Abstract Many poikilotherms exhibit adaptive changes in cellular membrane fluidity (homeoviscous adaptation) to counter potentially disruptive effects of thermal variation in their environments. The process of homeoviscous adaptation was studied in an intertidal snail, Littorina keenae, which encounters wide variation in body temperature due to differences in microhabitat conditions and the effects of tidal rhythms. This study investigated the effects of thermal variation arising from microhabitat differences between two geographic sites and from tidal cycle effects within a single site on membrane fluidity. L. keenae, a species widespread along the North American Pacific Coast, was studied at two sites on Cabrillo Point at Hopkins Marine Station in Central California, one site being wind and wave exposed (exposed site), the other relatively protected by its cove-like orientation (protected site). Protected site individuals exhibited body temperature fluctuations as great as 20°C within a six-hour period (10 to 30°C). whereas the exposed population exhibited only a 10°C range in the most extreme case (10 to 20°C). Fluidities of membranes of gill and mantle tissue were quantified by measuring fluorescence anisotropy using the membrane probe 1,6-diphenyl-1,3,5- hexatriene (DPH). Samples from the exposed site population collected prior to and following midday low tide exhibited no differences in membrane viscosity. However, in one instance exposed individuals collected at the same time showed two distinct levels of membrane fluidity, suggesting disparate thermal histories. In addition, comparisons with previously published data on membrane fluidity for other species may indicate that L. keenae is compensating toward its upper thermal range. Introduction Biological membranes are structurally and functionally essential to all living organisms. The cell membrane is generally depicted as comprising a dynamic, fluid phospholipid bilayer, held together by weak hydrophobic interactions among the non¬ polar acyl-chain regions of the phospholipids. These interactions in part allow mobility in the phospholipids as well as other integral membrane molecules (e.g., glycoproteins, channels, cholesterols) (Singer and Nicholson, 1972). The fluidity of the membrane depends on several factors, including composition the lipid acyl chains and temperature. Membrane fluidity is positively correlated with the degree of lipid acyl chain desaturation, because the double bonds introduce kinks that prevent efficient packing of the lipids into the bilayer. More saturated acyl chains favor denser compaction of the membrane lipids, composing a more ordered membrane (Cossins and Prosser, 1977; Farkas and Roy, 1989). Increasing temperature also promotes membrane fluidity by enhancing molecular motion (for review see Hochachka and Somero, 2002). The weak hydrophobic interactions leave the membrane vulnerable to the possibly deleterious effects of extreme temperatures. Thermal disruption of the membrane state can critically impact the essential structure and function of the membrane, indicating a need for a limited temperature range in which membrane constituents can most effectively function (for review see Hazel, 1995). For organisms inhabiting different thermal niches. adjustments to membrane composition may be required to maintain an effective membrane state in spite of thermal perturbations. One way an organism could respond to the threat of thermal disruption to physical membrane state is by altering the ratio of saturated to unsaturated lipid molecules in the bilayer. For example, organisms coping with extreme heat would need to stabilize their membranes, restructuring lipid composition to a higher proportion of saturated molecules; whereas an organism challenged by extremely cold conditions would want to increase the fluidity (lipid desaturation) of its membranes to counteract the stabilizing effect of the low temperatures. This adaptive conservation of a biologically optimal range for membrane fluidity is referred to as homeoviscous adaptation (Sinensky, 1974). Many studies have investigated the existence and extent of homeoviscous adaptation through comparisons among poikilothermic microorganisms and across vertebrate taxa (Sinensky, 1974; Cossins et all, 1984; Behan-Martin et al., 1993; Logue et al., 2000). Nevertheless, few studies have explored structural compensation in membranes for thermal stress among invertebrate species (Dahlhoff and Somero, 1993; Williams and Somero, 1996). Most of the studied vertebrate organisms tend to live in environments of relative thermal stability. However, for invertebrate poikilotherms inhabiting marine intertidal habitats, extreme temperature fluctuations can occur on a daily basis during low tide periods, and these organisms face an