Discussion
Dr. Slekar's ΔZMS1 is a S. cerevisiae mutant that lacks a gene which encodes a zinc-finger transcription factor and is involved in countering oxidative stress. The purpose of this microarray study was to compare the transcriptomes of ΔZMS1 to the wild type. It was hypothesized that differences in gene expression would be observed primarily in genes encoding transcription factors related to oxidative stress.
Normalization
Six of the nine genes that were found to be induced in the S. cerevisiae ΔZMS1 mutant appear to all be involved in transcription and translation. YPL167C encodes the catalytic subunit of DNA polymerase zeta which is involved in DNA repair. The ΔZMS1 mutant is susceptible to oxidizing agents, thus having a DNA repair gene upregulated may be an indication that S. cerevisiae is attempting to repair damaged DNA resulting from free radicals. In a similar vein, YDL005C is also of particular interest because it codes for a subunit of the RNA polymerase II mediator complex which is essential for transcriptional regulation through association with other core polymerase subunits in order to form the RNA polymerase II holoenzyme. The upregulation of this RNA polymerase subunit is a probable indicator that the ΔZMS1 mutant is attempting to respond to oxidative stress through increased transcription of various genes. Also of note is YPL181W which relieves transcriptional repression by binding to a co-repressor and preventing the repressor from binding to the promotor. This transcriptional repression is interesting because it is another indicator that the yeast is attempting to either compensate for the lack of the ZMS1 gene or respond to oxidative stress through the increased transcription of other genes. YPL190C encodes a single strand DNA-binding protein involved in efficient splicing and YBR116C codes for an essential protein in spliceosome assembly and exocytosis. Since both of these genes code for proteins that are involved in post translational modifications, primarily in the splicing of exons, upregulation of these genes would be another indication that S. cerevisiae is attempting to respond to either the lack of the ZMS1 gene or that it is attempting to respond to oxidative stress. Lastly, YBR118W encodes EF-1α, which is an elongation factor which binds the aminoacyl-tRNA to the ribosome. If all six of the above-mentioned genes are taken together, they indicate across-the-board upregulation of genes involved in every step of transcription and translation of proteins that are partially responsible for regulating S. cervisiae's response to oxidative stress. In addition, YPL188W is also upregulated and it codes for a mitochondrial NADH kinase which is required for the cellular response to oxidative stress.
The last two induced genes do not have a discernible connection to oxidative stress. YPL204W is an isoform of casein kinase I while YBR116C encodes for an unknown protein for which the purpose is currently unknown.
On the other hand, four genes were found to be reduced by the ΔZMS1 mutant: YDL065C, YPL278C, YER089C, and YJR042W. The degree of reduction for each of these genes can be found here. YDL065C in S. cerevisiae encodes an import receptor for newly synthesized class I peroxisomal membrane proteins (PMPs). The YDL065C gene products bind these PMPs in the cytoplasm and delivers them to the peroxisome for subsequent insertion into the peroxisomal membrane. Peroxisomes are important in eukaryotic cells for dealing with oxidative stress as they contain the anti-oxidant enzymes catalase and oxidase. A reduction in anti-oxidant activity carried out by the peroxisomes would make an organism more likely to be susceptible to damage from oxidative stress. Due to ΔZMS1 which has been shown to be a multi-copy suppressor of ΔZMF1 which when present encodes the anti-oxidant factor G6PD, it is reasonable that ΔZMS1 result in a reduction in other mechanism that protect S. cerevisiae from reactive oxygen species. YPL278C could possibly be involved with oxidative stress as it encodes a hypothetical protein that has yet to appear in any research publications to date. However because the function of the gene product of YPL278C is completely unknown, YPL278C could also not be involved with oxidative stress in S. cerevisiae. The reduction of YER089C and YJR042W were probably not related to oxidative stress in ΔZMS1. YER089C encodes for a type 2C protein phosphatase involved with osmostress and DNA checkpoint inactivation. YJR042W encodes the subunit of the Nup84p subcomplex of the nuclear pore complex (NPC) and is involved in nuclear trafficking. There are three plausible explanations for their reduced expression in the ΔZMS1 mutant. First, as stated in the introduction there is some studies suggesting that ΔZMS1 is a transcription factor that governs the expression of many nuclear genes. Secondly, oxidative stress in the ΔZMS1 mutant could have damaged the regions of the DNA containing the genes YER089C and YJR042W. Thirdly, there could have been error with the microarray slide even after normalization was conducted to reduce the probably of error.
SAM
Another program for analyzing this kind of data would be SAM (Significance Analysis of Microarrays). An experiment such as the one performed in this project is an example of an One Class data set. Prior to entering this data into SAM it needs to be normalized because SAM does not normalize the data values. What SAM does is find significant genes by computing a statistic, di, for each gene, i, and measures the strength of the relationship between gene expression and the response variable (i.e. changes in gene expression should lead to a change in the signal ratio). SAM then takes this value and runs a permutation, comparing it to all of the other di values from every other gene in the microarray. This methodology allows the program SAM to determine the relationship between each and every pair of genes. The user can then set a cutoff for significance based on his or her interpretation or use a “fold change in expression” based on the permutations in order to define the genes of interest (much like using standard deviations to eliminate irrelevant background signals). SAM, using this information and the di values can visualize this data into tables and graphs.
Problems and Future Considerations
In this study, a dye-swap comparison was used utilizing only the Cy3 (green) dyes in order to minimize errors caused by differential photobleaching between the Cy3 and Cy5 (red) dyes. However, differential photobleaching was not the only source of error in this experiment. The RNA purities were questionable as noted by multiple factors including the unusually high RNA A260/A280 ratios of 4.97 and 2.78 when normal RNA values tend to be about 1.8-2.0 (see Results). The poor RNA purity can be seen in the RNA gel, as illustrated by the degree of smearing and the appearance of contaminant bands (such as DNA). The low quality of the microarray image may be due to several different factors. The high background noise may have resulted from the low hybridization temperature used when annealing the cDNA to the microarray chips. Increasing the temperature would have reduced the amount of non-specific binding and thus would have increased the overall quality of the signals and ratios. Overall, in order to make the data better, the experiment would have to be repeated several times for two major reasons. First and foremost, analysis of this data is statistical in nature and therefore larger data sets allow for more statistically significant results. Secondly, repetitive trials of this experiment would enhance the quality of the data by improving the skill sets of the researchers.
Home Introduction Methods/Materials Results References
Contact us: