Accepted Posters List


Roger N. Rosenberg, MD
Professor of Neurology
University of Texas Southwestern Medical Center at Dallas

A regulatory immune response after DNA vaccination against Amyloid beta 42

Doris Lambracht-Washington, Qu B-X, Min Fu, Todd N. Eagar, Olaf Stüve, Roger N. Rosenberg

Background: Aβ immunotherapy for Alzheimer's Disease (AD) harbors the danger of an inflammatory autoimmune response, and a clinical trial in which AD patients received Aβ42 peptide immunizations was stopped when participants developed encephalitis due to an inflammatory Th1 response targeting Aβ in brain. We have used DNA Aβ42 trimer immunizations in mice and found a Th2-type antibody response indicative of a non-inflammatory immune response and the disappearance of Aβ42 specific T cells. This low level of T cell proliferation in DNA Aβ42 trimer immunized mice might be due to regulatory T cells (Tregs). Methods: CD4+CD25+ cells (Tregs) were depleted by injection of a CD25 antibody (clone PC61) in half of the mice which received DNA Aβ42 trimer immunizations. Humoral and cellular immune responses were compared by ELISA, ELISPOT, and CFSE proliferation. Results: Depletion of CD25 positive cells did lead to increased antibody titers: 1:700 in CD25 depleted mice compared to 1:350 in DNA Aβ42 trimer immunized mice after 4 immunizations. In both groups, the IgG1 antibody isotype indicated a Th2 response. After one immunization we found increased CD4 proliferation index in the CD25 depleted mice: 1.7± 0.19 compared to 1.1± 0.06 in DNA vaccinated mice (P = 0.0433). After four immunizations the T cell response in the DNA immunized mice disappeared while in CD25 depleted mice an increased but rather unspecific proliferation was found. Parallel ELISPOT analyzes showed no IFNy, IL-17, or IL-5 secretion in the DNA immunized mice but increased levels for the regulatory cytokines IL-10 and TGFβ1 after Aβ42 peptide re-stimulation (P = 0.01 and 0.004, respectively). IL-10 was likely produced by T cells as the re-stimulation assays showed increased IL-10 levels for T cell peptides Aβ7-22, 9-27, 10-26, 12-28, but not Aβ1-15 which is a B cell peptide. Conclusions: Depletion of CD25 positive T cells increased the specific antibody response in our model. Data obtained thus far are consistent with the disappearing effector T cell response in DNA Aβ42 trimer immunized mice (no proliferation, no IFNγ secretion) and the proposed regulatory T cell response (IL-10 secretion after DNA Aβ42 immunization) demonstrating safety for this immunization approach as potential AD therapy.


Wendy Qiu, M.D., Ph.D
Associate Professor
Boston University School of Medicine

The Interaction between the ApoE Genotype and Angiotensin Converting Enzyme Inhibitors on the Risk of Developing Alzheimer’s disease

Wei Qiao Qiu1, 2, 3*, D. Mkaya Mwamburi4, Neil Kowall3, Robert Stern3, Leslie E. Phillips5 and Haihao Zhu1

Departments of Pharmacology & Experimental Therapeutics1, Psychiatry2, and Alzheimer’s disease Center3, Boston University School of Medicine; Department of Public Health and Family Medicine9 , Tufts University 4 ; Urban Indian Health Institute, Seattle, WA5

Objective: The effect of Angiotensin converting enzyme (ACE) inhibitors on Alzheimer’s disease (AD) remains unclear, with conflicting results reported. We investigated the interaction of the Apolipoprotein E (ApoE) genotype and ACE inhibitors on AD.

Methods: We used cross-sectional data from the Nutrition and Memory in the Elderly (NAME) with an AD diagnosis and documentation of medications taken as well as longitudinal data from the National Alzheimer’s Coordinating Center (NACC). ApoE genotype in both studies was determined. The ACE inhibitors in the NACC study were divided into central and peripheral ACE inhibitors; ACE inhibitors in the NAME study were pooled.

Findings: Cross-sectionally using the NAME study, we found that ApoE4 carriers treated with ACE inhibitors had a greater frequency of AD diagnoses compared with those who did not have the treatment (28% vs. 6%, p = 0.01) or with ApoE4 non-carriers treated with an ACE inhibitor (28% vs. 10%, p = 0.03). AD was significantly associated with an interaction between ApoE4 and ACE inhibitor use. In the NAME study, subjects with both the ApoE4 genotype and using ACE inhibitors were at significantly increased risk of having AD (OR = 20.85, 95% CI = 3.08, 140.95, p = 0.002) after adjusting for age, gender, ethnicity and education. We thus further used the longitudinal the NACC study to determine two bi-directional possibilities for ACE inhibitors on AD in ApoE4 carriers: delaying or accelerating the AD onset. We found that among ApoE4 non-carriers, both central (OR = 1.45, 95% CI = 1.16, 1.81, p = 0.001) and peripheral (OR = 1.77, 95% CI = 1.22, 2.56, p = 0.003) ACE inhibitors were associated with a reduced risk of AD onset. In contrast, among those ApoE4 carriers, both subclasses of ACE inhibitors were not associated with the risk of AD onset significantly although the central ACE inhibitors had a trend to reduce the onset of AD (OR = 1.31, 95% CI = 0.96, 1.78, p = 0.09).

Conclusion: ACE inhibitors may reduce the risk of developing AD but the effect may depend on ApoE4 genotype. ACE inhibitors, especially peripheral ones, may not delay AD development effectively when ApoE4 allele is present.


