The coral Platygyra verweyi exhibits local adaptation to long-term thermal stress through host-specific physiological and enzymatic response

Climate change threatens coral survival by causing coral bleaching, which occurs when the coral’s symbiotic relationship with algal symbionts (Symbiodiniaceae) breaks down. Studies on thermal adaptation focus on symbionts because they are accessible both in vitro and in hospite. However, there is little known about the physiological and biochemical response of adult corals (without Symbiodiniaceae) to thermal stress. Here we show acclimatization and/or adaptation potential of menthol-bleached aposymbiotic coral Platygyra verweyi in terms of respiration breakdown temperature (RBT) and malate dehydrogenase (MDH) enzyme activity in samples collected from two reef sites with contrasting temperature regimes: a site near a nuclear power plant outlet (NPP-OL, with long-term temperature perturbation) and Wanlitong (WLT) in southern Taiwan. Aposymbiotic P. verweyi from the NPP-OL site had a 3.1 °C higher threshold RBT than those from WLT. In addition, MDH activity in P. verweyi from NPP-OL showed higher thermal resistance than those from WLT by higher optimum temperatures and the activation energy required for inactivating the enzyme by heat. The MDH from NPP-OL also had two times higher residual activity than that from WLT after incubation at 50 °C for 1 h. The results of RBT and thermal properties of MDH in P. verweyi demonstrate potential physiological and enzymatic response to a long-term and regular thermal stress, independent of their Symbiodiniaceae partner.

The menthol-bleaching method 53 was used to create aposymbiotic corals. Briefly, small fragments of corals were incubated in 300 ml menthol-supplemented artificial sea water (ASW, Instant Ocean, Aquarium Systems, Sarrebourg Cedex, France). The menthol/ASW medium was prepared by diluting a 20% (w/v) menthol stock (in ethanol) with ASW. Incubation in menthol was ended (at 3 weeks) when the coral was completely bleached (Fig. S1). The bleached coral host samples were used to determine the respiration breaking temperature within a week.
In order to collect enough enzyme for final kinetic analysis, more than 10 runs of purification process were carried out. At each run of purification, nubbins (approximate size = 7 × 7 cm) from 7-8 different colonies were used. Since enzyme kinetic analysis requires large amount of purified enzyme, which is almost not possible for a coral colony to provide, and each purification process is time consuming process, the experiment did not have any biological replicates as well as controls.
Respiration breaking temperature (RBT) of coral host. Aposymbiotic coral nubbins were subjected to heat treatment by incrementally increasing its temperature. Coral nubbins were incubated in a respiratory chamber immersed in a water bath at room temperature for 10 min. Heat treatment was initiated by transferring coral nubbins from room temperature directly into another respiration chamber with preheated (about +2 °C) and aerated artificial seawater (ASW) medium. At each temperature treatment, coral nubbins were incubated for 15 min (5 min for temperature equilibrium and 10 min for oxygen consumption measurement) before being transferred to the next higher temperature treatment. The heating process was ended at 42-43 °C. ASW temperature was measured to 0.01 °C precision with a digital thermometer (TM-907A, Lutron, Taiwan) connected to a thermocouple (TP-100, Lutron, Taiwan). While measuring oxygen consumption, ASW medium in the respiration chamber was mixed with an underwater magnetic stirrer. Changes in oxygen content were measured with an optical dissolved oxygen instrument (YSI ProODO, USA) to avoid electrode polarization, and the signal was channeled to a computer for data collection. RBT was determined by calculating the intercept of two linear regression lines developed respectively from the increase and decrease in respiration rate when the temperature was elevated (Fig. 1A).

Purification of malate dehydrogenase (MDH).
To answer if the differences in thermal sensitivity of respiration physiology between the coral samples from NPP-OL and WLT was linked to the modification of enzymes, MDH, a key enzyme in aerobic energy metabolism in P. verweyi, host tissue extract was analyzed with isozyme composition and Kinetic performance to temperature. MDH was isolated by blasting symbiotic coral with ice-cold ASW and homogenizing the tissue slurry with a hand-held glass tissue grinder. The resulting homogenate was centrifuged at 4 °C (x15,000 rpm) for 15 min to separate symbiotic algae and tissue debris, then the fraction was collected in 30-80% saturation of ammonium sulfate, re-suspending in 3.2 M (NH 4 ) 2 SO 4 containing 80 mM phosphate buffer (pH 7.5) and 0.1 M EDTA, and storing at 4 °C before use.
The isoforms of MDH were examined with a diethylaminoethanol (DEAE) Sepharose TM (Fast Flow, GE healthcare) column (2.6 × 30 cm), and the major MDH fraction was further purified following the modified methods described in 57,58 . Practically, MDH was purified with a serial column chromatography, including DEAE, CL-6B column, hydroxyapatite, and Sephacryl S-200 column. The detail chromatography condition is described in the supplement. MDH activity in the eluate was measured at 25 °C by adding 50 µl enzyme solution into a substrate solution mixed with 50 µl 3 mM NADH, 50 µl 1 mM oxaloacetate (OAA) and 900 µl 100 mM phosphate buffer pH 7.0. The enzyme activity was determined by the decrease in 340 nm absorbance recorded by time scanning mode in a spectrophotometer (Hitachi U1900) and standardized with protein concentration determined with Bradford Assay. The changes in 280 nm absorbance and salt concentration of the eluate gradient were determined with a spectrophotometer (Hitachi U1900) and a refractometer, respectively. SDS-PAGE in 12.5% gel visualized with coomassie blue staining was used to determine the purity and molecular weight of MDH in the Sephacryl S-200 eluate. The molecular weight of native MDH was determined by Sephacryl S-200 gel filtration after comparing with a set of purified protein markers included ribonuclease A (13.7 kDa), chymotrypsinogen A (25 kDa), ovalbumin (43 kDa,) and albumin (66.5 kDa). Protein concentration was determined by a Bradford Assay (Bio-Rad Protein Assay, Bio-Rad) according to the manufacturer's protocol.
Effects of temperature on enzyme kinetic parameters and stability. The enzyme reaction was determined as described in the purification section, except that OAA concentration was varied from 14 to 29 μM with a fixed NADH concentration (3 mM) to determine OAA's substrate concentration (K m ) and saturating substrate concentration (V max ), and NADH varied from 38 to 95 µM with a fixed OAA concentration (1 mM) to determine NADH's K m and V max . All reaction rates were obtained from the mean of triplicate reactions. The reaction temperature in the cuvette was maintained using a water jacket cuvette holder connected to a thermostatic bath (YIH DER, BL-720). Prior to each thermodynamic analysis, a digital thermometer was used to confirm temperature stability in the cuvette three times when mixing 100 µl substrate, 50 µl enzyme and 900 µl buffer (maintained in a thermostatic bath at the reaction temperature).
Obtained V max values were further used to calculate the activation energy required for MDH catalysis (Ea cat ) and that for inactivating MDH activity (Ea inact ) by temperature with the Arrhenius equation. Ea cat and Ea inact values were usually used to evaluate the thermostability of enzymes in different organisms. The Ea cat was determined by plotting natural-logarithmically (ln) transformed V max on T −1 in °K, in which the V max data were obtained from pre-optimum temperature. The Ea inact was calculated by plotting the ln-transformed inactivation rate constants k on T −1 in °K. The k values were determined by subtracting each V max obtained at the post-optimum temperature from the V max at optimum temperature and converted to per second. The fitness of the kinetic data to the www.nature.com/scientificreports www.nature.com/scientificreports/ The thermal stability of MDH isolated from NPP-OL and WLT coral samples were compared by incubating the diluted enzyme solution (in 100 mM phosphate buffer pH 7.0) containing activity of about 20 nmol min −1 at 50 °C for 60 min. Residual activities of MDH in percentages were determined with the ratio of MDH activity at each time interval compared to the original. Triplicate measurements of enzyme activity at every time interval were made under the condition described in the purification section.
Statistical analysis. Respiration performances of P. verweyi coral hosts between NPP-OL and WLT were compared with a Student's t-test. Differences in the slopes of linear regression of ln(V max ), ln(K m ) and ln(k) on temperature in °K were analyzed by Minitab statistical software.

MDH chromatography pattern and purity. Both NPP-OL and WLT samples displayed a consistent
profile that contained only two MDH fractions (Fig. 2). The two MDH fractions could not be further separated using a linear gradient of different NaCl concentrations (data not shown). Thus, only the major MDH isozyme (peak II) which contained >75% of total activity was collected for further purification. MDH in the DEAE eluate from both NPP-OL and WLT samples displayed only one peak and almost identical elution conditions in the  (Figs S2-S4). In the Sephacryl S-200 gel chromatography, the MDH activities of both NPP-OL and WLT samples were eluted synchronously with A 280 absorbance, indicating that the enzymes had been highly purified at both sites. MDH purity in the Sephacryl S-200 eluate was estimated to be >90% by SDS-PAGE and coomassie blue stain, as shown in Fig. S4, and only one clear band was visualized at about 39 kDa for the NPP-OL sample and 38 kDa for the WLT one. With calibration to standard proteins by Sephacryl S-200 gel filtration (Fig. S5), the molecular weight of the native MDH form was about 67 kDa and 64 kDa for the NPP-OL and WLT samples, respectively, indicating that the major MDH of P. verweyi was a typical dimer. Finally, the overall purification process described above increased the specific MDH activities from about 2 µmol min −1 mg −1 in the tissue extract to 77 (from NPP-OL) and 111 µmol min −1 mg −1 (from WLT) in the Sephacryl S-200 eluates; these purified enzyme solutions were used for further kinetic analysis at the different temperatures.
Enzyme kinetics and thermal susceptibility. The apparent K m and V max of MDH varied around 27 °C for the samples from NPP-OL and WLT; this was 7.5 and 12.7 µM for the K m of NADH, 9.4 and 5.7 µM for the K m of OAA, 24.6 and 22.9 nmol min −1 for the V max of NADH, and 31.2 and 28.9 nmol min −1 for the V max of OAA. When the reaction temperature rose, the Arrhenius plot with ln(V max ) on T −1 in °K showed a linear increase and then decrease at some higher temperatures, as shown in Fig. 3. The data's fitness to linear regression model and regression parameters are described in Table S1. According to the intersection of the Arrhenius plot in Fig. 3, samples from NPP-OL performed at a 4.7~10.6 °C higher optimum temperature than those from WLT, depending on substrates. However, the activation energy for MDH catalysis (Ea cat ), calculated from the data in Fig. 3, were as shown in Table 1 for the Ea cat of OAA and NADH, respectively, and did not display significant differences between the two sampling locations (P > 0.05). When plotting ln(K m ) for temperature in T −1 (°K) (Fig. S6), K m initially displayed a linear increase, but, different from the performance of V max in Fig. 3, came to a plateau after the optimum temperature. Therefore, the sensitivity of K m to rising temperatures was examined with the data sets obtained from pre-optimum temperatures. The K m dataset's fitness to linear regression model and regression parameters are described in Table S2. In Fig. 4, the slopes of the MDH K m values did not differ significantly between sites (P > 0.05), regardless of whether they were plotted with ln(K m ) on T −1 (°K) or K m on T (°C) (in inset), implying that samples from NPP-OL and WLT had comparable temperature sensitivities.
The activation energy for thermal inactivation (Ea inact ) and the enzyme stability at 50 °C were investigated to further compare the differences in thermal sensitivity of MDH between NPP-OL and WLT. Ea inact of MDH was  Table S3). MDH displayed significantly different slopes between NPP-OL and WLT for both substrates (Fig. 5, P < 0.05). Calculated Ea inact indicated that MDH from NPP-OL samples required almost two times higher activation energy to inactivate the enzyme than MDH from WLT, as shown in Table 1. When MDH was incubated at 50 °C, the enzyme activity of the samples from WLT decreased with time faster than that of samples from NPP-OL, as shown in Fig. 6. Moreover, residual activity for the MDH from NPP-OL samples was 64 ± 9% (mean ± S.D., n = 3), which was nearly two times higher than that from WLT (32 ± 2%).

Discussion
This study clearly demonstrates that coral hosts have the potential to acclimatize and/or adapt physiologically and enzymatically to long-term thermal stress. The 3rd nuclear power plant, operating since 1984, has been discharging thermal effluent from the outlet (NPP-OL), resulting in 2.0-3.0 °C warmer summer water temperatures in the adjacent reef compared to other nearby reefs, such as WLT. The RBT of aposymbiotic Platygyra verweyi in NPP-OL was also nearly 3.0 °C higher than that in WLT, suggesting that coral hosts in NPP-OL, despite the high percentage of individuals that associate with thermal-tolerant Durusdinium spp 20 , might have also physiologically adapted and/or acclimatized after being impacted by the consistent warmer SST for over 30 years. The critical temperature RBT (also named Arrhenius Breaking Temperature, ABT 43 ) for the mitochondria of aquatic organisms is shown to be highly correlated with their maximum habitat or acclimation temperature, and the species adapted to warmer temperatures would perform the RBT closer to that temperature 43 . Consistent with the other  www.nature.com/scientificreports www.nature.com/scientificreports/ aquatic organisms, the higher RBT values in aposymbiotic P. verweyi from NPP-OL than from WLT might be a response to the variation in local temperature between two conspecific populations.
Results of MDH bioassays provide some insights, although preliminary, into the protein adaptations that underpins performance differences between the thermal physiologies of Platygyra verweyi collected from different temperature regimes. Results clearly indicated that MDH of P. verweyi from NPP-OL was more thermally stable than that from WLT. The kinetic analyses showed that the MDH isolated from P. verweyi in NPP-OL displays a 4.7 or 10.6 °C higher optimum temperature and 1.8 fold higher Ea of thermal inactivation than that from WLT, depending on substrates. In addition, the MDH of P. verweyi from NPP-OL showed 2 times higher residual activity than that from WLT after 1 h at 50 °C. When the enzyme's adaptation to temperature was examined, the ligand affinity as measured by apparent Michaelis-Menten constants (K m ) was usually modified to respond to environmental temperatures 38,[40][41][42]44,51,59,60 . The MDH was also proposed to be functionally and structurally more sensitive to temperature perturbation than another ATP-generating related enzymes in mussels 50 . However, K m 's responses to temperature were comparable between the purified P. verweyi MDH in NPP-OL and that in WLT. Consistent with the response of K m to temperature, the catalytic Ea of MDH from P. verweyi in NPP-OL was not significantly different from that of WLT samples. When the V max of MDH from P. verweyi in NPP-OL was calculated from the equations derived from Fig. 3 with the optimum temperature of that from WLT (42.4 and 38.4 °C for that of NADH and OAA, respectively), the results indicated that the V max values were very close between two sites (NPP-OL/WLT: 50/53 and 91/97 for NADH and OAA, respectively). A study on Antarctic notothenioid fish showed that A 4 -lactate dehydrogenase (LDH) decreased activation energy and substrate affinity (i.e., increase in K m ) for adapting to cold temperature 51 , this suggests that, in order to provide comparable V max by balancing between stability and flexibility, adaptation to temperature might be achieved by modifying enzyme structure to increase K m and decrease catalytic Ea. Similar to the response of K m to temperature, catalytic Ea and V max at the same temperature between the MDH from NPP-OL and WLT samples suggest that higher thermal resistant performance in the MDH from NPP-OL than WLT samples might not be attributed to temperature adaptation mechanisms described in the literature 6,[40][41][42]44,51,59,60 .
