Temperature Induced Changes in Thylakoid Membrane Thermostability of Cultured, Freshly Isolated, and Expelled Zooxanthellae from Scleractinian Corals

Coral bleaching events are characterized by a dysfunction between the cnidarian coral host and the symbiotic dinoflagellate algae, known as zooxanthellae (genus Symbiodinium). Elevated temperature and intense light induce coral bleaching, where zooxanthellae are expelled from the host tissue. The primary cellular process in zooxanthellae which leads to coral bleaching is unresolved, and here, we investigated the sensitivity of the thylakoid membrane in a Symbiodinium culture and in genetically identified freshly isolated and expelled Symbiodinium cells. The fluorescence-temperature curve technique was used to measure the critical temperature (Tc) at which irreversible damage to the thylakoid membrane occurred. The accuracy of this technique was confirmed through the collection of scanning transmission electron micrographs which demonstrated the clear relationship between Tc and thylakoid membrane degradation. Analysis of 10 coral species with a diverse range of genetically distinct Symbiodinium communities showed a decline in Tc from summer to winter. A Symbiodinium culture and fragments of Pocillopora damicornis (Linnaeus, 1758) were exposed to a series of light and temperature treatments, where Tc increased from approximately 37 °C to 42 °C upon exposure to elevated temperature. Under bleaching conditions, the thermostability of the thylakoid membrane increased within 4 hrs by 5.1 °C, to a temperature far above bleaching thresholds, in both freshly isolated and photosynthetically competent zooxanthellae expelled from P. damicornis under these conditions. It is demonstrated that the thermostability of the thylakoid membrane increases in cultured, freshly isolated, and expelled zooxanthellae exposed to bleaching stress, suggesting it is not the primary site of impact during coral bleaching events. Unicellular dinoflagellates, known as zooxanthellae (genus Symbiodinium) reside within the gastrodermal tissue of scleractinian corals (Titlyanov and Titlyanova, 2002). This obligate symbiotic relationship is sensitive to environmental changes, such as elevated sea temperatures as small as 1–2 °C above average (Hoegh-Guldberg, 1999). Coral bleaching events occur in conjunction with high light intensity, which results in the dysfunction of the coral-algal symbiosis, leading to the expulsion of zooxanthellae, as well as a reduction in photosynthetic pigments (Kleppel et al., 1989). The measurement of chlorophyll a fluorescence is a widely used tool for studying the photosynthetic activity of zooxanthellae as it is a rapid and non-invasive method of analysis (Iglesias-Prieto, 1995; Warner et al., 1996; Jones et al., 1998; Fitt et al., 2001; Jones and Hoegh-Guldberg, 2001; Hill et al., 2004). Previous studies have indicated that bleaching causes a decline in the health of in hospite zooxanthellae (Jones et al., 2000), and much research has concentrated on identifying the location of bleaching-associated damage within the photosynthetic apparatus. Photosystem II (PSII) has been suggested as the primary site of impact (Warner et al., 1999; Hill et al., 2004; Takahashi et al., 2004; Hill and Ralph, 2006), along with dark-reactions CORAL REEF PAPER BULLETIN OF MARINE SCIENCE, VOL. 85, NO. 3, 2009 224 (Jones et al., 1998) and disruption of the thylakoid membrane (Iglesias-Prieto et al., 1992; Tchernov et al., 2004). Bound to the thylakoid membrane are chlorophyll molecules, which are the photosynthetic pigments responsible for light absorption. Damage to the membrane’s integrity has been shown to cause the degradation of the entire photosynthetic apparatus (Schreiber and Berry, 1977). Tchernov et al. (2004) found thermal damage to occur within the membrane after 168 hrs of exposure to elevated light and temperature, a time period which falls outside the usual definition of bleaching (paling expected within 6 hrs of exposure to bleaching conditions; Dove et al., 2006). In contrast, impacts to the photosynthetic health of zooxanthellae have been shown to occur within several hours by measurements of maximum quantum yield (Fv/Fm) (Warner et al., 1996; Hill et al., 2004, 2005). Schreiber et al. (1975) and Schreiber and Berry (1977) developed a bioassay which monitors the state of the thylakoid membrane through the measurement of fluorescence-temperature (F-T) curves. This method involves the measurement of minimum fluorescence (Fo) while temperature is ramped at a user-defined speed, generally between 0.5 and 6 °C min–1 (Smillie, 1979; Bilger et al., 1984; Seeman et al., 1986; Kuropatwa et al., 1992; Lazár and Ilik, 1997). Analysis of the resulting F-T curve, generates the following parameters: the critical temperature (Tc), the temperature of peak fluorescence (Tp), the ratio of the initial Fo to the maximum Fo (Finitial/ Fmaximum), the temperature at which Fv/Fm reaches 50% of its initial (T50) and the temperature at which Fv/Fm reaches zero (T0). These have been shown to correlate well to the heat sensitivity of the thylakoid membrane in numerous studies on higher plants (Schreiber and Berry, 1977; Raison and Berry, 1979; Smillie, 1979; Berry and Bjorkman, 1980; Smillie and Gibbons, 1981; Monson and Williams, 1982; Bilger et al., 1984; Nauš et al., 1992; Havaux, 1993) and in one case, on a green alga (Kouřil et al., 2001). See Figure 1A for graphical explanation of all F-T curve parameters. The thermal tolerance of the thylakoid membrane (indicated by Tc) has been shown to vary between species of higher plants inhabiting the same environment (Smillie and Nott, 1979; Bilger et al., 1984; Thomas et al., 1986; Havaux et al., 1988). Furthermore, the capacity for rapid acclimation of the thylakoid membranes to changing ambient temperatures has been demonstrated (Berry and Bjorkman, 1980; Smillie and Gibbons, 1981; Raison et al., 1982; Downton et al., 1984; Havaux, 1993; Lazár and Ilik, 1997; Kouřil et al., 2001). In these studies, shifts in the heat tolerance have been reported where the Tc has increased by several degrees Celsius in a matter of hours following exposure to elevated temperatures. Seasonal and spatial variation in thylakoid membrane thermostability, induced by varying temperatures has also been established through measurements of F-T curves. Seeman et al. (1986) and Knight and Ackerly (2002) demonstrated, using the F-T technique, that species inhabiting warmer regions exhibited greater photosynthetic tolerance to elevated temperature, and that seasonal increases in thermostability occurred from winter to summer. This plasticity has been attributed to changes in the composition of the thylakoid membrane lipids and proteins (Pearcy, 1978; Raison et al., 1982; Hugly et al., 1989; Ivanova et al., 1993). Scleractinian corals have been shown to harbor a number of genetically distinct Symbiodinium types (e.g., Baker, 2003; Ulstrup and van Oppen, 2003; Sampayo et al., 2008) and attempts have been made to correlate these with specific physiological traits. The most conclusive evidence of Symbiodinium-specific physiologies in hosHILL ET AL.: THERMOSTABILITY OF ZOOXANTHELLAE THYLAKOID MEMBRANES 225 pite has been presented for the coral genus Pocillopora (Rowan, 2004) and Acropora millepora (Ehrenberg, 1834) (Berkelmans and van Oppen, 2006) where associations with a clade D Symbiodinium genotype conferred higher thermotolerance than associations with a clade C Symbiodinium genotype. However, the ITS1-determined Symbiodinium C1 and D infected into juveniles of Acropora tenuis (Dana, 1846) showed contrasting results where Symbiodinium C1 was found to confer higher thermotolerance of the holobiont than Symbiodinium D (Abrego et al., 2008). In combination, these results suggest that holobiont physiologies are not contingent on Symbiodinium type alone but also depend on the host examined. This study explored the thermal tolerance of the thylakoid membrane of a Symbiodinium culture as well as genetically identified freshly isolated and expelled zooxanthellae. Further studies were done to evaluate the degree of acclimatization of the thylakoid membrane to light and temperature treatments in the coral, Pocillopora damicornis (Linnaeus, 1758). By comparing corals which harbored genetically distinct zooxanthellae, we attempted to identify relationships between genotype and physiology. Material and Methods Coral Specimens.