Measurements of Respiration and Photosynthesis as In-vivo Oxygen Flux:

Respiration and photosynthesis of individual anemones [oral disk diameter >4mm for natural E. pallida (n = 53), 4–6mm for re-infected animals (n = 19 per Symbiodinium species)] was measured in sealed glass scintillation vials fitted with an internal stir bar and an oxygen sensitive optode (Fibox 4, PreSens Gmbh, Regensburg, Germany). Vials were immersed in a constant 26◦C water bath and maintained in darkness for 15min. Illumination was provided for 20 min at an irradiance of 200 μmol photons m−2 s−1 after which, the LEDs were turned off for a final 30-min period to allow measurement of steady-state dark respiration. Background O2 flux was determined using vials containing 1-μm-filtered, UV-sterilized seawater, and was found to be negligible. After respirometry assays, each anemone was snap-frozen in liquid nitrogen, and stored at −80◦C. Rates of holobiont dark respiration were calculated as moles O2 consumed hour−1. Dark respiration rates were then subtracted from net photosynthetic rates (moles O2 produced hour−1 during the light-phase) to generate values for gross photosynthesis. Respiration- and gross photosynthesis rates were normalized to soluble animal protein and symbiont cell number, respectively.

Anemone Processing and Symbiodinium Density Analysis:

Anemones were thawed in their 2-mL screw cap vials and 0.6–1mL ice-cold lysis buffer (25mM Tris, pH 7.8, 1mM EDTA, 10% glycerol [v/v]) was added to each vial. Anemones were then homogenized in a chilled bead-beater (FastPrep -24, MP Bio, Santa Ana, CA, USA) for 60s at a speed of 6m s−1 with a 5-mm-diameter stainless steel bead. The homogenate was inspected visually (100× magnification) to confirm anemone tissue disruption and Symbiodinium cell integrity. A 100-μL sample was removed, fixed with 5μL glutaraldehyde (8% [w/v] stock solution in water), and stored at 4◦C in the dark for later quantification of Symbiodinium cell densities. The remaining homogenate was centrifuged at 3000 × g for 30 s. Two hundred microliters of the supernatant were removed for animal DNA extraction (see below) and the pellet was re-suspended. The sample was then centrifuged for 5min at 700 × g to separate the Symbiodinium cells from the remaining anemone material. The Symbiodinium pellet was immediately snap-frozen in liquid nitrogen and transferred to a −80◦C freezer. The supernatant (“anemone fraction”) was centrifuged at high speed (16,100 × g, 20min) to remove particulates, and aliquots of the clear supernatant were snap- frozen and stored at −80◦C.

In order to remove residual animal protein, Symbiodinium pellets were thawed, re-suspended in 1 mL ice-cold wash buffer (as lysis buffer above, but with the addition of 0.01% [v/v] Triton X-100), and centrifuged for 5 min at 700 × g. The supernatant was discarded and the pellet re-suspended in fresh ice-cold wash buffer. This procedure was repeated four times, after which the supernatant was clear and the pellet dark brown, with little evidence of contaminating anemone material. Symbiodinium cells were finally re-suspended in 300μL ice-cold lysis buffer in a clean 1.5-mL tube containing a 200-μL-volume of 0.5- mm-diameter acid-washed glass beads. Cells were lysed in a chilled bead-beater (see above) for 3 min at a speed of 6.5 m s−1. The lysate was inspected visually (100× magnification), before centrifugation to remove particulates (16,100 × g, 20 min). Aliquots of the supernatant were then snap-frozen in liquid nitrogen and stored at −80◦C.

Total soluble anemone and Symbiodinium protein were determined using a linearized Bradford assay (Ernst and Zor, 2010). To test the effectiveness of the algal washing steps described above, a standard curve was constructed by spiking 12 Symbiodinium pellets from similar-sized anemones (in triplicate) with 24–1200 μg anemone protein “contamination” originating from crude homogenates that had been gently centrifuged (500 × g for 5 min) to remove algal cells. Algal pellets were then washed and a 50-μL sample was removed and fixed for cell counts. The remaining cells were lysed as described above. Symbiodinium protein content was then measured for each pellet, normalized to cell number (see below), and plotted against the respective amount of anemone material added.

Symbiodinium densities were quantified using an Improved Neubauer hemocytometer and a fluorescence microscope to visualize cells’ chlorophyll a fluorescence (see above). Field of view was determined using the EVOS operating software (4× objective), and cells were counted using the “Analyze Particles” tool in ImageJ (NIH, Bethesda, MD, USA). At least 6 independent images were analyzed for each sample, and cell numbers were normalized to anemone protein content.

Biochemical Analysis of Mitochondrial Function:

Anemone and Symbiodinium aerobic capacity was quantified as the activity of the TCA cycle rate-limiting enzyme citrate synthase (CS), measured according the methods of Srere (1969) modified for use with small marine invertebrates. All samples were analyzed within 1 month of freezing. Representative results of assay optimization and validation procedures are provided in ESM Figure S1. Briefly, 20μL of thawed anemone or Symbiodinium supernatant (diluted to yield 2–6μg of protein) was added in triplicate to wells in a 96-well plate (Greiner Bio- One, Monroe, NC, USA) alongside triplicate blanks (20 μL lysis buffer) and positive controls [20 μL citrate synthase (1 U mL−1 in lysis buffer; Sigma-Aldrich, St. Louis, MO, USA)]. One hundred and seventy microliters of assay buffer (111mM Tris, pH 7.8, 0.11% [v/v] Triton X-100) containing 294μM 5,5′-dithiobis- (2-nitrobenzoic acid) (DTNB; Sigma-Aldrich, see above) and 588 μM acetyl-coenzyme A (Sigma-Aldrich, see above) were then added to all wells. DTNB stock solutions (5 mM) were made fresh in 0.1 mM Tris buffer, pH 8.0. Acetyl-coenzyme A solutions were prepared at a concentration of 12.35 mM in distilled water, stored in aliquots at −80◦C, and used within 6 months.

