Waiting
Login processing...

Trial ends in Request Full Access Tell Your Colleague About Jove
JoVE Journal Biology

Biology

Gas Chromatography-Mass Spectrometry-Based Targeted Metabolomics of Hard Coral Samples

Published: October 13, 2023 doi: 10.3791/65628
Automatic Translation
1, 1, 2, 3, 2
1Climate Change Cluster (C3), University of Technology Sydney, 2Metabolomics Australia, Bio21 Molecular Science and Biotechnology Institute, The University of Melbourne, 3School of Biological Sciences, Victoria University of Wellington

Summary

Here, we present the extraction and preparation of polar and semi-polar metabolites from a coral holobiont, as well as separated coral host tissue and Symbiodiniaceae cell fractions, for gas chromatography-mass spectrometry analysis.

Abstract

Gas chromatography-mass spectrometry (GC-MS)-based approaches have proven to be powerful for elucidating the metabolic basis of the cnidarian-dinoflagellate symbiosis and how coral responds to stress (i.e., during temperature-induced bleaching). Steady-state metabolite profiling of the coral holobiont, which comprises the cnidarian host and its associated microbes (Symbiodiniaceae and other protists, bacteria, archaea, fungi, and viruses), has been successfully applied under ambient and stress conditions to characterize the holistic metabolic status of the coral.

However, to answer questions surrounding the symbiotic interactions, it is necessary to analyze the metabolite profiles of the coral host and its algal symbionts independently, which can only be achieved by physical separation and isolation of the tissues, followed by independent extraction and analysis. While the application of metabolomics is relatively new to the coral field, the sustained efforts of research groups have resulted in the development of robust methods for analyzing metabolites in corals, including the separation of the coral host tissue and algal symbionts.

This paper presents a step-by-step guide for holobiont separation and the extraction of metabolites for GC-MS analysis, including key optimization steps for consideration. We demonstrate how, once analyzed independently, the combined metabolite profile of the two fractions (coral and Symbiodiniaceae) is similar to the profile of the whole (holobiont), but by separating the tissues, we can also obtain key information about the metabolism of and interactions between the two partners that cannot be obtained from the whole alone.

Introduction

Metabolites represent the end products of cellular processes, and metabolomics - the study of the suite of metabolites produced by a given organism or ecosystem - can provide a direct measure of organismal functioning1. This is particularly critical for exploring ecosystems, symbiotic interactions, and restoration tools, as the goal of most management strategies is to preserve (or restore) specific ecosystem service functions2. Coral reefs are one aquatic ecosystem that demonstrates the potential value of metabolomics for elucidating symbiotic interactions and linking coral physiological responses to community-level and ecosystem-level impacts3. The application of high-throughput gas chromatography-mass spectrometry (GC-MS) is especially valued due to its capacity to rapidly analyze a broad range of metabolite classes simultaneously with high selectivity and sensitivity, provide rapid compound identification when spectral libraries are available, and provide a high level of reproducibility and accuracy, with a relatively low cost per sample.

Corals are holobionts consisting of the coral animal, photosynthetic dinoflagellate endosymbionts (family: Symbiodiniaceae4), and a complex microbiome5,6. Overall, the fitness of the holobiont is maintained primarily through the exchange of small molecules and elements to support the metabolic functioning of each member7,8,9,10. Metabolomic approaches have proven especially powerful for elucidating the metabolic basis of symbiosis specificity9,11, the bleaching response to thermal stress7,8,12,13, disease responses14, pollution exposure responses15, photoacclimation16, and chemical signalling17 in corals, as well as aiding in biomarker discovery18,19. Additionally, metabolomics can provide valuable confirmation of the conclusions inferred from DNA- and RNA-based techniques9,20. There is, therefore, considerable potential for the use of metabolomics for assessing reef health and developing tools for reef conservation3, such as through the detection of metabolic biomarkers of stress18,19 and for examining the potential of active management strategies such as nutritional subsidies21.

Separating the host and symbiont cells and analyzing their metabolite profiles independently, rather than together as the holobiont, can yield more information about the partner interactions, independent physiological and metabolic statuses, and potential molecular mechanisms for adaptation11,12,22,23,24. Without separating the coral and Symbiodiniaceae, it is almost impossible to elucidate the contribution and metabolism of coral and/or Symbiodiniaceae independently, except for with complex genome reconstruction and metabolic modeling25, but this has yet to be applied to the coral-dinoflagellate symbiosis. Furthermore, attempting to extract information about the individual metabolism of the host or algal symbiont from the metabolite profile of the holobiont can lead to misinterpretation.

For example, until recently, the presence of C18:3n-6, C18:4n-3, and C16 polyunsaturated fatty acids in extracts from coral and holobiont tissues was thought to be derived from the algal symbiont, as corals were assumed to not possess the ωx desaturases essential for the production of omega-3 (ω3) fatty acids; however, recent genomic evidence suggests that multiple cnidarians have the ability to produce ω3 PUFA de novo and further biosynthesize ω3 long-chain PUFA26. Combining GC-MS with stable isotopic labeling (e.g., 13C-bicarbonate, NaH13CO3) can be used to track the fate of photosynthetically fixed carbon through coral holobiont metabolic networks under both control conditions and in response to external stressors27,28. However, a critical step in the tracking of 13C fate is the separation of the coral tissue from the algal cells-only then can the presence of a 13C-labeled compound in the coral host fraction be unequivocally assigned as a Symbiodiniaceae-derived metabolite translocated to the coral or a downstream product of a translocated labeled compound. This technique has demonstrated its power by challenging the long-held assumption that glycerol is the primary form in which photosynthate is translocated from symbiont to host29, as well as elucidating how inter-partner nutritional flux changes during bleaching27,28 and in response to incompatible Symbiodiniaceae species11.

While the decision to separate tissues is primarily driven by the research question, the practicality, reliability, and potential metabolic impacts of this approach are important to consider. Here, we provide detailed, demonstrated methods for the extraction of metabolites from the holobiont, as well as the separate host and symbiont fractions. We compare the metabolite profiles of the host and symbiont independently and how these profiles compare to the holobiont metabolite profile.

Subscription Required. Please recommend JoVE to your librarian.

Protocol

NOTE: The experimental design, sample collection and storage have been described in detail elsewhere2,30,31. Permit approval for the collection of wild corals must be obtained prior to collection and experimentation. The samples here were collected from colonies of Montipora mollis (green colour-morph) imported from Batavia Coral Farms (Geraldton, WA), originally collected from a reef off the Abrohlos Islands (Western Australia; 28°52'43.3"S 114°00'17.0"E) at a depth of 1 m under the Aquaculture License AQ1643. Prior to sampling, the colonies were held in an 800 L aquarium at 35 PSU, under blue and white light at 150 µmol photons·m−2·s−1, and at 25 °C ± 0.5 °C for 3 months. The coral fragments (~5 cm2, N = 6) were snap-frozen in liquid nitrogen and stored at −80 °C until processing.

1. Preparation of the extraction solutions and equipment

  1. At least 1 day prior to coral tissue removal, prepare the extraction solutions and equipment.
  2. Prechill ultrapure water in clean, detergent-free glassware at 4 °C.
  3. Mix 100% LC-grade methanol with a 10 µg·mL−1 final concentration of appropriate internal standard(s) (e.g., 13C6 sorbitol).
  4. Create a 50% methanol extraction solution using half 100% LC-grade methanol and half ultrapure water. Store both methanol solutions at −20 ˚C.
    NOTE: To help prevent degradation of the metabolites, it is recommended to perform the sample processing steps in batches of five coral fragments at a time, with one additional biological (water only) blank (total samples N = 6). Once each coral sample has been separated into the two fractions (coral host tissue, henceforth "Host", and microalgal cells, henceforth "Symbiont"), the total number of samples in one processing batch will be 12.

2. Coral metabolism quenching

NOTE: The experimental design, sample collection and storage have been described in detail elsewhere2,30,31. However, it should be noted that the time taken to quench metabolism (i.e., the time between coral sample collection and preservation) is critical to capture the original response30. Preserve the sample as quickly as possible after collection to prevent changes in the metabolite composition due to sample degradation or non-target physiological responses32.

  1. Place coral fragment in a sterile sample collection bag, and drain any excess seawater as much as possible. Submerge the sample in liquid nitrogen for a minimum of 30 s. Move the samples as soon as possible to a −80 ˚C freezer for storage.
    NOTE: Samples can be frozen at −80°C in light-blocked containers until processing, avoiding freeze-thaw cycles.

3. Coral tissue removal from the skeleton

NOTE: The samples should be kept on ice (4 °C) at all times to ensure they are simultaneously in liquid form whilst preventing ongoing metabolism.

  1. Place a clean, sterile sample collection bag on ice so that the bag is stable and open on top of the ice in a shallow well but is not submerged in the ice. Add 10 mL of cold (4 ˚C) ultrapure water to the bag.
    NOTE: This will help to avoid repeated freeze-thawing of the coral fragment due to the cold pressurized air and surrounding ice.
  2. Select a coral fragment with sterilized tweezers, and rinse with cold (4 ˚C) ultrapure water using a sterile Pasteur pipette until no seawater residue remains. Submerge the rinsed coral fragment in the bag containing the 10 mL of ultrapure water.
    NOTE: This rinse is critical to remove any residual salts that would interfere with the downstream analysis. Avoid hand contact with the water or coral fragment through the bag to maintain the sample at 4 ˚C.
  3. Tape a sterile 1 mL pipette tip over the end of an air gun with electrical tape, with ~5 mm cut off the end of the tip (Figure 1A).
  4. Aim the air gun onto the coral fragment with the bag half-sealed and the air flow on low-medium to gently remove the tissue by encouraging a circular movement of the water over the coral fragment.
  5. After ~3 min, or when all the tissue appears to have been removed from the skeleton, turn off the air, and remove the airbrush. Completely seal the bag.
  6. Squeeze all the removed coral tissue to a bottom corner of the bag. Cut off the opposite corner and gently pour the contents of the bag into a 15 mL tube on ice.

4. Optional homogenization

NOTE: Some coral species are more viscous than others, meaning the air-brushing will remove the tissue in clumps instead of in a slurry. If clumps of tissue are visible in the air-brushed homogenate, a homogenization step at 4 °C can be added for all the samples.

  1. Clean a mechanical saw-tooth homogenizer twice with 4 ˚C 70% methanol and finally with 4 °C ultrapure water.
  2. Homogenize the coral sample in a 15 mL tube for ~1 min until the sample is fully homogenized and no clumps are visible.
  3. Clean the homogenizer as in step 4.1 between each sample. Keep the homogenization time consistent across the samples.

5. Sample collection for normalization

  1. Collect a 1,000 µL aliquot from the homogenized tissue for Symbiodiniaceae cell counts, coral host tissue protein content analysis, and chlorophyll a estimation. Store at −20 ˚C until ready to analyze (section 10).

6. Optional coral host tissue-Symbiodiniaceae cell separation

  1. Centrifuge the coral homogenate at 2,500 × g for 5 min at 4 °C using a refrigerated centrifuge.
    NOTE: This speed is optimal to separate the heavier Symbiodiniaceae cells, while keeping their cell walls intact, from the host tissue, which is suspended in the supernatant.
  2. Remove the supernatant containing the host material, and place in a new 15 mL tube.
    NOTE: The lipids from the host tissue typically form a narrow pink/white layer on top of the symbiont cells. This layer can be collected along with the soluble host supernatant via pipetting (Figure 1B).
  3. Vigorously vortex the host for exactly 1 min. Keep the algal pellet sample and host supernatant sample on ice.
  4. Add 2 mL of ultrapure water at 4 °C to the algal pellet. Vigorously vortex for exactly 2 min to resuspend the pellet.
    NOTE: If individual fragments of 1 cm were not collected from the coral colony for Symbiodiniaceae genotyping, a 200 µL aliquot of the Symbiodiniaceae cell suspension can be collected here, preserved in the preferred DNA buffer solution, and stored as described in Thurber et al.30 for Symbiodiniaceae genotyping (e.g., as per González-Pech et al.12).
  5. Repeat steps 6.1-6.4 once more.
    NOTE: The reliable separation of the host and symbiont depends on the coral biomass and species, as some species can be more viscous than others. A minimum of three wash steps is recommended, but this can be increased depending on the separation success. Repeat wash steps 4.7-4.9 until no Symbiodiniaceae cells can be seen in the bottom of the host fraction and until the Symbiodiniaceae fraction is visibly free of host material (e.g., no white layer on top) (Figure 1).
  6. Remove the supernatant containing the host material, and place in a new 15 mL tube.
  7. Retain the symbiont pellet in the 15 mL tube.

7. Sample drying

  1. Freeze either the holobiont homogenate or both the separated host and Symbiodiniaceae fractions, at −80 °C for ~120 min. Lyophilize the samples overnight with a 0.01 mbar vacuum at −85 ˚C.
    NOTE: To avoid sample loss during lyophilization, it is recommended to use a lid cut from another sterile tube, or sterile parafilm, with a ~2 mm hole punched through carefully using a sterile 25 G needle.
  2. When dry, using a laboratory balance, weigh one of the following: 1) 25 mg of the holobiont; 2) 15 mg of the symbiont fraction; or 3) 30 mg of the host tissue from each sample into separate 2 mL plasticizer-free microcentrifuge tubes.
    NOTE: Critical step: The optimization of the biomass for extraction is essential to ensure that the GC-MS is not overloaded while ensuring sufficient signal. The dried coral material is very static. To avoid sample loss, use anti-static devices to eliminate electrostatic charges from the samples and weighing vessels. A simple and cost-effective alternative is to place a laundry dryer sheet under the sample tube. The dried symbiont pellet can be cut to the desired weight using a sterile blade.

8. Intracellular metabolite extractions

  1. Intracellular metabolite extraction from lyophilized holobiont:
    1. Add 400 µL of 100% cold (−20 ˚C) methanol with internal standard/s (IS; 13C6 sorbitol and/or 13C5-15N valine, at 10 µM) to each tube.
    2. Add a small number of 710-1,180 µm acid-washed glass beads (~10 mg) to each sample. Place in a bead mill at 50 Hz for 3 min in a prechilled (−20 ˚C) bead mill insert.
    3. Add an additional 600 µL of 100% cold (−20 ˚C) methanol with ISs (13C6 sorbitol and/or 13C5-15N valine, at 10 µM) to each tube.
    4. Vortex to mix for 1 min. Place on a rotisserie shaker at 4 ˚C for 30 min.
  2. Intracellular metabolite extraction from separated lyophilized Symbiodiniaceae cells:
    1. Add 200 µL of 100% cold (−20 ˚C) methanol with ISs (13C6 sorbitol and/or 13C5-15N valine, at 10 µM) to the dried Symbiodiniaceae material.
    2. Add a small number of 710-1,180 µm acid-washed glass beads (~10 mg). Place in a bead mill at 50 Hz for 3 min in a prechilled (−20 ˚C) bead mill insert.
    3. Add a further 800 µL of 100% cold (−20 ˚C) methanol with ISs, and vortex for 30 s.
  3. Intracellular metabolite extraction from separated lyophilized host tissue:
    1. Add 1 mL of 100% cold (−20 ˚C) LC-grade methanol containing ISs (13C6 sorbitol and/or 13C5-15N valine, at 10 µM) to the dried host material.
    2. Vortex to mix for 20 s. Place in a floating tube holder in a sonication bath set at 4 ˚C for 30 min.

9. Metabolite extract purification

  1. Centrifuge the samples (holobiont/host/symbiont) at 3,000 × g for 30 min at 4 °C.
  2. Transfer all the supernatant to a new 2 mL microcentrifuge tube, being careful not to disturb the cell debris pellet.
    NOTE: These are the semi-polar extracts. These can be kept on ice temporarily but stored long-term at −80 ˚C in the dark.
  3. To the remaining cell debris, add 1,000 µL of 50% cold (−20 ˚C) methanol. Vigorously vortex for 1 min to resuspend.
  4. Centrifuge the samples at 3,000 × g for 30 min at 4 °C.
  5. Collect and pool the supernatant (polar extracts) with the semi-polar extracts from the same sample.
    NOTE: The cell debris can be stored at −80 ˚C and used for protein content normalization (section 11).
  6. Centrifuge the pooled extracts at 16,100 g for 15 min to remove all the precipitates, and move the supernatant to a new plasticizer-free microcentrifuge tube (2 mL).
    NOTE: The sample extracts can be stored at −80 ˚C in the dark.
  7. When ready to analyze, aliquot 50 µL of each extract into a glass insert. Concentrate for 30 min at 30 ˚C using a vacuum concentrator. Repeat four more times (for 250 µL of total dried extract).
    NOTE: The dried samples can be stored at room temperature under desiccant conditions until analysis.

10. Metabolite derivatization

NOTE : A two-step online derivatization process is used for the methoximation and trimethylsilylation of the polar metabolites.

  1. Add 25 µL of methoxyamine hydrochloride (30 mg/mL in pyridine) to each sample.
  2. Agitate at 37 °C on an orbital shaker set at 750 rpm for 2 h.
  3. Add 25 µL of N,O-bis (trimethylsilyl)trifluoroacetamide + trimethylchlorosilane to each sample.
  4. Agitate again at 37 °C and 750 rpm for 1 h.
  5. Allow the samples to equilibrate at room temperature for 1 h before injecting 1 µL in a 1:10 split ratio onto the GC.

11. Gas chromatography-mass spectrometry analysis

NOTE: The mass spectrometer should be tuned according to the manufacturer's recommendations using tris-(perfluorobutyl)-amine (CF43).

  1. Use ultra-high purity helium as the carrier gas at a constant column flow rate of 1 mL/min.
  2. Use a 30 m DB-5 column with a 1 µm film thickness and a 0.25 mm internal diameter.
  3. GC oven program
    1. Set the inlet temperature to 280 °C.
    2. Start at the injection with an oven temperature of 100 °C, and hold for 4 min.
    3. Increase the temperature by 10 °C/min to 320 °C, and then hold for 11 min.
  4. Mass spectrometer parameters
    1. Set the MS transfer line to 280 °C, and adjust the ion source to 200 °C.
    2. Use argon as the collision cell gas to generate the multiple reaction monitoring (MRM) product ion.
    3. Achieve metabolite detection relative to a time-segmented MRM library containing MRM targets.

