Title Page

Introduction

Results

Discussion

References

Total RNA Isolation From Yeast

          Saturated overnights of yeast strains zms1Δ, zms2Δ, zms1Δ zms2Δ, ZMS1++, ZMS2++, along with a wild type strain were inoculated in SD liquid (minimal media) and -leu SD liquid for ZMS1++ and  ZMS2++.  Cultures were incubated for 24 hours at 30°C with shaking.  These cultures were transferred to 90mL of SD liquid at an OD600nm of 0.3.  They were incubated for seven hours until mid-log phase was reached at 30°C with shaking at an OD 600nm of 1.03 for ZMS1++ and 0.894 for wildtype. Yeast cells were harvested and transformed into spheroplasts to facilitate cell lysing. RNA was isolated using the RNAsafe kit from Obiogene. The RNA concentration and purity were determined using a spectrophotometer at 260nm and both 260nm and 280nm for purity. RNA quality was assessed using gel electrophoresis on a 1.2% agarose gel with 2 ug sample of RNA.  Gel was stained in ethidium bromide and visualized with UV light.

Labeling Procedure For Microarray Analysis   

         The Genisphere 3DNA Array 350 Detection Kit was used in order to make our labeled cDNA probe. 10µl of RNA from both wildtype and ZMS1++ were used to begin reverse transcription. mRNA from each product was first reversed transcribed into cDNA using primers provided. Before combing our cDNA products, wild-type product was labeled with the green dye (Cy3) and our ZMS1++ product was labeled with red dye (Cy5). Both wild-type and ZMS1++ cDNA products were combined and concentrated using WM30 microconcentrators. 

 Hybridization Of The DNA Chip            

           Slide used for this experiment was #129.  DNA chips were steam treated in order to remove excess salt and to denature the oligos on the chip. In order for the cDNAs to stick to only the sequences on the chip, hybridization steps and washes were performed using different concentrations.

            The slide was pre-hybridized using 3X SSC, 0.1% SDS, .1 mg/ml sonicated salmon DNA at room temperature.  The slide was then hybridized with a hybridization solution, which included the concentrated cDNA. It was then incubated at 45º C for 24 hours. The slide was then washed at room temperature with 2X SSC/0.2% SDS for 15 seconds, with 2X SSC/0.2% SDS at 42º C for the next 15 minutes, 2X SSC for the next 15 minutes at room temperature, and .2X SSC for the last 15 minutes at room temperature.

            The dyes were then hybridized onto the slide and incubated at 45º C for approximately 12 hours. These steps were performed in the dark in order to prevent the dye from fading. A final was performed at 50º C with 2X SSC/ 0.2% SDS/ 1mM DTT for 15 minutes, 2X SSC/1mM DTT at room temperature for the next 15 minutes, and .2X SSC in room temperature for the last 15 minutes. These steps were also performed in the dark.   

Microarray Analysis

Slides were sent to Davidson College for processing. When slide results came back, useful results were not present on our slide. The top half of Slide #104 (ZMS1∆ red, WT green) from Fall 2004 was then used for analysis instead. Results were analyzed using Scanalyze software. Gridding and segmentation through the program allowed us to study the amount of red/green pixels for each spot. Data was transferred into an excel worksheet. A Godlist was also used to label each spot to its corresponding gene and their function. A ratio of mutant to wildtype expression was calculated using the equation ^(CH2I-CH2B)/_(CH1I-CH1B). These ratios were converted into their log base 2 value for better analysis.

Clustering Analysis

             Using Magic Tool Software, the fifteen most under-expressed and over-expressed genes were found for both top and bottom on Slide #104 (zms1∆ red, WT green). The fifteen most under-expressed and over-expressed genes from the bottom half of Slide #106 (zms1∆ zms2∆ red, wild-type green) was also found and compared to those found in Slide #104.  Each gene's function was recorded and analyzed for relevance to oxidative stress in yeast.  

Normalization of Microarrays

            Normalization techniques were performed for Slide #104 data in order to determine whether the varying expressions of genes were due to background noise or due to actual differences between zms1∆ and wild-type. Two types of normalization was used: mean normalization and median normalization. Monday and Tuesday’s data for Slide #140 was analyzed, giving four different data sets for each gene and the first 500 genes were normalized. The data was sorted into a matrix. An example is in Figure 1.   

