Ronald K. H. Liem, PhD - US grants

This is the second renewal of a training grant on Cellular and Neurobiological Aspects of Aging. The training program covers research in the departments of Pathology and Anatomy & Cell Biology, the Center for Neurobiology and Behavior, as well as the Alzheimer's Disease Research Center (ADRC). Many of the faculty, who are part of this training grant are also members of the ADRC, which is headed by Dr. Michael Shelanski. Although there is an emphasis on Alzheimer's Disease, it is not the only focus of this training program. The research interests of the participating faculty include the study of other neurodegenerative diseases, such as Parkinson's Disease, Alexander's Disease and Amyotropic Lateral Sclerosis, as well as cell biological problems, such as the mechanisms of cell death, the dynamics of the cytoskeleton, the cessation of the cell cycle and the action of growth factors and their receptors. One of the strengths of the training program is the wide variety of research interests of the participating faculty, which has enabled us to attract good candidates. We are asking for continued support for four pre-doctoral and four post- doctoral trainees. The pre-doctoral trainees will be chosen from the pool of graduate students who are being trained by the participating faculty and will be selected on the basis of their research interests and qualifications. A Unified Program involving all the graduate programs at Columbia University College of Physicians and Surgeons has recently been instituted. As Part of the Unified Program, all the first year students will be supported by the Office of Graduate Affairs. We are now able to choose among the students who already have selected their mentors. The pool of students available to the participating faculty is large and the qualifications of the students are excellent. In addition to paying for all the first year students, Columbia University is demonstrating their support of the training program by supplementing the salaries of the trainees in order for them to be able to meet the high cost of living in New York. Postdoctoral trainees apply directly to the laboratory of the participating faculty and subsequently apply to be supported by the training grant. The participating faculty members are all funded and have well equipped laboratories with all the research equipment necessary to do high quality cell and molecular biology.

We have recently shown that astrocytes require the glial-specific intermediate filament protein GFAP, in order to form stable processes in response to neurons by constitutive expression of anti-sense GFAP mRNA in the U251 glioma cell line. These experiments provided the first proof that intermediate filaments have tissue-specific functions as well as tissue-specific distributions. Although we have now shown that glial-specific intermediate filaments are necessary for the formation of stable astrocytic processes in response to neurons, they are clearly not sufficient. In order to study the mechanism by which glial processes are formed, we propose a series of experiments that are designed to investigate the mechanism(s) by which this protein can affect the shape changes of glial cells when neurons are added. For these studies, we will use molecular biological methods to determine whether the N-terminal head or C-terminal rod region of GFAP is the functional domain of the molecule. We have recently isolated fully encoding cDNAs for both GFAP and vimentin, the non-specific intermediate filament protein. In order to determine the specific domain of the GFAP molecule, we propose to finish sequencing these cDNAs and transfect the GFAP cDNAs cloned in eukaryotic expression vectors into fibroblasts. These studies will allow us to determine whether the protein will co-assemble with the endogenous vimentin intermediate filaments. We will also transfect these cDNAs into the previously prepared anti-sense GFAP cell lines, to show that we can rescue the ability of these cells to respond to neurons by forming astrocytic processes. However, to be able to pinpoint the functional domain of GFAP, we will have to prepare additional anti-sense GFAP cell lines transfected only with either the 5' or 3' end of the cDNAs, prepare vimentin-GFAP hybrid constructs, and determine which hybrid molecule can restore the ability of the glial cells to respond to neurons by the formation of stable astrocytic processes. We expect to be able to identity the important sequence(s) in the GFAP molecule by deletion and mutagenesis studies. Ultimately, we hope that we can delineate the series of events that occur to cause the formation of astrocytic processes when neurons are added to glial cells, as well as in the developing nervous system.

