Transcriptional Response of Two Core Photosystem Genes in Symbiodinium spp. Exposed to Thermal Stress

Mutualistic symbioses between scleractinian corals and endosymbiotic dinoflagellates (Symbiodinium spp.) are the foundation of coral reef ecosystems. For many coral-algal symbioses, prolonged episodes of thermal stress damage the symbiont's photosynthetic capability, resulting in its expulsion from the host. Despite the link between photosynthetic competency and symbiont expulsion, little is known about the effect of thermal stress on the expression of photosystem genes in Symbiodinium. This study used real-time PCR to monitor the transcript abundance of two important photosynthetic reaction center genes, psbA (encoding the D1 protein of photosystem II) and psaA (encoding the P700 protein of photosystem I), in four cultured isolates (representing ITS2-types A13, A20, B1, and F2) and two in hospite Symbiodinium spp. within the coral Pocillopora spp. (ITS2-types C1b-c and D1). Both cultured and in hospite Symbiodinium samples were exposed to elevated temperatures (32°C) over a 7-day period and examined for changes in photochemistry and transcript abundance. Symbiodinium A13 and C1b-c (both thermally sensitive) demonstrated significant declines in both psbA and psaA during the thermal stress treatment, whereas the transcript levels of the other Symbiodinium types remained stable. The downregulation of both core photosystem genes could be the result of several different physiological mechanisms, but may ultimately limit repair rates of photosynthetic proteins, rendering some Symbiodinium spp. especially susceptible to thermal stress.


Introduction
The widespread and long-term persistence of coral reef ecosystems dominated by calcifying cnidarians is attributed to the mutualistic symbiosis between scleractinian corals and endosymbiotic members of the dinoflagellate genus, Symbiodinium [1]. These dinoflagellates provide the host with a source of photosynthetically derived simple carbon compounds (e.g. glycerol) that account for a significant proportion of the energy required for their daily metabolic demands [2]. Coral reef ecosystems have been an enduring feature of warm tropical waters for millions of years, but in recent decades, a dramatic decline in coral coverage and diversity has threatened their existence worldwide due to an increasing occurrence of coral bleaching events [3,4]. Coral bleaching is a phenomenon characterized by a significant loss in photosynthetic pigments and/or expulsion of the symbiotic algae, ultimately reducing the translocation of photosynthetically fixed carbon to the animal host [2]. Although coral bleaching can be the result of many physical and biological stressors [5], elevated sea surface temperatures due global climate change promote a breakdown in the coral-algal symbiosis and are generally considered a significant threat to the persistence of coral reefs [4,6,7]. The long-term impacts of bleaching include reduced rates of calcification and growth, lowered fecundity, increased prolifer-ation of disease, and mass mortality, which eventually degrade the ecosystem's diversity and function [8][9][10][11].
The underlying molecular and physiological mechanisms that initiate temperature-induced coral bleaching remain an area of considerable research. Extensive evidence suggests that a collapse in the symbiosis can be attributed to an impairment of the photosynthetic machinery within the symbiont [5]. Elevated temperatures reduce the efficiency of photochemistry by imposing damage to several photosynthetic proteins, including photosystem II (PSII) [12,13], light harvesting complexes [14], the Calvin cycle [15,16], or the thylakoid membrane [17, but see 18]. Understanding the proximal site or exact driver of thermal damage is confounded by the considerable diversity within the genus Symbiodinium, Currently, this group is divided into nine divergent phylogenetic clades (designated A-I) [19,20]. Furthermore, each clade consists of numerous distinct Symbiodinium genotypes that display functional differences in ecological and physiological attributes, including bleaching susceptibility [21,22]. Recent evidence also suggests that there is variability within the initial location of thermal-damage between different types of Symbiodinium [23].
