Assessment of metabolic modulation in free-living versus endosymbiotic Symbiodinium using synchrotron radiation-based infrared microspectroscopy

Abstract

The endosymbiotic relationship between coral hosts and dinoflagellates of the genus Symbiodinium is critical for the growth and productivity of coral reef ecosystems. Here, synchrotron radiation-based infrared microspectroscopy was applied to examine metabolite concentration differences between endosymbiotic (within the anemone Aiptasia pulchella) and free-living Symbiodinium over the light–dark cycle. Significant differences in levels of lipids, nitrogenous compounds, polysaccharides and putative cell wall components were documented. Compared with free-living Symbiodinium, total lipids, unsaturated lipids and polysaccharides were relatively enriched in endosymbiotic Symbiodinium during both light and dark photoperiods. Concentrations of cell wall-related metabolites did not vary temporally in endosymbiotic samples; in contrast, the concentrations of these metabolites increased dramatically during the dark photoperiod in free-living samples, possibly reflecting rhythmic cell-wall synthesis related to light-driven cell proliferation. The level of nitrogenous compounds in endosymbiotic cells did not vary greatly across the light–dark cycle and in general was significantly lower than that observed in free-living samples collected during the light. Collectively, these data suggest that nitrogen limitation is a factor that the host cell exploits to induce the biosynthesis of lipids and polysaccharides in endosymbiotic Symbiodinium.

1. Introduction

The marine Cnidaria–Symbiodinium endosymbiosis is an intriguing phenomenon resulting from the interaction between the gastrodermal cells of an animal host (corals and sea anemones) and their intracellular dinoflagellates of the genus Symbiodinium (i.e. endosymbionts). This mutual relationship not only plays a critical role in maintaining the health of individual hosts but also ultimately lays the foundation of the tropical marine ecosystem [1]. In the nutrient-poor tropical oceanic environment, cnidarian hosts provide inorganic nutrients (CO₂, NH₃ and ) to its endosymbionts. In return, dinoflagellates can release up to 95 per cent of their photosynthetically produced carbon and other photosynthate to their hosts [2]. Metabolite tracing using stable isotope ¹⁴C has detected that various metabolites, including sugar, carbohydrates, amino acids, glycerols and lipids are released from freshly isolated Symbiodinium into extracellular media upon the addition of host-free amino acids and ‘host-release factors’ [3–5]. Although these results have demonstrated a role of the host in the regulation of metabolite translocation from Symbiodinium, the extent of diel metabolome variability in endosymbiotic and free-living Symbiodinium metabolome remains unclear.

High throughput metabolomic screening using various mass spectrometries could be an important approach for elucidating the molecular mechanisms underlying stable endosymbioses [5], yet such techniques have never been applied to endosymbiotic anthozoans. Furthermore, these techniques do not allow simultaneous assessment of molecules belonging to different structural categories, such as lipids, carbohydrates, polysaccharides and proteins [6,7]. Here, we report the first application of a synchrotron radiation-based infrared microspectroscopy (SR-IMS) to compare diel changes of molecular composition in Symbiodinium between endosymbiotic and free-living states. Fourier transform infrared (FT-IR) spectroscopy is non-destructive, rapid and quantitative owing to its high sensitivity in simultaneously discriminating changes in functional groups of whole molecules (metabolites) in biological samples, and has been previously used as the first-round screening tool to provide preliminary metabolite data [7]. The recent coupling of a synchrotron radiation-based light source to FT-IR spectroscopy has greatly increased its sensitivity and accuracy [8]. In the present paper, SR-IMS is used to compare diel changes in the molecular composition of endosymbiotic and free-living Symbiodinium in order to provide insights into metabolic regulation in this important endosymbiosis.

2. Material and methods

Sea anemones (Aiptasia pulchella) were cultured in tanks with aerated sea water at ambient temperature (25°C) with a 12L (40 µmol m⁻² s⁻¹) : 12D photoperiod. Animals were fed Artemia nauplii twice weekly. The free-living Symbiodinium used in this study were originally isolated from the same A. pulchella stock and cultured according to a published procedure [9]. These cultures were replenished with fresh medium every five days and maintained in the laboratory for at least two years. The genetic identities of both endosymbiotic and free-living Symbiodinium were examined by PCR–RFLP analysis, and all were shown to be from clade B (electronic supplementary material, figure S1), indicating there was no detectable change in the cultured Symbiodinium. Furthermore, these free-living Symbiodinium were able to re-infect bleached hosts (electronic supplementary material, figure S2), demonstrating their viability and the feasibility of using them as the asymbiotic counterpart.

