known substrates of the MNB/DYRK1A kinase colocalize with MNB/DYRK1A on the apical side of growing dendrites in several groups of neurons [14,16,49]: dynamin 1 [50,51], an important element in membrane trafficking; and septin 4 [49], a cytoskeletal scaffolding component implicated in neurodegeneration [52].

    There are also some indications that MNB/DYRK1A might be involved in synaptic functions. At the molecular level, it has been shown that MNB/DYRK1A binds to, phosphorylates and/or modulates the interaction of several components of the endocytic protein complex machinery, such as amphiphysin, dynamin 1, endophilin 1 and synaptojanin 1 [50,51,53–55], suggesting that it is involved in synaptic vesicle recycling. Transgenic mice overexpressing Mnb/Dyrk1A exhibit altered synaptic plasticity associated to learning and memory defects [56], whereas haploinsufficient Dyrk1A+/− mice have a reduced number of spines in the dendrites of cortical pyramidal cells [41] and show alterations in the pre- and postsynaptic components of dopaminergic transmission [57]. Thus, although these phenotypes may be due to changes in synaptic plasticity related to MNB/DYRK1A function in the adult brain, we should not rule out that these phenotypes might reflect impaired synapse formation during development, particularly as dendritogenesis and synaptogenesis are two processes that are tightly co-ordinated during brain development [58].

    Finally, we must stress that although MNB/DYRK1A is widely expressed in the developing CNS, there are clear indications that MNB/DYRK1A does not affect neuronal proliferation/differentiation in all CNS structures. For instance, regional morphological phenotypes have been reported in the brain of Mnb/Dyrk1A mutant flies [10] and mice [11]. Furthermore, the effect of Mnb/Dyrk1A loss of function and gain of function in the developing mouse retina indicates that the main role of MNB/DYRK1A in this tissue may be related to cell death/survival rather than to cell proliferation/differentiation [59].

    Possible implications of MNB/DYRK1A in the neurodevelopmental alterations associated with DS

    The human MNB/DYRK1A orthologue was initially localized in the so-called DS critical region [3,4], the minimal region of chromosome 21 that when triplicated confers most DS phenotypes [60]. This finding, together with its overexpression in fetuses with DS [5], initially suggested the implication of MNB/DYRK1A in a broad range of DS phenotypes. However, a recent more refined genetic analysis of numerous HSA21 segmental trisomies has generated a high-resolution genetic map of DS phenotypes [61]. According to this study, there is not a single DS critical region, but rather different ones for the diverse phenotypic features. Thus, the extra dosage of MNB/DYRK1A appears to be associated with a more restricted repertoire of DS phenotypes than previously thought, including mental retardation but excluding congenital heart disease.

    The brains of individuals with DS are characterized by their reduced size and a decrease in neuronal density in certain regions (reviewed in [62]). This neuronal deficit most probably originates through alterations in neurogenesis during development, as it is already detected in fetuses and children with DS [63,64]. Accordingly, altered neural proliferation and neurogenesis have been found in the forebrain of fetuses with DS and in trisomic DS mouse models [65–67].

    Based on the previously described functions of MNB/DYRK1A in the transition from proliferation to differentiation during neurogenesis, we predict that overexpression of MNB/DYRK1A in the developing brain of fetuses with DS could contribute to this neuronal deficit in several ways. First, through its role as an asymmetric determinant of neurogenesis, the overexpression of MNB/DYRK1A may cause the precocious onset of neurogenesis in progenitors and the concomitant depletion of the proliferating progenitor pool (Fig. 2). Second, due to its role in regulating the cell cycle exit of neurons, the overexpression of MNB/DYRK1A may induce premature cell cycle arrest of neurogenic progenitors leading to a decrease in the number of neurons generated by each progenitor. Thus, the combined effects of impairing these two activities could result in a decrease in the production of neurons (Fig. 2). Considering the effect of MNB/DYRK1A on cell cycle regulators like cyclin D1 [26] and p21CIP1 [27], a third possible effect of the overexpression of MNB/DYRK1A might be to modulate the cell cycle of neuronal progenitors. For instance, extended cell cycles have been found in a DS mouse model [65,66]. This may be relevant as neurogenic progenitors have a longer cell cycle than proliferative progenitors, and a lengthening cell cycle could contribute to a switch from proliferative to neurogenic divisions [68]. Further work will be required to assess these hypotheses.

    Details are in the caption following the image

    A working model for the involvement of MNB/DYRK1A overexpression in the neuronal deficit of DS. A schematic representation of the pattern of progenitor division and neuronal generation in a normal brain, and the possible consequences that MNB/DYRK1A overexpression might cause during neurogenesis in the DS brain. During normal neurogenesis, the transient expression of Mnb/Dyrk1A in preneurogenic progenitors triggers the onset of neurogenic divisions and consequently the production of neurons. The increase in the level of Mnb/Dyrk1A expression in DS may produce the precocious onset of neurogenic progenitors and a concomitant loss of proliferating progenitors, leading to a reduction in the total number of neurogenic lineages. Additionally, the overexpression of MNB/DYRK1A might induce premature cell cycle arrest of neurogenic progenitors, leading to a decrease in the number of neurogenic divisions undertaken by each neurogenic progenitor. Thus, the consequences of these alterations in neurogenesis would be a decrease in the production of neurons.

