Jiali Li

Nuclear accumulation of HDAC4 in ATM deficiency promotes neurodegeneration in ataxia telangiectasia

Ataxia telangiectasia is a neurodegenerative disease caused by mutation of the Atm gene. Here we report that ataxia telangiectasia mutated (ATM) deficiency causes nuclear accumulation of histone deacetylase 4 (HDAC4) in neurons and promotes neurodegeneration. Nuclear HDAC4 binds to chromatin, as well as to myocyte enhancer factor 2A (MEF2A) and cAMP-responsive element binding protein (CREB), leading to histone deacetylation and altered neuronal gene expression. Blocking either HDAC4 activity or its nuclear accumulation blunts these neurodegenerative changes and rescues several behavioral abnormalities of ATM-deficient mice. Full rescue of the neurodegeneration, however, also requires the presence of HDAC4 in the cytoplasm, suggesting that the ataxia telangiectasia phenotype results both from a loss of cytoplasmic HDAC4 as well as its nuclear accumulation. To remain cytoplasmic, HDAC4 must be phosphorylated. The activity of the HDAC4 phosphatase, protein phosphatase 2A (PP2A), is downregulated by ATM-mediated phosphorylation. In ATM deficiency, enhanced PP2A activity leads to HDAC4 dephosphorylation and the nuclear accumulation of HDAC4. Our results define a crucial role of the cellular localization of HDAC4 in the events leading to ataxia telangiectasia neurodegeneration.

Cytoplasmic ATM in Neurons Modulates Synaptic Function

ATM is a PI 3-kinase involved in DNA double-strand break repair. ATM deficiency leads to ataxia-telangiectasia (A-T), a syndrome of cancer susceptibility, hypersensitivity to ionizing radiation, immune deficiency, and sterility [1, 2]-phenotypes that can straightforwardly be attributed to a defective response to DNA damage. Yet patients with A-T also suffer from ataxia, speech defects, and abnormal body movements [3-5]-neurological phenotypes whose origins remain largely unexplained. Compounding the discordance, Atm mutations in mouse interfere with DNA repair but have only mild neurological symptoms [6-9], suggesting that the link between DNA damage and the death of neurons can be broken [10-12]. We find that in neurons, ATM protein has a substantial cytoplasmic distribution. We show that in Atm(tm1Awb) mice, hippocampal long-term potentiation is significantly reduced, as is the rate of spontaneous vesicular dye release, suggesting a functional importance of cytoplasmic ATM. In the cytoplasm, ATM forms a complex with two synaptic vesicle proteins, VAMP2 and synapsin-I, both of which must be phosphorylated to bind ATM. Also, cytoplasmic ATM physically associates with the homologous PI 3-kinase, ATR. The neurological symptoms of ataxia-telangiectasia may thus result from defective nonnuclear functions of ATM not associated with DNA repair.

The role of ATM and DNA damage in neurons: Upstream and downstream connections

ATM (ataxia-telangiectasia mutated) is a large protein kinase whose best-known function is as a participant in the process of DNA damage repair, specifically lesions that result in double strand breaks. In the cells of the nervous system, however, the symptoms of children with ataxia-telangiectasia and the phenotypes of mice with engineered mutations in their ATM gene argue for a broader range of protein functions. ATM is now appreciated to play a role in vesicle dynamics as well as in the maintenance of the epigenetic code of histone modifications. Finally, the decline of ATM levels with age suggest that late onset neurodegenerative diseases may owe part of their pathogenesis to deficits in ATM signaling. Evidence from the location of HDAC4 in the hippocampal pyramidal cells of the Alzheimers disease brain supports this hypothesis. These multiple functions of the ATM protein are in keeping with the complex multi-system nature of the symptoms of ataxia-telangiectasia and encourage us to look beyond DNA damage for the full understanding of the disease and its consequences.

