Back to the Original Topic!

Differential Gene Expression in Δzms2 Yeast

William Cannon

Introduction:

Microarrays of Δzms2 yeast were analyzed for differentially expressed genes on grids 5 and 6 on slides 13760725 (group 6), 13760726 (group 7), and 13760728 (group 8).  Grids 21 and 22 were also analyzed on slide 13760726.  Four genes from Δzms2 yeast met the criteria for differential gene expression compared to wild type yeast.

Methods:

The same methods were used as the group project.  Magic tool was used for the analysis and normalization of slides 13760725 (group 6), 13760726 (group 7), and 13760728 (group 8).  The top grids (5 and 6) were used for slides 13760725 (group 6), 13760726 (group 7), and 13760728 (group 8).  The bottom grids (21 and 22), which contain the same DNA sequences as grids 5 and 6, of slide 13760725 were analyzed as well.

Results: 

The excel files with the microarrays dye intensity values from groups 6, 7 and 8 was analyzed. Grids 5 and 6 were analyzed from slides 13760725 (group 6), 13760726 (group 7), and 13760728 (group 8) and grids 21 and 22 were also analyzed on slide 13760726.  (Grids 21 and 22 are the corresponding grids 5 and 6 for the bottom grids of the microarray slides.)  All of these slides contained Δzms2 and wt yeast RNA.  Groups 7 and 8 labeled the wt with Cy5 (red) and Δzms2 with Cy3 (green).  Group 6 had the dyes reversed, with the Δzms2 labeled with Cy3 (green) and the wt labeled with Cy5 (red).  The Cy3 and Cy5 dye intensities were plotted against each other to assess dye bias on the microarray slides to be analyzed (Figure 1).  The slope of the best-fit line was 0.2715, 0.409, and 0.5825 for slides 13760725, 13760726, and 13760728 respectively.  A theoretical microarray with no dye bias would have a slope of 1.  The background intensity was subtracted from the foreground intensity and ratios of mutant over wild type were calculated.  A log₂ transformation was applied to the ratios in magic tool.  Further normalization was performed via magic tool by setting the mean to 0.  This normalization produced a more normal distribution for analysis.  Box plots created via Magic Tool showed that the different microarrays had similar means and distributions after normalization (Figure 2).

Figure 1.  The green and red dye intensities were plotted against each other for slides 13760726, 13760725 and 13760728.  Slides 13760726) and 13760728 labeled the wt with Cy5 (red) and Δzms2 with Cy3 (green).  Slide 13760725 had the dyes reversed, with the Δzms2 labeled with Cy3 (green) and the wt labeled with Cy5 (red).  Best fit lines were added.

Figure 2. Box plots were created via Magic Tool.  The plot on the left represents the four microarrays before mean normalization, while the plot on the right represents the four microarrays after mean normalization. The box plot with mean normalized data displayed similar means and distributions between the different microarrays.  The columns correspond to the different microarray trials:  Column 1 (Δzms2 stained red on the top grids of slide # 13760725), column 2 (Δzms2 stained green on top grids of slide # 13760726), column 3 (Δzms2 stained green on top grids of slide # 13760728), and column 4 (Δzms2 stained red on the bottom grids of slide # 13760725).

Genes in grids 5 and 6 were analyzed in Magic tool to determine which genes are differentially expressed in Δzms2 yeast.  Genes were considered differentially expressed if there was a three-fold increase or decrease in 3 out of the four array trials.   A ratio value after log₂ transformation that is greater than 1.6 or less than -1.6 indicates at least a 3-fold increase of decrease.  Four genes in grids 5 and 6 fit these criteria (Table 1).  Differentially expressed genes were compared to the yeast genome from the GCAT website (Table 2).

Table 1.   Genes that are differentially expressed in grids 5 and 6 in Δzms1 yeast are displayed in the Magic Tool format.  Genes were considered differentially expressed if the gene had a three-fold increase or decrease in three of the four array trials.

Table 2. The genes that were considered differentially expressed were compared to the yeast genome gene list available on the GCAT website.  Red color corresponds to upregulated genes, while green color corresponds to downregulated genes in Δzms1 yeast.

Discussion:

This project seeks to analyze several microarray slides in order to investigate differentially expressed genes in Δzms2 yeast when compared to wild type yeast.  Prior research has shown that over expression of ZMS1 or ZMS2 can suppress ZMF1 mutations.  ZMF1 mutations cause a phenotype of extra sensitivity to oxidative stress¹.  ZMS2 also may be a transcription factor because it has a zinc finger domain¹.  Hopefully this microarray analysis of Δzms2 yeast will highlight genes that may be regulated by ZMS2.

            Analysis of microarrays is many times hindered by dye bias.  Cy3 dyes are known to fade so microarrays that use C3 and Cy5 often have a Cy5 dye bias.  All of the slides analyzed in this project had a Cy5 (red dye) bias because the slope of red dye vs green dye plots was less than one (Figure 1).  This red dye bias includes slide 13760725, which had a dye swap where the wt was labeled with Cy3 (green) and the Δzms2 was labeled with Cy5 (red).  This red dye bias will cause more genes to seem under expressed in Δzms2 on slides 1376026 and 13760728, while causing more genes to seem over expressed on slide 13760725. Both the top and bottom grids were analyzed on slide 13760725, which had the dye swap. Three out of the four microarray trials had to have a 3-fold increase for the gene to be considered differentially expressed.  This differential gene criterion should hopefully account for dye bias because there are two trials for each dye scenario and at least three trials have to agree.

            Four genes met the criterion to be considered differentially expressed (Table 2).  One of these genes has no known biological, molecular, or cellular function.  None of the genes have obvious oxidative stress functions.  One of these up-regulated differentially expressed genes is YDR447C, or RPS17B.  RPS17B is involved in protein biosynthesis and is a structural component of the ribosome.  RPS17B is similar to an up-regulated differentially expressed gene in the Δzms1 microarray, RPS19B.  RPS19B is also involved in protein biosynthesis and is a structural component of the ribosome.  These genes could be upregulated in the Δzms1 and Δzms2 yeast because ZMS1 and ZMS2 repress their transcription.  They could also be upregulated as a compensation for the ZMS1 and ZMS2 knockouts.

            Further analysis of this microarray experiment should be performed.  Maybe a better normalization of the data needs to be performed to better adjust for the red dye bias in the slides.  The Δzms1Δzms2 microarrays need to be analyzed and correlated to the Δzms1 and Δzms1 microarrays because prior studies have shown that overexpression of ZMS1 or ZMS2 (but not both) masks ZMF1 mutations¹. It could also be fruitful to identify several genes that are known to be involved in oxidative stress or the Glucose 6-phosphate dehydrogenase (G6PDH) pathway and to investigate whether these microarray experiments suggest differential gene expression from wild type yeast.  The controls also need to be analyzed.  For example, ZMS1 should be completely downregulated in the Δzms1 and Δzms1Δzms2 yeast, while ZMS2 should be completely downregulated in Δzms2 and Δzms1Δzms2 yeast.  The blanks on the microarray should also not reveal any differential expression if the arrays worked correctly.  Housekeeping genes could also be analyzed as controls.  Housekeeping genes should probally have similar expression levels in both wildtype and the single knockout yeast strains (Δzms1 and Δzms2).  Future microarray experiments should probably use more microarray trials for each yeast strain.

            Ultimately more experiments have to be completed to determine if the genes identified are truly differentially expressed.  These experiments could include real-time PCR and northern blotting.

References:

1.  Slekar, K.  A Genetic Study of Anti-Oxidant Factors in Yeast.  Lecture 2008 Oct.