Funding opportunities

High throughput modeling of human neurodegenerative diseases in embryonic stem cells

Funding Type: 
New Faculty I
Grant Number: 
Funds requested: 
$2 257 218
Funding Recommendations: 
Not recommended
Grant approved: 
Public Abstract: 
An important class of neurological diseases predominantly affects spinal motor neurons, the neurons that control muscle movement. The most well known of these motor neuronopathies is Amyotrophic Lateral Sclerosis (ALS), commonly referred to as Lou Gehrig’s disease for the famous Yankee first baseman who died of the disease. The first symptoms of ALS are usually increasing difficulty walking or speaking clearly. People with ALS progressively lose their ability to initate and control movements, and may become totally paralyzed during the late stages of the disease. There are no cures or effective treatments for these diseases. Riluzole (Rilutek), the only FDA approved medication for ALS, only modestly slows disease progression. Consequently, ALS is usually fatal within one to five years from onset, with half dying within eighteen months. Although genetic studies have identified many mutations that cause these diseases, it is not understood why these mutations kill motor neurons. This lack of understanding about the root causes of motor neuron diseases currently hinders the development of effective treatments. We seek to study motor neurons carrying these mutations in cell culture dishes to understand how these diseases sicken and kill these cells. To generate these motor neurons, we will use embryonic stem cells. Embryonic stem cells can become any cell in our body, including motor neurons. We have developed a new technology that allows us to quickly replace healthy genes with mutant genes in mouse embryonic stem cells. We will use this technology to insert both normal and disease-associated versions of genes into embryonic stem cells. Study of the healthy and mutant mutant motor neurons derived from these embryonic stem cells will shed light on the ways in which the mutations cause harm. The development of cell based models of human diseases is likely to have additional benefits as well. For example, diseased motor neurons grown in cell culture dishes can be quickly and efficiently screened with potential drugs to discover agents that slow, halt or reverse the cellular damage. It is our hope that these experiments will both deepen our understanding of important neurodegenerative disorders, and lead to new directions for the development of effective therapies.
Statement of Benefit to California: 
Over 6,000 Americans are diagnosed each year with motor neuronopathies, about the same as are diagnosed with multiple sclerosis. One form of this illness, ALS, is responsible for about one in every 800 deaths, and cause many lengthy and costly hospital admissions. We propose using stem cells to model these diseases so that we can gain a deeper understanding of their root causes. It is our expectation that this deeper understanding will lead to new and better approaches to the treatment of these disorders. In addition, our technology for developing embryonic stem cell-based models of human diseases is likely to have applications in the biotechnology sector. Although our technology is most applicable for modeling simple dominant genetic diseases, it can be adapted to model recessive and complex disorders. Beyond increasing our understanding of human diseases, these cellular models represent useful screening tools for testing novel pharmacological treatments. Identification and development of these new therapies may support new companies or new products for existing companies. We hope that using stem cells to model neurodegenerative disorders will lead to progress in the fight against these diseases, as well as provide the tools and examples for those in academia and industry who hope to create stem cell models of other clinically important disorders.