Erik S. Musiek, MD, PhD

Washington Univ. School of Medicine

The circadian clock gene BMAL1 regulates brain bioenergetics and redox homestasis: implications for neurodegeneration

Erik S. Musiek, Miranda M. Lim, Adam Q. Bauer, Jee Hoon Roh, Joseph P. Culver, Benoit I. Giasson, Garret A. FitzGerald, and David M. Holtzman

Dept. Of Neurology, Knight Alzheimer’s Disease Research Center, Hope Center for Neurologic Disease, and Dept. of Radiology, Washington Univ. School of Medicine. Institute for Translation Medicine and Therapeutics, University of Pennsylvania School of Medicine.

Circadian dysfunction is commonly observed in neurodegenerative conditions, including Alzheimer’s Disease (AD), though its impact on disease pathogenesis is unknown. Circadian clock proteins regulate many aspects of cellular biochemistry and metabolism, and deletion of the master clock gene BMAL1 in mice causes complete loss of circadian rhythms, systemic oxidative injury, peripheral metabolic dysfunction, and accelerated aging. We hypothesized that the core circadian clock might regulate metabolism and redox homeostasis in the brain, and thus examined the impact of BMAL1 deletion on brain pathology. Deletion of BMAL1 in mice caused severe age-dependent astrogliosis and microglial activation starting as early as 1.5 months of age which was most evident in cerebral cortex and hippocampus. Neuropathology was associated with significant neuronal lipid peroxidation, impaired expression of the redox defense gene mitochondrial aldehyde dehydrogenase (ALDH2), ATP depletion, and activation of the metabolic sensor AMPK, all consistent with cerebral metabolic stress. BMAL1 deletion also caused perturbed mitochondrial function in cells and mice. Optical imaging of resting state functional connectivity demonstrated neural network dysfunction in vivo. Similar neuropathology and biochemical changes were observed in mice with brain-specific Bmal1 deletion (Nestin-Cre::Bmal1-floxed mice), which have intact sleep architecture and behavioral circadian rhythms, suggesting that these changes were not due to sleep deprivation or peripheral circadian dysregulation. Our findings implicate BMAL1 as a novel regulator of cerebral bioenergetics and neuronal lipid peroxidation, and suggest that circadian dysfunction may promote neurodegeneration.


Jonathan Jackson, M.A. (ABD)

Doctoral Student
Washington University

There has been considerable emphasis on potential preclinical markers of Alzheimer disease (AD). Cognitive tasks may potentially provide an advantageous method of noninvasive, inexpensive testing that can be administered in a variety of settings. Few cognitive tasks, however, have shown preclinical sensitivity to AD. The present study explored whether performance on a 5-minute Sustained Attention to Response Task (SART), a standard measure of mind-wandering, is sensitive to the earliest stages of dementia, as well as to the accumulation of biomarkers in non-demented individuals. 


David Ruvolo, M.S.

Washington Univ. School of Medicine

A Solution-Based Model for Effective Ethnic Recruitment in Alzheimer’s Research

Underrepresentation of ethnic groups is a tremendous hurdle throughout Alzheimer’s research (Williams et al, 2010; Dilworth-Anderson, 2011; Jefferson et al., 2011). Even though the 1993 NIH Revitalization Act, which required all NIH-supported research to report demographic data (Rice, 2008), little progress has been made (Geller et al., 2011). 

Current literature focuses on identifying factors, or barriers (Williams et al, 2010), that prevent individuals from participation (Williams et al., 2010; Wendler et al., 2006; Stood & Stahl, 2011), and yet, effective recruitment strategies and solutions are absent from current research (Stood & Stahl, 2011). 

The project goals were to (a) identify solutions from data collected by Williams et al. (2010) and (b) create a model for solution-based recruitment strategies.


Siekyeong Kim, M.D., Ph.D.

Visiting Scholar
Washington Univ. School of Medicine

Cortical Thickness and Head Size: Are they related?

Kim, S.1,2,10, Benzinger, T.1,4,9, Xiong, C.3, Morris, J.C.1,2,5-7, Hassenstab, J.1,2,8

1Charles F. and Joanne Knight Alzheimer’s Disease Research Center, 2Department of Neurology, 3Division of Biostatistics, 4Neurological Surgery, 5Program on Occupational Therapy 6Department of Pathology and Immunology, 7Program on Physical Therapy, 8Department of Psychology, and 9Department of Radiology, Washington University School of Medicine, St. Louis, MO, 10Department of Psychiatry, Chungbuk National University College of Medicine, Cheongju, Korea

Standardization for differences in head size is a routine procedure in volumetric neuroimaging studies. Surface-based statistics from programs like FreeSurfer increasingly are being used in studies of normal aging and neurodegenerative disease, where it remains an open question whether head size standardization is necessary when cortical thickness is an outcome variable. We assessed the relationship between head size and mean cortical thickness globally and in each lobe of the cortex, including the frontal, temporal, parietal, cingulate, and occipital lobes in participants with no dementia symptoms (Clinical Dementia Rating [CDR] 0, n = 201), very mild symptomatic Alzheimer’s disease (AD; CDR 0.5, n = 87), and mild symptomatic AD (CDR 1, n = 48). Across all participants, the thickness of the cingulate cortex decreased with increasing head size, after accounting for the effects of age, gender, comorbid conditions, and Apolipoprotein E genotype. In CDR 0 participants, there was an inverse relationship between head size and thickness across the entire cortex, and specifically in the frontal, temporal, and cingulate cortices. In the middle-aged and oldest CDR 0 participants, head size accounted for 21% and 23% of the explained variance in cortical thickness, respectively. There were no associations between gender and cortical thickness, regardless of CDR rating or age. All results were confirmed with vertex-wise analyses. Our findings suggest that studies of cortical thickness in normal aging and dementia populations should include standardization for age and should consider standardization for head size, particularly in studies with a focus on limbic and frontal regions.


