Objective: Child abuse has devastating and long-lasting consequences, considerably increasing the lifetime risk of negative mental health outcomes such as depression and suicide. Yet the neurobiological processes underlying this heightened vulnerability remain poorly understood. The authors investigated the hypothesis that epigenetic, transcriptomic, and cellular adaptationsmayoccur in theanterior cingulate cortex as a function of child abuse. Method: Postmortem brain samples from human subjects (N = 78) and from a rodent model of the impact of early-life environment (N = 24) were analyzed. The human sampleswere from depressed individuals who died by suicide, with (N= 27) or without (N = 25) a history of severe child abuse, as well as frompsychiatricallyhealthycontrol subjects (N = 26). Genomewide DNA methylation and gene expression were investigated using reduced representation bisulfite sequencing and RNA sequencing, respectively. Cell type-specific validation of differentiallymethylated loci was performed after fluorescenceactivated cell sorting of oligodendrocyte and neuronal nuclei. Differential gene expression was validated using NanoString technology. Finally, oligodendrocytes andmyelinatedaxons were analyzed using stereology and coherent anti-Stokes Raman scattering microscopy. Results: A history of child abuse was associated with cell type-specific changes in DNA methylation of oligodendrocyte genes and a global impairment of the myelin-related transcriptional program. These effects were absent in the depressed suicide completers with no history of child abuse, and they were strongly correlated with myelin gene expression changes observed in the animal model. Furthermore, a selective and significant reduction in the thickness of myelin sheaths around small-diameter axonswasobserved in individuals with history of child abuse. Conclusions: The results suggest that child abuse, in part through epigenetic reprogramming of oligodendrocytes, may lastingly disrupt corticalmyelination, a fundamental feature of cerebral connectivity.

Niemann-Pick type C (NPC) disease, an autosomal recessive disorder caused primarily by loss-of-function mutations in NPC1 gene, is characterized neuropathologically by intracellular cholesterol accumulation, gliosis and neuronal loss in selected brain regions. Recent studies have shown that NPC disease exhibits intriguing parallels with Alzheimer's disease (AD), including the presence of tau-positive neurofibrillary tangles (NFTs) and β-amyloid (Aβ)-related peptides in vulnerable brain regions. Since enhanced cholesterol level, which acts as a risk factor for AD, can increase Aβ production by regulating amyloid precursor protein (APP) metabolism, it is possible that APP overexpression can influence cholesterol-regulated NPC pathology. We have addressed this issue in a novel bigenic mice (ANPC) generated by crossing heterozygous Npc1-deficient mice with mutant human APP transgenic mice. These mice exhibited decreased lifespan, early object memory and motor impairments, and exacerbated glial pathology compared with other littermates. Neurodegeneration observed in the cerebellum of ANPC mice was found to be accelerated along with a selective increase in the phosphorylation/cleavage of tau protein. Additionally, enhanced levels/activity of cytosolic cathepsin D together with cytochrome c and Bcl-2-associated X protein suggest a role for the lysosomal enzyme in the caspase-induced degeneration of neurons in ANPC mice. The reversal of cholesterol accretion by 2-hydroxypropyl-β-cyclodextrin (2-HPC) treatment increased longevity and attenuated behavioral/pathological abnormalities in ANPC mice. Collectively, our results reveal that overexpression of APP in Npc1-deficient mice can negatively influence longevity and a wide spectrum of behavioral/neuropathological abnormalities, thus raising the possibility that APP and NPC1 may interact functionally to regulate the development of AD and NPC pathologies.

The TgCRND8 mouse model of Alzheimer's disease exhibits progressive cortical and hippocampal β-amyloid accumulation, resulting in plaque pathology and spatial memory impairment by 3 months of age. We tested whether TgCRND8 cognitive function is disrupted prior to the appearance of macroscopic plaques in an object recognition task. We found profound deficits in 8-week-old mice. Animals this age were not impaired on the Morris water maze task. TgCRND8 and littermate controls did not differ in their duration of object exploration or optokinetic responses. Thus, visual and motor dysfunction did not confound the phenotype. Object memory deficits point to the frontal cortex and hippocampus as early targets of functional disruption. Indeed, we observed altered levels of brain-derived neurotrophic factor (BDNF) messenger ribonucleic acid (mRNA) in these brain regions of preplaque TgCRND8 mice. Our findings suggest that object recognition provides an early index of cognitive impairment associated with amyloid exposure and reduced brain-derived neurotrophic factor expression in the TgCRND8 mouse.