intriguing homeostatic challenge. One study of membrane fluidity in intertidal mussels illustrated seasonal acclimatization in gill phospholipid structure associated with wide ranges in body temperature, and alteration of membrane order was also observed during a laboratory simulation of a hot low tide (Williams and Somero, 1996). An investigation of membrane fluidity in animals sampled directly from intertidal microhabitats following high and low tide periods would provide further insight into the energetic compromises that organisms must make in the face of environmental stressors. To explore the existence and scope of acclimatory membrane alterations in intertidal organisms facing myriad physiological challenges in the field, I studied membrane fluidities in the intertidal snail L. keenae, an invertebrate abundant in the marine intertidal of the Central Coast of California. The microhabitats of L. keenae examined in this study expose the animals to a range of convective heat loss from wind and wave forces, and render them vulnerable to varying degrees of solar heat radiation, intensified by reflected heat emanating from the granite boulders inhabited by the snails. Being small poikilotherms, L. keenae are at the mercy of their thermal environments. In this study, I examined the membrane fluidity of L. keenae occupying two specific microhabitats on Cabrillo Point at Hopkins Marine Station on California’s Central Coast: a protected site and an exposed site. Because of its geographic orientation, the protected site is sheltered from wind and swell, but faces the sun; whereas the exposed site is frequently susceptible to wind, spray, and waves. At both sites, L. keenae are situated in the upper regions of the intertidal zone, about three to four feet above the mean low water level, and are not regularly submerged (pers. obs.). Both sites are subject to seawater spray during high tide, and both are air-exposed during low tide (pers. obs.). Environmental variations between the two sites suggest a possibility for two different levels of thermal acclimatization, while thermal variations correlated with tidal rhythms provide a second point of comparison. I hypothesized that snails situated on the exposed site or snails recently exposed to high tide seawater spray, might have higher membrane fluidities over a range of temperatures than would snails from the protected site following air and heat exposure during a warm low tide period. Materials and Methods Field Collection Two variables were considered for this experiment: time and location. To test for differences in membrane fluidity within one site at different times of day, specimens were collected from the exposed site just before sunrise following a high tide, and again from the exposed site in the mid-afternoon immediately following a midday low tide. To test for variation between the exposed site and the protected site, samples from each location were collected at the same time of day: mid-afternoon, following a midday low tide. For both comparisons, specimens were collected from the field using forceps to minimize thermal conduction that might affect body temperature measurements. Body temperatures were obtained by inserting a type-K thermocouple probe into the foot tissue of the organism, and the specimens were then flash-frozen and kept on solid CO2 until returned to the lab where they were stored at -80°C. Tissue preparation Each specimen was dissected in iced seawater for the extraction of gill and mantle tissues. Following dissection, the sample tissues were stored at -80°C until tested. Frozen tissue from each snail was homogenized in 3.5mL of 50mM phosphate buffer, using a ground glass homogenizer. All procedures were performed on ice. I then divided the homogenized tissue suspension between two 2mL centrifuge tubes and centrifuged both at 80g (1000 RPM) for three minutes at 4°C. The supernatant from both tubes was transferred to a 3mL cuvette, and the absorbance was measured at a wavelength of 364nm. The supernatant was diluted to A364-0.15 and placed in a plastic fluorescence cuvette, into which 2uL of 1.8mM DPH solution was added for a final DPH concentration of 3.6nM in the cuvette (see below for DPH preparation). Cuvettes were then sealed with a thin layer of parafilm to prevent dust or lint contamination, and stirred for thirty minutes in the dark, allowing for the diffusion of the DPH molecule into membranes. DPH preparation I dissolved 2.3 mg 1,6-diphenyl-1,3,5-hexatriene (DPH) crystals into 5 mL N'-N'- dimethylformamide for a 1.8mM DPH concentration and covered the solution in foil to prevent the photo-degradation of DPH. The solution was then placed on a stir bar to mix for one hour prior to being introduced into the tissue preparation. Fluorescence Anisotropy Measurements All cuvettes