Introduction Methods Results Discussion References Photos
Microarray Analysis of 10 genes with altered expression patterns of ΔZMS1 mutants in ΔZMS2 and ΔZMS1/ΔZMS2 mutants
By: Karl Gorzelnik
Introduction:
Saccharomyces cervisiae, yeast, is a model organism for studying eukaryotes because it is a single celled organism whose entire genome has been sequenced and the effects of mutations are easier to identify than if it were a multi-celled organism. Dr. Slekar, here at James Madison University, is studying the effects of oxidative stress on yeast. Oxidative stress is caused by free radicals, atoms without stable rings of electrons which take electrons from DNA, RNA, proteins, and other cellular structures. Over time this will cause mutations in that line of cells, which would eventually eliminate that cells descendents because they would no longer be able to compete at the same level. Dr. Slekar is studying antioxidant stress mechanisms that encode transcription factors, protecting the cell from free radicals. On the zmf1 gene she is studying she isolated two different mutant strains, ΔZMS1 and ΔZMS2 which are knockouts for different characteristics. When bred to be ΔZMS1/ΔZMS2 they have both characteristics knocked out, but display a wild-type phenotype. Previously our group looked at the ΔZMS1 mutants she isolated, and identified five genes that were up-regulated and five genes that were down-regulated. In this study I will examine the ten genes identified while studying ΔZMS1, and see how they were effected in ΔZMS2 mutants and ΔZMS1/ΔZMS2 mutants. I hypothesize that the genes that are up-regulated and down-regulated would be similar in ΔZMS2 mutants and ΔZMS1 mutants, and if there are genes that differ it would be between ΔZMS1 and ΔZMS1/ΔZMS2 mutants. There should be similar differences between ΔZMS1 and ΔZMS1/ΔZMS2 as between ΔZMS2 and ΔZMS1/ΔZMS2 mutants.
Methods:
The methods used were exactly the same as in the previous study (Methods). The only change was the slides I looked at. In the previous study we examined the slide we hybridized cDNA to, slide #1360694, which had the ΔZMS1 mutant labeled with Cy3 green dye and the wild type labeled with Cy5 red dye. We also examined slides from other groups, #1360695, #1360722, and #1360727. Within each slide there were two copies of every gene for us to examine, except #1360722 because it only had grids on the top half. In this experiment I looked at 5 genes which I had previously found to be up regulated in ΔZMS1 mutants: YDR134C, YJL052C, YMR219W, YOR343C, YPL051W; and 5 genes which I had previously found to be down regulated in ΔZMS1 mutants:YAL017W, YDL118W, YDR165W, YML013W, and YML120C. I examined them in ΔZMS2 and ΔZMS1/ΔZMS2 mutants, using slides: #1360725, #1360726, and #1360728 for ΔZMS2 and #1360731, #1360723, and #1360727 for ΔZMS1/ΔZMS2. The pixel intensity values were prepared for Magic Tool by subtracting the background from the pixel and dividing the mutant by the wild type and then put into Magic Tool and analyzed by taking the base 2 logarithm and then standardizing to minimize the effects of dye bias.
Results:
With the dye intensity bias minimized results appeared similar for many of the genes for ΔZMS2 (Figure 10, below) as it did in the previous experiment for ΔZMS1 (Figure 8, in Results). There were two genes YDR165W and YML013W which came back inconclusive in this experiment, with 3 of the 6 slides coming back positive while the other 3 came back negative. Two genes showed opposite functions, YDR134C in ΔZMS1 is up regulated whereas in ΔZMS2 it is down regulated. Similarly YOR343C was up regulated in ΔZMS1 but down regulated in ΔZMS2. Figure 10, below, is a circle diagram which shows correlations between gene expression patterns in ΔZMS2, which varied from ΔZMS1.
Figure 10. Gene expression ratios for ΔZMS2 mutants. Negative numbers mean the gene is down regulated or express less compared to wild type, while positive numbers mean the gene is up regulated compared to wild type. From left to right the columns are of the following arrays: slide #1360725 top, slide #1360725 bottom, slide #1360726 top, slide #1360726 bottom, slide #1360728 top, and slide #1360728 bottom.
Figure 11. Circle diagram of up and down regulated genes from ∆ZMS1 studied using ∆ZMS2 mutants.
