Distinguishing the molecular diversity, nutrient content, and energetic potential of exometabolomes produced by macroalgae and reef-building corals

Significance Marine dissolved organic matter (DOM) is one of the most complex and abundant chemical mixtures on earth, comprising thousands of different molecules. The molecular structure of these compounds is one factor structuring the community of microorganisms that metabolize them; in turn, this microbial metabolism mediates the composition of DOM. Decades of coral reef research has established the fundamental importance of microbial biogeochemistry in ecosystem function. This study unveils coral reef DOM by identifying a myriad of specific metabolite classes released into the surrounding waters by reef-building corals and algae, further characterizing their energetic and nutrient content and providing a foundation for linking benthic ecology with microbial processes that influence both the livelihood and demise of coral reefs.

. Effects of seawater control incubations on all biogeochemical parameters. Parameters include DO, pH, DOC, DON, DOP, N+N (nitrate plus nitrite), NH4, PO4 and SiOX; all units in μM except DO as noted. We used 2-way ANOVA to evaluate differences between starting (Ambient) and ending seawater controls (Water) in the day and night. No interaction terms were significant (p > 0.05) and only oxygen differed significantly between day and night samples (p diel = 0.0043) with higher daytime means (8.06) than nighttime means (7.93) in the ambient starting water. Statistics for t-tests are shown above each comparison. Oxygen increased in the daytime incubation control. Both DOC and SiOx increased slightly in the seawater controls over the incubation periods of both daytime and nighttime samples; in the nighttime samples pH and N+N increased while Ammonium decreased. There was no change in PO4, DOP or DON in controls.   Figure S2: Covariation of DOC produced with oxygen, pH and bulk DOM compositional change across daytime and nighttime treatments. Model statistics are noted at the top of each panel and flagged as significant (*) or nonsignificant (n.s.) at alpha of 0.05. Note that DOC production tracks oxygen evolution (a), pH increase (b) and production of both humic fDOM (c) and exudate metabolite features (d) in daytime incubations, but not at night (f, g, h) except nighttime exudate feature production is related to DOC (i). Ambient feature production does not correlate with DOC release in day or night incubations (e, j) Figure S3. Effects of benthic producers on macronutrient parameters. All concentrations are in micromolar; N+N refers to nitrate+nitrate. Asterisks designate significant enrichment relative to the seawater controls using ANOVA with Dunnet's post hoc tests. Ambient lines represent the mean starting seawater concentrations measured at the start of the experiments before incubation.  Figure S4: Fluorescent DOM characteristics of daytime exudates. Principal Components Analysis (a) shows each water sample as a point from which 6 fDOM components were derived, shown as biplot arrows overlaid on a graph of the first two PCA components. Note that, consistent with Quinlan et al. 2018, algae are enriched in humic-like components (b; comprising Ultraviolet, Marine and Visible humic-like components shown as 3 adjacent bars, respectively), corals are enriched in proteinaceous components (c,d,e; comprising Tryptophan, Tyrosine and Phenylalanine-like components, the latter nonsignificant). HIX stands for the widely used fDOM humification index. Statistically significant parameters in 1-way ANOVA tests are shown (b-f) with means and standard error whiskers at right as well as asterisks denoting treatments significantly enriched relative to the control with Dunnet's post hoc tests.  Figure S6. Mean proportion of spectral peak area (XIC, MS1) allocated to feature types in different treatment categories. Ambient features are those statistically enriched in Ambient samples and never exceeding twice the reef water values as defined in the methods. Exudate features are twice the reef water values in at least one incubation treatment as defined in the methods; those enriched in the Control are considered planktonic and those enriched only in benthic organismal treatments are considered benthic. Exometabolite features are statistically significantly enriched in one of the five benthic organismal treatments relative to the Control as defined in the methods. Due to the stringent pairwise false discovery rate controls applied to the statistical differentiation of exometabolite features (1,667 released consistently by a benthic producer) from exudate features (8,936 enriched above the ambient reef water) the vast majority of the latter (81%, or 7,269) were classified as those enriched in any of the five benthic macroorganisms (4,846 benthic exudate features twice the mean peak area of ambient seawater in one or more organismal incubation treatments) as well as those potentially produced by the remaining planktonic community after 0.2 µm pre-filtration (2,047 planktonic exudate features twice the mean peak area of ambient seawater in either Day or Night Controls) or likely artifacts of incubation (376 features that increased in both Day and Night Controls over Benthic_Exud Figure S7. Clustering of exometabolite features. Relative abundances of exometabolite features (n=1667) were internally standardized using z-scores (heat map) and hierarchically clustered using Ward's Minimum Variance method (a). Five feature clusters were resolved (dendrogram colors at left and 2 letter codes at right: Porites lobata = PL, CCA = CC, Turf = TR, Dictyota = DT, Pocillopora verrucosa = PV), each enriched most prominently in one of the five benthic producer incubation treatments (top), with replicate treatments clustering together (bottom, also Figure 2). Note that within each exometabolite type features are enriched either primarily in the daytime (upper subcluster) or both day and night (lower subcluster). Panel (b) shows the raw summed MS1 XIC peak area attributed to each exometabolite feature cluster (2 letter codes from a) in each experiment day and night, emphasizing the specificity of these metabolite clusters both in quality (a) and in quantity (b).
-3.6 -2.9 -2.  Figure S8. Pairwise Bray-Curtis dissimilarities in exometabolite relative abundance among samples from different treatments. The categories across the bottom group the pairwise dissimilarities (points) according to the type of comparison, and means (grey bars) with the same letter are not significantly different at alpha = 0.05 (ANOVA with Tukey post hoc test). The starting (Ambient) and ending (Control) water exhibited the smallest variation in exometabolome composition (mean bray-curtis dissimilarity 0.34); the mean diel difference within water treatments (0.44) or within benthic producer treatments (0.53) was significantly less than the differences between exometabolomes during the day (0.76) or the differences between exometabolomes and water controls during the day (  c. d.      Figure S14. Comparison of the effect of standardized vs. raw peak area summation on calculation of weighted mean NOSC. The blue values match those presented in Figure 3 while the red first convert all peak areas to standard scores (using the z-scoring approach) to eliminate the potential bias of differential ionization of exometabolite features.  Organismal collection and handling. All specimens were collected in triplicate from water depths of 2.0-3.5 m at backreef locations (17°28'55"S, 149°50'43"W) five days prior to the experiment (22-23 September) in the backreef habitat between the north facing Paopao and Ōpūnohu bays and transferred in coolers without air exposure to flowing seawater cultivation tanks with ambient temperature and light within 1 hour of collection. Specimen surface areas did not differ significantly among treatments and averaged 124 cm 2 . Exudate generation was conducted 27-28 September. Incubations were conducted separately over a daytime period (0900 -1700) and, after a water exchange, over a subsequent nighttime period (2300 -0700) to assess differences in daytime and nighttime organic matter release. Each specimen was placed into one 1.5 L polycarbonate incubation container (acid washed, leached for 24h in ambient flowing seawater, then rinsed with acid and deionized water) and incubated with a fitted lid and no agitation or bubbling filled with freshly collected "sterile" reef water (filtered through preflushed 142 mm polyethersulfone 0.2 µm pore size filters). The incubation containers were randomly placed on a shaded flowing water table to standardize light and temperature (temperature ranged from 25.9 to 26.8 degree C, light distributions are shown in Figure S15). To assess initial water parameters, triplicate samples of 1.2 L filtered reef water (seawater Control T0; technical replicates) were immediately processed and at the end of each incubation period, 1.2 L water was collected from each container (triplicate biological replicates) and processed for a suite of measurements of water chemistry. Extracts were dried in a Centrivap, re-dissolved in 80% MeOH:H20 with 1% formic acid and transferred into 300ul glass inserts. Extraction efficiencies were measured by evaporating solvent from a known fraction of the sample, re-dissolving in MilliQ water and analyzing for DOC on a Shimadzu TOC-V as described above, then comparing derived concentrations eluted from the PPL resin with those in the initial samples. Extraction efficiencies were then calculated by comparing the values to the DOC concentrations of the initial sample.

