A05729049 1 Short Cycle Adaptation in the Circadian Rhythm of Arabidopsis PHYS 173 Spring 09 TA Ilya Valmianski 6 11 09 David Yee A05729049 A05729049 2 Abstract Bioillumination from a luciferase transgenic mutant of Arabidopsis was used to assess protein production in plants illuminated by red LED light at decreasing cycle length to view adaptation A charged coupled device was used to measure and determine the circadian rhythm by measuring the illumination output of the plants The amount of illumination produced was found to be correlated with the length of the day night cycle given to the plants indicating that the circadian rhythm is light dependant and capable of undergoing cycles shorter than natural Introduction The use of bioillumination through the Luciferase Reporter System is commonly used to study gene expression and is outlined in The Use of the Luciferase Reporter System for in Planta Gene Expression Studies By Wessel Van Leeuwen of the Wageningen Univerisity Bioillumination in Arabidopsis involves the introduction of the gene coding for luciferase production from Photinus pyralis the North American firefly through the use of an agrobacterium The gene is inserted into the chromosomal DNA of the Arabidopsis and expresses at levels that are correlated with expression levels of proteins on the same locus within the DNA From this correlation the amount of luciferase can be taken to represent the amount of natural protein being produced at a given time Bioluminescence works through the decarboxilation of a substrate luciferine that is coupled with the release of a photon of about 662 nm in wavelength ATP is also required in this reaction but is assumed to be in excess in healthy growing plants The reaction is as follows luciferin ATP luciferyl adenylate PPi luciferyl adenylate O2 oxyluciferin AMP light Experiments to measure expression levels are set up such that the protein being produced is the limiting reagent in the reaction so that for every protein being produced a photon is released Any A05729049 3 number of methods to measure this output can be used In this case a charged coupled device will be used to measure photon output As discussed in Real Time Reporting of Circadian Regulated Gene Expression by Luciferase Imaging in Plants and Mammalian Cells by David K Welsh of the Kay laboratories at UCSD protein expression in Arabidopsis is guided by a circadian rhythm that undergoes cycles in even in complete light conditions In this experiment different lighting conditions were used to witness plant adaptation 12 hour day 12 hour night cycles were used as a standard to view natural cycles and 6 hour day 6 hour night cycles were used to view adaptation Materials and Methods Figure 1 Experimental setup used to quantify the amount of bioluminescence from the plants Figure one shows the general setup used in this experiment The experiment involves a EV037 CCD that is set to take images every two hours at a 20 minute exposure time The CCD is cooled to its minimum temperature to reduce black body radiation which may increase noise in the data Output triggers from the camera control both the luciferine delivery and lighting in the experiment The output triggers are set such that 30 seconds before imaging a spray of 25 ml of 5mM luciferine is delivered to A05729049 4 the plants Luciferine is kept in dark conditions suspended in 1 Triton X solution since the luciferine is light sensitive Triton X serves as a weak detergent which disrupts the cell wall of the Arabidopsis allowing luciferine to be directly absorbed into the plant The Arabidopsis were seeded in sterilized nutrient agar and grown for approximately three weeks before experiments were conducted Lighting must remain off while imaging but otherwise lighting cycles using 40 red LED lights were used to provide day cycles of 12 and 6 hours Resulting images are shown in results Results and Data Analysis Data taken with no plants present showed pixels with non random higher illumination indicating the presence of hot spots or in this case lines in the CCD used for this experiment As discussed in Real Time Reporting of Circadian Regulated Gene Expression by Luciferase Imaging in Plants and Mammalian Cells this is common for CCDs and can be removed through a number of data analysis methods as discussed later Figure 2 20 minute exposure with no plant present Note the consistent lines of hot spots along with the large variation in saturation displayed on the color bar Data was taken with Arabidopsis for three 4 days with 12 hour day 12 hour night cycles and the mean image is shown below A05729049 5 50 1600 100 1400 150 1200 200 1000 250 300 800 350 600 400 400 450 200 500 100 200 300 400 500 Figure 3 Shows the mean image of 4 days of measurements with a 12 hour cycle Note that because the hot spots force the color bar to be of a wide range it is impossible to see smaller variances within the picture Using a filter based on higher variance helps remove unwanted pixels A05729049 6 Figure 4 Variance filtering Shows the log of the variance of each pixel which allows pixels that varies above a certain threshold to be removed In this case the threshold of a variance of about 5 was used and pixels with a variance above this were replaced with the median value Bottom shows the new mean image after the variance filter has been applied Note the change in the scale of the colorbar such that a wider smaller range of variances is represented Once invalid pixels have been removed a technique known has singular value decomposition SVD can was used to remove orthogonal modes which contain noise or do not contain any signal This technique factorizes the data as matrices to pick out a number of orthogonal modes equal to the number of pictures taken and represents them along with their intensity over time In this way shapes or patterns that vary together will remain in the same mode The separation into the first 15 modes of highest intensity is shown in figure 5 Figure 5 The first 15 modes of the first SVD decomposition Note the similarities between the first mode and noise data and the presence of plant signal in modes 2 through 10 A05729049 7 Plotting the intensities of these modes shows the relative impact each mode has on the final image We would expect to see much higher intensity for the first mode as it clearly dominates the original mean image and a decrease in intensity after mode 10 because the loss of any signal from plants after that