The ΔZMS1/ΔZMS2 mutants had three genes come back inconclusive, with 3 of the 6 slides having positive regulation while the other 3 had negative regulation, see Figure 12. Excluding those three genes, there were no others which differed from the ΔZMS1 mutant gene expression levels. Figure 13 shows gene expression correlations in a circle diagram for ΔZMS1/ΔZMS2, which showed many similarities to the ΔZMS1 gene expression correlations. Table 3, in the Results, shows the functions, processes involved in, and where the genes are in the cell, of the ten genes studied.
Figure 12. Gene expression ratios for ΔZMS2 mutants. Negative numbers mean the gene is down regulated or express less compared to wild type, while positive numbers mean the gene is up regulated compared to wild type. From left to right the columns are of the following arrays: slide #1360731 top, slide #1360731 bottom, slide #1360723 top, slide #1360723 bottom, slide #1360727 top, and slide #1360727 bottom.
Figure 13. Circle diagram of up and down regulated genes from ∆ZMS1 studied using ∆ZMS1/∆ZMS2 mutants.
Discussion:
The ten genes examined were genes which had consistent expression level changes in at least 5 of 7 slides when we tested for ΔZMS1. When I examined those ten genes for ΔZMS2 and ΔZMS1/ΔZMS2 mutants my results were different from my hypothesis. I hypothesized that the gene expression levels would be similar in ΔZMS1 and ΔZMS2, whereas only genes to differ in the experiment was different for ΔZMS2. ΔZMS1/ΔZMS2 mutants I thought would be more likely to have gene expression levels that differed because it displays a wild type phenotype, in terms of oxidative stress response, and so I thought gene expression levels would differ compared to ΔZMS1 and ΔZMS2 mutants because they showed mutant phenotypes in response to oxidative stress. The two genes that were down regulated in ΔZMS2 when they were up regulated in ΔZMS1 did not appear to be correlated together, see Figure 11. The circle diagram analysis of the genes showed correlation between certain genes, the ones which showed the most correlation for all of the mutants were on the same chromosomes: YDR165W, YDL118W, and YDR134C showed correlations between the three mutants, as they were on the same chromosome.
The two genes that were down regulated in ΔZMS2 which were up regulated in ΔZMS1 were YOR343C and YDR165W. YOR343C has an undetermined molecular function and location. YDR165W is involved with tRNA methylation of Guanine. More experiments will have to be run to determine how its function differs from ΔZMS1 and ΔZMS2, why it would be up/down regulated in each mutant and how its regulation affects the cell. Most of the genes had unknown functions, so future experiments could limit the genes down by function instead of by change in gene expression. The ΔZMS1/ΔZMS2 genes also had relatively low levels of change in gene expression, being at or below a 2 fold change, which is unusable for drawing conclusions. Since they didn't differ from ΔZMS1 anyway, the results were inconclusive. We could repeat the experiment to possibly get a larger change in gene expression.
When looking at the 10 genes I restricted this experiment to, my hypothesis was incorrect, as I thought ΔZMS2 mutants would have their genes expressed at the same levels as ΔZMS1 mutants, and ΔZMS1/ΔZMS2 mutants would have variation in there gene expression compared with ΔZMS1. If I were to look at the entire genome I suspect that ΔZMS2 mutants would behave more like ΔZMS1 mutants, as they have similar phenotypes. ΔZMS1/ΔZMS2 mutants have a wildtype phenotype, so it is likely that they would have more differences in gene expression than ΔZMS1, when looking at the entire genome. Future experiments could look at the entire genome to see these differences.
Since the experiment was just looking at 10 genes, which may not be affected at all by the mutant phenotype, we do not know if the 10 genes were really affected by the mutation, or if we just chose ones whose expression randomly changed the most. In order to verify our results we would need to run replicates not only of this experiment, but of the previous experiment looking at ΔZMS1 mutants.
The dye intensities were far from the 1:1 ratio which is ideal. Performing more experiments could detect the source of this error, which was applied by each of the groups. Other sources of error could be the incorrect aligning of grids. Since each group aligned grids for their own data, in Scanalyze, it might be that each group lined them up a different way from the other groups, with systematic under representation for portions of dots by one group, but not another. To ensure there was not a bias from multiple people aligning the grids, it would be most consistent if one person aligned the grids for the different groups, even though they didn't perform the experiment. Because microarray data can only establish trends and correlational expression patterns among yeast genes, the data must verified by the use of other molecular techniques. The microarray results can be verified by Southern blot, to confirm probe construction, and RT-PCR to confirm relative abundance of specific transcripts in samples.
Noble Egekwu - egekwuni@jmu.edu Karl Gorzelnik - gorzelkv@jmu.edu Jonathan Baugher - baughejl@jmu.edu