In all oxygenic photosynthetic organisms, there is a tightlycoupled balance between the rate of light-induced damage to photosynthetic proteins and the rate of subsequent cellular repair [24,25]. Specifically, in order to avoid an accumulation of damaged PSII reaction centers, photosynthetic organisms have evolved several repair cycles that involve the proteolytic breakdown, synthesis, and re-insertion of the D1 subunit (PsbA, which is encoded by the chloroplast gene psbA) into the PSII protein complex [25,26]. As the D1 is one of the key structural proteins of the PSII reaction center core and responsible for providing the plastoquinone (Q B ) binding pocket, it has a pivotal role in maintaining the intact photosynthetic electron transport chain. Of particular importance for Symbiodinium, periods of elevated temperature disrupt the cellular maintenance of the D1 protein, either by inhibiting the overall rate of protein repair [12,27] or by inducing an accelerated rate of PSII photo-inactivation [13]. This imbalance triggers a condition known as photoinhibition, in which a rapid accumulation of photodamaged D1 ultimately reduces the content of functional PSII reaction centers, thereby prompting photosynthetic dysfunction. These episodes of chronic photoinhibition are acknowledged as a driving factor responsible for many instances of large-scale coral bleaching [12,28,29]. While there is less evidence for photoinhibition of photosystem I (PSI) and subsequent degradation of the two core PSI reaction center core heterodimer proteins that are encoded by the chloroplast genes psaA and psaB [30], chilling stress can lead to photoinhibition and PSI protein loss in some plants [31,32]. In contrast, a recent study has noted a significant decline in the stromal PSI protein, PsaC, in Symbiodinium within the reef coral Turbinaria reniformis exposed to 32uC. However, this decline was attributed to a possibly protective reduction in PSI content, as PSI electron flow was not hampered while PSII electron transport was [33].
Establishing how Symbiodinium spp. regulate their photosynthetic machinery under elevated temperatures is vital towards understanding the underlying mechanisms contributing to coral bleaching. While previous work has focused on the breakdown and repair of some photosynthetic proteins [13,[34][35][36], no studies have examined the expression of any chloroplast-encoded genes within Symbiodinium in response to thermal stress. Furthermore, the chloroplast of peridinin-dinoflagellates (e.g. Symbiodinium) is unique from all other photoautotrophs, as its genome is highly reduced and organized on individual ''minicircles'' (,2-3 kbp) [37], emphasizing that the functional control of these genes in these dinoflagellates may not be analogous to other photosynthetic systems with regards to their maintenance or expression [38,39]. Therefore, this study investigated the possibility of thermal alteration of the transcriptional response of two plastid genes, psbA and psaA, responsible for encoding core structural reaction center proteins located within photosystems II and I, respectively. The experiments monitored the photophysiology and transcript abundance of each gene using four cultured isolates and two in hospite Symbiodinium populations during simulated thermal stress events. This investigation provides evidence that, in some cases, the response to thermal stress is modulated at the transcriptional level, yet alternative points involving translational processes may also play a role within the unique chloroplasts of different Symbiodinium spp.

Culturing Conditions
Four Symbiodinium cultures from three divergent clades (A, B, and F) were grown in artificial seawater media (ASP-8A) [40] at 28uC on a 14:10 light:dark cycle in an environmental incubator (Percival Scientific, Inc.). Denaturing gradient gel electrophoresis (DGGE) of the ribosomal ITS2 region identified these cultures as A13 (formerly A1.1), A20, B1 and F2 (Culture IDs: 80 (A13), RD04 (A20), 13 (B1), and 133 (F2)) [41]. DGGE analysis was performed before and after each experiment to ensure that each Symbiodinium genotype remained consistent throughout the study. A20 is a putatively free-living Symbiodinium isolated from Caribbean environmental samples (LaJeunesse, personal communication; GenBank Accession: EU449053). Replicate flasks (n = 4) of each culture were gently bubbled with air passed through a carbon column to stabilize pH [42] and maintained semi-continuously in log-phase growth (fresh media replenished every 3-4 days) for at least two-weeks prior to the experiment. The environmental incubator simulated a sinusoidal light curve that gradually elevated the light intensity until mid-day, followed by a slow decline until the end of the cycle (75-165 mmol photons?m 22 ?s 21 ). These conditions attempt to replicate the natural progression of light intensity observed throughout a typical day from 08:00 to 22:00 (14:10 light:dark cycle; Figure S1).