    Surprisingly, despite all the evidence pointing to various roles of MNB/DYRK1A in neural proliferation, neurogenesis and neuronal differentiation, no strong CNS developmental phenotypes have so far been described for most transgenic mice overexpressing Mnb/Dyrk1A. Nevertheless, all these transgenic mice exhibit learning/memory impairments [25,56,69,70]. It is possible that moderate increases in MNB/DYRK1A could produce subtle phenotypes that would require a more detailed analysis to detect. However, we should not rule out the possibility that due to the activities of MNB/DYRK1A in several sequential phases in proliferation/neurogenesis/differentiation, a maintained overexpression in the trangenic mice could result in compensatory phenotypes. Strikingly, the brains of 152F7 mice, which carry a YAC mouse line with three copies of at least two neighbouring HSA21 genes in addition to MNB/DYRK1A, are enlarged [25,69], a phenotype that apparently contradicts with the expected antiproliferative effect of MNB/DYRK1A [26,27].

    It is also well known that cortical neurons of brains with DS exhibit dendritic shortening or atrophy (reviewed in [71]). Thus, another developmental process that could be impaired through the overexpression of MNB/DYRK1A in DS is dendritogenesis. Indeed, cultured cortical neurons of Mnb/Dyrk1A transgenic mice exhibit poorer dendrite arborization [45]. Moreover, overexpression of MNB/DYRK1A in wild-type primary mouse cortical neurons leads to similar changes [45], strongly suggesting that MNB/DYRK1A triploidy can impair dendrite development in DS.

    Increased cell death is also associated with DS. For instance, cultured human cortical DS neurons exhibit intracellular oxidative stress and increased apoptosis [72]. Furthermore, increased cell death has been observed in the forebrain of fetuses with DS [67]. The involvement of MNB/DYRK1A in the regulation of caspase 9-mediated apoptosis in differentiating neurons of the developing retina has generated some speculation about the effects of MNB/DYRK1A gene-dosage imbalance in deregulating the apoptotic response in DS [59]. However, it seems unlikely that the overexpression of MNB/DYRK1A can contribute to the neuronal deficit of DS by stimulating developmentally regulated cell death as several studies have related increased MNB/DYRK1A levels to antiapoptotic or cell survival effects rather than to the induction cell death [59,73,74].

    Concluding remarks and perspectives

    As summarized in Table 1, many proteins have been identified as possible substrates and/or interacting proteins of the MNB/DYRK1A kinase. Nevertheless, we know very little about the actual physiological substrates/interacting partners of MNB/DYRK1A in neuronal development. In large, this is due to the fact that most molecular studies have been carried out in non-neuronal cells. Thus, efforts should be made to address the true specificity of these putative MNB/DYRK1A-related proteins in adequate neuronal systems and in suitable functional contexts. Also, given the wide molecular repertoire of substrates (transcription factors, translation factors, cytoskeletal proteins, membrane receptors, regulators of membrane dynamics, etc), it is possible that MNB/DYRK1A kinase could act at several levels in a multifaceted manner, integrating several cellular responses within a given neuronal process.

    MNB/DYRK1A also displays a rather varied subcellular distribution during neurodevelopment [14–16]. The early literature classified MNB/DYRK1A as a nuclear protein kinase because it contained a bipartite nuclear translocation signal and MNB/DYRK1A-tagged peptides indeed localized in the nucleus of transfected cell lines [75]. However, immunocytochemical analysis by high-resolution confocal microscopy has since shown that the endogenous MNB/DYRK1A protein has a mainly cytoplasmic and perinuclear localization in differentiating mammalian neurons [14]. Nevertheless, MNB/DYRK1A has also been detected in the form of speckles in neuronal nuclei at given developmental stages [14,16]. Thus, a working hypothesis is that MNB/DYRK1A is normally concentrated in the perinuclear area and that it translocates into the nucleus to regulate transcription factors in response to certain stimuli. It will therefore be very interesting to study the mechanisms that regulate this translocation process (see also the interesting comments about the distribution of MNB/DYRK1A in the adult mammalian brain in the accompanying review [9]).

    As previously discussed, there is also compelling evidence for the very precise spatiotemporal regulation of Mnb/Dyrk1A expression during brain development [13–16], which appears to be crucial for MNB/DYRK1A function. For example, it has been reported that the transient expression/activation of MNB/DYRK1A induces neuronal differentiation [38,39], but this is impaired by its stable overexpression [76]. Furthermore, it should be noted that the only well-known mechanism to activate the MNB/DYRK1A kinase is through a transient Tyr-kinase activity that autophosphorylates tyrosine residues in the activation loop during protein translation [77]. This implies that the upregulation of MNB/DYRK1A kinase can be indirectly controlled by regulating its expression, making the observed transient expression of MNB/DYRK1A in specific neurodevelopmental contexts (Fig. 1) even more relevant functionally. However, only a few molecules have been found to modulate Mnb/Dyrk1A gene expression in cell lines (reviewed in [2], see also Table 1) and almost nothing is known about the mechanisms regulating its expression during brain development. Thus, studies in true neurodevelopmental systems will be required to dissect out the mechanisms that actually regulate Mnb/Dyrk1A expression and their implication in brain development.

    Acknowledgements

    We are grateful to the Ministerio de Ciencia e Innovacion, the Generalitat Valenciana and the Fondation Jérôme Lejeune for their support of our MNB/DYRK1A research, and to former and present laboratory members for their contributions. We also thank Walter Becker for comments and suggestions.