Review Summary: 
SYNOPSIS: The applicant proposes to establish a high throughput system that models human neuronopathies in mouse ES cells, with possible extension to human ES cells. The applicant takes advantage of the gene trap screens he had helped established in mouse ES lines. Reiter will use a somewhat high throughput method (Floxin technology) to create knock-ins in human neurodegenerative disease genes in mouse embryonic stem cells. He will use Cre-mediated recombination to insert human mutant disease gene cDNAs into gene trap insertions and then study the physiology of neurons in vitro in differentiated ES cells. The proposal is organized into four aims: Aim 1. Generate ES cell models of human neuronopathies. He will use gene traps that contain Lox sites that allow reversion of, and subsequent modification using the Floxin system. He will then replace the orthologous mouse loci with the human wild-type and disease alleles for selected autosomal dominant motor neuronopathies. Aim 2: Explore the genetic and cell biological bases of neurodegeneration. He will generate motor neurons in vitro from each of the human disease allele-bearing ES cell lines and compare disease allele-bearing neurons and wild-type neurons in assays of the four most widely studied cellular properties associated with neurodegeneration: oxidative stress, neurofilament disorganization, unfolded protein aggregate toxicity, and glutamate excitotoxicity. Aim 3: Define the motor neuron interactomes for wild-type and disease-associated gene products. Systematic identification of both the subcellular localization and protein interaction of each wild-type and disease tagged protein will be carried out. Aim 4: Validate findings in human ES cell models. He will express mutant alleles in motor neurons derived from human ES cells. This will not be done by gene targeting, so the transgene expressed proteins must function dominantly over the endogenous protein expression. STRENGTHS AND WEAKNESSES OF THE RESEARCH PLAN: The proposal involves the modeling of motor neuronopathies by ESCs. The hypothesis is that ESCs that carry mutations responsible for these neurodegenerative diseases display phenotypes. The PI will use Floxin - Flanked Lox Site Insertion - in which ESCs have a single gene trap in an identified locus - in order to carry out high throughput modification of mammalian genes. Unlike traditional transgenes, alleles are expressed in their appropriate chromosomal context under the control of their endogenous regulatory elements. The method is thought to be much more rapid compared to methods involving homologous recombination of ES cell lines, and has not been widely used in mouse ES cells..The PI plans to use this system to create models of motor neuronopathies in ES cell lines. The PI will focus initially on four genes that are mutated in motor neuronopathies: Rab7, Bscl2, Dynactin-1, and Gars. In total, he plans to make twenty three different cell lines, each expressing a loss-of-function gene trap allele, a revertant allele, or a human wild-type or missense allele. The PI presents preliminary data in which the Floxin system was used to mutate Ofd1, the gene mutated in oral-facial-digital syndrome 1. In general, the proposal has some great strenghts. One valuable direction with respect to ESC research is the use of stem cells as tools to better understand neurodegenerative disease. This proposal has that as its goal. There are exciting Preliminary Data demonstrating the feasibility of the Floxin approach which may be of value in the proposed experiments as well as other studies noted by the PI (e.g., insertion of Cre, tTA, GFP; development of transgenic mice). The PI's new method is a creative one that appears to provide a valuable and innovative technique. The research plan, namely in the first 3 aims, is quite well elaborated and clearly well thought through. The PI is aware of potential limitations to his approach, namely the ability to recapitulate a human phenotype on a mouse ES cell development background, with in vitro readouts. Overall, the general concerns could not dampen the enthusiasm for the proposal. The planned studies related to mutliple alleles of 4 genes sounds a bit overly ambitious (and too much like a catalogue), especially if one envisions phenotypic characterization of all the mutant ESCs for all of the parameters that the PI describes. This characterization is further complicated by the fact that these are new techniques for the PI and require a number of collaborations. One has some concerns as to whether these collaborations will really be productive ones. There are additional issues related to the possibility that no phenotype will be found in the MNs (e.g., because of non-cell autonomous degeneration). Despite these concerns, the PI has an impressive track record and one suspects that at least one or more of the mutant genes will indeed have a phenotype. This proposal will support the entry of this talented and productive individual into the SC field with a focus on neurodegenerative disease. The PI's scientific environment is outstanding. Strenghts and weaknesses for specific aims are detailed below. In specific aim 1, the PI will use the Floxin system to knock-in ~20 different human mutant alleles into the orthologous mouse loci of mouse ES cell lines. The PI does not anticipate problems in carrying out this specific aim. The Preliminary Data related to the use of the Floxin system in the case of the Ofd1 locus provides reasonable assurance that this aim can be carried out successfully. However, he has chosen a list of genes for which appropriate gene trap insertions exist, and expects to make more than twenty different cell models. Also, there