Thermal Stress: Cultured Symbiodinium
In addition to their genetic identity, three of the Symbiodinium isolates were previously categorized as either thermally tolerant (F2) or thermally sensitive (A13 and B1) based on previous evidence of either stability or disruption in PSII photochemistry as well as protein repair of the D1 protein within PSII [43]. Additionally, type A20 (originally isolated by J. Rodgers (EPA)) was chosen as another thermally tolerant isolate based on preliminary temperature stress experiments. The cultures were treated by elevating the temperature from 28uC to 32uC at a rate of 1uC per day. Once the temperature reached 32uC, replicate flasks (n = 4) were sampled for RNA analysis (see details below) twenty-four hours later and then, every other day (48 h) for one week. All samples were removed at 10:00, which represented the 85 mmol photons?m 22 ?s 21 phase of the light cycle. Additionally, chlorophyll a fluorescence was measured at each time point to assess the photophysiology of each culture in response to the thermal stress. Following this experiment, new flasks of untreated Symbiodinium A13 were exposed to the same experimental conditions (i.e. 28uC to 32uC at a rate of 1uC per day); however, samples for RNA transcript analysis were only removed every 24 h (at 10:00 each day) during the initial rise in temperature to investigate the possibility of initial changes in transcript abundance during the thermal ramping. All gene expression profiles are shown relative to a set of untreated controls that were held at 28uC.

Thermal Stress: In Hospite Symbiodinium
Coral fragments from different colonies of Pocillopora (verrucosa morphospecies; [44]) were collected from Punta Galeras Reef, La Paz, MX (24u219159N, 110u179050W) and transported back to the laboratory at the Universidad Autónoma de Baja California Sur in July 2007. Colony selection was based on the symbiont composition of each coral, which was previously characterized by ITS-2 rDNA fingerprinting in 2006 [45], resulting in a total of 8 distinct colonies harboring either Symbiodinium C1b-c or D1 (n = 4 colonies for each symbiont). Further identification of the dominant Symbiodinium type from each colony was analyzed (DGGE-ITS2) again immediately before and after the experiment and no change in the specific symbiotic combinations were noted (data not shown). Six fragments from each Pocillopora colony were mounted to labeled 1.5-inch PVC pipe couplers with marine epoxy and divided among replicated control and treatment tanks (i.e. three fragments per colony in each treatment). Each tank was continually filled with natural seawater pumped from a shallow bay nearby and mechanically filtered through a sand bed. After a short 36 h acclimation period, the two experimental tanks were heated approximately by 1.5uC per day for 4 days and maintained at 32uC for 7 days. Conversely, the control tanks remained at the ambient seawater temperature (,26uC) for the duration of the heating treatment. During the experiment, all coral fragments were exposed to similar midday light levels as those measured at the reef site (950-1000 mmol photons?m 22 ?s 21 ) by applying an overhead shade cloth. Total RNA was extracted (see details below) from four fragments in each treatment on days 1, 6, and 7 (final sampling point) at 32uC. To account for any differences in the diel expression of Symbiodinium genes, all samples were isolated/ processed beginning at 09:00 each sampling day. All gene expression profiles are shown relative to the unheated controls (ambient tank), sampled at the same time and day as the thermally treated corals for each Symbiodinium type.