Michael R. Nichols, Ph.D.

Associate Professor
University of Missouri, St. Louis

Neuroinflammation is a characteristic feature of the Alzheimer’s disease (AD) brain. Significant inflammatory markers such as activated microglia and cytokines can be found surrounding the extracellular senile plaques. Recent evidence suggests that the exterior of the plaques, which are primarily composed of insoluble amyloid-β protein (Aβ) fibrils also contain a measurable level of soluble Aβ oligomers. These and other observations led us to examine what specific form of Aβ is the most effective microglial proinflammatory mediator in vitro and to investigate some of the mechanisms involved in microglial activation by Aβ. Size exclusion chromatography purification of freshly reconstituted Aβ(1-42) in cell culture medium revealed classical 100 nm long curvilinear protofibrils which were very effective at stimulating TNFα secretion from primary murine microglia. In contrast to Aβ(1-42), Aβ(1-40) protofibrils took longer to form and were much less effective at stimulation of microglia. Neither of the Aβ protofibril preparations was toxic to microglia. Freshly-purified Aβ(1-42) or Aβ(1-40) monomer did not stimulate microglia, but surprisingly, neither did Aβ(1-42) fibrils even though they exhibited extensive thioflavin-T fluorescence compared to protofibrils. Confocal microscope images confirmed cell-surface binding of Aβ(1-42) protofibrils to BV-2 microglia while Aβ(1-42) monomers did not bind to the cells. Messenger RNA transcripts for a variety of innate immune mediators and products were investigated in primary microglia exposed to Aβ(1-42). Protofibrils significantly increased TNFα, IL-1β, Toll-like receptor 4 (TLR4) and CD47 mRNA levels while monomers had little effect. Previous work identified a role for TLRs in mediating Aβ-induced inflammatory. We utilized primary microglia from knockout mice to investigate the participation of the TLR adaptor protein MyD88 in the process.


Paul Barrett, Ph.D. Candidate

Graduate Student
Vanderbilt University

3D Structural Properties of the Amyloid Precursor Protein and the Etiology of Alzheimer's Disease

Paul J. Barrett1, Yuanli Song1, Wade D. Van Horn1, Eric J. Hustedt2, Johanna M. Schafer1, Arina Hadziselimovic1, Andrew J. Beel1, and Charles R. Sanders1

1Dept. of Biochemistry Center for Structural Biology, and Institute of Chemical Biology  2Dept. of Molecular Physiology and Biophysics Vanderbilt University School of Medicine, Nashville, TN 37232, USA.

The amyloid precursor protein (APP) is a transmembrane protein that is implicated in the onset of Alzheimer's disease (AD).  The biological role of APP is not well understood, leading to many questions concerning its role in AD pathogenesis.  One potential interaction that could impact AD onset is the binding of cholesterol by APP, as it has been shown that there is an epidemiological correlation between elevated levels of cholesterol and occurrence of AD.  It is known that APP is subject to alternative pathways of proteolytic processing; one pathway involves the enzymes α and γ secretase and produces benign fragments.  The other pathway involves the enzymes β and γ secretase and the resulting fragment, commonly known as the Aβ peptide, is believed to cause Alzheimer’s disease.  Cleavage of APP by β-secretase results in the formation of the transmembrane fragment referred to as C99 (or CTF-β), which is the subject of these studies.  We not only present the novel structure of C99 determined by Nuclear Magnetic Resonance (NMR) spectroscopy, but we also show a novel mechanism for binding cholesterol via an unanticipated cholesterol binding motif.


Yuanli Song, Ph.D.

Vanderbilt University Medical Center

Cholesterol binds to Amyloid Precursor Protein (APP) --Implication of a novel strategy to reduce Aβ peptides by targeting APP

Alzheimer’s disease (AD) is an irreversible, progressive brain disease that slowly affects memory, thinking, and behaves.  Although there are competing hypothesis, more and more evidence shows that the genesis of Aβ peptides is an early event. The toxicity of Aβ peptides triggers the development of AD. Aβ peptide is generated from a protein called amyloid precursor protein (APP) via an amyloidogenic pathway. Alternatively, APP can also be processed in a non-amyloidogenic pathway. Research has shown the amyloidogenic pathway is present in lipid rafts, membrane regions with more cholesterol, sphingomyelin, and ganglioside (GM1). The non-amyloidogenic pathway is outside of lipid rafts. Mounting evidence suggests that elevated levels of cholesterol promote the production of Aβ peptides. However, the mechanism of this effect is not yet clear. We hypothesize that cholesterol promotes APP processing via amyloidogenic pathway upon binding to APP. To prove the hypothesis, we applied solution NMR to detect the binding site and binding affinity of cholesterol to APP. Our results provide molecular information for a novel strategy to reduce Aβ production by targeting APP. If APP is kept outside of the lipid raft, APP should be processed in a non-amylodogenic pathway. The advantage of targeting APP to reduce Aβ production is that this avoids the severe side effects caused by general inhibition of β- or γ-secretase because of the broad substrate specificities of both β- and γ-secretases.


Norbert Seidler, Ph.D.

Kansas City University of Medicine and Biosciences

Proposed role of GAPDH pseudogene P44 in Alzheimer disease

Sara O. Mason, Norbert W. Seidler

Systematic analysis of GAPDHP44 (official symbol for a pseudogene of the conserved glycolytic enzyme, glyceraldehyde 3-phosphate dehydrogenase) was conducted with the aim of proposing mechanisms to explain the significance of a beneficial single nucleotide polymorphism (SNP) for Alzheimer disease that was identified (Li et al., PNAS 2004;101:15688) in this gene. There are over 60 GAPDH pseudogenes in the human genome, many of which are derived from retrotransposition resulting in an inserted DNA copy that does not contain parent gene intronic sequences. The GAPDHP44 pseudogene is located on chromosome 12 (specifically, 12q14.2) on the forward strand. We carefully examined the sequences in the putative coding region as well as upstream elements. While the literature does not contain evidence of its expression, studies have indicated the possibility of multiple isoforms of GAPDH and that GAPDH pseudogenes are preferentially spared disabling mutations. We observed that a TATA box exists 24 nucleotides upstream from a putative transcription start site. Additionally, the upstream elements include enhancer-like sequences that show homology to parent GAPDH gene intronic sequences. Furthermore, a Kozak consensus sequence was identified. The putative protein product would contain 139 amino acids due to the presence of a stop codon, and is surprisingly similar to the dinucleotide domain of 151 amino acid residues found in the parent gene product. The location of the beneficial SNP appears in a region (residues 80-120) that binds to GOSPEL (for, GAPDH’s competitor of Siah protein enhances life), preventing nitrosylated GAPDH from entering the nucleus and promoting neuronal apoptosis. By regulating GAPDH-GOSPEL interaction, we propose that the putative GAPDHP44 truncated protein from the beneficial allele may have an ameliorating effect on the pathogenesis of Alzheimer disease.


Amber Watts, Ph.D.

Assistan Professor
University of Kansas

Metabolic Syndrome (MetS) is a cluster of risk factors (i.e., abdominal obesity, hypertension, dyslipidemia, glucose and insulin dysregulation) that is associated with cardiovascular disease, diabetes, and dementia. Recent studies addressing the association of MetS with cognitive performance and risk for dementia report mixed results. An important step in clarifying these conflicting results is determining whether cognition is influenced by the effects of individual MetS components versus the additive or synergistic effects of multiple components. We assessed the effect of MetS on cognitive performance and decline over two years in 58 cases of early AD and 64 healthy older adult controls in the Brain Aging Project. Using factor analytic techniques, we compared the effect of a combined MetS factor to the effect of individual MetS components on change in attention, verbal memory, and mental status. In healthy controls, a combined MetS factor did not significantly predict cognitive performance, though individual MetS indicators, higher insulin and lower body mass index, predicted poorer cognitive performance outcomes. In the AD group, higher scores on a combined MetS factor predicted better cognitive outcomes. Evaluation of individual MetS indicators revealed that higher insulin predicted better cognitive outcomes, while higher blood pressure predicted poorer performance. We conclude that the cluster of biomarkers that make up metabolic syndrome is not equivalent in individuals with early AD and healthy older adult controls and it shows different relationships with cognitive outcomes in each. This suggests that the concept and definition of MetS should not be used interchangeably in older adults with and without AD.


Jason Ulrich, Ph.D.

Washington University School of Medicine

Dynamic in vivo measurement of apoE2, apoE3, and apoE4 levels and lipidation within the brain by microdialysis

Ulrich JD, Verghese PB, Cirrito JR, Holtzman DM

Alzheimer’s disease (AD) is characterized by the deposition of aggregated forms of amyloid  (A) plaques within the brain, which is hypothesized to be a critical instigator of synaptic loss and neurodegeneration. The strongest genetic determinant for the development of the most common form of AD, late onset AD, is possession of the apolipoprotein E 4 allele (apoE4). ApoE is a critical lipid transport protein in the CNS where it is expressed mainly by astrocytes. There are three common apoE variants (apoE2, apoE3, and apoE4) which differentially regulate A clearance from the CNS and A plaque formation. In addition to isoform-dependent effects on A metabolism, studies in mouse models of AD suggest that apoE expression level and the degree of apoE lipidation also influence A clearance and deposition. However, little is known about the physiological regulation of apoE within the CNS, particularly in regard to the regulation of the levels and lipidation of the extracellular pool of apoE. To directly assess apoE within the brain interstitial fluid (ISF), we performed in vivo microdialysis using a 1Million Da cut-off microdialysis probe (AtmosLM, Eicom) in awake, freely moving mice expressing apoE2, apoE3, or apoE4. We were able to dynamically measure ISF apoE on an hourly basis over the course of multiple days. Analysis of recovered ISF apoE by native gel electrophoresis indicated apoE particles were lipidated and were heterogeneous in size with lipoprotein size diameter of 8.1-17nm, similar to that of HDL-like particles isolated from mouse or human CSF. Analysis of ISF apoE levels and lipidation could provide novel insight into the mechanistic basis by which apoE influences AD onset and progression.


Kiran Yanamandra, Ph.D.

Washington University School of Medicine

Intracerebroventricular injection of anti-tau antibodies reduce tau pathology in P301S human tau transgenic mice

Kiran Yanamandra, Najla Kfoury, Hong Jiang, Tom Mahan, Marc I. Diamond, David M. Holtzman.
Department of Neurology, Hope Center for Neurological Disorders, Alzheimer’s Disease Research Center, Washington University School of Medicine, St. Louis, Missouri, 63110.

Tau aggregation occurs in several neurodegenerative diseases including Alzheimer’s disease (AD) and frontotemporal dementia (FTD) - termed tauopathies. Under disease conditions, tau aggregates in neurofibrillary tangles in neurons and in dystrophic neurites. The development of this pathology is associated with progressive neuronal and synaptic loss as well as cognitive decline. In the tauopathies, pathology progresses in a predictable manner, but the mechanism is unknown. One hypothesis is that tau aggregates propagate from one cell to another. We are testing this hypothesis by evaluating anti-tau antibodies that were selected for their activity in blocking trans-cellular propagation in cultured cells. These anti-tau antibodies also blocked the seeding effects of misfolded tau aggregates applied to cultured cells.
In transgenic mouse models of AD or Parkinson’s disease, passive immunization against Aβ or α-synuclein can reduce the amount of Aβ and α-synuclein aggregate accumulation. Additionally, recent passive immunization studies using monoclonal antibodies against different forms of tau reduced tau pathology, and in some cases improved behavior. In this study we used monoclonal antibodies that were selected for activity in cell-based assays that monitor trans-cellular propagation of misfolded tau. We chronically infused these anti-tau antibodies, control antibody, or vehicle into the lateral ventricle of P301S human tau transgenic mice (n = 16/group) by intracerebroventricular injection starting at 6 months of age via osmotic pump for 12 weeks. We performed both biochemical and histological analysis to assess the effects of these antibodies on tau pathology. Preliminary quantitative data from immunohistchemical staining shows that abnormally phosphorylated tau immunoreactivity assessed with the anti-tau antibody AT8 was strongly reduced in anti-tau antibody-treated mice compared to vehicle-treated mice. In cell culture assays, we have found that these anti-tau antibodies are not getting inside of cells. Biochemical analyses and behavioral tests are ongoing. Positive results would further support the concept that immunotherapy targeting tau may be a feasible approach in the treatment of tauopathies. In particular, the CNS infusion, perhaps via increased antibody availability in the CNS appears to be a very potent approach. Finally, the results implicate trans-cellular movement of tau aggregates in pathogenesis.


Brandon B. Holmes, B.A.

Washington University School of Medicine

Tauopathies are a class of neurodegenerative disorders characterized by the pathological accumulation of microtubule-associated protein tau in the human brain. Tau aggregate accumulation is observed in many diseases, such as Alzheimer’s disease and Frontotemporal Lobar Degeneration. In the early stage of disease, tau pathology is often restricted to discrete and stereotyped regions of the brain. With disease progression, however, the pathological changes typically spread through the nervous system according to specific anatomical patterns. Emerging evidence demonstrates that tau aggregates are capable of transcellular spread to co-cultured cells where they convert natively folded tau protein into an aggregated, fibrillar form. Thus, misfolded tau aggregates capable of transcellular spread may directly serve as an agent of disease progression. However, the mechanism by which tau aggregates enter cells to induce misfolding of native tau protein remains unknown. In the present research, we demonstrate that heparan sulfate proteoglycans mediate cellular tau fibril uptake both in vitro and in vivo. Biochemical, flow cytometry, and automated microscopy approaches revealed that harmacological and genetic inhibition of heparan sulfate proteoglycans suppressed tau fibril uptake in multipotent neural cell lines and murine primary cortical neurons. Additionally, inhibition of heparan sulfate proteoglycans led to a complete abolishment of the propagation of tau protein misfolding by inhibiting cellular entry of pathogenic tau seeds. Finally, we have
used heparin mimetics to inhibit the in vivo internalization of recombinant tau fibrils in the hippocampus of non-transgenic mice. Thus, the present study delineates a critical cellular pathway required for tau fibril entry into neuronal cells and demonstrates that inhibition of this pathway can abrogate the propagation of tau protein misfolding.


Sarah DeVos, B.S.

Graduate Student
Washington University School of Medicine

Alzheimer’s Disease (AD), the most common form of dementia, is characterized by two pathological hallmarks: tau neurofibrillary tangles (NFT) and amyloid-β (Aβ) plaques. In addition, it has also been noted that AD is a clinical risk factor for late onset seizures. Multiple AD mouse models recapitulate this increased seizure susceptibility and cognitive decline. Within the last 5 years, many of these AD models have been studied in the setting of mouse tau knockout (tau-/-). Both the increased seizure susceptibility and cognitive deficits were ameliorated in these amyloid-depositing lines when tau levels were reduced. Further, tau-/- alone interestingly appeared to be protective against chemically induced seizures. A similar knockout strategy, however, is unlikely to be practical in humans. To circumvent this, we propose using antisense oligonucleotide (ASO) technology to decrease endogenous tau expression levels in the adult mouse in an effort to study the effects on seizure susceptibility.


Shannon L. Macauley-Rambach, Ph.D.

Postdoctoral Research Scholar
Washington University School of Medicine

Understanding the link between diabetes and Alzheimer’s disease: the effects of hyperglycemia on amyloid-beta levels within brain interstitial fluid

S. L. Macauley1, R. Perez1, A.Q. Bauer2, T.E. Mahan1, J.P. Culver2, J. M. Lee1, and D. M. Holtzman1,3

1Department of Neurology and the Hope Center for Neurological Disorders, 2Department of Radiology, and 3Charles F. and Joanne Knight Alzheimer's Disease Research Center, Washington University School of Medicine, St. Louis, Missouri 63110;

Since cerebral glucose metabolism is tightly linked to neuronal activity and elevated neuronal activity increases Aβ production, we seek to understand how alterations in cerebral glucose regulate Aβ levels in the brain’s interstitial fluid (ISF). To study the acute effects of systemic hyperglycemia on ISF Aß levels, transgenic mice (APP/PS1) co-expressing the Swedish mutation in human amyloid precursor protein (APPswe) and a missense mutation in the human presenilin protein (PS1ΔE9) were simultaneously subjected to glucose clamps and in vivo microdialysis in both young and aged mice. This combination of techniques allows us to acutely manipulate systemic blood glucose levels in awake, freely moving mice while assessing changes in glucose, lactate, and Aß concentrations within the hippocampal interstitial fluid (ISF). First, we observed an age dependent decrease in ISF glucose by approximately 25% between 3 and 10 months of age, while there was an age dependent increase in ISF lactate of 33%. Next, by teaming hyperglycemic clamps and microdialysis in 3 mo. APP/PS1 mice, we were able to increase blood glucose levels from 78.3±8.3 to 156.2±6.3mg/dL, a 102% change from baseline, which resulted in a concurrent 58% increase in ISF glucose levels (3.2±0.16 to 5.0±0.22mg/dL). The 58% increase in ISF glucose concentration occurred concomitantly with with a 27% increase in hippocampal ISF Aß. Once the glucose clamp was removed, ISF glucose and Aß returned to baseline levels. Interestingly, in 14 mo. old APP/PS1 mice, a similar ~30% increase in ISF glucose and Aß occurred during the hyperglycemic clamp but in contrast to the younger mice, ISF Aß remained elevated by 34% once the clamp was removed and euglycemia was restored. This data suggests that increases in cerebral glucose significantly increase ISF Aß in both the young APP/PS1 mice without AD pathology as well as middle aged APP/PS1 mice with AD pathology. It also demonstrates that there is age or pathology-dependent changes in the energy metabolism within the brain that could further increase Aß levels. Taken together, our preliminary data suggests that chronic perturbation in glucose metabolism, via metabolic disorders such as T2DM, could result in altered Aß levels and subsequently impact the progression of AD.


Courtney Sutphen, B.S.

Graduate Student
Washington University School of Medicine

Longitudinal Changes in CSF Biomarkers of AD During Middle-Age in Cognitively Normal Individuals

Courtney Sutphen, Aarti R. Shah, Matthew Amos, Shengmei Ma, Chengjie Xiong, Xianyun Xie, Andrei G. Vlassenko, Tammie Benzinger, David M. Holtzman, John C. Morris, Anne M. Fagan

The Washington University Adult Children Study (ACS) is a longitudinal clinical and biomarker study of cognitively normal (Clinical Dementia Rating of 0), middle-aged individuals (age 45-74) with positive or negative family history of sporadic AD that is designed to investigate the time course of AD pathologic biomarker changes during the preclinical period. Levels of CSF Ab42, tau and ptau 181 at baseline and follow-up (mean ~3 years) were obtained by ELISA. Positive family history is defined as one or both parents diagnosed with AD dementia before age 80 whereas negative family history is defined as neither parent diagnosed with AD while living to be at least 70 years of age. We reasoned that AD biomarker changes would be more pronounced in individuals with positive family history.

Analysis of longitudinal CSF samples in the ACS cohort demonstrates annual decreases in Ab42 and increases in tau species beginning in middle-age, in some cases as early as early middle-age (45-54 yrs). Our data show that Ab42 typically begins deviating from the norm earlier in the disease trajectory than tau species. Comparison of these CSF changes with longitudinal amyloid imaging will shed light on the possible time course of and relation between the various pathologic markers during the preclinical period, though a larger sample size is needed to make firm conclusions.


Clarice Wang, M.A.

Graduate Student
The University of Kansas

Healthy younger adults (YA), healthy older adults (HOA), and adults with very mild AD (AD; CDR=0.5) were included in this study. We used a 3.0 Tesla head-only scanner to collect gradient echo blood oxygen level dependent (BOLD) scans during participants’ comprehension of two different prose genres, expository and narrative. All stories were binaurally presented. Identical stories were also presented in Korean, used as a baseline activation control for audition of similar speech sounds and patterns.

We found both a significant age and dementia effect dependent on prose genre. Comprehension of expository prose discriminated young adults from healthy older adults.
Comprehension of narrative prose discriminated healthy older adults from AD adults. Aging involves noncompensatory overrecruitment of cognitive control areas for long term memory and AD involves specific attentional deficits. This study further supports the sensitivity of a prose comprehension task as a diagnostic tool for AD.


Jee Hoon Roh, M.D., Ph.D.

Postdoctoral Fellow
Washington University School of Medicine

Amyloid-β (Aβ) aggregates in the brain years prior to the clinical onset of Alzheimer’s disease (AD). Prior to Aβ aggregation, levels of extracellular, soluble interstitial fluid (ISF) Aβ, which are regulated by neuronal activity and the sleep-wake cycle, correlate with the future amount of Aβ deposition in the brain. The amount and quality of sleep declines with aging and to a greater extent in AD. How sleep quality amount as well as the diurnal fluctuation in Aβ change with Aβ aggregation are not well understood. We report that there is a normal sleep-wake cycle and diurnal fluctuation of ISF Aβ present in different brain regions of APPswe/PS1δE9 mice before Aβ plaque formation. Following plaque formation, the sleep-wake cycle markedly deteriorates and diurnal fluctuation of ISF Aβ dissipates. Virtual elimination of Aβ deposits in the mouse brain by active immunization with Aβ42 completely normalized the sleep-wake cycle and the diurnal fluctuation of ISF Aβ. These data suggest that Aβ aggregation disrupts the sleep-wake cycle as well as the diurnal fluctuation of Aβ. Conversely, to investigate whether changes in the sleep-wake cycle affect Aβ accumulation in the mouse brain, we genetically modulated the neuropeptide orexin (hypocretin), which regulates arousal and wakefulness, in mouse models of β-amyloidosis. Preliminary results from experiments in APPPS1-21 in which the orexin gene is knocked out showed a decreased amount of Aβ and amyloid plaques in the brain. Overexpression of orexin via lentivirus in the bilateral hippocampus in APPswe/PS1δE9 mice from 5 to 9 months did not increase the amount of Aβ plaques in the brain. Taken together, our preliminary data suggest that the sleep-wake cycle and Aβ metabolism have a reciprocal relationship and modulation of orexin or secondary changes in sleep-wake cycle possibly affects Aβ aggregation and its downstream consequences in the brains of mice. Sleep-wake behavior and diurnal fluctuation of Aβ in the CNS appear to be functional and biochemical markers respectively of Aβ-associated pathology that should be explored in humans both diagnostically and in response to treatment.


