Sarah Bumaid |
 |
Research Advisor & Department
Jun Liu - Molecular Biology & Genetics
Name of Project
Mapping the jj32 Mutation, a Mutation Affecting Postembryonic Mesodermal Development in C. elegans
Abstract
The Liu lab is interested in the molecular mechanisms regulating the diversification and differentiation of the mesodermal germ layer, which gives rise to many functionally important cells in metazoans, such as muscles, blood and bones. We are using the nematode Caenorhabditis elegans as a model system. In C. elegans the postembryonic mesodermal lineage, the M lineage, is derived postebmryonically from a single progenitor cell, the M mesoblast. In hermaphrodites this gives rise to 14 body wall muscles (BWM), 2 sex myoblats (SMs) and 2 coelomocytes (CC). The Liu lab had previously isolated a number of mutations that affect the development of the M lineage, including the jj32 mutation, which led to the production of additional SMs and the loss of M-derived CCs.
jj32 has been previously mapped to chromosome II to the right of unc-4 and rol-1. I will be using snip-SNP mapping to further narrow down its position in chromosome II. snip-SNP mapping is based on single nucleotide differences that exist between the two known strains of C. elegans, the N2 and the Hawaiian strain. I will cross hermaphrodites that are homozygous for the jj32 mutation as well as for unc-4 and rol-1 and a gfp marker that would make it possible to observe the number of coelomocytes under UV light, with wild type Hawaiian males that are homozygous for the gfp marker only. Their wildtype progeny (F1) will then be picked and placed on individual plates to allow them to self-fertilize. The Non-Unc and Non-Rol F2 animals that are homozygous for jj32 are then picked and placed onto individual plates and are allowed to grow. When enough offspring is produced the plates are washed and then the worms are lysed. Their DNA is then PCRed and digested with different enzymes that can recognise different SNPs on chromosome II. and the results from these analysis would tell us how close the jj32 mutation is to the snip markers. The goal is to be able to identify the corresponding wild type gene of jj32.
|
Lily Cao |
 |
Research Advisor & Department
Tim Huffaker - Molecular Biology & Genetics
Name of Project
Effects of C-terminal Truncations of Stu1 in Saccharomyces cerevisiae
Abstract
Stu1 is an essential microtubule binding protein in S. cerevisiae responsible for the assembly and dynamic function of mitotic spindles. It has been shown to localize at the plus-ends of microtubules as well as the kinetochores on chromosomes. Stu1 helps mobilize chromosomes and stabilizes the spindle structure during mitosis. Mutations in the protein lead to spindle collapse and degradation of cell growth. In order to identify domains that allow Stu1 to bind to kinetochores and microtubules, various lengths of the C-terminus coding region of the gene were deleted. Cell viability, protein levels, and protein localization were analyzed to determine domains that allow Stu1 to interact with other proteins to carry out its functions. Five constructs were made using techniques such as PCR amplification, bacterial transformation, and DNA sequencing. The truncations ranged from 194 amino acids to over 991 amino acids. The truncated gene was then transformed into a strain of yeast with STU1 previously deleted from its genome. These cells also have a uracil vector containing a wildtype copy of STU1 to allow the cells to live. The cells were plated onto 5’FOA drug to induce the loss of the uracil plasmid, while maintaining the vector with the truncated copy of STU1. This plasmid shuffling technique allowed the truncated Stu1 to be the only determining factor for viability. The 5’ FOA plating results seems to indicate that none of the five constructs are able to sustain growth. Western blotting using GFP antibody to detect GFP attached to the C-terminal end of the truncated protein showed that possibly due to their instability, the truncated proteins were not being produced. Similarly, no or very little GFP signals were detected under the microscope. The results seem to indicate that deleting 194 amino acids compromises Stu1 functions. New constructs with smaller deletions (approximately 20 amino acids for each successive construct) were made and similar analysis will be performed to narrow down the region of Stu1 crucial for cell viability. The results from this project seem to show that the C-terminus of Stu1 is essential for the proper functioning of the protein.
|
Christopher Castorena |
 |
Research Advisor & Department
Bryan Danforth - Entomology
Name of Project
Rooting Phylogenies Using Gene Duplication Events
Abstract
Phylogenetic rooting methods using outgroups have thus far failed to establish a robust root node for the phylogeny of the bees. The goal of this project is to use gene paralogues as an alternate means of determining this node.