RNA Extractions and cDNA Synthesis
Total RNA was isolated from all Symbiodinium samples with a commercial extraction kit (TRI Reagent Solution, Ambion) that included the optional steps in the manufacturer's protocol to remove additional contaminating proteoglycans and polysaccharides. All cultured Symbiodinium samples were filtered onto 5.0 mm polycarbonate filters (15-30 mL) during their exponential growth phase, and then immediately submerged in 1 mL of TRI Reagent and stored at 280uC. Pocillopora tissue was removed from the skeletons using a WaterPikH with ,50 mL of 0.45 mm filtered seawater. The resulting homogenate was centrifuge at 5000 g for 5 min to isolate the Symbiodinium pellet. The supernatant was discarded and the pellet was resuspended and washed twice in 5 mL of filtered seawater. Each Symbiodinium pellet was frozen in liquid nitrogen and stored at 280uC until the RNA extractions. In order to fully lyse the Symbiodinium cells, samples were homogenized in the TRI Reagent with 1.0 mm glass beads for 2 minutes at full speed in a Mini Beadbeater (BioSpec Products). The resulting RNA pellets were resuspended in 50 mL nuclease-free H 2 O and quantified on a NanoDrop 2000 (Thermo Scientific). RNA samples (,500 ng total RNA) were treated with RQ1 RNase-Free DNase (Promega) according to the manufacturer's instructions to remove any residual DNA and reverse transcription reactions were performed with a High-Capacity cDNA Reverse Transcription kit (Applied Biosystems) using random primers to initiate cDNA synthesis. Additionally, reactions without the Reverse Transcriptase enzymes were carried out for each sample and subjected to PCR to ensure the removal of all residual DNA from each RNA extraction. None of these samples were successfully amplified, thus, confirming the efficacy of the DNase treatment. The resulting cDNA was diluted 1:5 with nuclease-free H 2 O and used as the template in the real-time PCR analysis. Table 1 presents a list of the real-time PCR primers used in this study. The Proliferating Cell Nuclear Antigen (PCNA) primers were previously established for Symbiodinium [46], while Sadenosyl-L-methionine synthetase (SAM) primers were selected due to their high expression stability in other Symbiodinium types exposed to both temperature and light stress [47]. Primers for psbA and b-Actin were designed based on sequence alignments of cDNA from multiple Symbiodinium types, as well as, several other dinoflagellates obtained from the NCBI GenBank database (www. ncbi.nlm.nih.gov). Multiple sequence alignments were performed in BioEdit [48] using the ClustalW Multiple Alignment tool [49]. Primer sets were designed from these alignments within conserved regions identified for each gene using DNASTAR (Lasergene Primer Select). Primer compatibility was tested against DNA isolated from distinct Symbiodinium cultures that span five different clades (ITS2-types: A13, A20, B1, C1b-c, D1 and F2). An additional psbA primer set was designed for the in hospite Symbiodinium analysis due to the amplification of a second product (also derived from the psbA gene; data not shown) in C1b-c samples. Furthermore, the b-Actin primers failed to amplify within Symbiodinium A20, while the PCNA primers only weakly amplified cDNA from the D1 alga.

Primer Design
Given the absence of any psaA sequence data for Symbiodinium, the psaA amino-acid sequence from the following microalgae were obtained from GenBank to create a multiple-sequence protein alignment: Glaucocystis nostochinearum (AAX82908.

Real-Time PCR and Gene Expression Analysis
Real-time PCR assays were performed on an ABI Prism 7500 Sequence Detection System (Applied Biosystems). Three technical replicates of each cDNA sample were amplified using 2 mL H 2 O, 5 mL SensiMix SYBR mastermix (Bioline), 1 mL of each forward and reverse primer (optimized final concentrations of 0.9 mM and 0.3 mM, respectively, for all primers), and 1 mL cDNA in 10 mL total volume reactions. PCR conditions consisted of 50uC for 2 min, 95uC for 10 min, followed by 40 cycles of 95uC for 15 s, 53uC for 30 s (48uC for the psaA primers), and 72uC for 1 min. Following each reaction, a dissociation step (melt curve analysis) from 60uC to 95uC was included to ensure that primer dimers did not significantly contribute to the fluorescence signal. In addition to the experimental samples, a five-step ten-fold dilution series of cDNA from a common control sample was included on each plate for every primer set to account for differences in priming efficiency between plates. A fixed fluorescent threshold of 0.05 was applied for all reactions to determine the cycle threshold (C T ) values. The resulting C T values of the technical replicates were averaged and used only if their standard deviation was less than 0.50 (typically less than 0.25). Additionally, the expression stability of the three reference genes under each set of experimental conditions was evaluated using the open source software geNorm [50] (http:// medgen.ugent.be/,jvdesomp/genorm/). All housekeeping genes received an Expression Stability value (M) of less than 1.5 and were thereby considered adequate reference genes for a particular experiment [50]. All psbA and psaA values were normalized to the geometric mean of three reference genes (SAM, Actin and PCNA) and expressed relative to untreated (control) samples taken at the same time as the treatment. As previously mentioned, the b-Actin and PCNA primers failed to amplify Symbiodinium A20 and D1, respectively. Therefore, these Symbiodinium types were only normalized against two of the three housekeeping genes.