12. Symbiodiniaceae cell counts, coral host tissue protein content analysis, and chlorophyll a estimation

  1. Symbiodiniaceae cell counts:
    1. Take an aliquot of the coral tissue homogenate.
    2. Centrifuge the samples at 2,000 × g to pellet the algae.
    3. Remove the ~200 µL supernatant from the algal pellet, and place in a new microcentrifuge tube.
      NOTE: This will be the protein sample which will be used to normalize the data; store it at −20 ˚C before analyzing, if necessary.
    4. Resuspend the algal pellet in 1 mL of filtered sea water by gently pipetting up and down. If necessary, dilute the algal suspension to facilitate the cell counting.
    5. Conduct a cell count using a hemocytometer under a light microscope by adding 10 µL to one of the chambers. Complete 8-10 counts per sample.
      NOTE: Alternative methods for counting the algal cells can also be applied where available (e.g., flow cytometry, high-throughput confocal microscopy).
    6. Calculate the concentration of the symbiont cells (mL−1), taking into account any dilution factors used.
  2. Assay for the protein content
    1. Quantify the sample protein content (e.g., via Bradford's colorimetric assay, as initially described by Bradford et al.33, or the Lowry assay34,35, the protocol for which has been described for cnidarians elsewhere36).
  3. Chlorophyll a extraction
    1. Use a cell pellet of ~200, 000 cells, frozen or fresh.
    2. Transfer each algal pellet into 2 mL of dimethylformamide (DMF) in a glass scintillation vial, and incubate in the dark at 4 ˚C for 48 h.
      NOTE: DMF is toxic and carcinogenic, so the sample preparation must be completed under a fume hood that is as dark as possible and on ice. If there are <200,000 cells, use less DMF.
    3. Centrifuge for 3 min at 16,000 × g.
    4. Transfer 200 mL into a UV-96 well plate for photometric measurements. Run each sample in triplicate with DMF as the blank.
    5. Measure the absorbance at wavelengths (E) 663.8 nm, 646 nm, and 750 nm. Subtract the absorbance at 750 nm from the absorbance at both of the other wavelengths.
      NOTE: Measuring at 750 nm corrects for any scattering or turbidity in the sample.
    6. Calculate the chlorophyll a concentration (µg/mL) using equation (1):
      Chl a concentration(µg/mL) = (12.00 × E663.8) - (3.11 × E646.8)    (1)

13. Quantification of the cell biomass following metabolite extractions for normalization

NOTE: There are two options for the quantification of cell biomass described below: the quantification of protein related to biomass using a modified Bradford colorimetric method and the measurement of the cell debris dry weight. Either method is appropriate to use, as both offer accurate quantification of the cell biomass.

  1. Protein content of the cell debris
    1. Resuspend the frozen cell debris with 1 mL of 0.2 M NaOH, and incubate the samples at 98 ˚C for 20 min.
    2. Cool the samples on ice for ~10 min, and centrifuge at 3,000 × g for 5 min at an ambient temperature.
    3. Quantify the sample protein content (e.g., via Bradford's colorimetric assay, as initially described by Bradford et al.33 and modified by Smart et al.37).
  2. Measurement of cell debris dry weight
    1. Resuspend the cell debris from the intracellular metabolite extraction in double-distilled water (~10 mL).
    2. Filter the solution under a vacuum using a preweighed membrane filter (0.22 µm pore, 47 mm).
    3. Wash the tubes containing biomass twice with ultrapure water to ensure the complete transfer of biomass to the membrane filter.
    4. Remove the membrane filter containing the biomass and dry it using a microwave oven (low power; ~250 W for 20 min).
    5. Store the filter paper in a desiccator overnight. Record the dry weight of the filter paper and calculate the dry weight of the biomass by subtracting the weight of the dry membrane filter (using a clean dry membrane filter dried alongside the sample filter) from the total weight.

14. Data analysis

  1. Analyze metabolite targets using metabolite databases where each target is comprised of a quantifier and qualifier MRM.
  2. Visually inspect the detected metabolite targets and manually integrate them as required.
  3. Use a metabolite peak area to calculate relative abundance of each sample for each group. Values are blank corrected and normalized to sample internal standard peak area, and then to sample cell debris protein content as per Smart et al.37.
  4. Discard metabolites with a relative standard deviation greater than 35% in all the treatment groups (N = 23 metabolites).
  5. Transform the data (e.g., cube root), and mean-center them; confirm a normal distribution and homogeneity of variance.
  6. Perform the data analysis (ANOVA and heatmap construction; e.g., using https://www.metaboanalyst.ca)38. Cluster the samples to examine within-treatment variability using the packages “cluster”, “factoextra,” and “klustR”. Calculate the gap statistic (a method to determine the optimal number of clusters39) using the "clusGap" function in R and plots using the R package "tidyverse". Perform PERMANOVAs to examine the significance in the separation between the treatment metabolite profiles (e.g., in Primer).

Subscription Required. Please recommend JoVE to your librarian.

Representative Results

All the data produced during this work are available in the supplementary information.

Host-symbiont separation

Figure 1
Figure 1: Setup and validation of the separation of coral host tissues and Symbiodiniaceae cells. (A) The air gun setup for the removal of coral tissue from the coral skeleton. A pipette tip is attached to the air gun with electrical tape, and a small (~5 mm) section is cut from the tip to allow for more air to escape without dislodging the tip. (B) Examples of the holobiont and separated host tissue (supernatant) and Symbiodiniaceae cells (pellet). The value represents number of centrifugation steps. The arrow points to the narrow host lipid layer on top of the symbiont pellet that can be collected and combined with the host fraction. (C-E) Brightfield (top) and chlorophyll autofluorescence (bottom) microscopy images of an aliquot of the (C) holobiont with both host tissue and symbiont cells, (D) the host fraction without symbiont cells as verified by the absence of any chlorophyll autofluorescence, and (E) intact symbiont cells, demonstrating the removal of the host tissue. Scale bars = 100 µm. Please click here to view a larger version of this figure.