were then placed into a water bath at 10°C for approximately 15 minutes for thermal equilibration. I placed each cuvette into the water-jacketed cell of the spectrofluorometer. Parameters for fluorescence polarization measurements were as follows: excitation, slit width = 3 nm, A = 364.0 nm; emission, slit width = 10 nm, X - 430.0 nm; response time = 8 s; repetitions = 5; recording range = 0 to 1000. After checking the temperature of the mixture within the cuvette (using a K-type thermocouple), I took polarization measurements at the following polarized filter angles: 0°,0°; 90°,0°; 90°,90°; 0°,90°; for excitation, emission angles respectively. This series of measurements composed one trial. For each sample, two trials were repeated at each temperature, and the actual temperature within the cuvette was re-measured with the thermocouple immediately prior to each trial. This procedure was repeated at each of the following temperatures: 10°C, 15°C, 20°C, 25°C, 30°C, and 35°C. Values given for the emission of polarized light at each filter angle were recorded and used to determine the fluorescence polarization of the sample for that trial (Litman and Barenholz, 1982). Fluorescence polarization values (P) - which are inversely proportional to membrane fluidity - were calculated using the following equation (Litman and Barenholz, 1982): 190,0 (0,0)- 10,90 0,90 P- 10,0 (0,0) 10,90 (J0,90 The polarization value was calculated for each replicate and averaged to obtain the fluorescence polarization value for the sample. Statistics Regression lines from the relationship of fluorescence polarization and temperature were generated for each sample. The slopes of these regression lines were used for statistical comparisons (i.e., exposed site versus protected site, and exposed site morning collection versus exposed site afternoon collection) using a single-factor analysis of variance (ANOVA). The ANOVA was also used to compare estimated polarization values at common temperature (20°C) for the same comparisons. In the case of the exposed vs. protected site comparison, a one-tailed student's t-test was also applied. In all cases, a S 0.05. Results Body temperatures for L. keenae fluctuated with tide and time of day at both sites, but organisms at the protected site were found to experience more extreme fluctuations than those at the exposed site. Body temperatures (Tp) recorded at the exposed site ranged from 9°C to 19°C within a six-hour period; whereas measurements at the protected site exhibited a wider fluctuation, from 10°C to 32°C. Fluorescence polarization values for every sample fell between 0.22 and 0.31 across the tested temperature range. As the temperature increased, polarization values declined (enhanced membrane fluidity). Exposed site samples collected in the morning and afternoon showed similar membrane fluidities. Specimens collected before sunrise from the exposed site following high tide had a mean Tp of 13°C, and membrane fluidities were similar for all snails collected. The second group of exposed specimens collected in the afternoon following a midday low tide had a mean Tp of 18°C and resulted in two pairs of differing membrane fluidities. One pair closely matched the fluidities of the morning collection, while the other pair displayed lower polarization values across the temperature range. (Fig. 1). The effect of temperature on fluidity was significantly different between the morning and afternoon samples (ANÖVA, p-0.02), but this difference disappeared when I excluded the two individuals having considerably lower polarization values (ANÖVA, p-0.07). (Fig. 2). Estimated fluorescence polarization values at 20°C (disregarding the two individuals having low polarization values) of morning and afternoon samples were not significantly different (ANOVA, p-0.92). Although the comparison between morning and afternoon collections at the exposed site yielded no significant variation, membrane fluidities between exposed and protected snails were different. Fluorescence polarization values for samples from the exposed site following a midday tide (average Tp =16°C) were lower than those collected from the protected site at the same time on the same day (average Tp -25°C). (Fig.3). The effect of temperature on fluidity was not significantly different between sites (ANOVA). However, the membrane fluidities of protected and exposed snails were statistically different across the temperature range: using 20°C as a point of comparison, estimated fluorescence polarization values for that temperature were compared using two statistical methods. The one-tailed t-test showed a significant difference between the exposed site and protected site fluidity measurements, p-0.014, while the more conservative ANOVA gave a barely non-significant p of 0.075. Discussion Littorina