Felton 2 ABSTRACT The study of superoxide dismutase and Catalase is important due to their role in detoxifying reactive oxygen species. Our understanding of the functioning of these enzymes is forwarded through the assaying the activity of these enzymes in Mytilus californianus over small spatial and temporal scales. Superoxide dismutase appears to be extremely variable from individual to individual, and may not be a useful tool to generalize about reactive oxygen species trends. While statistical significance is lacking, the patterns which emerge from the Catalase activity yield compelling arguments to continue the use of the intertidal as a study system to determine the scale on which these enzymes vary and what factors drive this variation. Felton 3 INTRODUCTION Reactive Oxygen Species Reactive oxygen species (RÖS) are oxygen containing molecules like O2 (superoxide radical), H2O2 (hydrogen peroxide), and ÖH(hydroxyl radical) that include unpaired electrons. The study of these molecules, and their role in cells began in earnest in 1969 with the discovery of the function of erythrocuprein (hemocuprein) by Irwin Fridovich (McCord & Fridovich 1969). Since then, it has been established that reactive oxygen species can drive a variety of useful processes in the cell including retrograde redox signaling to other parts of the cell from the mitochondria (Miriyala et al. 201 1), as a modulator of cellular apoptosis (Fulda et al. 2010), and as a cellular defense against infection (Slauch 201 1). Reactive oxygen species also have great potential to cause irreversible damage to membrane lipids, ribosomal machinery, and DNA (Liu et al. 2011; Caporossi 2003; Roberts et al. 1997). The damage caused by RÖS either directly or through erroneous regulation of a physiological process (Miriyala et al. 201 1) is speculated to contribute to conditions like hypertension (Hirooka 2011), epilepsy (Waldbaum & Patel 2010) and Alzheimer’s Disease (Lakatos et al. 2010). The balance between RÖS production and ROS detoxification is therefore extremely important to the healthy functioning of cells. Two of the biggest cellular sources of the superoxide radical are the oxidative energy generating processes of photosynthesis (Collén et al. 2007) and aerobic respiration (Selivanov et al. 2011). Superoxide is formed when electrons are “leaked" to oxygen from the electron transport chains or photo system two (Han et al. 2001). Superoxide itself' is not a very reactive molecule, and so causes little damage itself. It does however reduce cytochrome c (a key component in metabolism) and participates in the formation of more reactive species like hydrogen peroxide, to which DNA is particularly vulnerable (Caporossi 2003). It is predicted Felton 4 that between 1 and 2% of the oxygen consumed by organisms through oxidative metabolism is converted to ROS (Roberts et al. 1997). Reactive oxygen species are an unavoidable, potentially toxic, side product of ATP production. The cells must therefore have a way to prevent RÖS from causing significant damage. Cells have a variety of defenses against RÖS, including antioxidant compounds and antioxidant enzymes. Among the cellular enzymes employed to detoxify RÖS are superoxide dismutase (SÖD) and Catalase (CAT). These enzymes work sequentially to take O2’ and convert it to H2O2 and then to water and free oxygen. This detoxification process is a key, energy consuming process, in preventing damage to the cell (Roberts et al. 1997). In order to clarify the potential for ROS damage and what conditions it might lead to, it is imperative that we study the processes that drive ROS production as well as their detoxification within a cell. The Intertidal A rocky intertidal zone appears to be an ideal place to work on understanding ROS. Since ROS is produced during aerobic respiration in heterotrophic organisms (Selivanov et al. 2011), it is reasonable to suspect that perturbations to aerobic respiration might drive perturbations in ROS production, inciting a response by the cells detoxifying machinéry (Fridovich 1986). The intertidal is often lauded as a model study system for ecological and physiological studies because of large physical gradients over short vertical distances. The intertidal is also favored for experiments due to accessibility and rapid turnover. The intertidal zone has been demonstrated to have sufficient variability in physiologically relevant factors from location to location to drive changes in aerobic respiration (Dahlhoff et al. 2001). Recent work found