In vivo assessment of mitochondrial respiratory alternative oxidase activity and cyclic electron flow around photosystem I on small coral fragments

The mutualistic relationship existing between scleractinian corals and their photosynthetic endosymbionts involves a complex integration of the metabolic pathways within the holobiont. Respiration and photosynthesis are the most important of these processes and although they have been extensively studied, our understanding of their interactions and regulatory mechanisms is still limited. In this work we performed chlorophyll-a fluorescence, oxygen exchange and time-resolved absorption spectroscopy measurements on small and thin fragments (0.3 cm2) of the coral Stylophora pistillata. We showed that the capacity of mitochondrial alternative oxidase accounted for ca. 25% of total coral respiration, and that the high-light dependent oxygen uptake, commonly present in isolated Symbiodiniaceae, was negligible. The ratio between photosystem I (PSI) and photosystem II (PSII) active centers as well as their respective electron transport rates, indicated that PSI cyclic electron flow occurred in high light in S. pistillata and in some branching and lamellar coral species freshly collected in the field. Altogether, these results show the potential of applying advanced biophysical and spectroscopic methods on small coral fragments to understand the complex mechanisms of coral photosynthesis and respiration and their responses to environmental changes.

Coral reefs owe most of their high diversity and productivity to energetic relationships occurring in Scleractinian corals 1 . Coral symbiotic nature is characterized by a complex metabolism integration between the cnidarian host and its photosynthetic endosymbiont (Symbiodiniaceae, dinoflagellate) 2,3 . The coral physiology is thus determined by multiple metabolic traits 4 from which respiration and photosynthesis are the utmost driving processes 5 . Coral bioenergetics has been investigated through gas exchange analysis thanks to the development of different systems which allow measurements in situ (e.g. 6,7 ) or in laboratory (e.g. 8,9 ). Coral respiration is largely fuelled by photosynthetically-derived carbon molecules 10 , by active heterotrophy 11 , and by oxygen released during photosynthesis 12 . It is carried out in both animal and algal mitochondria by the respiratory complexes (Complex I, II, III, IV and F 1 F O ATP synthase) which fulfils most of the cellular ATP demands. Host mitochondrial electron transport activity is directly related to respiratory substrates availability 13 . However, it has been suggested that the way in which corals consume storage molecules, as lipids, may be related to their sensitivity to bleaching conditions 14 . In the model sea anemone Exaiptasia pallida, thermal stress causes a decrease in gene expression of cytochrome c and ATP synthase subunit a, and ultimately to mitochondria degradation and apoptosis 15 . Cnidarian cellular physiology seems to be intimately linked to mitochondrial redox state 16 . Beyond this, the role of non-proton pumping (and consequently non-ATP yielding) alternative respiratory enzymes, such as alternative NADH dehydrogenases and alternative ubiquinol oxidase (AOX) which genes are present in the cnidarian genome 17,18 , is still unknown. Differently from the host, Symbiodiniaceae mitochondria lack the canonical NADH:ubiquinone oxidoreductase (Complex I) but possess an AOX 19,20 . Understanding the contribution of all these components to symbiont and host respiration will help to characterize the physiological responses of corals to natural oxygen variations in the reefs and its interaction with stress 21  www.nature.com/scientificreports/ conducted on coral colonies of Pachyseris speciosa, Pocillopora damicornis, Acropora formosa and Pavona cactus. These colonies were collected by SCUBA diving in the area of Malakal in the Palau Archipelago (7°19′28.3″ N 134°28′01.6″ E), during a sampling campaign conducted in January 2018. After collection, colonies were transferred to the facilities of the Palau International Coral Reef Center (PICRC) and suspended from a nylon wire inside a tank supplied with running filtered natural seawater. Temperature in the tank was 30 ± 1 °C and the maximum Photosynthetic Photon Flux Density (PPFD) at noon was about 100 μmol photons m −2 s −1 . Symbiodinium microadriaticum culture (CCMP 2467) originally isolated from shallow water coral S. pistillata from the Red Sea, was kindly provided by Prof. Oren Levy (Bar Ilan University, Israel). The culture was maintained in 250 mL Fernbach flasks containing 75 mL of artificial sea water (Coral Pro Salt, Red Sea Fish LTD) at 34 PSU and enriched with Guillard's (F/2) Marine Water Enrichment Solution (Sigma-Aldrich). The culture was grown under cool white fluorescent lamps at 75 µmol photons m −2 s −1 on a 12 h:12 h light:dark cycle and it was subcultured weekly. For spectroscopic measurements cells were adjusted to 10 µg of total chlorophyll per mL and supplemented with 10% (w/v) of Ficoll to avoid sedimentation.