The precise mechanisms underlying the memory-blocking properties of ethanol are unknown, in part because ethanol targets a wide array of neurotransmitter receptors and transporters. The aim of this study was to determine whether the memory loss caused by ethanol is mediated, in part, by α5 subunit-containing γ-aminobutyric acid subtype A receptors. These receptors have been implicated in learning and memory processes and are targets for a variety of neurodepressive drugs. Also, since these receptors generate a tonic inhibitory current in hippocampal pyramidal neurons, we examined whether concentrations of ethanol that block memory in vivo increased the tonic current using whole-cell patch-clamp recordings in hippocampal neurons. Null mutant mice lacking the α5 subunit (Gabra5-/-) and wild-type mice were equally impaired in contextual fear conditioning by moderate (1mg/kg) and high (1.5mg/kg) doses of ethanol. The higher dose of ethanol also reduced auditory delay fear conditioning to the same extent in the two genotypes. Interestingly, wild-type mice were more sensitive than Gabra5-/- mice to the sedative effects of low (0.5mg/kg) and moderate (1mg/kg) doses of ethanol in the open-field task. Concentrations of ethanol that impaired memory performance in vivo did not increase the amplitude of the tonic current. Together, the results suggest that the α5-subunit containing γ-aminobutyric acid subtype A receptors are not direct targets for positive modulation by ethanol nor do they contribute to ethanol-induced memory loss. In contrast, these receptors may contribute to the sedative properties of ethanol.

Purpose The low-density lipoprotein receptor (LDLr) mediates the uptake of LDL particles enriched with cholesterol, into several tissues. In contrast to other tissues, the brain is thought to obtain cholesterol solely by de novo synthesis, yet certain brain regions such as the brainstem are highly enriched with the LDLr. The goal of the present study was to assess the role of the LDLr in maintaining cholesterol concentrations in the brainstem of wildtype and LDLr knockout (LDLr−/−) mice. Cholesterol concentrations were also measured in the cortex, which served as a reference point, due to the lower expression of the LDLr, as compared to the brainstem. Methods LDLr−/− and wildtype mice consumed an AIN-93G diet ad libitum until 7 weeks of age. After microwaving, the cortex and anterior brain stem were isolated for cholesterol analysis. Cholesterol was extracted into chloroform/methanol, derivatized in trimethylsilyl chloride and measured by gas chromatography/mass spectrometry. Results Concentrations of cholesterol in the brainstem did not differ statistically between LDLr−/− (18.8 ± 1.6 mg/g wet weight brain) and wildtype (19.1 ± 2.0). Cortical cholesterol concentrations also did not differ statistically between LDLr−/− (11.0 ± 0.4 mg/g wet weight brain) and wildtype (11.1 ± 0.2) mice. Conclusion The LDLr is not necessary for maintaining cholesterol concentrations in the cortex or brainstem, suggesting that other mechanisms are sufficient to maintain brain cholesterol concentrations.

The brain cannot synthesize n-6 or n-3 PUFAs de novo and requires their transport from the blood. Two models of brain fatty acid uptake have been proposed. One requires the passive diffusion of unesterified fatty acids through endothelial cells of the blood-brain barrier, and the other requires the uptake of lipoproteins via a lipoprotein receptor on the luminal membrane of endothelial cells. This study tested whether the low density lipoprotein receptor (LDLr) is necessary for maintaining brain PUFA concentrations. Because the cortex has a low basal expression of LDLr and the anterior brain stem has a relatively high expression, we analyzed these regions separately. LDLr knockout (LDLr−/−) and wild-type mice consumed an AIN-93G diet ad libitum until 7 weeks of age. After microwaving, the cortex and anterior brain stem (pons and medulla) were isolated for phospholipid fatty acid analyses. There were no differences in phosphatidylserine, phosphatidylinositol, ethanolamine, or choline glycerophospholipid esterified PUFA or saturated or monounsaturated fatty acid concentrations in the cortex or brain stem between LDLr−/− and wild-type mice. These findings demonstrate that the LDLr is not necessary for maintaining brain PUFA concentrations and suggest that other mechanisms to transport PUFAs into the brain must exist.