increasing abundance of proteins that generate NADPH, which is Felton 5 implicated in ROS defense, with acute temperature stress in blue mussels (Tomanek & Zuzow 2010). These findings suggest that the congener species, Mytilus californianus, the intertidal ribbed mussel, is well situated to be developed into a study system for understanding the balance between ROS production and RÖS detoxification. However, careful temperature measurements reveal that the scale of variability of physiologically relevant factors may be much smaller than long assumed (on the scale of centimeters to meters), and that mussels in the same bed can vary in minimum and maximum body temperatures, the rate of increase (and decrease) in temperature, and the accumulated sub-lethal thermal stress (Mark W. Denny et al. 2011). It is unclear however whether these micro-scale variations are sufficient to affect the balance of ROS. In order to determine whether temperature and other abiotic variations inherent to the intertidal affect the balance of RÖS, it must be established whether micro-scale variations affect ROS detoxification, and whether the magnitude of this effect is greater than the magnitude of the difference between mussel beds (a larger spatial scale). Enzyme activity assays like SÖD and Catalase lend themselves to determining whether micro-scale variations affect ROS detoxification since they are relatively straightforward (Beers & Sizer 1952; updated by Weydert & Cullen 2010), have been shown to be good indicators of overall RÖS concentration (Fridovich 1986). My question therefore is do four mussel beds, located near each other, always have the same relationship to each other in terms of enzyme activity? Does this activity change from one time to another? Hypothesis Given that RÖS detoxification is an energy requiring process (Roberts et al. 1997), and the resources available to a cell are limited, I believe it is reasonable to expect cells to regulate Felton 6 the potential for ROS detoxification based on their need; rather than maintaining constitutively high levels in the cell. Previous work I conducted last summer, indicates that there is sufficient variability between mussel beds (compared to between individual mussels within a bed) to drive significant changes in the activity of the metabolic enzymes malate dehydrogenase and citrate synthase. Both enzymes show significant changes in activity parsed by time and location (ANÖVA, p-value = 0.05). Assuming the link between metabolism and ROS production is robust (as suggested by Tomanek and Zuzow), I believe it is reasonable to expect a similar pattern in the activity levels of RÖS detoxifying enzymes. METHODS Study Site: China point at Hopkins Marine Station in Pacific Grove Ca. is on the leeward side of the Monterey peninsula. The geography of the point is such that the west side is more wave exposed than the east side, and this is reflected by a gradient in the distribution of classically "wave exposed" and “wave protected" species from west to east along the shore (Bell & M.W. Denny 1994). On the more wave exposed site, the rocky substratum in the mid intertidal is dominated by Mytilus Californianus, the ubiquitous ribbed mussel, which is found in temperate zones along the western coast of the United States (Morris et. al. 1980). My study site is located near the tip of the point, on either side (north and south) of a set of boulders (Fig. 1). I selected two mussel beds on the north side of the boulders, expected to be cool, one lower in the intertidal, and having an approximately horizontal aspect, and one higher in the intertidal and approximately vertical in aspect. On the south side of the boulders, expected to experience warmer temperatures than the north facing sites, I selected two mussel beds, one lower in the intertidal Felton 7 and approximately horizontal, and the other higher, on an angle, and in a surge channel. I had a total of four mussel beds in this study. Experimental Design: The sampling scheme for collecting samples to be analyzed consisted of ten sampling time points over five days of spring tides, during the first week of July, 2010. At each time point, six mussels were taken from the mussel beds, selected in a haphazard manner. They were returned to the lab, dissected, and the gill tissue frozen in liquid nitrogen, and stored in a-80°C freezer. The sampling took place each day before and after the lower low tide, which occurred in the early morning, when the water level was 3.Oft above the mean lower low water mark. The