Aiptasia pulchella were starved for one week prior to experimentation to avoid interference from food particles in the compositional analysis. Symbiosomes containing endosymbiotic Symbiodinium were then isolated as previously described [10]. Free-living Symbiodinium were also treated with the same procedure of symbiosome isolation.

Endosymbiotic and free-living Symbiodinium were collected at the sixth hour of each light and dark photoperiod in a 12L : 12D cycle and then fixed with cold 10 per cent trichloroacetic acid (TCA) for 30 min at 4°C. Afterwards, the fixative was removed by centrifugation at 800g for 5 min. Symbiodinium pellets were washed with ddH₂O three times and then resuspended in ddH₂O at an approximate concentration of 2.5 × 10⁶ cells ml^–1. Eight to 10 aliquots of 2 µl cell suspension were separately spotted onto a low-e microscope slide (MirrIR, Kevley Technologies, Chesterland, OH, USA), followed by air-drying with a fast super dryer (less than 10% RH, SDC-100, Taiwan Dry Tech Corporation).

The synchrotron radiation-based infrared microspectroscope used in this study was located at Beamline BL14A1 of the National Synchrotron Radiation Research Centre (NSRRC) in Hsinchu, Taiwan. Infrared spectra were collected on a Magna-IR 860 FT-IR spectrometer (Thermo Nicolet) fitted with an IR microscope with a liquid-nitrogen-cooled mercury–cadmium–telluride detector. The procedure for the spectral acquisition and the profile comparison between the endosymbiotic and free-living Symbiodinium is outlined in electronic supplementary material, figure S3.

3. Results and discussion

Three experimental steps were critical for the success of the present study. First, after confirming their utility as an asymbiotic counterpart by demonstrating their genetic identity (electronic supplementary material, figure S1) and infection capability (electronic supplementary material, figure S2; see §2), free-living Symbiodinium were treated with the same procedure as that of the endosymbiotic Symbiodinium isolation [10] in order to serve as an appropriate experimental control. Second, to fix and preserve the biological composition of sampled Symbiodinium and prevent the formation of artificial chemical bonds during the sampling process, a coagulant (precipitating) fixative, TCA, was used. Coagulant fixatives reduce the solubility of protein molecules by disrupting hydrophobic interactions yet do not covalently bind to proteins as do the cross-linking, aldehyde-based fixatives [11]. Third, for spectral analyses, the amide II band (A₁₅₁₀) of each spectrum was used as the internal control for qualitative and quantitative analysis (electronic supplementary material, figure S3).

Owing to the fact that the absorption bands yielded by the Symbiodinium cells were quite complex in terms of their ability to be accurately identified, only the bands with a defined assignment and whose ‘area ratio’ differed significantly between free-living and endosymbiotic Symbiodinium (typically fold changes either greater than 1.6 or less than 0.4) were selected for analysis (figures 1 and 2). The absorption at peak 1 (A₃₀₁₂, Olefinic=CH stretching), which corresponds to the spectrum of unsaturated lipids (table 1), was 3.53-fold and 4.27-fold higher in endosymbiotic cells during light and dark photoperiods (figure 2a), respectively, than in free-living cells during the same periods. The absorption at peak 2 (A₁₇₃₃, C=O stretching, the ester functional groups in lipids), which represents the spectrum of lipids (table 1), was 1.66-fold and 1.93-fold higher in endosymbiotic cells during the light and dark photoperiods (figure 2a), respectively. This relative accumulation of lipids in endosymbiotic cells was further confirmed by cellular staining with a fluorescent BODIPY lipid probe (electronic supplementary material, figure S4).

Figure 1.

Averaged infrared spectra of the endosymbiotic (solid lines) and free-living (dotted lines) Symbiodinium in (a) 3600-2750 cm⁻¹ region, (b) 1750-950 cm⁻¹ region. A₁₅₁₀ the absorption of amide II (protein) used as the normalization control for qualitative and quantitative analysis. The assignments of peak numbers that are highlighted are defined in table 1.