The brain cannot synthesize n-6 or n-3 PUFAs de novo and requires their transport from the blood. Two models of brain fatty acid uptake have been proposed. One requires the passive diffusion of unesterified fatty acids through endothelial cells of the blood-brain barrier, and the other requires the uptake of lipoproteins via a lipoprotein receptor on the luminal membrane of endothelial cells. This study tested whether the low density lipoprotein receptor (LDLr) is necessary for maintaining brain PUFA concentrations. Because the cortex has a low basal expression of LDLr and the anterior brain stem has a relatively high expression, we analyzed these regions separately. LDLr knockout (LDLr(-/-)) and wild-type mice consumed an AIN-93G diet ad libitum until 7 weeks of age. After microwaving, the cortex and anterior brain stem (pons and medulla) were isolated for phospholipid fatty acid analyses. There were no differences in phosphatidylserine, phosphatidylinositol, ethanolamine, or choline glycerophospholipid esterified PUFA or saturated or monounsaturated fatty acid concentrations in the cortex or brain stem between LDLr(-/-) and wild-type mice. These findings demonstrate that the LDLr is not necessary for maintaining brain PUFA concentrations and suggest that other mechanisms to transport PUFAs into the brain must exist.

Victor Y. Cheng,1* Loren J. Martin,1* Erin M. Elliott,1 John H. Kim,2,3 Howard T. J. Mount,1,2,3 Franco A. Taverna,4 John C. Roder,4 John F. MacDonald,2 Amit Bhambri,1 Neil Collinson,5 Keith A. Wafford,5 and Beverley A. Orser1,2,6 1Institute of Medical Science, 2Department of Physiology, and 3Centre for Research in Neurodegenerative Diseases, Department of Medicine, Division of Neurology, University of Toronto, Toronto, Ontario, Canada M5S 1A8, 4Mount Sinai Hospital Research Institute, University of Toronto, Toronto, Ontario, Canada M5G 1X5, 5Merck Sharp and Dohme Research Laboratories, Neuroscience Research Center, Terlings Park, Harlow, Essex CM20 2QR, United Kingdom, and 6Department of Anesthesia, Sunnybrook and Women’s College Health Science Center, Toronto, Ontario, Canada M4N 3M5

TgCRND8 mice express a transgene encoding a double mutant (KM670/671NL plus V717F) form of human APP under control of the Syrian hamster prion (haPrP) promoter. These animals develop spatial reference memory impairments that can be discerned from 12 wks of age in Morris water maze tasks. It has been suggested that TgCRND8 plaque pathology begins after 9 wks and that cognitive impairments coincide with the onset of this plaque deposition (Hyde et al., 2005, Behav Brain Res 160:344–355). However, we typically find that the onset of plaque pathology begins only after the emergence of spatial deficits at 12 weeks and we have suspected that TgCRND8 mice might be impaired by other cognitive measures at even earlier ages. We examined object recognition memory in TgCRND8 and α–synuclein transgenic mice, in which expression of the mutant or wild type human proteins were under control of the haPrP promoter. Onset of progressive deficits in object recognition memory preceded plaques and spatial memory impairment in TgCRND8 mice by > 4 wks. Profound disruption of 3 h object recognition memory was observed before animals reached 8 wks of age. It seems unlikely that this phenotype arises because of a transgene insertional artifact, as other agreggating proteins, when expressed under this promoter, differentially affect the object recognition task. In particular, mice producing wild type, or missense mutant (A53T) human α–synuclein displayed intact 3 h object memory at 17 wks of age, while A30P mutant α–synuclein–Tg mice were impaired. In 1 yr old A53T and A30P mice, no 3 h object memory could be discerned, but performance in a subsequent 6 h trial remained intact. Both memory reactivation (Kelly et al., 2003, J Neurosci 12:5354–60) and lengthening of the testing delay (Hammond et al., 2004, Neurobiol Learn Mem 82:26–34) increase involvement of hippocampus relative to rhinal cortices in object recognition memory. Thus, these data support the notion that mutant human α–synuclein produces a relative sparing of hippocampus. The early impairment of object recognition memory in TgCRND8 mice suggests that rhinal cortices as well as hippocampus are early targets of pathology in this model.