odd time points represent approximately the beginning of emersion, and even ones represent the beginning of immersion. Over a period of five days, the duration between time points remains relatively constant (approximately 6 hours between time point 1 and 2, 18 between time point 2 and 3, and again, approximately 6 hours between time point 3 and 4). Later in July, up to 0.1g of tissue (when available) from each mussel was homogenized using a TissueLyzer6 in 9 parts phosphate buffer at a pH of 7.8, yielding 1:9 dilutions homogenates. The resulting supernatant was aliquotted into two tubes, and frozen. In August of 2010, the tubes were defrosted in batches and MDH and CS activity assays were performed on the samples. The homogenates were frozen again at -80°C. In May of 2011 after establishing that SÖD and CAT are robust to at least one freeze¬ thaw cycle, the remaining homogenates were once again thawed, the aliquots of each mussel were pooled, and SOD and CAT activity assays were performed (see below for assay procedure). As a procedural control against day to day variations in the assay solution, two homogenates Felton 8 from each time point were thawed in sets of between 10 and 15 until the seventh time point was reached. The process was then repeated, yielding an n—4 for each mussel bed for the first seven time points. Äfter being assayed, any remaining homogenate was noted to have been thawed a second time, and refrozen. Superoxide Dismutase Activity Assay: This assay is adapted from McCord and Fridovich (1969). Given that the O2 radical is unstable under metabolically relevant pHs, this assay couples together a number of reactions. The generation of O2 is achieved by the oxidation of xanthine by xanthine oxidase. In the absence of SÖD, cytochrome c scavenges the O2 radicals and is so reduced. The rate at which this occurs is considered the “control rate", and this change in oxidation state of cytochrome c drives the change in absorbance at 550 nanometers, recorded by the spectrophotometer. The introduction of any SÖD present in a volume of sample lowers the rate of cytochrome c reduction by “competing" with the cytochrome c to scavenge the O2' and convert it to H2O2. The hydrogen peroxide has the potential to interact with other enzymes present in the sample, or in impure preparations of the xanthine oxidase, driving the re-oxidization of cyctochrome c. Catalase is added to the sample to break down the hydrogen peroxide and prevent confounding. For my assays, I determined that 8.7 uL of xanthine oxidase in 1.2 mL of cocktail (composed of phosphate buffer + EDTA pH 7.64, 2.5 mL of 1mM xanthine, 500 uL ImM cyctochrome C, and 100 uL 1mM Catalase) yielded a rate of cytochrome c oxidation of about 0.0200 absorbance units increase over a span of one minute. The addition of 10 uL of sample was sufficient to yield on average, inhibition between 30 and 70%, which given the hyperbolic Felton 9 nature of units of SÖD vs. % inhibition, is the range of inhibitions over which the assay is most accurate (McCord & Fridovich 1969). Each day, the average background rate of cytochrome c oxidation was determined to account for innate daily variation. Each sample was assayed three times. The data were collected in excel, and a linear regression was run on the first 20 seconds of each data file in order to avoid possible curvature due to slowing down of reactions as they run out of substrate. A visual inspection of a scatter plot of the data from each assay determined whether any small jumps or dips in the data (1 or 2 points) were likely to be significantly affecting the p-value. These values were discarded and the regression re-run. The slope of the regression was converted to units of SÖD using the relationship: units of inhibition = (standardized activity of sample) / (100 - standardized activity of sample) and standardized activity of sample = ((rate of change in absorbance of daily average background - rate of change in absorbance of individual assay) / rate of change in absorbance of daily average background rate) * 100. Given this relationship, one unit is defined as 50% inhibition of the “control rate" (rate of change in absorbance of daily average background) of cytochrome c inhibition. The p-values for each regression were checked, and those below 0.95 were discarded. The choice to reject p-values below 0.95 was arbitrary, but driven by the consistent ability to attain p-values above 0.98 (data not shown). Very few samples were rejected (less than 5%). The remaining two assays for