Neurons in Vulnerable Regions of the Alzheimer’s Disease Brain Display Reduced ATM Signaling

Abstract Ataxia telangiectasia (A-T) is a multisystemic disease caused by mutations in the ATM (A-T mutated) gene. It strikes before 5 years of age and leads to dysfunctions in many tissues, including the CNS, where it leads to neurodegeneration, primarily in cerebellum. Alzheimer’s disease (AD), by contrast, is a largely sporadic neurodegenerative disorder that rarely strikes before the 7th decade of life with primary neuronal losses in hippocampus, frontal cortex, and certain subcortical nuclei. Despite these differences, we present data supporting the hypothesis that a failure of ATM signaling is involved in the neuronal death in individuals with AD. In both, partially ATM-deficient mice and AD mouse models, neurons show evidence for a loss of ATM. In human AD, three independent indices of reduced ATM function—nuclear translocation of histone deacetylase 4, trimethylation of histone H3, and the presence of cell cycle activity—appear coordinately in neurons in regions where degeneration is prevalent. These same neurons also show reduced ATM protein levels. And though they represent only a fraction of the total neurons in each affected region, their numbers significantly correlate with disease stage. This previously unknown role for the ATM kinase in AD pathogenesis suggests that the failure of ATM function may be an important contributor to the death of neurons in AD individuals.

Alteration in 5-hydroxymethylcytosine-mediated epigenetic regulation leads to Purkinje cell vulnerability in ATM deficiency

A long-standing mystery surrounding ataxia-telangiectasia is why it is mainly cerebellar neurons, Purkinje cells in particular, that appear vulnerable to ATM deficiency. Here we present data showing that 5-hydroxymethylcytosine (5hmC), a newly recognized epigenetic marker found at high levels in neurons, is substantially reduced in human ataxia-telangiectasia and Atm(-/-) mouse cerebellar Purkinje cells. We further show that TET1, an enzyme that converts 5-methylcytosine (5mC) to 5hmC, responds to DNA damage and manipulation of TET1 activity directly affects the DNA damage signalling and ATM-deficient neuronal cell cycle re-entry and death. Quantitative genome-wide analysis of 5hmC-containing sequences shows that in ATM deficiency there is a cerebellum- and Purkinje cell-specific shift in 5hmC enrichment in both regulatory elements and repeated sequences. Finally, we verify that TET1-mediated 5hmC production is linked to the degenerative process of Purkinje cells and behavioural deficits in Atm(-/-) mice. Taken together, the selective loss of 5hmC plays a critical role in driving Purkinje cell vulnerability in ATM deficiency.

Stable Brain ATM Message and Residual Kinase-Active ATM Protein in Ataxia-Telangiectasia

The gene that is mutated in ataxia-telangiectasia (A-T), ATM, is catalytically activated in response to DNA damage. Yet a full accounting for the CNS deficits in human A-T or its mouse models remains elusive. We have analyzed the CNS phenotypes of two mouse Atm alleles—Atmtm1Bal (Bal) and Atmtm1Awb (Awb). Neither mutant has detectable mRNA or protein in peripheral tissues. In brain, although Bal/Bal mice have no ATM protein, they have nearly normal amounts of Atm mRNA. Bal/Bal neurons exhibit extensive cell cycle reentry and degeneration in both cortex and cerebellum. Unexpectedly, in Awb/Awb mice a novel mRNA is found in which the engineered mutation is excised. This mRNA is apparently translated and produces a catalytically active ATM protein that responds to DNA damage by phosphorylating p53 and Chk2. Prompted by these results, we examined eight cases of human A-T and found evidence for residual ATM protein in seven of them. These findings offer important new insights into the human disease and the role of brain ATM activity in the severity of the neurological symptoms of A-T.

The Interaction of the Atm Genotype with Inflammation and Oxidative Stress

In ataxia-telangiectasia (A–T) the death of neurons is associated with the loss of neuronal cell cycle control. In most Atm−/− mouse models, however, these cell cycle anomalies are present but the phenotype of neuronal cell loss found in humans is not. Mouse Atm −/− neurons re-enter a cell cycle and replicate their DNA, but they do not die – even months after initiating the cycle. In the current study, we explore whether systemic inflammation or hypoxia-induced oxidative stress can serve as second stressors that can promote cell death in ATM-deficient neurons. We find that after either immune or hypoxic challenge, the levels of cell cycle proteins – PCNA, cyclin A and cyclin B – are significantly elevated in cerebellar Purkinje cells. Both the number of cells that express cell cycle proteins as well as the intensity of the expression levels in each cell is increased in the stressed animals. The cell cycle-positive neurons also increasingly express cell death markers such as activated caspase-3, γ-H2AX and TUNEL staining. Interestingly, nuclear HDAC4 localization is also enhanced in Atm −/− Purkinje neurons after the immune challenge suggesting that both genetic and epigenetic changes in Atm−/− mice respond to environmental challenges. Our findings support the hypothesis that multiple insults are needed to drive even genetically vulnerable neurons to die a cell cycle-related cell death and point to either inflammation or oxidative stressors as potential contributors to the A−T disease process.