Mauricio Garnier-Villarreal

Ph.D. Student, Research Assistant
University of Kansas, Center for Research Methods and Data Analysis

Differences in BADL, IADL and Cognitive Status in a Costa Rican Sample of older adults

M. Garnier-Villarreal ; D. Johnson; M. Salazar-Villanea

Basic Activities of Daily Living (ADL) are the most fundamental components of self-care (e.g., feeding ourselves, bathing, toileting, dressing, and grooming); the Instrumental Activities of Daily Living (IADL; Lawton and Brody, 1969; e.g., household chores, balancing a checkbook, navigation skills) were designed to tap a range of complex functional abilities that require reasoning, good judgment, and problem solving abilities.

The cumulative findings from a series of studies that contrasted ADLs versus IADLs show that (a) cognitive status did not contribute to ADL performance (Kempen & Ormel; 1998); (b) cognitive ability was differentially associated with IADL but not ADLs (McGuire et. al., 2006); (c) factor analysis of the determinants of independent living yielded a two-factor solution, one physical and the other cognitive (Ng et al., 2006); and (d) ADLs were predicted by physical mobility, whereas IADL was predicted by Intellectual disability (Hilgenkamp et al., 2011) .
After an analysis of longitudinal data from the Costa Rican Study of Longevity and Healthy Aging (CCP-CRELES), we found that the data collection process resulted in multiple reasons for nonresponse. From the older adults at the first wave of data collection, we found three different circumstances that happened at Wave 2: (a) a group of participants who were able to answer both waves (n = 1818), (b) a group of participants who were able to answer the first wave but were categorized as unable to answer the second wave (n = 206), and (c) a group of participants who were able to answer the first wave but were deceased by the second wave (n = 102). This research addresses whether there are differences between these three groups in their first wave data in BADL, IADL and Cognitive Status.


Jan Duchek, Ph.D.

Associate Professor
Washington University in St. Louis

Relationship Between Stroop Performance and Resting State Functional Connectivity in Cognitively Normal Older Adults

A Stroop task (along with standard neuropsychological measures) and rs-fcMRI were obtained in separate sessions in 237 cognitively normal older adults. We compared the relationship between Stroop performance, including measures from reaction distributional analyses, and composite scores from four previously defined resting state networks (RSNs) [default mode (DMN), salience (SAL), dorsal attention (DAN), and sensory motor networks (SMN)].

A larger Stroop effect in errors was associated with reduced functional connectivity within the DMN and SAL networks. Reaction time distributional analyses indicated that the slow tail of the reaction time distribution was related to reduced functional connectivity within the SAL. Standard psychometric measures (e.g., Logical Memory) were not related to composite scores from the RSNs. In addition, the relationship between Stroop performance and DMN connectivity (but not SAL connectivity) was stronger in CSF Aβ42 positive individuals.

These findings suggest a link between rs-fcMRI composite scores and specific attentional performance measures, both of which have recently been shown to be sensitive biomarkers for AD.


Nancy L. Baenziger, Ph.D.

Associate Professor
Washington University School of Medicine

Stress signaling networks on the path to neurodegeneration modeled in human cells.

Pooja Vijairania, Somi Kim, Grazie Mendonsa, Kalen Olson, and Nancy L. Baenziger.

Stress of many types is a risk factor for developing neurodegenerative disease such as AD. We model stress mechanisms in human skin fibroblasts that manifest cellular aging (WI-38) or Alzheimer's Disease (presenilin mutant), as well as in SHSY5Y human neuronal cells. Two stress responses that we model are:

Exaggerated inflammatory GPCR signaling
Cyclical neurite retraction and sprouting

The inflammatory neuropeptide bradykinin (BK) can be generated by traumatic brain injury, a risk factor for AD. The BK ligand/GPCR receptor system is highly expressed in AD-vulnerable brain regions. We model this inflammatory stress over the course of cellular aging in WI-38 fibroblasts, and show that AD fibroblasts show upregulation of an AD-related BK signaling pathway whereas normal fibroblasts do not.

Psychosocial stress, such as restraint, isolation, and inescapable abuse, is also a risk factor for AD. This type of stress can reorganize individual neuronal architecture and broad elements of neuronal connectivity. Under psychosocial stress some neurites retract while others sprout wildly as "lawless dendrites" apposing to non-synaptic regions. This profile is replicated in the neurodegeneration associated with AD. We model this process via "psychosocial stress in a culture dish," comprising tandem thermal stress of cooling and heating SHSY5Y human neuronal cells. This strategy provokes waves of neuronal retraction and subsequent excessive sprouting.

Both our inflammatory and psychosocial stress models offer significant translational implications, for identifying new biomarkers, drug targets, and coping mechanisms.


Magdalena Szumska, M.A.

Ph.D. Student
University of Kansas

Disclosure of Amyloid Status to Cognitively Normal Older Adults

Demonstration of brain accumulation of fibrillar amyloid beta protein via positron emission tomography (PET) with amyloid specific ligands may support the diagnosis of Alzheimer’s disease (AD). There is increasing recognition of the potential use of amyloid imaging to detect in vivo the pathology of AD in individuals with no ostensible cognitive impairment. Though clinical use of amyloid PET in cognitively normal patients is currently not justified, research use will be key to pursuit of therapies able to delay cognitive impairment and dementia due to AD.
In the proposed study, we will investigate the socio-emotional response of participants who have received the results of their research amyloid PET scans.