keenae living in the intertidal microhabitats of Cabrillo Point on California’s Central Coast regularly experience changes in ambient temperature as large as 20°C in a six hour period. I recorded body temperature fluctuations spanning a range of 10°C for L. keenae inhabiting the exposed site, while L. keenae at the protected site experienced a range of 20°C for the same time period. This pattern makes intuitive sense, as the exposed site is more consistently vulnerable to heat loss through wind convection and to seawater spray, even during hot low tides. Conversely, organisms inhabiting the protected site are sheltered from both wind and swell. On hot days, these animals would be unable to cool themselves, either convectively to wind or conductively to water. During high tides, both sites are subject to seawater spray, so one might expect both sites to experience similar minimal body temperatures. The Tp values recorded over the course of this study precisely mirror this expectation. The recorded Tp values for the exposed site ranged from 10°C at high tide to 19°C at low tide; the Tp values for the protected site varied between 11°C and 32° For the exposed and protected samples that were collected following a midday low tide period, one might expect the protected site collection to have had a warmer thermal history than the exposed site collection, and indeed the average body temperatures for individuals in that collection from the protected site were 25°C, while the average Tp for exposed site individuals was nine degrees cooler at 16°0. Thus I predicted that the protected snails, with a relatively warm thermal history, would compensate for the heat-induced increase in membrane fluidity by stabilizing its tissue membranes. The membrane fluidities of the exposed site collections taken at different times of day reflect the relatively narrow range of thermal fluctuation observed at the exposed site. Although inclusion of all of the afternoon exposed site samples in statistical comparisons returned significant differences between morning and afternoon exposed-site collections, when the two afternoon samples displaying lower polarization values were removed, the differences disappeared. The rationale for the removal of these data follows. In evaluating the slopes of each individual tested over the course of the study, every tissue preparation demonstrated a decrease in fluorescence polarization as temperature increased. The low polarization values for the two individuals in the afternoon collection group not only were inconsistent in their raw polarization values, but the slopes for both individuals were also nearly zero (-0.0004 and -0.0007 fluorescence polarization per °C, respectively), indicating the physically unlikely lack of change in membrane fluidity with respect to temperature. This insensitivity to temperature led me to suspect that the preparation of the sample was flawed. Therefore, I included two graphical and statistical analyses for the same comparison, one including and one excluding the inconsistent data. (Figs. 1 and 2). However, for the purposes of discussion, 1 will refer to the data that exclude the anomalous samples. No significant differences in membrane fluidity could be seen between the exposed group collected in the early morning following high tide and the exposed group collected later that afternoon following low tide (ANÖVA, p-0.07). Based on the assumption that any adjustments in membrane composition would occur in response to differing thermal histories, a lack of alteration may indicate a lack of thermal disparity between the two groups. The Tp values for the morning collection averaged 13°C and the mean Ty for the collection was 18°C, so the average body temperature range remained relatively narrow. Moderate changes in body temperature in such a short period may present less of a cost to the organism than would an energy expenditure to restructure its membranes during that time. The disparity between the two thermal environments (exposed and protected) is reflected in the membrane orders of organisms taken from each site. (Fig.3). Over a range of temperatures, samples collected from the protected site exhibit higher membrane stability than do those individuals from the exposed site. To quantify this difference, I used both a single-factor ANÖVA and a student's t-test for the estimated polarization values at 20°C. Because 20°C lies in the middle of the data set, using this x-value would 11 show the least amount of error in the estimation of the slope, giving the most accurate prediction of a common polarization value. The one-tailed t-test showed a significant difference between the two estimated polarization values (p-0.014), but the conservative ANOVA gave p-0.07, marginally insignificant. Variation in membrane order