—Eight replicate fragments of upper, sun-exposed surfaces of 10 coral species were sampled during the Austral summer (January 2005) and winter (July 2005) from Heron Island reef flat, Great Barrier Reef, Australia (151°55 ́E, 23°26́ S). The species collected were from a depth of 1–2 m and consisted of A. millepora, Acropora nobilis (Dana, 1846), Acropora valida (Dana, 1846), Cyphastrea serailia (Forsskål, 1775), Goniastrea australensis (Milne-Edwards and Haime, 1857), Montipora digitata (Dana, 1846), Pavona decussata (Dana, 1846), P. damicornis, Porites cylindrica Dana, 1846, and Stylophora pistillata Esper, 1797. For the temperature and light manipulated experiments, four replicate P. damicornis colonies were also collected. Prior to experimentation, all corals were maintained in shaded aquaria (< 100 μmol photons m–2 s–1) at ambient lagoon temperature for 2 d. Fluorescence Measurements.—Coral fragments were air brushed in 15 ml of 0.45 μm filtered seawater to remove zooxanthellae and animal tissue from the skeleton. 10 ml of this Figure 1. (A) Representative fluorescence-temperature (F-T) curve, where the temperature was increased at a speed of 1 °C min. The location of minimum fluorescence (F o ), maximum quantum yield (F v /F m ), critical temperature (T c ), temperature of maximum fluorescence (T p ), and initial (F initial ) and maximum fluorescence (F maximum ) are shown. The temperature (°C) is indicated above the x-axis and time (min) indicated below. This example is of Acropora millepora during summer. (B) Average F-T curves of Pocillopora damicornis exposed to bleaching conditions (400 μmol photons m s and 32 °C) at 0, 2, and 4 hrs (n = 4). BULLETIN OF MARINE SCIENCE, VOL. 85, NO. 3, 2009 226 solution was then filtered through a 20 μm mesh to isolate the zooxanthellae which passed through the filter. A freshly isolated zooxanthellae solution was then placed in a Water-PAM fluorometer (Walz GmbH, Effeltrich, Germany) cuvette for analysis. A custom-made heat exchange chamber, with an internal cavity through which water was pumped, was inserted into the Water-PAM cuvette to control the temperature of the sample (2.2 ml). The water flowing through the heat exchange chamber was connected to a temperature regulated waterbath (Julabo Labortechnik, model EC, Germany) and was raised at a user-defined speed. 1 °C min–1 was found to be appropriate for these experiments (Table 1), where any faster rise in temperature (i.e., at 2 °C min–1) resulted in an over-estimation of the critical temperature and temperature at maximum fluorescence (Smillie, 1979). A temperature rise of 1 °C min–1 was an adequate speed, as decreasing the rate of heating to 0.5 °C min–1 did not alter the temperature values obtained (Table 1). The Water-PAM was placed on an oscillating table (set at approximately 90 revolutions min–1) to prevented the settling of zooxanthellae and to eliminate the formation of a temperature gradient in the solution within the cuvette. A K-type thermocouple was inserted into the cuvette to monitor the temperature over time on a digital thermometer (Fluke, 52 Series II, Washington State, USA). Once a zooxanthellae sample was placed into the Water-PAM, the recording of minimum fluorescence (Fo) began (λ > 710 nm), and 5 min of dark-adaptation was given at a constant temperature (ambient lagoon temperatures in summer and winter sampling and treatment temperature for the manipulative studies). Following this time, a saturating pulse was applied and the far-red LED (emission peak at 730 nm) was turned on. The far-red light served to reoxidize the electron transport chain, through the excitation of Photosystem I (PSI) (Schreiber et al., 1975; Bilger et al., 1984; Knight and Ackerly, 2002). This is necessary to suppress the fluorescence rise which can occur under hypoxic conditions in corals in darkness (Bukhov et al., 1990; Bukhov and Mohanty, 1999; Yamane et al., 2000), and has been shown to occur in experiments with in hospite zooxanthellae (Ulstrup et al., 2005). No significant difference was detected between the use of far-red and no far-red (Table 1), although this light source was still used to prevent any occurrence of dark-induced reduction of the electron transport chain between PSII and PSI (Hill and Ralph, 2008). The temperature inside the cuvette was then raised at a speed of 1 °C min−1 and the minimum fluorescence (Fo) was measured until a down-turn in Fo was observed (Fig. 1A). The critical temperature (Tc) was determined by the intersection of the steady minimum fluorescence (Fo) and the subsequent rise in Fo (Schreiber and Berry, 1977). The temperature of peak fluorescence (Tp) is also indicated on Figure 1A (Bilger et al., 1984). Additional parameters suggested to be of physiological importance (Smillie and Gibbons, 1981; Nauš et al., 1992; Lazár and Ilik, 1997; Kouřil et al., 2001; Knight