Chlorophyll a Fluorescence
Photosynthetic activity during each experiment was assessed by active chlorophyll a fluorescence, using a fluorescence induction and relaxation system (FIRe fluorometer, Satlantic), using either the internal cuvette or with a fiber optic probe (for the cultures and corals respectively). Samples were dark-acclimated for 10-20 min, prior to the FIRe measurement protocol, which consisted of a 120 ms saturating single-turnover excitation pulse, followed by 40 weak modulated pulses (60 ms separation) to record the relaxation kinetics of PSII. Ten iterations of each fluorescence trace were averaged into a single transient to increase the signal-to-noise ratio. These data were fit to a biophysical model [51] to determine the values of minimum (F o ) and maximum (F m ) single turnover fluorescence, as well as the effective absorption cross-section of PSII (s PSII ). s PSII describes a theoretical ''size'' of the lightabsorbing target, inferred through the probability of a photon being absorbed and used for photochemistry [52]. Maximum quantum yield of PSII ((F m 2F o )/F m = F v /F m ) for the cultured Symbiodinium was recorded each sampling day (10:00), whereas the fluorescence measurements of each coral fragment were taken approximately one-hour after sunset (21:00), on each day of the experiment.

Statistical Analysis
All relative changes in gene expression were analyzed using the relative expression software tool, REST 2009 [53] (http://www. qiagen.com/products/rest2009software.aspx). This software employs a statistical randomization technique to test for up or downregulation of a gene, based on a mean expression value normalized against multiple reference genes, with 10,000 iterations. For each Symbiodinium culture, differences in F v /F m and s PSII from the untreated controls (Day 0) were tested using a one-way analysis of variance (ANOVA) in SPSS version 19.0. Data were tested for assumptions of normality (Sharpiro-Wilk) and homogeneity of variance (Levene's test), and when applicable, a Tukey's multiple comparison test analyzed the intraspecific differences of each fluorescence metric. For the in hospite Symbiodinium measure-ments, sample variance was not homogeneous at all time points (Levene's test; p,0.05), and therefore, differences in F v /F m and s PSII between treatment and control corals for a particular Symbiodinium type were determined using the nonparametric Mann-Whitney U test.

Results and Discussion
The four cultured Symbiodinium isolates differed in their chlorophyll a fluorescence in response to elevated temperatures (32uC) (Figure 1). The quantum yield of PSII (F v /F m ) provides insight into the efficiency at which PSII reaction centers are operating within the functional chloroplasts. Exposure to 32uC induced a significant reduction in F v /F m for both thermally sensitive culture strains (A13 and B1) beginning on day 3 and continuing to decline with each subsequent sampling point (Figure 1a; p,0.05, ANOVA); whereas, this metric remained relatively stable over the course of the experiment in the thermally tolerant isolates (A20 and F2). Similarly, the in hospite chlorophyll a fluorescence measurements of Pocillopora fragments revealed significant differences between the photochemical competency of Symbiodinium C1b-c and D1 exposed to thermal stress ( Figure 2). F v /F m declined significantly in heated corals associated with Symbiodinium C1b-c (Figure 2a; p = 0.016, Mann-Whitney U test), whereas colonies with D1 maintained a high level of photochemical efficiency for the duration of the thermal treatment (p.0.05). Similar to thermal treatments tested by Robison and Warner (2006) with several of these same cultured isolates (i.e. reduced F v / F m in B1 and A13 by 29% and 44%, respectively), the significant decline in photochemical efficiency observed in the thermally sensitive types (i.e. A13, B1 and C1b-c) was most likely due to a decline in functional PSII reaction centers and a loss of total D1 protein content [12,13,24,27,54].
