Goldowitz Lab Research Projects

Our laboratory uses mouse mutants and experimental embryology and reference populations to study how single genes, or ensembles of genes, participate to support the normal development and function of the central nervous system.


The cerebellum is an important brain region for the coordination of motor and cognitive behaviors. Developmental abnormalities of the cerebellum have been linked to autism, schizophrenia, and other disorders of human neural function. This project aims to elucidate the gene networks underlying cerebellar development. To this end, we measured the gene expression over developmental time course in normal mouse cerebellum, as well as several mutant mouse strains that exhibit cerebellar abnormalities. From this we generated extensive data sets of gene expression and performed multiple bioinformatic analyses. We also developed and integrated web-based informatic and visualization tools for researchers to analyze our data sets, and test hypotheses about the cellular and molecular development of the cerebellum.  More recently, we have collaborated with the FANTOM5 consortium on the development of a new cerebellar time series utilizing the next generation sequencing technology. 

We have several specific aims:

  1. To measure and compile gene expression data sets from normal and mutant cerebellum over a defined developmental time course.
  2. To find genes important for cerebellar development with the identification of cliques in expression data, and latent semantic indexing of Medline references to mine data on the patterns, both in time and space, of expressed genes and cellular phenotypes.
  3. To validate transcriptome data with molecular(qRT-PCR), anatomical (in situ hybridization and immunocytochemistry) and experimental (siRNA and knockout) approaches.

The data that was obtained and the web-based tools that were constructed in this project are fully open to the research community. This project is also designed to interface with several of the currently funded Human Brain Projects that look at the anatomy and cell biology of the adult mouse brain and cerebellum. The phenotypic data that is gathered will contribute to the growing understanding of the molecular and cellular bases of cerebellar development. Such information may help understand and treat disorders of cerebellar origin, such as the most common form of childhood brain cancer, the medullablastoma, which is believed to emanate from the developing granule cells of the cerebellum. In the long term, we hope to use the tools developed in this project to make predictions about the molecular pathways and cellular programs that are important to the well-being of the central nervous system.


The three Specific Aims that comprise this application are:

  1. High-throughput acquisition of systems-level phenotypes related to drug, alcohol, and neurobehavioral traits in 80 recombinant inbred lines.
  2. Integration of systems level phenotypes and existing microarray data, QTL mapping, and determination of evidence for candidate gene pleiotropy.
  3. Testing the pleiotropic effects of candidate genes: discover and deliver mouse models for drug and alcohol related phenotypes in a rapid and cost-effective manner.

We will deliver several major resources to the drug and alcohol research community:

  1. A large set of phenotypic data on general behavior and alcohol- and drug-related phenotypes for BXD recombinant inbred (RI) lines of mice that have been expanded here at UTHSC to over 80 lines (Peirce et al, 2004, Appendix). This will add value to studies in other fields using these lines.
  2. The identification of candidate genes underlying alcohol and drug abuse related phenotypes based on the resulting QTLs.
  3. The further integration of the mouse phenotype repository, MuTrack (Baker et al, 2003, Appendix), with the sophisticated tool for complex trait analysis, WebQTL (Chesler et al, 2004, Appendix; Wang et al, 2003).
  4. Development of WebQTL tools to query the relationship of genes of interest to the phenotypes database and efficiently identify candidate genes that underlie aberrant behavioral repertoires.
  5. The identification of newly created or publicly available mutant mice for phenotypes of relevance to drug abuse and alcoholism.


The third trimester of pregnancy is characterized by a highly dynamic period for cerebellar development. During this time, the cerebellum undergoes its most rapid growth unparalleled by any other cerebral structure. Very preterm infants are particularly vulnerable to impairments in their developing cerebellum given that these critical phases of cerebellar growth occur without the protective intrauterine environment. Therefore, it is crucial to understand how we can minimize cerebellar alterations to maximize neurodevelopmental outcomes in extremely preterm infants. We have developed in the lab a diversity of neonatal translational mouse models to study the impact of different perinatal insults on the developing cerebellum. Intrauterine growth retardation, infection/inflammation and infratentorial haemorrhages are studied to further understand the underlying pathophysiological mechanisms and develop new potential neuroprotective strategies.


One key limitation to our understanding of autism is the limited access to studying developmental events in the human brain. The use of model organisms that have homologous genetic and anatomical underpinnings is critical. The mouse is recognized as the leader in “pre-clinical” models and our research employs a unique genetic and developmental model of brain pathology that is known to underlie autism to gain insights into the etiology of autism and serve as a platform to test interventional strategies. Our work proposes to use a mouse model that has varying losses of cerebellar Purkinje cells to model the role of this structure in behaviors that are “autistic-like” and explore hypotheses about how those losses affect the function of the CNS at the behavioral and anatomical levels. This systems approach should present researchers with a new understanding of cerebellar function and connectivity with the prefrontal cortex and the possible relationships this has with modeling autistic-like behaviors. We use a mutant mouse, Lurcher, that loses all of its Purkinje cells over the 2-3 weeks of life, akin to the last trimester and perinatal period in the human. By making experimental mouse chimeras with Lurcher and wildtype mice (the joining together of 8 cell embryos to make an organisms with cells from both genotypes), we produce mice that have varying percentages of Lurcher and wildtype cells, and hence varying numbers of Purkinje cells as the Lurcher mutation principly affects these cells. Behavioral and anatomical analyses of these chimeras are made to understand the relationship between a primary loss of Purkinje neurons to behavior and connectivity between the cerebellum and other key structures which are believed to underlie the autistic disorder.


Experimental mouse chimeras will be used in the analysis of the cellular target of neurodegenerative diseases. The well-documented strain differences in mice to kainic acid neurotoxicity will be used as the test-bed for this approach. C57BL/6 and FVB/N mice are remarkably resistant and sensitive, respectively, to the neurotoxic effects of kainic acid. In these experiments we will aggregate embryos from these two lines of mice to examine whether neuron-intrinsic or neuron-extrinsic mechanisms are responsible for the death of neurons in the hippocampal formation. The key to the analysis is the recent availability of genetic markers that are histologically demonstrable within cells of each of these strains. Control and C57BL/6-ROSA26 <–> FVB/N-GFP chimeric mice will be injected with kainic acid for 2 and 7 day survival. Sections of the fixed brains will be processed for the histological demonstration of cell genotype and cell phenotype (alive or dying). Three avenues of investigation will be pursued. First, it will be determined if a linear relationship exists between the number of living neurons and percentage of chimerism. Second, the non-neuronal cell populations (glia and endothelial cells) of chimeras will be examined to determine if the percentage chimerism of these cells exhibit any relationship to phenotypic outcome. Finally, the genotypic origin of non-neuronal cells in the vicinity of affected or unaffected hippocampal neurons will be examined for further evidence for a cell-intrinsic or –extrinsic explanation. The long term goal of this research is to establish a powerful approach to the analysis of the cellular targets of human neurodegenerative diseases modeled in the mouse. This will provide important insights for a rationally-based therapy for a variety of neurological diseases such as Parkinson’s, stroke, and epilepsy.