The two major pathological hallmarks of Alzheimer's disease are the neurofibrillary tangles and the senile plaques. Senile plaques consist of extracellular amyloid fibrils, composed of the beta-amyloid peptide, a proteolytic fragment of the beta-amyloid precursor protein surrounded by dystrophic neurites, activated microglia and astrocytes. Other proteins are also found in the AD plaques, including alpha1-antichymotrypsin and apolipoprotein E. The neurofibrillary tangles in neuronal cell bodies are composed of paired helical filaments (PHF). A large number of studies have now shown that PHFs are made up of the microtubule associated protein tau, which is abnormally phosphorylated. Recent studies have shown that lys-ser- pro (KSP) sequence on tau are among the sequences abnormally phosphorylated in Alzheimer's Disease. The high molecular weight neurofilament protein, NF-H is phosphorylated on similar KSP consensus sequences. This sequence is present more than fifty times in the tail of the NF-H molecule and most of these sites are normally phosphorylated in vivo. Recent studies have shown that cdk5, a kinase related to the cell cycle dependent kinase cdc2, is expressed in the brain and associates with neurofilaments, as well as microtubules. This kinase phosphorylates some, but not all of the KSP sites of NF-H and is also able to phosphorylate tau on some of the sites abnormally phosphorylated in AD. This proposal focuses on the specific function of this kinase in the nervous system and how it may relate to the abnormal phosphorylation of tau in AD. In addition, we will attempt to isolate other kinases which phosphorylate NF-H on the remaining KSP sites, and which may also act abnormally on tau in Alzheimer's Disease. The aims of this proposal are: 1. To study the effects of overexpression of cdk5 on the phosphorylation of NFH and tau in vivo by introducing its cDNAs cloned in a neuronal expression vector in transgenic mice. 2. To determine the effect of inhibition of cdk5 in transgenic mice by mutating the cdk5 cDNA clone in its active site t produce an inactive kinase, which will inhibit the endogenous ckd5. This mutant kinase will be introduced into transgenic mice with a neuron specific expression vector and we will determine the effect of the inhibition of phosphorylation of NF-H and tau. 3. To isolate other kinases which phosphorylate NF-H on the remaining KSP sites by protein chemical methods, as well as the yeast two-hybrid system.

DESCRIPTION This proposal builds on observations made during the current grant period that the over-expression of the neuronal intermediate filament protein alpha-internexin causes the formation of axonal swellings in transgenic mice, as well as observations made with GFP-intermediate filament protein chimeras which suggest an interaction between intermediate filaments and the microtubule cytoskeleton. It also builds on the recent characterization of BPAG1, which has both an intermediate filament-binding domain and an actin-binding domain. The emphasis of the current proposal is to examine proteins which interact with neuronal intermediate filaments, to identify motor proteins involved in the transport of neuronal intermediate filaments, and to make further progress in understanding the function and regulation of the neuronal intermediate filament protein a-internexin. The specific aims of this study are: (1) to study how members of the plakin family interact with neuronal intermediate filaments; (2) to study the mechanism of the transport of neuronal intermediate filaments in the axon; and (3) to continue studies on the function of alpha-internexin by generating a knockout mouse model and characterizing the resulting phenotype.

DESCRIPTION (provided by applicant): The Integrated Program in Cellular, Molecular and Biophysical Studies at Columbia University College of Physicians and Surgeons is a Ph.D. granting program that combines faculty from all seven basic science departments at the Medical Center. The program started out as the interdisciplinary nature of biomedical research became more obvious and the need for interdepartmental "core courses" was perceived specifically to educate graduate students in the basics of biochemistry, molecular biology, genetics and cell biology. The Integrated Program was formally established as a degree-granting program in 1986 and the program was funded by NIGMS in 1987. The program has a distinguished, well-funded faculty of 98 trainers, whose research expertise represents nearly all the areas of modern cellular and molecular biology and molecular biophysics. There are currently 67 students in the Program. Students take four core courses and a course in the Responsible Conduct of Research in the first year and complete three laboratory rotations. At the end of the first year, they choose their thesis mentor and take a qualifying examination. In subsequent years, the students take three elective courses and complete their thesis research. The Program sponsors a "Frontiers in Cell Biology" seminar series, a student research seminar series and a biennial student-faculty retreat. Most students complete the Program in 5-6 years. Fifty students have graduated from the Program in the past five years alone and nearly all have gone on to post-doctoral positions in outstanding laboratories. Based on the continued growth in the pool of training grant eligible applicants to this Program and the size and strength of the Training Faculty, we request support for 15 students each year.