A Positive Feedback Mechanism That Regulates Expression of miR-9 during Neurogenesis

MiR-9, a neuron-specific miRNA, is an important regulator of neurogenesis. In this study we identify how miR-9 is regulated during early differentiation from a neural stem-like cell. We utilized two immortalized rat precursor clones, one committed to neurogenesis (L2.2) and another capable of producing both neurons and non-neuronal cells (L2.3), to reproducibly study early neurogenesis. Exogenous miR-9 is capable of increasing neurogenesis from L2.3 cells. Only one of three genomic loci capable of encoding miR-9 was regulated during neurogenesis and the promoter region of this locus contains sufficient functional elements to drive expression of a luciferase reporter in a developmentally regulated pattern. Furthermore, among a large number of potential regulatory sites encoded in this sequence, Mef2 stood out because of its known pro-neuronal role. Of four Mef2 paralogs, we found only Mef2C mRNA was regulated during neurogenesis. Removal of predicted Mef2 binding sites or knockdown of Mef2C expression reduced miR-9-2 promoter activity. Finally, the mRNA encoding the Mef2C binding partner HDAC4 was shown to be targeted by miR-9. Since HDAC4 protein could be co-immunoprecipitated with Mef2C protein or with genomic Mef2 binding sequences, we conclude that miR-9 regulation is mediated, at least in part, by Mef2C binding but that expressed miR-9 has the capacity to reduce inhibitory HDAC4, stabilizing its own expression in a positive feedback mechanism.

The temporary and accumulated effects of transcranial direct current stimulation for the treatment of advanced Parkinson’s disease monkeys

Transcranial direct current stimulation (tDCS) is a useful noninvasive technique of cortical brain stimulation for the treatment of neurological disorders. Clinical research has demonstrated tDCS with anodal stimulation of primary motor cortex (M1) in Parkinson’s disease (PD) patients significantly improved their motor function. However, few studies have been focused on the optimization of parameters which contributed significantly to the treatment effects of tDCS and exploration of the underline neuronal mechanisms. Here, we used different stimulation parameters of anodal tDCS on M1 for the treatment of aged advanced PD monkeys induced with 1-methyl-4-phenyl-1, 2, 3, 6-tetrahydropyridine (MPTP) administration, and then analyzed the temporary and accumulated effects of tDCS treatment. The results indicated anodal tDCS on M1 very significantly improved motor ability temporarily; importantly, the treatment effects of anodal tDCS on M1 were quantitatively correlated to the accumulated stimulation instead of the stimuli intensity or duration respectively. In addition, c-fos staining showed tDCS treatment effects activated the neurons both in M1 and substantia nigra (SN). Therefore, we propose that long time and continue anodal tDCS on M1 is a better strategy to improve the motor symptoms of PD than individual manipulation of stimuli intensity or duration.

The first observation of seasonal affective disorder symptoms in Rhesus macaque.

Diurnal animals are a better model for seasonal affective disorder (SAD) than nocturnal ones. Previous work with diurnal rodents demonstrated that short photoperiod conditions brought about depression-like behavior. However, rodents are at a large phylogenetic distance from humans. In contrast, nonhuman primates are closely similar to humans, making them an excellent candidate for SAD model. This study made the first attempt to develop SAD in rhesus macaque (Macaca mulatta) and it was found that short photoperiod conditions could lead monkeys to display depressive-like huddling behavior, less spontaneous locomotion, as well as less reactive locomotion. In addition to these depression-related behavioral changes, the physiological abnormalities that occur in patients with SAD, such as weight loss, anhedonia and hypercortisolism, were also observed in those SAD monkeys. Moreover, antidepressant treatment could reverse all of the depression-related symptoms, including depressive-like huddling behavior, less spontaneous locomotion, less reactive locomotion, weight loss, anhedonia and hypercortisolism. For the first time, this study observed the SAD symptoms in rhesus macaque, which would provide an important platform for the understanding of the etiology of SAD as well as developing novel therapeutic interventions in the future.