         Figure 1. Matrix of Microarray Data."X' denotes a gene, "N" equals the number of microarrays and "P" equals the number of genes in each microarray.

              As a reminder, for all normalization done in this experiment, each red and green background was subtracted from their intensity and then transformed into their log base 2 value. For mean normalization, the mean value of red and green intensity for each microarrray (column)  was determined. Four red intensity means and four green intensity means would result. For example, when calculating red means:

Column 1 mean = M₁ . . . Column 4 mean = M₄

              After a  red mean was calculated for each microarray, that microarray's red mean was subtracted from each of that microarray's red intensity for each gene. This was also done for the green means and green intensities. For example, when subtracting red means from  red intensities:

X_1,1- M_{1 . . .} X_1,4- M₄

X_{1, 500}- M_{1 . . .} X_{1, 500}- M₄

               All new red intensity values after mean manipulation for all microarrays were then graphed for frequency in a histogram. The same was done for all the new green intensity values for all microarrays.

                Median normalization was also done for each microarray. Again, each red and green background was subtracted from their intensity and then transformed into their log base 2 value.  The values for red and green intensity for the same gene in all four  microarrays were grouped to calculate a median for each gene.  For example, when calculating the red intensity medians:

                            M₁ = red intensity median for (X₁₁ … X₁₄)

                            M₅₀₀ = red intensity median for (X_{500, 1} … X_{500, 4})

               A median for all the red intensity medians and a median for all  green intensity medians for all genes was calculated. For example, iwhen calculating the median for all red intensity medians:

M_m = Median for (M₁ … M₅₀₀)

                Each gene's red and green intensity was then multiplied by a ratio. For example, when calculating the new value of the red intensity for a gene in a microarray #1, the red intensity value was multiplied by a ratio and that ratio is:

ratio = (M_m / A₁)

 M_m = Median for all Red Intensity Medians

A₁ = Median for (X_{1, 1} . . . X_{500, 1})

*same ratio for each gene in micrarray #1*

The ratio for each gene in microarray #2 is:

ratio = (M_m / A₂)

 M_m = Median for all Red Intensity Medians

A₁ = Median for (X_{1, 2} . . . X_{500, 2})

*same ratio for each gene in microarray #2*

This was also be done for Slide #3 and #4.

                All red intensity values after median manipulation for all microarrays were then graphed for frequency in a histogram. The same was done for all the new green intensity values for all microarrays.

                Histograms were also made for each green and red intensity of all microarrays before log base 2 values were calculated and log base 2 values without normalization.

To normalize the gene expression of each of the microarrays individually, a mean and median normalization was done. To start this, the mutant to wild-type ratio was calculated for each gene by the following formula:

ratio = (CH2I-CH2B) / (CH1I-CH1B)

The log base 2 value of the gene expression ratio was calculated after. For each microarray, the mean and median of all 500 genes was calculated. For mean normalization for a microarray, the mean for all 500 genes for that microarray was subtracted from each of the log base 2 value of each gene's expression.  For example, calculation for gene 1 in microarray 1 was:

X_{1, 1} - M₁

X_{1, 1}= log base 2 value of  (CH2I-CH2B) / (CH1I-CH1B)

M₁ = Mean of (X_{1, 1} . . . X_{1, 500})

A histogram plot of the frequencies was generated. The same technique was done for median normalization except the median was found for all 500 genes in a microarray and subtracted from the log base 2 gene expression ratio of each gene in that microarray.  A histogram plot of the frequencies was again plotted.

Expression Determination

               Expression ratios of genes from the pentose phosphate pathway and ALD6 were determined using Scanalyze for  Slides 104, 106, and 5722 using Monday's section data.  Ratios were log transformed by base 2.  The averages and standard deviations were found for each slide. Ratios were subtracted from the average and then divided by the standard deviation to determine the number of standard deviations the expression ratio was from the average.  Anything with a ratio greater than 1 standard deviation was determined to be either induced or repressed.