DESCRIPTION (provided by applicant): MACF (Microtubule actin crosslinking factor) is a large protein (608 kDa) that can associate with the actin and microtubule networks. MACF has a complex domain structure. The N-terminal part of MACF is made up of a calponin-type actin-binding domain and a plakin-like domain. This domain is common to the plakin family of proteins that has been shown to connect cytoskeletal elements to each other and to the junctional complexes at the plasma membrane. The plakin-like domain is followed by a central rod domain composed of spectrin repeats and calmodulinlike EF hand motifs, similar to members of the spectrin superfamily. Finally, the C-terminal domain contains a novel microtubule-binding domain. The N-terminal domain exhibits a striking homology to a proposed neuronal isoform of BPAG1, the mutated gene product of the mouse mutant. dystonia musculorum (dt). The rod domain and the EF hand motifs of MACF are highly homologous to dystrophin. MACF is ubiquitously expressed in the mouse embryo with high expression levels in the nervous system and moderate expression levels in muscles. Kakapo, also known as short stop, is the Drosophila homologue of MACF and is essential for neuronal growth and adhesion between and within cell layers. Mutations in the kakapo/short stop gene in Drosophila cause defects in muscle-tendon cell differentiation, local development of neuronal processes and axonal outgrowth. The properties of kakapo/short stop make MACF a potential key player in axonal outgrowth and we therefore propose to study the function and interaction partners of MACF in more detail. In this proposal, we will determine the expression pattern and localization of MACF: antibodies against MACF will be raised and used to study the expression pattern of MACF. We will investigate loss-of-function phenotypes in C. elegans with type-specific GFP-labeled neurons using double-stranded RNA interference to a C. elegans MACF homologue and we will carry out a series of rescue experiments and dominant negative experiments on cultured dt and wild type neurons. We will define the association partners of MACF: the microtubule-binding domain at the C-terminus of MACF will be dissected in vivo by transfection studies and in vitro by microtubule-binding assays. Other association partners to different domains of MACF will be isolated by using the yeast two-hybrid system and by co-immunoprecipitation experiments. Finally, we will characterize a novel splice variant of MACF.

DESCRIPTION (provided by applicant): MACF (Microtubule actin crosslinking factor) is a large protein (608 kDa) that can associate with the actin and microtubule networks. MACF has a complex domain structure. The N-terminal part of MACF is made up of a calponin-type actin-binding domain and a plakin-like domain. This domain is common to the plakin family of proteins that has been shown to connect cytoskeletal elements to each other and to the junctional complexes at the plasma membrane. The plakin-like domain is followed by a central rod domain composed of spectrin repeats and calmodulinlike EF hand motifs, similar to members of the spectrin superfamily. Finally, the C-terminal domain contains a novel microtubule-binding domain. The N-terminal domain exhibits a striking homology to a proposed neuronal isoform of BPAG1, the mutated gene product of the mouse mutant. dystonia musculorum (dt). The rod domain and the EF hand motifs of MACF are highly homologous to dystrophin. MACF is ubiquitously expressed in the mouse embryo with high expression levels in the nervous system and moderate expression levels in muscles. Kakapo, also known as short stop, is the Drosophila homologue of MACF and is essential for neuronal growth and adhesion between and within cell layers. Mutations in the kakapo/short stop gene in Drosophila cause defects in muscle-tendon cell differentiation, local development of neuronal processes and axonal outgrowth. The properties of kakapo/short stop make MACF a potential key player in axonal outgrowth and we therefore propose to study the function and interaction partners of MACF in more detail. In this proposal, we will determine the expression pattern and localization of MACF: antibodies against MACF will be raised and used to study the expression pattern of MACF. We will investigate loss-of-function phenotypes in C. elegans with type-specific GFP-labeled neurons using double-stranded RNA interference to a C. elegans MACF homologue and we will carry out a series of rescue experiments and dominant negative experiments on cultured dt and wild type neurons. We will define the association partners of MACF: the microtubule-binding domain at the C-terminus of MACF will be dissected in vivo by transfection studies and in vitro by microtubule-binding assays. Other association partners to different domains of MACF will be isolated by using the yeast two-hybrid system and by co-immunoprecipitation experiments. Finally, we will characterize a novel splice variant of MACF.

