The primary focus of the lab is to develop tools for studying gene expression on a genome wide scale. We have developed a technique using high-throughput sequencing for rapidly mapping the integration sites of retroviruses and transposable elements and in collaboration with Dr. Shuo Lin at UCLA, we are in the second phase of a process to map thousands of proviral integrations in the zebrafish germline to create an archived zebrafish mutant resource. We inject high-titer pseudotyped MMLV retrovirus preparations into early zebrafish embryos (approximately the 1,000 cell stage) and raise the infected fish. Those """"""""founder"""""""" fish are outcrossed to wild-type fish and The F1 generation males are raised to sexual maturity. The males are then sacrificed and the sperm cryopreserved, with a matching tissue biopsy being collected for each fish. The genomic DNA is purified and used for PCR amplification of the genomic regions directly adjacent to the proviral integration site using linker-mediated PCR. Using an indexing technique we developed, 500 PCR samples are sequenced together using the HiSeq2000 platform. Genomic locations are determined using high-throughput sequence mapping Bowtie software and sample assignment extracted from the index. To date, we have mapped >17,000 retroviral integrations and demonstrated the efficacy of our mutagenesis strategy. We have generated >3,700 predicted mutations and will continue this effort in production mode at a rate of approximately 200 samples per month. This phase established the framework for generating the resource, and determine the number of total fish necessary to map 50,000 retroviral integrations (which will disrupt an estimated 8,000 genes). We have generated the necessary number of founder fish containing greater than 100,000 retroviral integrations. We have frozen 7,400 F1 fish for our archive and we are on pace to reach our goal of 8,000 in approximately two more years. We have recently expanded our mapping efforts through collaborations to map other insertional mutagens such as the transposable elements TOL2 and AC/DS. We have maped >1,000 of these integrations for our collaborators and will add another 2000 genes in the next two years, adding even more mutations to the zebrafish research community. We have pilot data working with Stephen Ekker, Karuna Sampath, and Harold Burgess with TOL2 and AC/DS gene traps that suggest the technique will port over perfectly and can be utilized for many other projects using insertional mutagenesis. We are also collaborating with two PI's interested in human gene therapy: Dr. Charles Venditti and Dr. Fabio Candotti to map AAV, Foamy Virus, MLV, and Lentiviral gene therapy vectors in human and mouse DNA to look for adverse integration events. Finally, in collaboration with other labs around the world, we have initiated efforts to systematically mutate and phenotype all the genes in the zebrafish genome, effectively mapping out the in vivo functions for all the known vertebrate genes. We are in an early phase to design and implement a phenotyping project. The initial international meetings involved over 70 scientists and resulted in a framework for moving a zebrafish phenotyping project forward. The phenotyping initiative will have two parallel efforts: 1) a traditional """"""""one-at-a-time"""""""" phenotyping of mutant alleles, where only one mutation is in the fish background and will be extensively phenotyped. The second approach will track up to 20 independent mutant alleles present in individual families of ENU mutagenized zebrafish. Each allele can be tracked independently and scored for phenotypes, allowing for a twenty-fold reduction in the number of animals that need to be screened. This increased density comes at the cost of not being able to perform the more detailed phenotyping of the """"""""one gene"""""""" approach, so phenotyping will be limited to the first 5 days of development, but it becomes possible to do low-resolution coverage of nearly all genes in the zebrafish genome. This longer-scale effort is projected to take 5-10 years to complete.
Showing the most recent 10 out of 48 publications