Microscopic visualization showed no Symbiodiniaceae cells in the host tissue samples following the three wash steps (Figure 1D). Similarly, there was minimal host tissue present in the symbiont fractions (Figure 1E). However, the holobiont homogenate (Figure 1C) showed that the intracellular Symbiodiniaceae may not have been adequately released from their symbiosomes by simple air-brushing and, thus, not as sufficiently lysed during the mechanical homogenization (protocol section 4) or solvent extraction (protocol section 8) compared to centrifugal separation (Figure 1E). This limitation could explain some of the large within-group variation between the holobiont samples and specific observations in the holobiont profiles. For example, two metabolites (glycolic acid and 1-hexadecanol) were significantly more abundant in the holobiont versus symbiont but not in the host versus symbiont; this may have been a result of the large within-group variation in the relative abundance of these metabolites in the holobiont profiles. In particular, holobiont sample 3 had relatively higher concentrations of dodecanoic acid, glycolic acid, and 1-hexdecanol than any of the symbiont or host samples individually. As the metabolite peak area data are normalized to sample protein content, if the symbiont cells were not cleaned sufficiently of host tissue, then the holobiont protein content may have been underestimated, thus influencing the biomass normalization and leading to the calculation of a higher metabolite abundance relative to the biomass in this sample. This further highlights the potential for increased variation in holobiont metabolomics.

Metabolite profile analysis
The choice of the mass spectrometer mode is dependent on the analysis being performed. For steady-state metabolite profiling, a comprehensive targeted methodology using a QqQ-MS in MRM mode enables improved metabolite detection and identification due to the elimination of the background noise that may be generated by high biological salt concentrations. For stable isotope-labeling approaches, a mass spectrometer (i.e., single quadrupole, triple quadrupole [QqQ], quadrupole-time of flight [QTOF]) running in scan mode allows for the detection of mass isotopologs that indicate stable isotopic enrichment patterns.

Targeted GC-MS analysis was completed using the Shimadzu Smart Metabolite Database (v3; which covers approximately 350 endogenous metabolites and multiple stable isotopically labeled internal standards) and the Shimadzu LabSolutions Insight software. Across all the treatments, 107 annotated metabolites were identified, including a suite of amino acids, organic acids, carbohydrates, fatty acids, and sterols (Supplementary Table S1). Qualitative and quantitative ion pairs are provided in Supplementary Table S2. Kmeans clustering identified three distinct clusters of samples (verified by a gap statistic test), with all the samples in separate, distinct clusters; the symbionts were in in Cluster 1, the host in Cluster 2, and the holobiont samples in Cluster 3 (Figure 2).

Figure 2
Figure 2: Kmeans cluster analysis of the metabolite profiles. (A) The sample metabolite profiles were clustered by Euclidean distance by the optimal number of clusters (N = 3), which was verified via a gap statistic calculation. (B) Parallel coordinate visualization of the average metabolite relative abundance (line, colored according to the cluster) and confidence interval (shading, colored according to the cluster) for each cluster (red = Cluster 1, green = Cluster 2, blue = Cluster 3). Please click here to view a larger version of this figure.

Host and symbiont metabolite profile comparison
The host and symbiont profiles were significantly distinct from each other (PERMANOVA, t = 16.909, p < .001; Supplementary Table S3), with 100 individual metabolites significantly different between the host and symbiont fractions (ANOVA, FDR < .05; Figure 3 and Supplementary Table S1). Of these, 13 metabolites were significantly more abundant in the symbiont than the host extracts, including the eicosanoids docosahexaenoic acid (C22:6[ω-6]; DHA), eicosapentaenoic acid (C20:5[ω-6]; EPA), and arachidonic acid (C20:4[ω-6]; ARA) (ANOVA, FDR < 0.05; Figure 3 and Supplementary Table S1). A total of 87 metabolites were less abundant in the symbiont than the host extracts (ANOVA, FDR < 0.05; Figure 3 and Supplementary Table S1).

Figure 3
Figure 3: Heatmap visualization of the metabolite relative abundances with post-hoc analysis results of the group comparisons. The host, symbiont, and holobiont samples (N = 5 per group) were hierarchical clustered via Ward's linkage, and the metabolites were clustered according by Euclidean distance measure. The colored squares indicate significant differences in the group comparisons detected via ANOVA with post hoc Tukey's HSD (Supplementary Table S1). Please click here to view a larger version of this figure.

Holobiont metabolite profiles compared to separated host and symbiont profiles
The holobiont metabolite profiles demonstrated large within-group variability, substantiated by the large separation of samples in the holobiont cluster in the Kmeans analysis; specifically, Sample 3 and Sample 4 were separated along dimension 2 from Sample 1, Sample 2, and Sample 5 (Figure 2A). The holobiont samples were intermediate between the separated host and symbiont fractions (Figure 2A). While the Kmeans cluster distributions (Figure 2A), parallel coordinates (Figure 2B), and heatmap visualization of the metabolite relative abundance (Figure 3), indicated that the holobiont profile more closely matched the host fraction profile, the holobiont profile significantly differed from both the host (PERMANOVA, t = 3.47, p < 0.001; Supplementary Table S3) and symbiont profiles (PERMANOVA, t = 11.29, p < 0.001; Supplementary Table S3). At the individual metabolite level, 66 and 82 metabolites in the holobiont were significantly different to the host and symbiont profiles, respectively (ANOVA, FDR < 0.05, Supplementary Table S1). Of these, 54 (81.8%) out of the 66 significant metabolites had significantly higher relative abundance in the host than the holobiont fraction, and 78 (95%) had significantly higher relative abundance in the holobiont than the symbiont fraction; four were more abundant in the symbiont fractions, including DHA, glycerol-3-phosphate, inositol phosphate, and phosphoric acid (Figure 3). Eight metabolites (including two fatty acids [linoleic (C18:1) and myristic (C14:0) acids], five dicarboxylic acids, and the amino acid guanine) were significantly different in abundance between the host and symbiont but not when compared to the holobiont fraction (Figure 3).

Supplementary Table S1: The relative abundance of each metabolite identified using GC-MS analysis in the holobiont and separated host and symbiont. The values are mean ± standard error, and the ANOVA results provided, including the post hoc analysis, are indicated by the colored cells (columns K, L, and M). Please click here to download this File.

Supplementary Table S2: Qualitative and quantitative ion pairs for the identified metabolites. Please click here to download this File.

Supplementary Table S3: PERMANOVA results. Please click here to download this File.

Subscription Required. Please recommend JoVE to your librarian.