Before enrollment, each potential participant completes a pre-scan counseling session (modeled on genetic counseling best-practices) with trained clinicians to gain better health literacy for AD and understand the basic risk/susceptibility factors for AD. If consented, then clinicians will disclose the result of amyloid PET imaging tests, regardless whether it is positive or negative (n=400).

The socio-emotional response to the disclosed diagnosis will be assessed at baseline (pre-scan counseling visit) and reassessed at the immediate (post-scan), at 6-week, and at 6-month follow-up. The participants’ psychological well-being will be assessed using state-of-the art instruments such as: Impact of Event Scale, Beck Anxiety Inventories, Geriatric Depression Scale, Impact of Amyloid Testing in Alzheimer's Disease and Perceived Risk Assessment. Individuals who report symptoms over diagnostic thresholds for anxiety and depression, or express the desire to see a specialist will be referred to mental health specialist. This study has important implications for patients with prodromal AD and for AD prevention.


Tyler Blazey, B.S.

Research Assistant
Washington University School of Medicine

Elevated PiB precedes dementia in autosomal dominant: PIB, FDG and atrophy in the DIAN cohort

Blazey T (1), Benzinger TLS (1), Koeppe R (2), Jack CR (3), Raichle M (1), Su Y (1), Snyder A (1), Marcus D (1), Xie X (1), Ercole L (1), Stevenson T (1), Farrar A (1), Hornbeck R (1), Christensen J (1), Johnson K (8), Sperling R (8), Ringman, J (5), Villemagne, V (10), Thompson PM (5), Ghetti BF (6), Saykin AJ (6), Schofield P (9), Masters, CL (10), Rowe CC (10), Salloway S (7), Correia S (7), Fox NC (11), Brickman AM (12), Mayeux R (12), Martins R (13), Rimajova M (13), Mathis C (4), Klunk W (4), McDade E (4), Weiner M (14), Bateman R (1), Fagan A (1), Goate A (1), Cairns, N (1), Xiong C (1), Buckles V (1), Moulder K (1), and Morris JC (1) for the Dominantly Inherited Alzheimer Network

1) Washington University School of Medicine, 2) University of Michigan, 3) Mayo Clinic, 4) University of Pittsburgh, 5) University of California Los Angeles, 6) Indiana University, 7) Butler University, 8) Harvard University, 9) University of New South Wales, 10) University of Melbourne, 11) University College London, 12) Columbia University, 13) Edith Cowan University, 14) University of California San Francisco,

191 participants representing a mix of non-carrier and carriers in both the pre-symptomatic and symptomatic stages of AD underwent PiB, FDG PET and MRI. Estimated years to onset (EYO) was calculated as the difference between the participant's age at evaluation and the age at which parental cognitive decline began. All imaging exams were processed using a combination of in-house software and FreeSurfer 5.1. FreeSurfer regions of interest were applied to volumetric T1-weighted MRI, FDG and PiB data. For each modality, a linear mixed effects model was used to test for differences at specific EYO.

Differences in PiB binding between carriers and non-carriers start to become significant up to 15 years prior to symptom onset. The cross-sectional estimates of PiB accumulation rates show considerable between region variability, with the precuneus and the stratium showing the steepest trajectories. Ten years before the development of symptoms, carriers begin to exhibit volume loss in the hippocampus and the accumbens as well as hypometabolism in the precuenus/posterior cingulate and lateral parietal lobe. Significant levels of cortical atrophy are not found until 5 years before symptom onset.


Celeste Karch, Ph.D.

Postdoctoral Associate
Washington University School of Medicine

Novel rare variants in amyloid precursor protein alter beta-amyloid metabolism

Celeste M. Karch, Amanda T. Jeng, Carlos Cruchaga, Alison M. Goate

Alzheimer’s disease (AD) is characterized by the accumulation of extracellular amyloid plaques and intracellular neurofibrillary tangles in the brain. The inherited form of AD is associated with mutations in amyloid precursor protein (APP), presenilin 1, and presenilin 2. Mutations in these genes are reported to alter APP splicing and beta-amyloid production. The etiology of the sporadic form of AD, late onset AD (LOAD), is poorly understood. Recent sequencing efforts have focused on identifying rare variants in families in which LOAD is common. From a screen of 439 families with 4 or more individuals affected with AD, 6 novel variants were identified in APP. To determine if these novel APP variants alter APP metabolism by a mechanism similar to known pathogenic APP mutations, novel variants were introduced into vectors containing human APP by site-directed mutagenesis and the vectors were expressed in HEK293 cells.  Using immunoblotting and ELISA techniques, we measured intracellular APP, extracellular Aβ40 and Aβ42, and APP half-life.  The novel APP variants failed to produce changes in Aβ42 or the Aβ42/Aβ40 ratio to the same extent as the pathogenic mutations APP-SWE and APP-V717L. However, APP-N660Y expressing cells produced significantly more Aβ42 and Aβ40 than APP-wild-type (WT) expressing cells.  Interestingly, APP-G191E expressing cells produced significantly less Aβ42 and Aβ40 than APP-WT expressing cells.  APP-G322A, APP-E599K, and APP-A673T expressing cells also produced significantly less Aβ42, without altering Aβ40 levels, than APP-WT expressing cells.  Together, these findings illustrate that the rare variants in APP that occur in families with LOAD have complex effects on APP metabolism that may influence risk for AD when combined with other genetic or environmental risk factors.


If you have any questions, please send an email to