Figure 2.

Quantitative analysis of Symbiodinium spectra in different states. (a) The endosymbiotic (solid lines) versus free-living (dotted lines) states. (b) The light versus dark photoperiods.

Table 1.

Band assignments of selected absorptions in IR spectra of Symbiodinium at 3600–950 cm⁻¹ region.

absorption peaks		definition
no.	frequency (cm−1)	frequency (cm−1)	assignments	biological molecules [references]
1	3012	3012	olefinic=CH stretching	unsaturated lipids, cholesterol esters [12]
2	1733	1739–1733	saturated ester C=O stretching	lipids (ester functional groups in lipids) [12]
3	1322	1370–1300	N=O stretching	nitrogenous compounds [13]
4	1151	1175–1140	C–O–C stretching	polysaccharides (glycosidic link) [14]
5	1118	1122–1118	C–C symmetric stretching, C–O–C stretching, C–OH bending	the polysaccharides of cell wall components [15]

The absorptions at peak 3 (A₁₃₂₂, N=O stretching), which represents the spectrum of nitrogenous compounds (possibly the nitrate-related compounds; table 1), was relatively lower in endosymbiotic cells, with an 0.12-fold decrease relative to free-living samples collected during the light photoperiod (figure 2a). The absorption at peak 4 (A₁₁₅₁, C–O–C stretching), which represents the spectrum of glycosidic link found within certain polysaccharides (oligosaccharides and disaccharides but not monosaccharides; table 1), was 1.76-fold and 2.03-fold higher in endosymbiotic cells sampled during the light and dark photoperiods (figure 2a), respectively, relative to free-living cells sampled at these times.

The absorption at peak 5 (A₁₁₁₈, C–C symmetric stretching, C–O–C stretching, C–OH bending) which was possibly composed of molecules associated with the polysaccharide of the cell wall (table 1) was relatively consistent in endosymbiotic samples over the light–dark photoperiod (ratio 0.94, figure 2b). Nevertheless, the absorption at this peak was significantly higher in free-living than in endosymbiotic cells sampled during the dark photoperiod (0.16 endosymbiotic/free-living ratio in figure 2a). These results are consistent with the occurrence of nocturnal cell wall biosynthesis in free-living Symbiodinium, an observation supported by prior cell cycle analyses [9].

In endosymbiotic Symbiodinium, the relative ratios of all selected absorption bands did not change significantly over time (figure 2b). However, in free-living cells, the relative ratio of peak 3 (A₁₃₂₂, corresponding to nitrogenous compounds; table 1), increased 6.24-folds in the light (figures 1b and 2b). This suggests that nitrogenous compounds were assimilated dramatically in free-living, but not endosymbiotic, cells during the light photoperiod. The fact that endosymbiotic cells appeared to be nitrogen-deprived compared with free-living cells is also evident as shown in figure 2a (0.12 and 0.76 endosymbiotic/free-living ratio in light and dark photoperiods, respectively).

Understanding the metabolic relationship between host and endosymbiont is of utmost importance for unravelling the cellular mechanisms underling stable endosymbiosis [5]. The present study has revealed several key metabolic differences between endosymbiotic and free-living Symbiodinium, including lower concentrations of nitrogenous compounds and higher levels of lipids, unsaturated lipids and polysaccharides in the former. The underlying of such variation in metabolite accumulation remains to be elucidated. However, accumulation of lipids and carbohydrates has been observed in the freshwater microalgae Chlamydomonas reinhardtii and Scenedesmus subspicatus grown in batch culture with limiting nitrogen concentrations [16]. In addition, we have observed (P. L. Jiang et al. 2011 unpublished results) that a dramatic accumulation of lipids can be induced in free-living Symbiodinium treated with nitrogen-deprived media. This strongly implies that nitrogen limitation may be the mechanism by which the cnidarian host induces the biosynthesis of lipids, unsaturated lipids and polysaccharides in its intracellular symbionts. Indeed, lipids may represent one of the major energy resources that corals derive from endosymbionts, given the recent finding of the importance of host lipid bodies in corals [17]. Although lipid accumulation owing to environmental stresses has also been observed in other symbiosis, such as rhizobia [18], its mechanism remains to be elucidated.