each sample were averaged to yield the SÖD activity. For those samples that all three assays had sufficiently high p-values, the two closer trials were averaged to yield overall activity. Catalase Activity Assay: Felton 10 This assay is adapted from Beers and Sizer (1952); updated by Weydert and Cullen (2010). In this assay, hydrogen peroxide (H2O2) is broken down by CAT present in the sample. yielding a decrease in absorbance at 240 nanometers (UV range). Since there is potentially significant background rates of CAT, breaking down hydrogen peroxide already present in the sample, each sample must have a background rate obtained, by adding 25 uL of the sample to 0.5 mL of phosphate buffer (pH 7.64) and recording absorbance over two minutes. Each sample is then assayed two times in a solution of H2O2 in phosphate buffer. In general, the CAT replicates tend to be closer than SOD replicates and so it is not necessary to do them in triplicate (data not shown). The data file for each individual assay was collected in Excel, and a scatter plot used to determine whether any data points (usually the first) would be dropped. In this case, the p-values for the background rates were consistently low, which would be expected when the rate is close to zero. The background rates were subtracted from the average of the two CAT trials, unless some significant anomaly required the exclusion of one of the trials, in which case no average was obtained (less than 5% of cases) Statistical Analysis: This experiment consisted of two fixed factors, time and mussel bed. Even though the mussel beds were each sampled repeatedly, the level of interest is the individual mussel. Combined with the decision to treat time as a fixed factor, this avoids the need for a repeated measure ANOVA. The linear model for the 2 factor, Model I ANÖVA is: x = time point + mussel bed + time point * mussel bed + g; where x = SOD or CAT activity. To answer whether the Felton 11 relationship between mussel beds over time stays consistent, the term of interest is the interaction term. If it is significant, then the relationship between the mussel beds does not stay constant through time. Äfter the ANÖVA, the data was checked for homoskedasticity using Cochran’s test, and a visual inspection of residuals for gross departures from normality. Having found no significant violations to the assumptions of homoskedasticity and normality of residuals, SNK tests were performed for CAT data on time points pooled over mussel beds and for mussel beds within each time point. Visual inspections of box and whisker plots and line plots yielded further information on trends in the relationship between the mussel beds. RESULTS Superoxide Dismutase: There was no significant interaction between time point and mussel bed and no significant effect of either main factor (see Table 1). Even with pooling the MSwithin (see Table II), there is no significance. At this point, I constructed box and whisker plots to visualize the not significance. Figs. 4 and 5 show the data grouped by time point and mussel bed respectively. Catalase: The ANOVA for CAT yielded a not significant interaction term, and a not significant effect of mussel beds. It did yield a significant effect of time point however. Given this, I ran an SNK test, with a critical value = Q.o25.m.7 * V(MSeror/ 16) on time points pooled over mussel beds (See Fig. 4 for a graph of this data). I also ran an SNK test with a critical value = Qo25.m.77 * V(MSerror / 6) on mussel beds within time points because while the interaction term was not Felton 12 significant, the p-value was small enough to suggest that there is some interaction, and that examining it might yield interesting results. The data that I currently have shows little clear patterning. When comparing mussel beds within time points (refer to Fig. 5), the set of tests which should clarify whether the relationship between mussel beds changes over time, there is very little significance. This is not surprising. given that the interaction term in the ANÖVA (see Table III) is non-significant. As noted in Fig. 10, the SNK test is significant for time points 2 and 3. It is significant only over all the 4 mussel beds however, and does not resolve at the 3 mean level where both ranges are non-significant. This does not lend itself to a clear interpretation, as no bed can be said to be significantly different from the others. Setting aside strict statistical significance, there is an interesting pattern between mussel beds and a distinct