between the exposed and protected sites, though apparent and consistent in graphical representation, can be considered at best marginally significant, with only four individuals from each site for replication, and only the least conservative statistical test giving significant results. However, the subtle implication of one significant result suggests the possibility that organisms inhabiting more protected geographical regions of Cabrillo Point may have adjusted the composition of their tissue membranes to compensate for the often extreme changes in thermal stress. The polarization values imply that at a given temperature within the tested range, protected- site L. keenae may have more stable membranes than do L. keenae at the exposed-site. A comparison between the estimated polarization values for the two groups at their respective average Th values (14°C and 24°C) reveals a striking similarity. The predicted value for the exposed site is 0.276, and 0.270 for the protected site, suggesting conservation of a specific, target membrane state even between differing thermal microhabitats. In context of previous interspecific vertebrate studies, the snails' conservation of a fluorescence polarization value of 0.27 at average Tp is much higher than other conserved fluidity ranges observed in vertebrates. In a 1993 study by Behan-Martin et al., membrane fluidities for several vertebrate species were compared by plotting the temperature at which they displayed a polarization value of 0.22 against their accustomed 12 body temperature. The resulting line is exceptionally conserved. (Fig. 4). The equivalent value for L. keenae was included as an *X" and is clearly divergent from the data of Behan-Martin (1993). (Fig. 4). The disparity between membrane fluidity conservation in L.keenae and fluidity conservation among the vertebrate data might be due to several factors, the most parsimonious of which would be to assume error in my own data collection. The illustrated vertebrate polarization values were obtained using purified lipid vesicles, while my data were obtained using crude homegenates of both gill and mantle tissues. The inconsistency could also be a result of phylogenetic differences between vertebrates and invertebrates, though more data concerning invertebrate membrane fluidities would be required to make that distinction. Nevertheless, membrane fluidities of L. keenae are not as divergent from other published data as suggested by the above analysis. Inclusion of my data on a plot of previously published polarization values as a function of temperature (Logue, et al., 2000) indicates that L. keenae seems to be conserving its membranes in a more ordered state than the given vertebrates. (Fig.5). The range for L. keenae’s polarization values is comparable to the illustrated endothermic vertebrates, suggesting that L. keenae is acclimatized to higher body temperatures. However, the slopes for the snail polarization values are less steep than the most similar vertebrate slopes. One explanation might be that L. keenae is routinely exposed to the upper thermal range included in this figure and would be adjusting membrane fluidity to these warm temperatures. Both the similarity of L. keenae to warm-adapted animals and the presence of a more gradual slope in the snail data suggest the possibility that keenae is compensating toward its upper thermal limits. Conclusions L. keenae inhabiting the marine intertidal environment at Cabrillo Point experience rapid and dramatic fluctuations in ambient and body temperatures. Snails in the microhabitat of the protected site see warmer, more extensive fluctuations than do the exposed site snails. Samples collected from the exposed site in the morning and later that afternoon showed no significant differences in the membrane states between the two groups. Conversely, the comparison between exposed and protected sites following a midday low tide revealed slightly significant differences that may imply acclimatory alterations in membrane compostion. At the very least, this subtle dissimilarity demands closer inspection. Acknowledgements I would like to thank my advisors George Somero and Jonathon Stillman for their extensive guidance and communication in working on this project. In addition I would like to express my gratitude to the Somero Lab group for their support and assistance, as well as to express my appreciation for the support and generosity of the faculty and staff of the Stanford Hopkins Marine Station. Literature Cited Behan-Martin, M.K. and G.R. Jones, K. Bowler, A.R. Cossins. 1993. A near perfect temperature adaptation of bilater order in vertebrate brain membranes. Biochim. Biophys. Acta. 1151: 216-222. Cossins, A. R. and C. L. Prosser. 1977. Evolutionary adaptations of membranes to temperature. Proc. Natl. Acad. Sci. USA. 75: 2040-2043. Dahlhoff, E., and G. N. Somero. 