The effective absorption cross-section of PSII (s PSII ) also exhibited different patterns amongst the Symbiodinium types as the experiment progressed (Figure 1b and 2b). Both A13 and C1b-c displayed significant increases in s PSII by Day 7 at 32uC (Figure 1b  rapidly lose the core of their PSII reaction centers (i.e. D1 protein) under thermal stress [43], its associated light-harvesting pigments may remain well 'connected' by continuing to absorb incoming photons and shuttling this energy to other nearby functional reaction centers. This data supports a 'lake' model of lightharvesting complex organization in Symbiodinium A13 and C1b-c, whereby light harvesting and energy transfer is distributed among several different PSII reaction centers [55, but see 36]. Importantly, an increase in s PSII during thermal stress will lead to greater susceptibility to photoinhibition by elevating the rate of photodamage to the remaining functional PSII [35]. Conversely, the F2 isolate exhibited the opposing pattern whereby F v /F m remained stable while s PSII declined (p,0.05, ANOVA; Figure 1). This trend was likely the result of either a reduction in chlorophyll a concentration [43] or a possible change in the xanthophyll content [56][57][58] as the alga begins to acclimate and employ different photoprotective mechanisms under elevated temperature [59]. Overall, the chlorophyll a fluorescence data indicates Symbiodinium B1, A13, and C1b-c were especially susceptible to thermal stress, undergoing a precipitous breakdown of photochemical efficiency under the elevated temperatures. Symbiodinium A13 and C1b-c also exhibited a unique trend in their expression of both psbA and psaA. By 24 h at 32uC, there was a significant decrease (p,0.05, REST) in the transcript abundance of both genes in the A13 isolate that persisted for the remainder of the experiment (Figure 3 and 4). This significant downregulation equates to a persistent loss of transcript abundance ranging from a 2.3-5.4 and a 1.8-4.6-fold decline of psbA and psaA, respectively ( Figure 3). Conversely, transcript levels remained stable for Symbiodinium B1, F2, and A20 ( Figure 3). Differences in transcript abundance under thermal stress were also observed for intact associations involving Pocillopora spp. and its associated symbionts. Here, Symbiodinium C1b-c had a considerably lower abundance of psbA transcripts on day 6 (3.5-fold decline) and both psbA and psaA on day 7 at 32uC (psbA: 2.6-fold decline; psaA: 2.1-fold decline; Figure 5; p,0.05, REST), while plastid transcript abundance in Symbiodinium D1 were indistinguishable from the controls (p.0.05, REST). The downregulation of psbA and psaA in in hospite Symbiodinium C1b-c confirms that this stress-induced decline is not solely an artifact from a single unique isolate of cultured Symbiodinium. However, the decline in core reaction center transcripts is also not a ubiquitous response that defines all 'sensitive' Symbiodinium types, as the B1 isolate maintained consistent transcript levels despite an increasing level of photochemical stress (Figure 1 and 3). Considering the differences in photosynthetic properties [60], as well as the variation in the initial site of thermally-induced damage [23], it is likely that the transcriptional response of these genes will differ between some Symbiodinium spp., independent of their thermal tolerance.
A closer inspection of Symbiodinium A13 during the temperature ramping process revealed that the reduction in the photosynthetic transcripts was only evident after 24 hrs at 32uC (p,0.05, REST; Figure 4). Prior to this time point, psbA and psaA transcripts were  actually upregulated during the temperature ramping process (i.e. 29, 31, and 32uC-2 hrs (psaA only); p,0.05, REST). An initial increase of chloroplast transcript abundance, including psbA and psaA, was also found in green pumpkin cotyledons grown under increasing temperatures until a upper stress-threshold was crossed, which marked a drastic decline in all chloroplast mRNA levels [61]. Elevating photosynthetic transcripts could be an initial acclimation mechanism employed by Symbiodinium A13 to adjust to increasing thermal stress. Prior exposure to elevated temperatures can improve the resiliency of both the coral host and the associated Symbiodinium community to a subsequent bleaching event [62,63]. Additionally, as the temperature increases, thermally sensitive Symbiodinium lose a significant portion of their functional photosynthetic proteins (e.g. D1 protein and light-harvesting antenna proteins) due to an increased rate of photodamage and/or a reduction in the repair rates of damaged photosynthetic machinery [13,14,27,35,36]. Perhaps, in addition to a suite of other molecular responses, Symbiodinium A13 initially acclimates to increased protein repair rates (or to counteract a reduction in protein repair rates) by upregulating the transcription of psbA and psaA. Nevertheless, this capacity for acclimation appears limited, as after 24 hrs at 32uC the transcript abundance for both genes was significantly reduced (Figure 3 and 4). A loss of plastid transcripts (including psbA and psaA) was noted in the green alga, Euglena gracilis, and wheat seedlings (Triticum aestivum) during heat stress, contributing to an impairment of protein synthesis [64,65]. Furthermore, the initial downregulation of both psbA and psaA (Day1 at 32uC; Figure 3) occurred prior to any observed differences in F v /F m (Day 3 at 32uC; Figure 1). The sudden decline of both genes suggests that A13 may have reached a critical level of stress prior to the commonly used physiological index of PSII photochemical efficiency. Again, the reduced availability of functional photosynthetic transcripts could greatly hinder the ability of some sensitive Symbiodinium spp. to handle periods of thermal stress by introducing another factor limiting the already impaired rates of photosynthetic protein repair [34][35][36].