To control for non-CS-specific reaction products following the addition of assay buffer, baseline absorbance (λ = 412 nm) was recorded for 3min using a microplate reader (Fluostar Omega, BMG, Cary, NC, USA) maintained at a temperature of 26◦ C. The CS-catalyzed reaction was initiated by adding 10 μL oxaloacetate to each well (OA; Sigma-Aldrich, see above; 12 mM stock solution prepared fresh in distilled water and stored on ice), giving a final concentration of 600 μM OA. Sample absorbance was monitored at 412 nm for a further 3 min, and CS enzyme activity was derived using the following equation:

CS Specific Activity U mg−1

= △412OA − △412blank × Vreaction × D / 13.6 ×L × Vsample ×P

Where 412 is the linear rate of change in 412-nm absorbance prior to and after the addition of OA, V is the volume (mL), D is the sample dilution factor, 13.6 (mM−1 cm−1) is the 412-nm extinction coefficient for the reaction product, L is the optical path-length (cm) and P is the sample protein concentration (mg mL−1). Total Symbiodinium and animal CS activities were calculated as the product of the respective specific activity and total protein contents.

Quantification of Anemone Mitochondrial Copy Number:

Anemone genomic DNA was extracted using two independent methods. For some animals (n = 33), DNA was extracted using a Wizard Genomic DNA Extraction kit (Promega Corporation, Madison, WI, USA). The DNA pellet was washed with ethanol, air-dried under sterile conditions at 30◦ C, and finally dissolved in 50 μL nuclease- free water (BioExpress, Kaysville, UT, USA). DNA from a second group (n = 20) was extracted and purified using a QiaAmp DNA Mini kit (Qiagen, Germantown, MD, USA). DNA concentration was measured using a Quant-iT PicoGreen assay (ThermoFisher [Invitrogen], Waltham, MA, USA) and purity was determined spectrophotometrically as the 260/230 nm and 260/280 nm absorbance ratios (NanoDrop , ThermoFisher, Waltham, MA, USA). The 260/280nm ratio was consistently >1.8, but the 260/230 nm ratio was variable, particularly in samples purified using the Wizard Promega kit (range = 1.4–2.2), however, this had a negligible effect on the efficiency of subsequent QPCR reactions.

Primer Design and Standard Curve Construction:

Oligonucleotide primers are provided in the referenced publication (Hawkins et al. 2016). Primer validation was undertaken using end-point PCR, with each reaction containing 20 ng anemone DNA in a 20-μL mix of 1× Standard Mg-free PCR buffer, 0.25 U Taq DNA polymerase, 0.25 μM (CO1 and ATP6) or 0.5 μM primers (EF- 1-α), 2.5 mM MgCl2, and 0.25 mM dNTPs. Cycling conditions were 94◦C for 2 min, followed by thirty cycles of 94◦C for 15 s, 60◦C for 30 s, and 72◦ C for 30 s, with a final elongation at 72◦ C for 10 min. Agarose gel electrophoresis confirmed a single PCR product for each primer set and, after product purification, amplicons were sequenced in both directions using the respective PCR primers. Electropherograms were inspected visually to confirm the reliability of base-calling and sequences were compared to NCBI Genbank or Aiptasia genome databases. In all cases the sequences aligned most strongly with those used for primer design. Longer sequences of each of the three target genes was amplified and subsequently cloned in order to create material for QPCR standard curves.

QPCR Analysis:

Duplicate 2-μL aliquots of extracted DNA (5–10 ng μL−1) were added to a 18-μL reaction mix (SensiMixTM SYBR Hi-ROX; Bioline, Taunton, MA, USA) such that the final mix contained 0.25 μM CO1 or ATP6 primers, or 0.5 μM EF-1-α primer. Gene- fragments were amplified using an AB-7500 real-time QPCR system [ThermoFisher (Applied Biosystems), Waltham, MA, USA], with the following cycling conditions: 94◦ C for 10 min, followed by 40 cycles of 94◦C for 15 s, 60◦C for 1 min, and 72◦C for 15 s. A melt-curve analysis (60–94◦C in 0.3◦C increments, 30s per step) was carried out in order to detect non-specific amplification products. A single PCR product was detected in all cases, with a melting temperature within 1◦C of the theoretical melting temperature of the sequenced amplicon as determined by a web-based tool (OligoCalc; Kibbe, 2007). Baseline values were determined automatically and threshold value was set manually at 0.04 (maintained across all samples and standards). Amplification efficiencies were 93–96% in all instances. The number of CO1, ATP6, and EF-1-α sequences per 20-μL reaction was determined by comparing mean Ct values for each sample to the respective log-dilution standard curve, and CO1/EF-1-α and ATP6/EF-1-α ratios were then calculated.

naturalKLpop.csv