was no mention of whether any of these genes normally have alternative splicing, and thus what the impact of only expressing one splice form might be. Although the basic approach in mouse ES cells is straightforward, there are a number of critical considerations that were not mentioned. For example, in the long run the modified ES cells will be most useful if they can be used to generate mouse models. However, the cells will go through 3 subclonings, thus what percentage of modified ES cells will go through the germline? Since germline transmission is the most costly part of an ES cell experiment, the efficiency of this after 3 subclonings is critical. Further, since the endogenous 3’ UTR is not included in the knock-in allele, protein level expression from the modified allele could be different than from the normal allele, and when the first exon is coding, the cloning will be more complicated for making the insertion vector. If there are internal promoters, this could interfere with interpretation of the results of the experiment. Attempts to insert Floxin constructs have succeeded at rates of 30-50%. This is not any higher than many gene targeting experiments. In addition, the approach only allows only one splice form of a gene to be expressed, not all genes or introns of a gene are accessible to gene trap instertions, and in vitro studies of neurons will be limited to the kinds of neurons that can be induced, and do not include the critical process of establishment of functional circuits. In specific aim 2, the PI will generate motor neurons (MNs) in vitro from the human disease allele-bearing ESCs. These MNs will be characterized re. oxidative stress, neurofilament organization, unfolded protein aggregate toxicity, and glutamate excitotoxicity in an effort to clarify the mechanisms by which the mutations lead to MN death. He will be assisted in these experiments by Drs. Ferriero, Papa, and Diamond. The PI notes that by comparing the phenotype seen in the MNs expressing human mutant alleles with that found in MNs derived from the gene trap ESCs (which are a loss of function), he will be able to clarify whether there is a loss or gain-of-function. A potential problem with these studies, as noted by the PI, is that the neurodegeneration may be non-cell autonomous, i.e., there will be no phenotype by just examining MNs; for example, MNs derived from ESCs that have been obtained from mutant SOD1 transgenic mice have no phenotype unless they are co-cultured with mutant glial cells. The PI states that that he will be able to test this by culturing wild type MNs with mutant glia; however, it is unclear how he will generate the mutant glia. This issue is an important one since some of the neurodegenerative processes he plans to study may be non-cell autonomous. Another issue is that this specific aim is very over-ambitious. It sounds like the PI is planning too many studies on two many alleles on too many genes with too many collaborators. Nevetheless, data regarding the phenotype of many mutant proteins come from huge overexpression studies in which the mutant genes are transfected into MN cell lines or primary MNs. The planned studies are certainly preferable because of they involve the normal transcription control elements and a bona fide MN. In specific aim 3, the PI will define the MN interactomes for wild-type and disease-associated gene products. In order to carry this out, the PI will use the Floxin system to produce alleles with TAP tags (which involves two purification tags) in order to biochemically characterize mutant protein complexes in MNs by mass spec, comparing the mutant with the wild type interactomes; the PI will work in collaboration with Drs. Fisher and Krogan. These are interesting experiments that may provide information about mutant protein binding partners that explains disease pathogenesis - since the reason for cell-specificity of the neurodegenerative diseases may be related to binding and sequestration of cell-specific proteins to the mutant allele's protein; the PI's planned method provides a screen for binding of the mutant protein to the cell of interest, the MN. In some neurodegenerative diseases, it appears that a disruption of protein binding early in development may be key to the later neurodegeneration; however, this may not be true in the case of the mutant proteins that the PI plans to examine - in which case the binding proteins found will not be important to disease pathogenesis. It is also possible that expression of the mutant protein in other cells besides MNs is key to the neurodegeneration, so that the mutant protein's binding partners that the PI finds in MNs will not be the key ones in disease pathogenesis. The definition of the motor neuron "interactome" as stated by the PI is quite broad and the anticipated data in this aim is likely to be extensive and difficult to mineThe PI’s ability to define a "motor neuron interactome" as laid out in this aim is not well supported, specifically his ability to define the biological relevance of the purified proteins. Despite these risks, these experiments are clearly worth performing because of the important data that may be found. Specific aim 4, the last aim addressing validation in human ES cells, is the weakest. The PI invokes an array of human ES cells and potential techniques that might allow the (random) introduction of mutant alleles in human ES cells followed by differential and in vitro "functional" analyses. The techniques listed are inefficient as the PI is well aware and the aim as stated is overambitious. Nonetheless, the applicant does succeed in conveying an exciting innovative plan that is likely to yield interesting data. He definitely