DESCRIPTION (provided by applicant): The mutant mouse dystonia musculorum (dt) suffers from a severe hereditary sensory neuropathy. The mice display progressive loss of limb coordination starting in the second week of life. Dorsal root ganglia of dt mice are considerably smaller in size than those of wild type mice with organelles and empty vacuoles accumulating within sensory axons. Focal axonal swellings filled with neurofilaments, mitochondria and membrane bound dense bodies are hallmarks of the pathology of these mice. The disorganization of the cytoskeleton precedes neurodegeneration in the mutant mice. The gene that is mutated in dt mice is called dystonin, and is also known as bullous pemphigoid antigen 1 (BPAG1). BPAG1 is a member of the plakin family of cytoskeletal linker proteins. Plakin family members have a characteristic plakin domain that can interact with a variety of cell adhesion molecules, as well as members of the armadillo-repeat domain containing protein family (catenin family). In addition, various plakins have domains that can bind to actin microfilaments, microtubules and intermediate filaments. In epithelial cells, BPAG1-e is involved in anchoring keratin intermediate filaments to the hemidesmosomes. Why does the deletion of a hemidesmosomal protein cause neuronal degeneration? To investigate this question, we have analyzed the BPAG1 locus in detail. We found that by tissue specific splicing, it encodes a variety of isoforms with different combinations of the various interacting domains. There are two isoforms expressed in the nervous system: the major isoform, BPAG1-a and BPAG1-n, which is expressed at lower levels than BPAG1-a. BPAG1-b is the largest isoform and is highly expressed in muscles. Alternatively spliced 5' and 3' ends that affect the ABD and MTBD respectively can lead to additional isoforms of BPAG1-a, BPAG1-b and BPAG1-n. In this proposal, we will study the function of the neuronal isoforms in more detail to understand why mutations of this protein result in neuronal degeneration. This proposal focuses on the cause for the mouse mutant dystonia musculorum (dt) and is therefore in the realm of the Program Announcement initiative PA- 02-156: "Studies into the causes and mechanisms of dystonia." In particular, the proposal meets the criteria of the second research area targeted by the PA: "Identification of proteins that interact with dystonia-related cellular factors (genes, proteins) and determination of their coordinated function."

DESCRIPTION (provided by applicant): This application is designed to identify compounds to treat Charcot-Marie-Tooth type 2E (CMT2E). CMT is the most commonly inherited neurological disorder with a reported prevalence of 1 in 2,500 people worldwide. It is found in all races and ethnic groups. CMT is slowly progressive and CMT patients suffer from degeneration of the peripheral nerves that control sensory information of the foot/leg and hand/arm. The nerve degeneration causes the subsequent degeneration of the muscles in the extremities. Among the symptoms of CMT are foot-drop, steppage gait, high arches, foot bone abnormalities, and basic problems with hand function, as well as sometimes breathing difficulties. CMT is not usually life threatening, but can cause severe disabilities. Although the genes mutated in CMT are also expressed in the central nervous system, the disorder almost never affects brain function. CMT is divided in two major types, CMT1 and CMT2, based on nerve conduction velocities, which are reduced in CMT1 and relatively normal in CMT2. In general, CMT1 is a demyelinating neuropathy caused by mutations in genes important in myelin formation, whereas CMT2 is axonal. Mutations in the neuronal intermediate filament gene, NEFL have been shown to cause a subtype of CMT2, called CMT2E. NEFL encodes the neurofilament light (NFL) protein that we have shown to be a necessary component for the assembly of neuronal intermediate filaments. Mutations in this gene are responsible for approximately 2% of all CMT cases. Neuronal intermediate filaments form the intermediate filament network in neurons and are the predominant cytoskeletal structure in the axon. Neurofilamentous aggregates both in the neuronal cell bodies and axons are seen in patients with mutations in NEFL, as well as in other neurodegenerative diseases, such as amyotrophic lateral sclerosis. We have shown that the mutant NFL proteins form aggregates in transfected neuronal and non-neuronal cells. Misassembled neurofilament aggregates are found in all transfected cells expressing the pathogenic CMT-associated NFL mutant proteins, but not in cells expressing several polymorphic variants that are non-pathogenic. We hypothesize that inhibitors of neurofilament misassembly will lead to therapies for CMT. Therefore, the goals of this proposal are to identify small molecules that inhibit misassembly and to test their effects in a mouse model of CMT2E. We will focus our attention on two of the first described mutant NFL proteins, P8R NFL and Q333P NFL that we have characterized in the most detail. Using high-throughput visual screening methodology, we will identify small molecules that inhibit neurofilament misassembly. For the second and related part of the project, we will generate knock-in mouse models of CMT2E with mutations in the Nefl gene. These mice will be characterized phenotypically and we will then test compounds identified in the visual screens for their ability to ameliorate peripheral neuropathy in these mouse models. The proposed research will identify lead compounds for the development of drugs to treat human subjects with CMT, as well as potentially other neurodegenerative diseases and it will also generate mouse models of a human hereditary neuropathy that will be of value to many other investigators in the field. PUBLIC HEALTH RELEVANCE: CMT is the most commonly inherited form of peripheral neuropathy. The goal of this project is to use chemical screening to identify novel drugs to treat one type of CMT and test them in mouse models of the disease.