Second, configuration analysis also indicated that the genes coding the MDH of Platygyra verweyi may be different from those in published studies in other marine invertebrates 50,59 . Both MDH samples from NPP-OL and WLT consistently showed two distinct MDH fractions with about 25 and 75% of total activity, according to www.nature.com/scientificreports www.nature.com/scientificreports/ the DEAE profile, suggesting that the different respiration physiology performances might not be attributed to the variations between the two fractions. In addition, only one MDH isoform was obtained in the sample from NPP-OL and WLT, respectively, when the major MDH fraction (peak II) was purified. The purified MDH from  www.nature.com/scientificreports www.nature.com/scientificreports/ both sites were all in the dimer form. In current available coral genome 60,61 , Stylophora pistillata contains two MDH isoforms with very similar monomer molecular weights (33.6 and 36.1 kDa) within at least 4 MDH genes 61 . However, the genome data could not manifest if these monomers were constructed into dimer as found in the MDH sample from P. verweyi. Although in the literature, MDH adaptation was shown to be achieved with 1 or 2 amino acid mutations. A nearly 1 kDa difference in MDH molecule between that from NPP-OL and WLT might be attributed to two alternative possibilities. One possibility is acclimatization scenario with 2 MDH enzymes with very similar molecular weight, but with different thermal stability in the P. verweyi from NPP-OL and WLT. Another possibility is adaptation scenario in which MDH from NPP-OL contains roughly 9 amino acid mutations compared to that from WLT. The proposed hypothesis can be tested through direct protein sequencing or cloning the target gene for further DNA sequencing. At present, we do have analysis carried out using MALDI-TOF mass spectrometry and this information will be using in the future studies to clone MDH gene for further sequencing. Therefore, genetic analysis of the genes encoding the MDH should be conducted before we can conclude whether MDH genes of P. verweyi from NPP-OL already show signs of acclimatization and/or adaptation.
Even though changes in the algal symbiont from thermal sensitive to tolerant genotypes (e.g., Durusdinium sp.) were highlighted as a potential mechanism of coral acclimatizing to high temperature 15,24 , the species compositions of corals in NPP-OL were dramatically changed with long-term exposure to high temperatures and concurrent associations with tolerant Durusdinium spp 20 . P. verweyi is one of few coral species in NPP-OL that has not only survived, but prospered 19 . If the more thermally stable MDH isoform found in NPP-OL inhabiting P. verweyi was a result of selection from the regular thermal stress in the summer, would it be possible to acclimatize a population in WLT to survive in NPP-OL? Reciprocal transplantations of P. verweyi between NPP-OL and WLT 54 suggest that heating period duration was also important to the P. verweyi survival rate when transplanted from WLT to NPP-OL. We conducted two reciprocal transplantations in April of 2014 and 2015 so the transplanted corals could have 3~4 months to acclimatize to the warmer water occurring in the summer. The results of the transplantation indicated that the survival rates of P. verweyi transplanted from WLT to NPP-OL depends highly on the duration of the warm period 54 . If the accumulating heat stress is prolonged beyond a threshold-for example, 10.43 degree of heating week (DHW)-P. verweyi would barely survive, even if shuffled from the thermally liable symbiont to stress tolerant D. trenchii 54 . Conversely, the NPP-OL colonies transplanted to WLT displayed a higher growth rate than conspecific NPP-OL and WLT native populations. Therefore, in order to understand how coral will adapt to changing climate and increased occurrences of high seawater temperatures, it is important to resolve the puzzle involving mechanisms behind aposymbiotic coral host acclimatization and/or adaptation, symbiont population dynamics and its contribution to coral stress response and the duration of heating period. To completely understand if a given coral species can survive the intensity and duration of temperature stress, insights from aposymbiotic host mechanisms will pay way for a better understanding of the nature of stress resistance and contribution to the same from the coral host perspective.

Data Availability
All additional data from the experiment are provided in "Supplementary Materials".