change at time point four. At time one, the high sites show higher activity than the low sites, but the high and low sites switch ranking at time 2, while maintaining rank in relation to each other. From time 2 to time 3, the cool high and warm low sites converge in activity, and the remaining sites switch rank again, becoming the extreme high and low values. At time 4, all the sites are close to each other, and through times 5, and 6, remain relatively close, tracking an increase in activity. At time 7, the cool low, and warm high sites drop in activity compared to the other two which drop only a little. Post-hoc comparisons of CAT activity between sampling times, averaged over mussel beds, reveals a significant difference between time point 6 and the group of time points 1,2, and 3. There was no significant difference between time points 4, 5 and 7, and either time point 6 or the group of time points 1, 2, and 3 (Fig. 4). DISCUSSION Felton 13 Superoxide Dismutase: The activity level of superoxide dismutase is extremely variable from mussel to mussel, even within a single time point, from a single bed. It therefore appears to be no ranking of mussel beds, and no change over time (Figs. 4 and 5). Rather, the four beds appear to be part of a single population, despite the predicted differences in factors we might otherwise expect to drive differences in the activities of this enzyme. This may be an artifact of the inherent variability of the assay, or an accurate reflection of true variation. In order to determine this, it is necessary to quantify the inherent variation in SÖD. To approach this question, a large sample of mussels should be brought into the lab and held under common garden conditions for 4 weeks to acclimate before assaying SOD activity. Of further interest is the cause of the variation between mussels. It is unclear whether my assumption that a relationship between temperature and enzyme amounts is unsound, or whether in this case, enzyme concentration is a poor proxy for enzyme activity. Assuming the relationship is sound, the inter-mussel variation may be driven by variation in the phenotypic plasticity of individual mussels (Helmuth et al. 2010), or by inherent differences in ability to handle ROS (genetic differences) (Roberts et al. 1997). It is also possible that the variability is simply the steady state for each mussel, and at no time during the sampling where mussels stressed to a level high enough to trigger an RÖS detoxification response. Existing studies on the effect in vivo exposure to high temperatures on the in vitro activity levels of SOD is limited. Studies in soybean seeds suggest that a departure of 25°C from optimal temperature is necessary to insight a detectable detoxification response that led to increased survivorship and successful germination (Posmyk et al. 2001). Studies in the Antarctic limpet Nacellla concinna suggest that an increase of even 9°C above 0°C is sufficient to drive Felton 14 significant changes in SOD activity, explained at least partially by Q1o effects (Abele et al. 1998). The question remains open then as to whether the mussels in this experiment experienced a strong enough stimulus to generate an RÖS detoxification response. Recent work in mussels found changes in the amount of RÖS detoxifying enzymes present in Mytilus trossolus and Mytilus galloprovincialis driven by the max temperature in an acute heat stress event (24 - 32°C) (Tomanek & Zuzow 2010). Further work indicates that there are differences in gene expression of enzymes involved in oxidative stress between cold and warm adapted congeners in the genus Mytilus (Lockwood et al. 2010) when exposed to temperatures up to 32°C. Given current understanding of the relationship between temperature and oxidative stress, it may be advantageous to conduct heat ramps in the laboratory to determine the response of SÖD in order to establish levels at which a coherent response is elicited. Catalase: Given the intriguing pattern that appears to be present for the first 3 time points, that is then lost for times 4, 5 and 6, and appears to be changing yet again at time 7, it is imperative that I add the remainder of the data set. This will increase the N to 6 (from 4) and the number of time points to 10. Assuming that the patterns visible now reflect what is really going on, and additional data will serve to simply strengthen the pattern, we can speculate now on the drivers behind the patterns. Fig. 10 demonstrates that the activity levels of the mussel beds are changing in