and Ackerly, 2002) include the calculation of the initial fluorescence over the maximum fluorescence (Finitial/Fmaximum), the temperature at which Fv/Fm reaches 50% of its initial (T50) and the temperature at which Fv/Fm reaches zero (T0). For the seasonal study, saturating pulses were applied every 2 min during the measurement of the F-T curve on half of the replicates. This was performed in order to calculate T50 and T0. In the remaining replicates, no saturating pulses were applied following the initial Fv/Fm measurement, in order to calculate Tc and Tp with greater precision and accuracy. Table 1. Effect of heating rate and presence/absence of far-red light on initial Fv/Fm, Tc, and Tp of fluorescence-temperature curves performed on cultured Symbiodinium sp. (CS-156). Averages ± standard error of mean shown (n = 6). P-values (one-way ANOVAs) and Tukey’s post hoc comparison tests are shown as superscript letters. Light exposure Ramping speed Fv/Fm Tc (°C) Tp (°C) Darkness 1 °C min–1 0.503 ± 0.006 37.0 ± 0.1 a 43.7 ± 0.7 a Far-red 0.5 °C min–1 0.504 ± 0.001 36.9 ± 0.2 a 43.8 ± 0.2 a Far-red 1 °C min–1 0.505 ± 0.004 37.0 ± 0.2 a 42.7 ± 0.4 a Far-red 2 °C min–1 0.488 ± 0.004 38.9 ± 0.3 b 47.1 ± 0.3 b P-value 0.069 < 0.001 < 0.001 HILL ET AL.: THERMOSTABILITY OF ZOOXANTHELLAE THYLAKOID MEMBRANES 227 Experimental Protocol.—Freshly isolated zooxanthellae from the 10 coral species examined, were raised in temperature from 29 °C during summer and from 22 °C during winter (ambient lagoon temperatures) at a rate of 1 °C min–1 until after Tp was reached (Fig. 1A). Measurements were performed on the CS-156 Symbiodinium culture from the Commonwealth Scientific and Industrial Research Organisation (CSIRO) microalgae research center in Hobart, Tasmania, Australia. CS-156 was isolated from Montipora verrucosa (Lamarck, 1816) in Hawaii, USA, and grown in f/2 media at 25 °C and 40 μmol photons m–2 s–1, with a light to dark ratio of 12:12 hrs. A second dinoflagellate culture, Amphidinium carterae Hulburt, 1957, was also analyzed to compare its physiological traits to those found in the Symbiodinium culture. This strain, CS-21 (also from CSIRO), isolated in Halifax, Canada, was grown in the laboratory under the same conditions as CS-156. The impact of increased light intensity and elevated temperature on Fv/Fm, Tc, and Tp was investigated in the Symbiodinium culture (CS-156) as well as for freshly isolated and expelled Symbiodinium cells from P. damicornis. Symbiodinium culture solutions (n = 4) were exposed to four treatments (see Table 2) for 9 hrs, with hourly determinations (0–9 hrs) of Fv/Fm, Tc, and Tp. The four treatments were, (i) control (low light and low temperature), (ii) high light and low temperature, (iii) low light and elevated temperature, and (iv) high light and elevated temperature (Table 2). Similarly, freshly isolated zooxanthellae from fragments of P. damicornis (n = 4) were exposed to four treatments (Table 2) for 8 hrs, with Fv/Fm, Tc, and Tp determined at 0, 1, 2, 4, and 8 hrs. Figure 1B shows the average F-T curves of P. damicornis exposed to high light and elevated temperature (conditions which initiated a bleaching response; see Hill and Ralph, 2006) at 0, 2, and 4 hrs. The rapid and acute stress imposed by the high light and elevated temperature treatment on P. damicornis fragments was sufficient to induce a bleaching response where symbionts were expelled from the host. However, it should be noted that this treatment may not directly reflect the conditions experienced by corals during bleaching episodes in nature. In addition to the measurements on freshly isolated zooxanthellae, Fv/Fm, Tc, and Tp of zooxanthellae which were expelled under bleaching conditions of high light and elevated temperature (400 μmol photons m–2 s–1 and 32 °C) after the time periods of 0–1 hr, 1–2 hrs, 2–4 hrs, and 4–8 hrs were measured. Coral fragments were placed in filtered seawater and the expelled zooxanthellae were collected by filtering the aquaria seawater through 0.45 μm filter paper after each time interval. Zooxanthellae were isolated by gently brushing the surface of the filter paper into 10 ml of filtered seawater. 2.2 ml of the solution was then placed into the Water-PAM for analysis. The same fragments were then returned to the experimental aquaria with filtered seawater to collect expelled zooxanthellae for the next time interval. Genetic Identification of Symbiodinium.