DESCRIPTION (provided by applicant): The Integrated Program in Cellular, Molecular and Biomedical Studies (CMBS) at Columbia University Medical Center is a Ph.D. granting program that combines faculty from all the basic science departments. The CMBS Program is an umbrella program that presents students with a unique opportunity to obtain individualized training in basic cell and molecular biology, microbiology, structural biology, biophysics, genetics, immunology, neurobiology, computational biology, as well as translational biomedical disease-related research. Our hope is to train the next leaders in the field of biomedical research and also to provide training for future leaders in other areas where a biomedical research background will be of great benefit. The CMBS program is an accredited degree-granting program that was first established in 1986 and has been supported by this Training Grant since 1987. The program has a distinguished, well- funded faculty of 135 trainers, whose research expertise represents nearly all the areas of modern cellular and molecular biology, neurobiology and computational biology. There are currently 69 students in this program. Seventy two students have graduated from the CMBS Program in the past five years and have gone on to postdoctoral positions in outstanding laboratories, careers in the pharmaceutical of biotechnology industry, or careers where they use their biomedical training to provide other societal benefits. Students take core courses in molecular genetics, molecular and cell biology as well as statistics during their first year and complete three laboratory rotations. In the secod year, students take their qualifying examination and a course in the Responsible Conduct of Research. The CMBS Program hosts a student research seminar series and a biennial student-faculty retreat. Most students graduated in 5-6 years. The CMBS Program remains the premier cellular and molecular biology graduate program at Columbia University Medical Center and support from this Training Grant is crucial for the continued success of this program. PUBLIC HEALTH RELEVANCE: The goal of this program is to provide students with a unique opportunity to obtain individualized training in basic cell and molecular biology, microbiology, structural biology, biophysics, genetics, immunology, neurobiology, computational biology, as well as translational biomedical disease-related research.