The first part of the project entails simulating gene paralogue evolution over a series of tree topologies. I programmed a simulation that generates dna sequences and evolves them according to a particular tree topology and model of evolution. My program then analyses these dna sequences with Parsimony and Bayesian analysis which try and recover the true topology. We analyze the accuracy of these two methods with the results of these simulations to see under what parameters each method performs well and where they perform poorly. Because of the computational intensity of these simulations (they would have taken almost 4 years on my Mac) I had to parallelize my simulations and submit them to Cornell’s Linux cluster.
Preliminary results have shown that when there is little divergence between the paraloges, Bayesian analysis outperforms Parsimony, but as the level of divergence increases, parsimony beats Bayes. When the Bayesian method is given more information about how the sequences evolved, its accuracy increases tremendously, and parsimony is no longer able to outperform it for reasonable branch lengths. We are also analyzing the effects of sequence length on the accuracy. Preliminary results have shown that increasing the sequence length always improves Bayesian analysis but Parsimony shows signs of statistical inconsistency (accuracy decreases with increasing data).
These results will be useful for biologists want to find the best method to analyze their data sets. Bayesian analysis can take months to do an analysis that would take parsimony a couple of days, so it’s good to know when you can get away with just Parsimony.
|
Patrick Chen |
 |
Research Advisor & Department
David Lin - Vet Biomedical Sciences
Name of Project
Retroviral infection as a Tool for Studying the Development of the Olfactory Epithelium
Abstract
Olfaction is an important mediator between information sent by the environment and the signals received by the brain. Disturbance of the organization of the epithelium in mice leads to an absence of innate fear response to fox odorant, indicating the importance of the topography. The olfactory epithelium can be broadly divided into zones which have been classically defined by the expression of odorant receptors. However, very little is known about how these zones are formed. We hypothesize that a molecular program of genes is differentially expressed in dividing progenitor cells. This differential expression results in the formation of distinct zones in the epithelium. We have identified a differentially expressed gene that is a potential candidate for zone specification. Testing this hypothesis has traditionally required generation of transgenic mice to express or delete candidate genes within the progenitor layer. A faster and easier system for manipulating genes in the nose would be useful. A novel tool that has not been commonly used is the use of viral infection to alter the gene expression in the epithelium.
The first goal of my project for this summer is to optimize a retroviral production and infection protocol. Retroviruses only infect and integrate their DNA into dividing cells, which allows for easy cell lineage tracing and a much faster way for genetic manipulation. The problem that labs have traditionally had with viral infection in the nose has been that infectivity is not uniform throughout the entire epithelium; as well, there is low infectivity. I worked on optimizing a viral infection protocol and was successful in improving infectivity as well as getting uniform infectivity. The project goal is to use this retrovirus protocol and test our candidate gene. By perfecting this tool, we will be able to test the effects of the gene in the nose.
|
Erin Chu |
 |
Research Advisor & Department
Nathan Sutter - Vet Clinical Sciences
Name of Project
See Spot Shrink: The Effect of Sequence Variation in Three Candidate Genes on Body Size Diversity in Canis familiaris
Abstract
The domestic dog (Canis familiaris) is the only vertebrate that displays a 60-fold range in body size. We took advantage of the unique population structure of the dog, namely its hundreds of highly inbred, phenotypically distinct breeds, to search for genes that contribute to body size diversity. Through a genome-wide scan for breed average mass, we identified three independent associated loci containing candidate genes we have labeled Hermes, Butters, and Jeremy. To test the hypothesis that causal mutations within these genes contribute to size differences in dogs, we have sequenced the exons and promoters in a total of 12 toy and giant dogs. Fifteen markers (mostly SNPs) were discovered in Hermes and for several the toy and giant dogs have perfectly disparate genotypes, for example, "AA" in toy dogs versus "GG" in giant dogs (where p < 7E-7, in a two-tailed Fisher’s exact test of allele count). Two of the markers discovered in Hermes occur in coding sequence, but both are silent mutations. Thus, the causal mutation(s) in the Hermes locus is likely to affect expression level of Hermes or impact a difference gene within the Hermes-containing linkage disequilibrium block. To date we have discovered one marker in Butters but it does not occur in coding sequence. We will continue sequencing the remainder of Butters’ protein coding region and all of Jeremy’s exonic regions to examine the candidacy of these two genes as contributors to body size in Canis familiaris.
|
| |
>>next previous<< |