—Upon completion of fluorescence measurements, a sub-sample of the original isolated zooxanthellae solution were centrifuged to a pellet and stored frozen (−80 °C) for subsequent genetic identification. Zooxanthella DNA was extracted using the DNeasy tissue extraction kit (Qiagen, USA) according to the manufacturer’s protocol. In order to distinguish between Symbiodinium genotypes the ribosomal DNA internal transcribed spacer 1 (ITS1) region was amplified as described in van Oppen et Table 2. Photon flux density (μmol photons m–2 s–1) and temperature (°C) of the four experimental treatments [control (low light and low temperature)], high light + low temperature, low light + elevated temperature, and high light + elevated temperate) for the Symbiodinium culture and freshly isolated Symbiodinium samples. Cultured zooxanthellae Freshly isolated zooxanthellae Treatment Photon flux density (μmol photons m–2 s–1) Temp. (°C) Photon flux density (μmol photons m–2 s–1) Temp. (°C) Control 40 25 100 22 High light + low temp. 400 25 400 22 Low light + elevated temp. 40 32 100 32 High light + elevated temp. 400 32 400 32 BULLETIN OF MARINE SCIENCE, VOL. 85, NO. 3, 2009 228 al. (2001). A forward primer fluorescently labelled with TET enabled detection of sequence variation using single stranded conformation polymorphism (SSCP) (Sunnucks et al., 2000) and the GelScan2000 system (Corbett Research, Australia). PCR (polymerase chain reaction) products mixed with formamide gel-loading dye (Sambrook et al., 1989) were denatured for 3 min at 95 °C and immediately snap-cooled on ice. 1 μL of each sample was loaded onto a 4% non-denaturing TBE-polyacrylamide gel and run on the Gelscan2000 (Corbett Research). PCR products that resulted in different SSCP profiles were re-amplified using a non-labelled forward primer and purified using ExoSAP-IT (USB, USA) and subsequently sequenced. Samples containing multiple DNA templates were cloned using the TOPO® Cloning Kit (Invitrogen). Colony inserts were re-amplified using rDNA ITS1 primers for SSCP analysis and subsequent purification was performed on products with matching SSCP profiles to those of the original sample. Colony PCRs that yielded new SSCP profiles were omitted from the analysis. Samples were sent to Macrogen (Korea) for sequencing and results were compared to existing sequences stored in GenBank (www.ncbi.nlm.nih.gov). In samples where multiple DNA templates were present, the genotype corresponding to the brightest band on the SSCP gel was classified as the dominant genotype. Following Symbiodinium identification, correlations between clusters of corals harboring the same Symbiodinium genotype and derived fluorometric parameters were tested. Scanning Transmission Electron Microscopy.—Symbiodinium culture suspensions (CS-156) of 18 ml were ramped from 25 °C to 45 °C at 1 °C min–1 with 2 ml aliquots extracted at 25, 32, 35, 36, 37, 38, 39, 40, and 45 °C. Upon extraction, samples were immediately centrifuged at 2000 × g for 1 min and fixed in 0.1M sodium cacodylate buffer (in 0.22 μm filtered seawater, pH = 8.2) and 4% glutaraldehyde for 1 hr at room temperature. Pellets were rinsed in a series of three, 10 min, 0.1M sodium cacodylate buffer solutions, followed by secondary fixing in 2% OsO4 for 1 hr at room temperature. After rinsing in two, 10 min, distilled water solutions, samples were dehydrated in a series of ascending acetone solutions, of 50, 70, 80, 90, 95, and 100%. Samples were then infiltrated overnight in a 1:1 solution of Spurr’s resin:acetone and subsequently infiltration by 100% Spurr’s resin for 24 hrs. A microtome (Ultracut T, Leica, Vienna, Austria) sectioned samples at 130 nm in thickness and these were viewed on a scanning transmission electron microscope (Quanta 200, FEI Company, Oregon, USA). Statistical Analysis.—One-way analysis of variance (ANOVA) and Tukey’s post hoc multiple comparison tests (α = 0.05) were used to detect significant differences between variables and to identify the location of these difference. The Kolmogorov-Smirnov normality test and Levene’s homogeneity of variance test were used to identify if the assumptions of the parametric one-way ANOVAs were satisfied. If these assumptions were not met, arcsine (Fv/Fm and Finitial/Fmaximum) and log10 (Tc, Tp, T50, and T0) transformations were performed. All analyses were carried out using the SPSS statistical software package (version 11.0.0, 2001, Chicago, Illinois, USA).