relation to one another. This suggests that whatever is driving the activity levels, is not applying the same pressure to each mussel bed through time. Öthers have shown that temperature and salinity correlate with changes in protein expression (Lockwood & Somero 2011; Tomanek & Zuzow Felton 15 2010). Temperature is variable in the intertidal on scales small enough to reasonably expect two mussel beds at the spatial scale of my experimental beds to vary (Helmuth et al. 2010). It is possible that differences in the temperature regime faced by the various mussel beds are the underlying driver for the changing enzyme activity levels. Of particular interest is the significant difference in mussel bed means, seen at times 2 and 3, which is lost at time 4 and not regained through time 7. This suggests that either some disturbance was present at time points 2 and 3, or some disturbance began at time point 4. Whether it was an acute event which elicited a response that took some time to ramp up (time 6 has higher activity than time 4), an acute event which elicited a response that remained high for some time afterward as a safe guard, or a chronic stress event that may have begun to alleviate at time 7 is unclear. The nature of this perturbation and the response has implications for cellular energy budgets (Roberts et al. 1997) which may lead to decreased fitness, decreased competitive ability and eventually changes in community structure (Society 2011). Interpreting the Catalase results is further complicated by the reality that Catalase is not the only enzyme to break down H2O2, and superoxide is not the only source. glutathione reductase also breaks down HpO2 to water and oxygen via a different pathway (Posmyk et al. 2001). Additionally, numerous peroxidases are present in high concentrations in certain organelles, and produce H2O2 (Roberts et al. 1997). Therefore, the activity levels observed are actually the end result of a number of processes that affect the concentration of the substrate (H2O2) and therefore the need for detoxification by Catalase. Further Study: Felton 16 Besides finishing this data set, further tests are needed to contextualize this study. Given the consistency in the assay, and the sensitivity to inter-individual variation, the Catalase assay should be the focus of further study. The advantages of working in the field are balanced by the disadvantage of being unable to control everything. There is significant potential for confounding, particularly in a design like this one, which controls for tidal height. Differences may be due to temperature, emersion, desiccation, tidal or diurnal cycles, etc. Patterns of enzyme activity that correlate with the tidal cycle have been found in metabolic enzymes (malate dehydrogenase and citrate synthase) in M. californianus (Laurie Kost, personal communication). A further study in the field would focus on sampling over a tidal cycle, to try to determine if there are also cyclic changes in the activity of Catalase. Laboratory experiments that would help clarify the potential effect of abiotic factors include heat ramps and ambient temperature emersion. Further limitations on this study include that we cannot determine whether the changes in activity level are due to changes in the concentration of the enzyme, or post translational modifications. Western blotting may be able to determine whether the concentrations of the enzymes are changing. There is little indication that post translational modification of either SOD or CAT is employed in vivo for activity regulation, but may be elucidated with careful 2-D gel work. As discussed above in the SÖD discussion section, it is unclear whether levels of stress were high enough to induce a response. One way to get at this question post-experiment is to assay DNA damage. Since RÖS target macromolecules including DNA if they are not detoxified, an increase in DNA damage might suggest that RÖS damage is occurring, and the detoxifying mechanisms are either not responding, or are overwhelmed. A lack of damage Felton 17 would suggest that the levels of SÖD observed are sufficient to prevent RÖS related DNA damage. Conclusion: This study reveals that in the intertidal, small spatial and temporal variations in abiotic factors are sufficient to drive significant changes in Catalase enzyme activity. These changes in activity are thought to be closely correlated with changing concentrations of the reactive oxygen species that are this enzyme’s substrate. These results therefore suggest changing concentrations of RÖS, which we know have significant capability to cause damage in cells if not detoxified. This study helps lay the groundwork to elucidate the relationship between RÖS production, RÖS damage, and ROS detoxification. Acknowledgements: I would like to thank Professors George Somero and Mark Denny for their guidance in designing this project and interpreting the results, and Dr. Wes Dowd, Laurie Kost, and Helen Heysmann for their help in collecting and processing samples and performing the lab work. I would also like to thank Professor Steve Palumbi and Drs. Sarah Lee and Vanessa Michelou for their guidance through the 175H course. Felton 18 WORKS CITED: Abele, D. et al., 1998. Exposure to elevated temperatures and hydrogen peroxide elicits oxidative stress and antioxidant response in the Antarctic intertidal limpet Nacella concinna. Comparative Biochemistry and Physiology Part B: Biochemistry and Molecular Biology, 120(2), pp.425-435. Beers, R.F. & Sizer, I.W., 1952. A SPECTROPHOTOMETRIC METHOD FOR MEASURING THE BREAKDOWN OF HYDROGEN PEROXIDE BY CATALASE. JOURNAL OF BIOLOGICAL CHEMISTRY, 195(1), pp.133-140. Bell, E.C. & Denny, M.W., 1994. Quantifying“ wave exposure": a simple device for recording maximum velocity and results of its use at several field sites. Journal of Experimental Marine Biology and Ecology, 181(1), p.9-29. Caporossi, D., 2003. Cellular responses to H202 and bleomycin-induced oxidative stress in L6C5 rat myoblasts. Free Radical Biology and Medicine, 35(1 1), pp.1355-1364. Collén, J. et al., 2007. Response of the transcriptome of the intertidal red seaweed Chondrus crispus to controlled and natural stresses. The New phytologist, 176(1), pp.45-55. Dahlhoff, E.P., Buckley, B.A. & Menge, B.A., 2001. Physiology of the rocky intertidal predator Nucella ostrina along an environmental stress gradient. ECÖLÖGY, 82(10), pp.2816-2829. Denny, Mark W. et al., 2011. Spreading the risk: Small-scale body temperature variation among intertidal organisms and its implications for species persistence. Journal of Experimental Marine Biology and Ecology. Felton 19 Fridovich, I., 1986. BIOLOGICAL EFFECTS OF THE SUPEROXIDE RADICAL. ARCHIVES OF BIOCHEMISTRY AND BIOPHYSICS, 247(1), pp.1-11. Fulda, S., Galluzzi, L. & Kromer, G., 2010. Targeting Mitochondria for Cancer Therapy. Nature Reviews Drug Discovery, 9(6), pp.447-464. Han, D., Williams, E. & Cadenas, E., 2001. Mitochondrial respiratory chain-dependent generation of superoxide anion and its release into the intermembrane space. Biochemical Journal, 353 (Pt 2), p.411. Helmuth, B. et al., 2010. Organismal climatology: analyzing environmental variability at scales relevant to physiological stress. The Journal of experimental biology, 213(6), pp.995-1003. Hirooka, Y., 2011. Oxidative stress in the cardiovascular center has a pivotal role in the sympathetic activation in hypertension. Hypertension research : official journal of the Japanese Society of Hypertension, 34(4), pp.407-12. Lakatos, A. et al., 2010. Association between mitochondrial DNA variations and Alzheimer’s disease in the ADNI cohort. Neurobiology of aging, 31(8), pp.1355-63. Liu, T. et al., 2011. Roles of reactive oxygen species and mitochondria in cadmium-induced injury of liver cells. Toxicology and industrial health, 27(3), pp.249-56. Lockwood, B.L., Sanders, J.G. & Somero, G.N., 2010. Transcriptomic responses to heat stress in invasive and native blue mussels (genus Mytilus): molecular correlates of invasive success. The Journal of experimental biology, 213(Pt 20), pp.3548-58. Felton 20 Lockwood, B.L. & Somero, G.N., 2011. Transcriptomic responses to salinity stress in invasive and native blue mussels (genus Mytilus). Molecular ecology, 20(3), pp.517-29. McCord, J.M. & Fridovich, I., 1969. SUPEROXIDE DISMUTASE AN ENZYMIC FUNCTION FOR ERYTHROCUPREIN (HEMOCUPREIN). JOURNAL OF BIOLOGICAL CHEMISTRY, 244(22), pp.6049-& amp;. Miriyala, S., Holley, A.K. & St Clair, D.K., 2011. Mitochondrial Superoxide Dismutase - Signals of Distinction. Anti-cancer agents in medicinal chemistry, 1 1(2), pp.181-190. Posmyk, M.M. et al., 2001. Ösmoconditioning reduces physiological and biochemical damage induced by chilling in soybean seeds. Physiologia Plantarum, 111(4), p.473-482. Roberts, D. a, Hofmann, G.E. & Somero, G.N., 1997. Heat-Shock Protein Expression in Mytilus californianus: Acclimatization (Seasonal and Tidal-Height Comparisons) and Acclimation Effects. Biological Bulletin, 192(2), p.309 Selivanov, V.A. et al., 2011. Reactive oxygen species production