Discussion

The separation of the host and symbiont is easily and rapidly achievable via simple centrifugation, and the results here show that separating the fractions can provide valuable information indicative of specific holobiont member contributions, which can contribute toward the functional analysis of coral health. In adult corals, lipid synthesis is primarily performed by the resident algal symbiont40, which supplies lipids (e.g., triacylglycerol and phospholipids)41 and fatty acids that can promote stress recovery11,42. Of the 13 metabolites that were more abundant in the symbiont versus host profiles in this work, 9 were fatty acids, including the biologically relevant eicosanoids DHA (C22:6[ω-6]), EPA (C20:5[ω-6]), and ARA (C20:4[ω-6]), which are implicated in inter-kingdom stress signaling in the coral-Symbiodiniaceae symbiosis9,10. Inositol phosphate plays a crucial role in diverse cellular functions, including cell growth, apoptosis, endocytosis, and cell differentiation. The relative abundances of inositol isoforms and derivatives have frequently been found to change during symbiosis dysfunction7,11,27,28, although the role of these isoforms and their derivatives in the coral-Symbiodiniaceae symbiosis remains unclear. Thus, the clear difference in relative abundance between host and symbiont profiles helps to contribute to this knowledge.

Many studies have used holobiont metabolite profiles to investigate coral metabolic interactions and performance under ambient and stress conditions13,43,44,45. Here, we show that the analysis of the holobiont homogenate can result in large within-treatment variability, and the resulting profiles can mask host- and/or symbiont-specific abundance patterns and metabolic contributions to the overall holobiont metabolome. For example, of the 13 metabolites that were significantly more abundant in the symbiont relative to the host fraction, only 4 were significantly more abundant in the symbiont relative to the holobiont. These included the potentially biologically relevant DHA and inositol phosphate7,9,10,11,27,28; however the distinct differences observed in the stress-signaling eicosanoids EPA and ARA between the host and symbiont were not significant in the holobiont versus symbiont samples. These metabolites are increasingly being recognized as important metabolites in the molecular language of symbiosis and as indicators of stress responses10, but examining holobiont profiles alone would not reveal the distinct changes in the relative abundance ratios of these metabolites in specific holobiont members. Thus, analyzing the holobiont can mask differences that are otherwise apparent when the host and symbiont are analyzed separately, and these differences might not be detected as significant in studies that compare abiotic or biotic treatments solely using holobiont profiles. This could make it challenging to infer specific partner contributions or functions7,8,9,11,12,22,28, particularly when attempting to answer biological questions for which symbiotic interactions underlie the observed holobiont phenotype46,47.

An alternative methodology would be to separate the host tissues and symbiont cells, take an aliquot of the homogenate or an extract of each fraction to recombine prior to extraction, and analyze this alongside the separated fractions; this method would ensure the release of symbionts from the host tissues but would increase the potential for human error and/or metabolite loss48. Analyzing the fractions separately and integrating the data post analysis may be possible, but the calculations would need to take into account the proportion of symbiont and host cells in the original coral holobiont.

While more information can be obtained by separating the host and symbiont fractions for metabolite analysis, especially in terms of individual member contributions to holobiont metabolism and function, whether to separate the host and symbiont rather than analyze the holobiont is ultimately a decision governed by the research question. For instance, analyzing the metabolite profile of the holobiont is relevant when other physiological measurements are taken with the holobiont (e.g., analysis of volatile metabolite emissions from coral colonies49,50,51) and when metabolomics profiles need to be integrated with holobiont datasets. In addition, the separation of the host and symbiont is not without limitations; for example, additional manipulation steps might interfere with metabolite stability and result in data loss or confounding effects48.

Furthermore, there are additional optimization steps involved in the host-symbiont separation procedure: 1) optimizing the number of washes for the specific coral species; and 2) identifying the optimal biomass to extract from both fractions for GC-MS analysis. Both steps need to be included before the final sample processing can commence, thus increasing both the consumable and analytical requirements for material, time, and costs. In this work, metabolite loss from the host and symbiont extractions due to the additional steps may have contributed to the higher relative abundances of some metabolites observed in the holobiont relative to either the host or symbiont, such as glycolic acid, 1-hexadecanol, and dodecanoic acid. However, the large within-group variation in the holobiont group for these metabolites is, as mentioned, an alternative reason for these observed patterns.

The application of metabolomics approaches, while relatively new, has had a profound impact on our capacity to elucidate the function of specific symbioses. For example, this approach has revealed the contribution of the phytoplankton growth-promoting hormone indole-3 acetic acid, which is synthesized by bacteria in the Pseudo-nitzschia multiseries-Sulfitobacter symbiosis. Moreover, this approach has elucidated host-derived and symbiont-derived translocation in corals11,27,28,29, which has exciting potential for coral reef conservation and restoration3, such as through biomarker detection19. Here, we have provided a procedure for both approaches with the hope that this will facilitate and accelerate the application of metabolomics for future coral reef investigations.

Subscription Required. Please recommend JoVE to your librarian.

Disclosures

The authors have no conflict of interests to disclose.

Acknowledgments

J.L.M. was supported by a UTS Chancellor's Research Fellowship.

Materials

Name Company Catalog Number Comments
100% LC-grade methanol Merck 439193 LC grade essential
2 mL microcentrifuge tubes, PP Eppendorf 30121880 Polypropylene provides high resistance to chemicals, mechanical stress and temperature extremes
2030 Shimadzu gas chromatograph Shimadzu GC-2030
710-1180 µm acid-washed glass beads Merck
G1152		 This size is optimal for breaking the Symbiodiniaceae cells
AOC-6000 Plus Multifunctional autosampler Shimadzu AOC6000
Bradford reagent Merck B6916 Any protein colourimetric reagent is acceptable
Compressed air gun Ozito 6270636 Similar design acceptable. Having a fitting to fit a 1 mL tip over is critical.
 DB-5 column with 0.25 mm internal diameter column and 1 µm film thickness Agilent 122-5013
DMF Merck RTC000098
D-Sorbitol-6-13C and/or 13C5–15N Valine Merck 605514/ 600148 Either or both internal standards can be added to the methanol.
Flat bottom 96-well plate Merck CLS3614
Glass scintillation vials Merck V7130 20 mL, with non-plastic seal
Immunoglogin G Merck 56834 if not availbe, Bovine Serum Albumin is acceptable
Primer v4
R v4.1.2
Shimadzu LabSolutions Insight software v3.6
Sodium Hydroxide Merck S5881 Pellets to make 1 M solution
tidyverse v1.3.1 R package
TissueLyser LT Qiagen 85600 Or similar
TQ8050NX triple quadrupole mass spectrometer Shimadzu GCMS-TQ8050 NX
UV-96 well plate Greiner M3812
Whirl-Pak sample bag Merck WPB01018WA Sample collection bag; Size: big enough to house a ~5 cm coral fragment, but not too big that the water is too spread