DESCRIPTION (provided by applicant): Defects in the endolysosomal system are increasingly viewed as key pathological features of neurodegenerative disorders, such as Alzheimer's disease (AD) and Parkinson's disease. An emerging hypothesis is that chronic endolysosomal defects occurring in these disorders compromise the degradative capacity of lysosomes, causing the aberrant accumulation of a variety of lysosomal cargoes that are targeted to these organelles generally through the endocytic or autophagy pathway. These cargoes not only include a range of aggregate-prone proteins or peptides (e.g., Abeta, aberrant alpha-synuclein and tau), but also lipids, such as cholesterol and sphingolipids. Recently, our lab has employed a systems-based approach called lipidomics to profile hundreds of lipids from healthy and diseased tissue using state-of-the-art mass spectrometry. With this technology, we identified a striking lipid alteration in vulnerable brain regions from patients with AD, in a mouse model of Niemann-Pick type C (NPC), an aggressive lysosomal storage disorder that shares some pathogenic processes in common with AD, as well as in other instances where apolipoprotein-derived cholesterol is aberrantly accumulating. This lipid is called lysobisphosphatidic acid (LBPA) or bis(monoacylglycero) phosphate, an atypical phospholipid that is specifically enriched in the multivesicular endosomes and lysosomes, where it is further concentrated on cholesterol-rich intralumenal vesicles (ILVs). LBPA has been previously suggested to promote the formation of ILVs in the endolysosomal compartment, to regulate the storage and distribution of apolipoprotein-derived cholesterol, and facilitate lysosomal degradation by stimulating hydrolases. Together with these studies, our lipidomic data not only identify LBPA as a candidate biomarker for disorders associated with endolysosomal dysfunction, but they also suggest a critical role for this phospholipid both in the physiology and the pathophysiology of these organelles as well as in the administration of apolipoprotein-derived cholesterol. Unfortunately, tools are currently lacking to understand the precise (patho)physiological roles of this endolysosomal lipid and manipulate its levels to assess its therapeutic potential. The primary reason behind this roadblock is that the enzymes mediating the synthesis and degradation of this elusive phospholipid are unknown. The main goals of this proposal are thus (i) to identify the LBPA-metabolizing enzymes and more generally, the genes positively or negatively regulating the levels of LBPA using a combination of genome-wide RNAi screen and lipidomics; and (ii) to assess the impact of various types of LBPA manipulations on lysosomal function in normal neurons as well as in neurons derived from Npc1 mutant mice. We anticipate that our studies will be critical to understand the role of LBPA in endolysosomal physiology with potentially major implications for disorders associated with lysosomal dysfunction.

DESCRIPTION (provided by applicant): Intracellular signaling lipids control a large variety of cellular processes, including membrane trafficking, cytoskeletal dynamics, transport across membranes and signal transduction. Not surprisingly, lipid signaling and alterations thereof are increasingly linked to human disease. Alzheimer's disease (AD) is one such disorder in which lipid dyshomeostasis and membrane trafficking defects are believed to play a critical role. This concept is easily reconciled with the fact that the main molecular players in AD, including amyloid precursor protein (APP) and the ?-, ß- and ?-secretases, are all transmembrane proteins (or protein complexes) that traffic in cells and exert their functions at or within cellulr membranes. Our working hypothesis is that specific lipid changes may drive or mediate fundamental aspects of AD pathogenesis. Systems-based approaches, such as lipidomics, are emerging as a powerful tool to profile cells, tissues or organisms in a diseased state, providing both an unbiased and comprehensive picture of lipid alterations potentially linked to pathogenicity. To better understand the link between lipid signaling defects and AD pathogenesis, we have recently conducted a lipidomic analysis of brain samples derived from three transgenic animal models of familial AD as well as three independent brain regions from patients with late-onset AD. We found that out of 330 lipid species analyzed, only one lipid species was significantly reduced in AD-affected brain regions in mice (forebrain) and men (entorhinal and prefrontal cortex): phosphatidylinositol-3-phosphate (PI3P). PI3P is a phosphoinositide primarily synthesized by lipid kinase Vps34 and acts as a master regulator of the endosomal and autophagy pathways. PI3P controls the recruitment of a variety of compartment-specific effectors harboring PI3P binding modules, such as FYVE or PX domains. We found that knocking down/out Vps34 recapitulates salient features linked to AD pathogenesis, namely (i) enlarged endosomes; (ii) aberrant endosomal trafficking and processing of the amyloid precursor protein (APP); and (iii) accumulation of autophagy substrates. Additionally, work from others shows that chronic lack of Vps34 in neurons produces neurodegeneration. Altogether, our results have identified PI3P deficiency as a key factor in AD pathogenesis. This proposal focuses on addressing the consequences of disrupting PI3P signaling on two processes that emerge as critical in AD pathogenesis, namely the endosomal trafficking and processing of APP (Aim 1) and neuronal autophagy (Aim 2). It will also assess the impact of PI3P deficiency on the Aß and Tau pathologies in vivo (Aim 3). We anticipate that our studies will provide key insights into the biology of APP and Tau as well as a better understanding of the role of lipid dysregulation in AD pathogenesis.