by forward and reverse electron fluxes in the mitochondrial respiratory chain. PLoS computational biology, 7(3), p.e1001115. Slauch, J.M., 2011. How does the oxidative burst of macrophages kill bacteria? Still an open question. Molecular microbiology, 80(3), pp.580-3. Society, E., 2011. The Influence of Interspecific Competition and Öther Factors on the Distribution of the Barnacle Chthamalus Stellatus Author (s): Joseph H. Connell Felton 21 Published by : Ecological Society of America Stable URL : http://www.jstor.org/stable/1933500. THE INFL. America, 42(4), pp.710-723. Tomanek, L. & Zuzow, M.J., 2010. The proteomic response of the mussel congeners Mytilus galloprovincialis and M. trossulus to acute heat stress: implications for thermal tolerance limits and metabolic costs of thermal stress. The Journal of experimental biology, 213(Pt 20), pp.3559-74. Waldbaum, S. & Patel, M., 2010. Mitochondria, oxidative stress, and temporal lobe epilepsy. Epilepsy research, 88(1), pp.23-45. Weydert, C.J. & Cullen, J.J., 2010. Measurement of superoxide dismutase, catalase and glutathione peroxidase in cultured cells and tissue. Nature protocols, 5(1), pp.S1-66. Felton 22 Table I: Superoxide Dismutase Analysis of Variance. A p-value of 0.05 is considered significant. Sum-of¬ F-ratio Source df Mean-Square Squares TIME POINT 0.039 0.217 0.007 0.970 MUSSEL BED 0.499 0.797 0.072 0.024 TIME POINT MUSSEL BED 0.308 18 0.017 0.568 0.912 Error 2.379 79 0.030 Felton 23 Table II: Superoxide Dismutase analysis of variance, error and interaction term pooled. The new error term mean square consists of the pooling of the previous error term with the interaction term. This entails adding the sums of squares, and dividing by the combined degrees of freedom. A p-value of 0.05 is considered significant F-ratioP Source Sum-of-Squares df Mean-Square 0.039 TIME POINT 0.236 0.964 0.007 0.024 MUSSEL BED 0.072 0.866 0.461 2.687 97 Error 0.028 Felton 24 Table III. Catalase activity analysis of variance. The interaction term is not significant, but too small to dismiss entirely and thus pool with the error. The main effect of mussel bed also not significant. The main effect of time point is highly significant however. Given the huge number of degrees of freedom, the ANÖVA is extremely sensitive. F-ratio Source Sum-of-Squares df Mean-Square TIME POINT 4.726 6 49682.012 0.000 298092.075 MUSSEL BED 0.409 0.976 3 10255.856 30767.567 TIME POINT 266397.759 18 14799.875 1.408 0.152 MUSSEL BED Error 809474.287 77 10512.653 Felton 25 Figure Legends: Fig. 1. An aerial view of Hopkins Marine Station in Pacific Grove Ca. The rectangle notes the approximate location of my two study sites. Fig. 2. Superoxide dismutase activity by time point. Each box and whisker plot represents one time point, with an N = 16. The values charted are the 16 mussels for the time point, pooled across all mussel beds. Fig. 3. Superoxide dismutase activity by mussel bed. Each box and whisker plot represents one mussel bed with an N=28. The data plotted are the 28 samples drawn from each mussel bed over the 7 time points. Fig. 4. Catalase activity at each time point, averaged over the four mussel beds. This figure shows the mean and one standard error of Catalase activity for each time point, pooled over the four mussel beds. There is significance between time points 1, 2 and 3, and time point 6 as indicated by the asterisks. Fig. 5. Catalase activity by time point for each mussel bed. The Catalase activity observed in the four mussel beds are plotted on the same graph, against time in order to visualize how the plots Felton 26 are changing in relation to each other through time. The asterisks indicate significant SNK results at the m=4 level (which was the only level to demonstrate significance). The SNK result for time point 1 was close to significant. Time points 4 and 6 were strongly not significant. Error bars represent standard error. "M“ in an SNK test refers to the number of means being compared in the range. M=4 indicates that all four mussel beds are included in the comparison. 4- I.. Fig. 1 Felton 27 1.0 0.9 0.8 0.7 0.6 0.4 0.3 0.2 0.1 0.0 1 Fig. 2 1 3 4 5 6 7 TIMEPOINT Felton 28 1.0 0.9 0.8 0.7 0.6 0.5 0.4 03 0.2 0.1 0.0 Fig. 3 2 Mussel Bed Identity. Felton 29 Legend 1- Warm High 2-Warm Low 3-Cool High 4-Cool low 600 500 400 300 200 100 -100 * 1 * 3 4 5 6 7 TIMEPOINT I Fig. 4 Felton 30 600 500 400 300 200 100 0 -100 Fig. 5 kaataaa- — I L * 1 2 3 4 5 6 7 TIMEPOINT Felton 31 Legend: COOLWARM, HIGHLOW Cool, High Cool, Low Warm, High Warm, Low