The relative change of these plastid-encoded genes, in both the cultured and in hospite samples, is consistent with other studies measuring the gene expression of several nuclear-encoded transcripts within Symbiodinium [46,66] and other dinoflagellates [67][68][69][70]. The relatively small differences observed in the mRNA expression of Symbiodinium and other dinoflagellates under vastly different conditions has led to the suggestion that most of the changes in protein expression occurs post-transcriptionally [71][72][73][74]; however, the aforementioned gene expression studies were largely focused on nuclear expression and did not thoroughly evaluate the unique nature of dinoflagellate plastid transcription [39]. Regardless of whether or not Symbiodinium possess a strict form of plastid transcriptional control, the reduction of photosyn-  thetic transcripts from two thermally sensitive genotypes, either prior to or simultaneously while other proxies of photoinactivation are noted (e.g. declines in F v /F m and D1 protein content), suggests that the availability of chloroplast mRNA could play an important role in the ability of a symbiont to maintain an adequate level of photochemical competency during elevated temperatures. A significant reduction in transcript abundance could act in conjunction with (or even a precursor of) the impaired repair rates of photosynthetic proteins by limiting the substrate availability of the translational machinery. This scenario is especially probable if one considers the exceptionally caustic environment created within a stressed chloroplast [75].
Overall, a stress-induced reduction in chloroplast encoded transcripts could be the result of at least three distinct (yet not necessarily mutually exclusive) processes: (1) an active downregulation of plastid transcription, (2) impairment and/or damage to the transcriptional machinery within the chloroplast, (3) or the degradation of the available functional transcripts. All three processes may act simultaneously and could be initiated by similar underlying factors. For example, under elevated temperatures, thermally sensitive Symbiodinium undergo a physiological breakdown and/or damage to several photosynthetic components, such as the D1 protein of PSII, the thylakoid membrane, or inhibition to Rubisco [15,17,27]. This impairment of efficient photosynthesis results in a drastic reduction of ATP synthesis, along with a concurrent elevation of reactive oxygen species (ROS) [17,42,[75][76][77]. Under such ''energy-starved'' conditions, Symbiodinium A13 and C1b-c could actively downregulate transcription of psbA and psaA due to a decline in the available ATP energy reserves, or in order to conserve the remaining ATP for other vital cellular processes. Furthermore, the concurrent production of excess ROS is extremely destructive to many cellular constituents, including lipids, proteins, and nucleic acids [75]. These conditions create a caustic environment within a stressed chloroplast that could impair the transcriptional machinery or damage the available photosynthetic transcripts. Regardless of the exact underlying mechanism, the decline in these photosynthetic transcripts (especially psbA) for Symbiodinium A13 or C1b-c could significantly impair their ability to cope with periods of thermal stress. The reduction in plastidencoded transcripts could limit the substrate availability of the translational machinery and work in conjunction with either lower repair rates [27,35,36,60] or increased damage of photosynthetic proteins [13] to render certain Symbiodinium spp. exceptionally susceptible to thermal stress.