has the training and the expertise to conduct the proposed work albeit he will have to trim away some of the subaims in order to remain within the timeframe of the application. QUALIFICATIONS AND POTENTIAL OF THE PRINCIPAL INVESTIGATOR: The PI has been Assistant Professor in the Department of Biochemistry and Biophysics in the Cardiovascular Research Insititute at UCSF since 2006. He received his MD/PhD at UCSF, graduating in 2001; his PhD involved investigations of zebrafish and resulted in 5 publications, three that were senior authored (and one in Genes and Development). From 2001-2003 he pursued a postdoctoral fellowship at UC Berkely with Dr. William Skarnes. He is the last author of a publication in 2005 in Nature that includes Stainier, and he is the first author of a publication in Genes and Development in 2006; these papers involved function of primary cilium and Tectonic, a regulator of the Hedgehog pathway. From 2003-2006, he was a fellow at UCSF in the Developmental and Stem Cell Biology Program. There are 3 additional papers either published or submitted since 2006 including a review in Science and a paper published in PNAS about cilium function. His present funding includes a Burroughs Welcome Fund Career Award in Biomedical Sciences on Tectonic and an NIH R01 through 2012 on Hedgehog signaling. The PI has a lab on UCSF's Mission Bay campus with dedicated space for 12 individuals. There are neighboring labs of two CIRM awardees, Drs. Pleasure and Stainier, and nearby labs of a number of other SC investigators. The PI has planned formal quarterly meetings with Drs. Sam Pleasure and Arnold Kriegstein. In addition, he has two mentors, Drs. Bruce Alberts and Patrick O'Farrell. He lacks experience in neuroscience and ES cell differentiation, but has recruited an excellent group of faculty members with experience in the neurodegenerative field and the genetic manipulation of ES cells. There does not appear to be a letter from anyone with expertise in ES cell biology proper (eg motor neuron derivation, etc.). A graduate student in his lab has some experience in ES cell culture from a rotation with Reijo-Pera. Input from a scientist with experience in ES cell biology would be very beneficial. The PI plans to chair a panel on Stem Cells and Differentiation at the annual UCSF Institute of Regenerative Medicine (IRM) retreat. These details and the letters in the proposal leave little question that the PI is collaborating with a large group of relevant scientists and neurologists. The PI presently has 4 graduate students and three post-doctoral fellows in his lab. The PI did not complete residency training and is therefore not applying for the CIRM New Faculty Award as a physician scientist; however, he does have some interaction with patients at the Human Genetics clinic. This proposal and the PI's venture into a new area is a testament to his "fearlessness." He has experience with gene traps and ES cell techniques, having contributed to the development of gene trap vectors during his post-doctoral fellowship with Dr. William Skarnes. In addition, the PI has experience in analyzing mutations in mammalian neural development. The present studies will change the PI's direction into new clinically relevant directions and neurodegenerative disease. They were partly prompted by the (unexpected) creation of a mouse model of motor neuronopathy in his lab. The PI's background lends itself to these translational aspects of the planned research. INSTITUTIONAL COMMITMENT TO PRINCIPAL INVESTIGATOR: The PI has been Assistant Professor in the Department of Biochemistry and Biophysics in the Cardiovascular Research Institute at UCSF since 2006. The letter from the Department Chair/Dean describes him as "a fearless, pioneering, interdisciplinary scientist" and notes that the Biochemistry Department joined forces with the new Program in Human Genetics and the Cardiovascular Research Institute to recruit him. The PI received $900,000 as start-up as well as some large equipment. He has a lab on UCSF's Mission Bay campus with dedicated space for 12 individuals. There are a number of Core facilities including the UCSF IRM Human Embryonic SC Research Center. A second SC facility is being built at the Gladstone Institute. Over the past 3 years, 8 new faculty were recruited to the stem cell program, and up to 6 additional faculty over the next few years are planned. The Chair/Dean states, "We cannot imagine a better academic environment for an outstanding faculty member." DISCUSSION: There was general agreement among reviewers that the PI is a strong candidate, with an excellent track record and past publications. The PI has cell models of neurodegeneration in important fields. In general, reviewers were positive about the proposal. Discussion occurred about the gene-targeting technique that the PI proposes to apply to hESCs. Two reviewers were enthusiastic about bringing the technique into the field, while one reviewer cautioned that perhaps the PI has over-stated the level of efficiency of the technique, and that comparable levels were available with other technologies. This reviewer further stated that the technique has been around for some time. One reviewer noted that the applicant failed to acknowledge problems of inserting cDNAs into genes including multiple spliceforms, the removal of 3'UTR, and the introduction of a neoR cassette. Panelists discussed the relevance of the proposed gene targeting techniques and the fact that most were not being done in hESC. Overall, the issues surrounding the technique did not overcome the panel’s enthusiasm for the proposal, and to recruit the candidate to the field.

© 2013 California Institute for Regenerative Medicine