DOWNLOAD MATERIALS LIST

References

  1. Bundy, J. G., Davey, M. P., Viant, M. R. Environmental metabolomics: A critical review and future perspectives. Metabolomics. 5 (1), 3-21 (2008).
  2. Matthews, J. L., et al. The metabolic significance of symbiont community composition in the coral-algal symbiosis. Applied Environmental Metabolomics. Beale, D. J., Hillyer, K. E., Warden, A. C., Jones, O. A. H. , Academic Press. Cambridge, MA. 211-229 (2022).
  3. Lawson, C. A., et al. Informing coral reef conservation through metabolomic approaches. Coral Reef Conservation and Restoration in the Omics Age. Coral Reefs of the World. van Oppen, M. J. H., Aranda Lastra, M. , Springer. Cham, Switzerland. 179-202 (2022).
  4. LaJeunesse, T. C., et al. Systematic revision of Symbiodiniaceae highlights the antiquity and diversity of coral endosymbionts. Current Biology. 28 (16), 2570-2580 (2018).
  5. Rohwer, F., Seguritan, V., Azam, F., Knowlton, N. Diversity and distribution of coral-associated bacteria. Marine Ecology Progress Series. 243, 1-10 (2002).
  6. Maire, J., et al. Intracellular bacteria are common and taxonomically diverse in cultured and in hospite algal endosymbionts of coral reefs. The ISME Journal. 15 (7), 2028-2042 (2021).
  7. Hillyer, K. E., et al. Metabolite profiling of symbiont and host during thermal stress and bleaching in the coral Acropora aspera. Coral Reefs. 36, 105-118 (2016).
  8. Hillyer, K. E., Tumanov, S., Villas-Bôas, S., Davy, S. K. Metabolite profiling of symbiont and host during thermal stress and bleaching in a model cnidarian-dinoflagellate symbiosis. Journal of Experimental Biology. 219 (4), 516-527 (2016).
  9. Matthews, J. L., et al. Optimal nutrient exchange and immune responses operate in partner specificity in the cnidarian-dinoflagellate symbiosis. Proceedings of the National Academy of Sciences of the United States of America. 114 (50), 13194-13199 (2017).
  10. Rosset, S. L., et al. The molecular language of the cnidarian-dinoflagellate symbiosis. Trends in Microbiology. 29 (4), 320-333 (2020).
  11. Matthews, J. L., et al. Partner switching and metabolic flux in a model cnidarian-dinoflagellate symbiosis. Royal Society. 285 (1892), 20182336 (2018).
  12. González-Pech, R. A., et al. Physiological factors facilitating the persistence of Pocillopora aliciae and Plesiastrea versipora in temperate reefs of south-eastern Australia under ocean warming. Coral Reefs. 41, 1239-1253 (2022).
  13. Williams, A., et al. Metabolomic shifts associated with heat stress in coral holobionts. Science Advances. 7 (1), (2021).
  14. Deutsch, J. M., et al. Metabolomics of healthy and stony coral tissue loss disease affected Montastraea cavernosa corals. Frontiers in Marine Science. 8, 1421 (2021).
  15. Stien, D., et al. A unique approach to monitor stress in coral exposed to emerging pollutants. Scientific Reports. 10 (1), 9601 (2020).
  16. Lohr, K. E., et al. Resolving coral photoacclimation dynamics through coupled photophysiological and metabolomic profiling. Journal of Experimental Biology. 222 (8), (2019).
  17. Jorissen, H., et al. Coral larval settlement preferences linked to crustose coralline algae with distinct chemical and microbial signatures. Scientific Reports. 11 (1), 14610 (2021).
  18. Roach, T. N., Dilworth, J., Jones, A. D., Quinn, R. A., Drury, C. Metabolomic signatures of coral bleaching history. Nature Ecology & Evolution. 5 (4), 495-503 (2021).
  19. Parkinson, J. E., et al. Molecular tools for coral reef restoration: Beyond biomarker discovery. Conservation Letters. 13 (1), 12687 (2020).
  20. Jiang, J., et al. How Symbiodiniaceae meets the challenges of life during coral bleaching. Coral Reefs. 40, 1339-1353 (2021).
  21. Guerra, F. D., Attia, M. F., Whitehead, D. C., Alexis, F. Nanotechnology for environmental remediation: materials and applications. Molecules. 23 (7), 1760 (2018).
  22. Matthews, J. L., et al. Metabolite pools of the reef building coral Montipora capitata are unaffected by Symbiodiniaceae community composition. Coral Reefs. 39, 1727-1737 (2020).
  23. Papina, M., Meziane, T., van Woesik, R. Symbiotic zooxanthellae provide the host-coral Montipora digitata with polyunsaturated fatty acids. Comparative Biochemistry and Physiology Part B: Biochemistry and Molecular Biology. 135 (3), 533-537 (2003).
  24. Kellogg, R., Patton, J. Lipid droplets, medium of energy exchange in the symbiotic anemone Condylactis gigantea: A model coral polyp. Marine Biology. 75, 137-149 (1983).
  25. Ankrah, N. Y., Chouaia, B., Douglas, A. E. The cost of metabolic interactions in symbioses between insects and bacteria with reduced genomes. mBio. 9 (5), e01433 (2018).
  26. Kabeya, N., et al. Genes for de novo biosynthesis of omega-3 polyunsaturated fatty acids are widespread in animals. Science Advances. 4 (5), (2018).
  27. Hillyer, K. E., Dias, D., Lutz, A., Roessner, U., Davy, S. K. 13C metabolomics reveals widespread change in carbon fate during coral bleaching. Metabolomics. 14 (1), 12 (2018).
  28. Hillyer, K. E., Dias, D. A., Lutz, A., Roessner, U., Davy, S. K. Mapping carbon fate during bleaching in a model cnidarian symbiosis: the application of 13C metabolomics. New Phytologist. 214 (4), 1551-1562 (2017).