Liquid Chromatography Tandem Mass Spectrometry (LC-MS/MS
). An aliquot (5 μL) of each sample was injected into a Vanquish ultra-high performance liquid chromatography system (UHPLC) coupled to a Q-Exactive Orbitrap Mass Spectrometer (Thermo Fisher Scientific, Bremen, Germany). Chromatographic separation was performed with a C18 core-shell column (Kinetex, 150 × 2 mm, 1.8 µm particle size, 100 Å pore size, Phenomenex, Torrance, USA) with a flowrate of 0.5 mL min -1 (Solvent A: H2O + 0.1% formic acid (FA), Solvent B: Acetonitrile + 0.1% FA). After injection, the samples were eluted for ten minutes, including 30 seconds of 5% Solvent B, across a linear gradient over 7.5 minutes of 5 to 50% Solvent B and 2 minutes 50 to 99% Solvent B, followed by a 2 min washout phase at 99% B and a 3 min re-equilibration phase at 5% B; ion data was collected through the washout phase. Carry over was controlled by MeOH blank injections between every 5 samples and mass and retention time accuracy and drift were monitored with a quality control mix consisting of 6 standard compounds (Sulfamethazine, Sulfamethizole, Sulfachloropyridazine, Sulfadimethoxine, Amitryptiline, Coumarin-314).
Retention time and mass accuracy shifts were negligible over the course of this continuous multiday run, manually monitored by running the quality control mix at start, midway and end and recalibrating the mass spectrometer. Mass spectra (MS1) were acquired in positive electrospray ionization (ESI+) mode and parameters were set to 52 AU sheath gas flow, 14 AU auxiliary gas flow, 0 AU sweep gas flow and 400•C auxiliary gas temperature. The spray voltage was set to 3.5 kV and the inlet capillary temperature was set to 320• C. The stacked-ring ion guide (S-Lens) was set to 50V. The maximum ion injection time was set to 100 ms with automated gain control (AGC) targets set to 1.0E6 for survey scans and 3.0E5 for MS/MS with minimum threshold of 10% C-trap AGC. Scan range for MS1 was set to 150-1,500 m/z with a m/z 200 resolution (Rm/z

Calculations of Nominal Oxidation State of Carbon and Gibbs energies of oxidation half
reactions. The NOSC in an organic compound is first derived from the following oxidation half reaction: Statistical Analyses. Before subsequent statistical analyses, all exudate features were relativized by dividing peak area intensity for each exudate feature by the total exudate peak area intensity within each sample to account for minor variation in DOM loading and retention. For linear modeling statistical analyses, relative abundances were angular transformed (asin(sqrt(X)) and    Table S1. List of Subnetworks enriched in exometabolomes. The relative abundance of all Exudate nodes in each subnetwork containing exometabolite features were summed and statistically compared among the six endpoint treatments followed by a post hoc Dunnet's test for 2X enrichment over the control. Network mean relative abundances listed in Bold Underline are significantly enriched. Enrichment Category denotes which treatments the network was enriched in.The ClassyFire Class derived from MolNetEnhancer is listed for each network.