1993b. Effects of temperature on mitochondria from abalone (genus Haliotis): adaptive plasticity and its limits. J. Exp. Biol. 185: 151- 168. Farkas, T. and R. Roy. 1989. Temperature-mediated restructuring of phosphatidylethanolamines in livers of freshwater fishes. Comp. Biochem. Physiol. 93B: 217-222. Hazel, J.R. 1995. Thermal Adaptation in Biological Membranes: Is Homeoviscous Adaptation the Explanation? Annu. Rev. Physiol. 57: 19-42. Hazel, J.R. and E. Williams. 1990. The Role of Alterations in Membrane Lipid Composition in Enabling Physiological Adaptation of Organisms to Their Physical Environment. Prog. Lipid. Res. 29: 167-227. Hochachka, P.W. and G. Somero. 2002. Temperature. pp. 290-379 in Biochemical Adaptation: Mechanism and Process in Physiological Evolution, Oxford University Press, New York, NY. Litman, B.J. and Y. Barenholz. 1982. Fluorescent probe: diphenylhexatriene. Meth. Enzymol. 81: 678-685. Logue, J. A. and A.L. De Vries, E. Fodor, A.R. Cossins. 2000. Lipid Compositional Correlates of Temperature-Adaptive Interspecific Differences in Membrane Physical Structure. J. Exp. Biol. 203: 2105-2115. Sinensky, M. 1974. Homeoviscous Adaptation - A Homeostatic Process that Regulates the Viscosity of Membrane Lipids in Escherichia coli. Proc. Nat. Acad. Sci. USA. 71: 522-525. Singer, S. J. and G. L. Nicholson. 1972. The Fluid Mosaic Model of the Structure of Cell Membranes: cell membranes are viewed as 2-dimensional solutions of globular proteins and lipids. Science. 175: 720-731. Figure Legends Fig. 1. Fluorescence polarization (inversely proportional to membrane fluidity) of DPH in lipid membranes of gill and mantle tissue extracted from two groups of L. keenae individuals from the exposed site, one group collected before sunrise following an early- morning tide (closed triangles, N=4), the other collected in the afternoon following a midday low tide (open circles, N=4). Single factor analysis of variance shows a significant difference between the two slopes (morning collection, y--0.0022x+0.3166, =0.9973; afternoon collection, y—-0.001 1x+0.2569, R=-0.8655), p-0.02. Data points depict individual samples. Fig. 2. Fluorescence polarization of DPH in lipid membranes of gill and mantle tissue of two groups of L. keenae individuals collected from the exposed site: one group collected prior to sunrise following an early-morning high tide (closed triangles, N-4), and the other collected in the afternoon following a midday low tide (open circles, N-2). Samples here are the same as those illustrated in Fig. 3, minus the two samples showing low polarization values. A single-variable analysis of variance shows no significant difference between the slopes for the two groups (morning collection, y- 0.0022x+0.3166, R==0.9973; afternoon collection, y—-0.0017x+0.3053, R°=0.9405), p-0.92. Data points depict means + 1 S.D. Figure 3. Fluorescence polarization of DPH probe (inversely proportional to membrane fluidity) in lipid membranes of gill and mantle tissue excised from Littorina keenae specimens collected from both the exposed site (open circles, N=4) and the protected site (closed triangles, N=4) in the afternoon immediately following a midday low tide. A single-factor analysis of variance showed no significant difference between the slopes (exposed, y—-0.0021x+0.3057, R°-0.9923; protected, y--0.0023x+0.3256, R°-0.9589) p=0.6. At 20°C, estimated polarization values show significant difference using a one- tailed student's t-test, p-0.014. Data points depict means + 1 S.D. Fig 4. Conservation of membrane fluidity across several vertebrate species (adapted from Behan-Martin, et al., 1993). Temperature at which the animal shows a polarization of 0.22 is plotted against the animal’s accustomed body temperature. Snail value is the“ symbol. Fig. 5. Fluorescence polarization (membrane fluidity) values shown as a function of temperature for vertebrates adapted from Logue et al., 2000, including exposed site and protected site data for L. keenae (from Fig. 1 for comparison) shown as patterned lines. Fig. 1 0.32 0.3 0.28 § 0.26 g 0.24 5 0.22 0.2 0.18 0.16 S 4 8 O 8 O O O 5 10 15 20 25 30 35 40 Temperature °0 Fig. 2 50.32 0.3 g 0.28 0.26 3 0.24 E 0.22 S 5 10 15 20 25 30 35 40 Temperature °0 Fig. 3 0.32 0.3 0.28 0.26 0.24 0.22 5 10 15 20 25 30 35 40 Temperature °0 Fig. 4 50 L. keenae 40 FPigeon Mouse Rat 1 Astarling MousenI 30 mC M o 20 4. arch Sia an G Gerbil H Hamster 10 Ta Co Cam Gi Goldfish Notothenia Tr Trout Ci Cichlid 40 10 30 20 Body temperature (°0) 50 Fig. 5 0.34 0.28 0.22 30.16 0.10 urkey L keenae Protected Site Rat 24°0 . Toadfish (24°C) Tilapia (25°c) Blue grunt (24°c) 40 20 Temperature (C) L. Keenae Exposed Site (14°c — Striped bass (17°C —Trout (8°0) Notothenia (0°0 Pagothenia (5 100