  29. Burriesci, M. S., Raab, T. K., Pringle, J. R. Evidence that glucose is the major transferred metabolite in dinoflagellate-cnidarian symbiosis. Journal of Experimental Biology. 215 (19), 3467-3477 (2012).
  30. Thurber, R. V., et al. Unified methods in collecting, preserving, and archiving coral bleaching and restoration specimens to increase sample utility and interdisciplinary collaboration. PeerJ. 10, 14176 (2022).
  31. Grottoli, A. G., et al. Increasing comparability among coral bleaching experiments. Ecological Applications. 31 (4), 02262 (2020).
  32. Mushtaq, M. Y., Choi, Y. H., Verpoorte, R., Wilson, E. G. Extraction for metabolomics: access to the metabolome. Phytochemical Analysis. 25 (4), 291-306 (2014).
  33. Bradford, M. M. A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Analytical Biochemistry. 72 (1), 248-254 (1976).
  34. Peterson, G. L., et al. A simplification of the protein assay method of Lowry et al. which is more generally applicable. Analytical Biochemistry. 83 (2), 346-356 (1977).
  35. Lowry, O. H., Rosebrough, N. J., Farr, A. L., Randall, R. J. Protein measurement with the Folin phenol reagent. Journal of Biological Chemistry. 193 (1), 265-275 (1951).
  36. Zamer, W. E., Shick, J. M., Tapley, D. W. Protein measurement and energetic considerations: Comparisons of biochemical and stoichiometric methods using bovine serum albumin and protein isolated from sea anemones. Limnology and Oceanography. 34 (1), 256-263 (1989).
  37. Smart, K. F., Aggio, R. B., Van Houtte, J. R., Villas-Boas, S. G. Analytical platform for metabolome analysis of microbial cells using methyl chloroformate derivatization followed by gas chromatography-mass spectrometry. Nature Protocols. 5 (10), 1709-1729 (2010).
  38. Pang, Z., et al. Using MetaboAnalyst 5.0 for LC-HRMS spectra processing, multi-omics integration and covariate adjustment of global metabolomics data. Nature Protocols. 17 (8), 1735-1761 (2022).
  39. Tibshirani, R., Walther, G., Hastie, T. Estimating the number of clusters in a data set via the gap statistic. Journal of the Royal Statistical Society: Series B (Statistical Methodology). 63 (2), 411-423 (2001).
  40. Chen, W. -N., et al. Diel rhythmicity of lipid-body formation in a coral-Symbiodinium endosymbiosis). Coral Reefs. 31 (2), 521-534 (2012).
  41. Imbs, A. Fatty acids and other lipids of corals: composition, distribution, and biosynthesis. Russian Journal of Marine Biology. 39 (3), 153-168 (2013).
  42. Rosset, S., et al. Lipidome analysis of Symbiodiniaceae reveals possible mechanisms of heat stress tolerance in reef coral symbionts. Coral Reefs. 38 (6), 1241-1253 (2019).
  43. Carreón-Palau, L., Parrish, C. C., Del Angel-Rodriguez, J. A., Perez-Espana, H. Seasonal shifts in fatty acids and sterols in sponges, corals, and bivalves, in a southern Gulf of Mexico coral reef under river influence. Coral Reefs. 40 (2), 571-593 (2021).
  44. Imbs, A. B., Dang, L. T. Seasonal dynamics of fatty acid biomarkers in the soft coral Sinularia flexibilis, a common species of Indo-Pacific coral reefs. Biochemical Systematics and Ecology. 96, 104246 (2021).
  45. Oku, H., Yamashiro, H., Onaga, K., Sakai, K., Iwasaki, H. Seasonal changes in the content and composition of lipids in the coral Goniastrea aspera. Coral Reefs. 22 (1), 83-85 (2003).
  46. Weis, V. M. Cell biology of coral symbiosis: foundational study can inform solutions to the coral reef crisis. Integrative and Comparative Biology. 59 (4), 845-855 (2019).
  47. Oakley, C., Davy, S. Cell biology of coral bleaching. Coral Bleaching. van Oppen, M., Lough, J. , Springer. Cham, Switzerland. 189-211 (2018).
  48. Lu, W., et al. Metabolite measurement: Pitfalls to avoid and practices to follow. Annual Review of Biochemistry. 86, 277-304 (2017).
  49. Lawson, C. A., et al. Heat stress decreases the diversity, abundance and functional potential of coral gas emissions. Global Change Biology. 27 (4), 879-891 (2021).
  50. Olander, A., et al. Comparative volatilomics of coral endosymbionts from one-and comprehensive two-dimensional gas chromatography approaches. Marine Biology. 168 (5), 76 (2021).
  51. Wuerz, M., et al. Symbiosis induces unique volatile profiles in the model cnidarian Aiptasia. Journal of Experimental Biology. 225 (19), (2022).

Tags

Gas Chromatography-mass Spectrometry Targeted Metabolomics Hard Coral Samples Metabolic Basis Cnidarian-dinoflagellate Symbiosis Temperature-induced Bleaching Steady-state Metabolite Profiling Coral Holobiont Symbiotic Interactions Metabolite Profiles Physical Separation Tissue Isolation Metabolite Extraction GC-MS Analysis Optimization Steps
Gas Chromatography-Mass Spectrometry-Based Targeted Metabolomics of Hard Coral Samples
Play Video
PDF DOI DOWNLOAD MATERIALS LIST

Cite this Article

Matthews, J. L., Bartels, N., Elahee … More

Matthews, J. L., Bartels, N., Elahee Doomun, S. N., Davy, S. K., De Souza, D. P. Gas Chromatography-Mass Spectrometry-Based Targeted Metabolomics of Hard Coral Samples. J. Vis. Exp. (200), e65628, doi:10.3791/65628 (2023).

Less
Copy Citation Download Citation Reprints and Permissions
View Video

Get cutting-edge science videos from JoVE sent straight to your inbox every month.

Waiting X
Simple Hit Counter