Coral fragment preparation.
Square fragments of about 5 by 5 mm (Fig. 1a) were cut from various coral colonies with a Dremel 8220 (Dremel, USA) equipped with a diamond disk. For the branching corals S. pistillata and A. formosa, two transversal cuts were done close to the end of similar terminal branches, one to remove the growing white tip, and the other which determine the length of the fragment (i.e. approximately 5 mm). A third longitudinal cut determined the thickness of the fragment. The longitudinal cut in the shrubby coral P. damicornis was made in the skeleton regions thicker than 3 mm. Laminar corals P. speciosa and P. cactus were cut perpendicularly to the plane of the skeleton. This procedure was performed out of the aquarium and took less than 30 s per fragment, after which it was rapidly immersed in another sea water container to remove debris. Coral fragments were analysed within one hour after being cut (termed 'day 0') or thereafter maintained in the same aquarium conditions as their parent colony.
Chlorophyll a fluorescence measurement. In vivo chlorophyll a fluorescence images were obtained by a SpeedZen imaging system (BeamBio/API, France). An array of blue LEDs (450-470 nm) was used to excite chlorophyll a during 400 μs, and images were obtained by a CCD camera (UI-3240CP-NIR-GL Rev.2, IDS, Obersulm, Germany) filtered by a long pass red filter (680 nm) (Fig. 1b). After incubating samples in darkness www.nature.com/scientificreports/ for at least 15 min, minimal chlorophyll a fluorescence yield (Fo) image was acquired. A saturating red actinic light pulse (3000 μmol photons m −2 s −1 ), provided by a LED array (660 nm), was applied for 200 ms and maximal chlorophyll a fluorescence yield (Fm) was acquired. Maximum quantum yield of photosystem II was calculated as Fv/Fm = (Fm − Fo)/Fm 43 . Coral fragments were exposed to different PPFD (15,68,172,354 and 755 μmol photons m −2 s −1 , provided by the same red LED array than the saturating pulse) during 3 minutes and fluorescence (Fs) was recorded every 5 s (Fig. 1d). A saturating red actinic light pulse was applied to obtain the maximum fluorescence yield under light acclimation (Fm′) every 30 s, and the effective quantum yield of PSII at steady state was calculated as ϕPSII = (Fm′ − Fs)/Fm′. The relative electron transfer rate through PSII (rETR-PSII) was then calculated as ϕPSII x PPFD, and the non-photochemical quenching of chlorophyll a fluorescence (NPQ) as (Fm − Fm′)/Fm′ 43 .
Fast induction curves of chlorophyll a fluorescence were recorded with a Handy PEA fluorometer (Hansatech Instruments, UK), by illuminating the sample during 10 s with a strong red light pulse (peak wavelength of 650 nm) of 2000 μmol photons m −2 s −1 . This fluorescence rise is characterized by a sudden increase from a basal fluorescence level (O), with two intermediate inflections (J and I), to a maximum fluorescence level (P) commonly referred as OJIP transient (see Chapter 12 in 34 ). O phase is taken at 50 μs, J appears at 2 ms, while I appears after 50 ms and P after 2 s in corals. Each coral fragment was set in the bottom of a standard spectrophotometer cuvette facing the head sensor and filled with 2 mL of artificial seawater and maintained in darkness for 15 min before the measurement.
Light-induced absorption changes. All measurements were performed with a JTS-10 spectrophotometer (BioLogic, France) piloted by a BeamBio/API (France) electronic device. When necessary, the final size of the coral fragments was adjusted to fit into standard 45 mm × 12.5 mm × 12.5 mm (H × W × D) glass spectrophotometer cuvette, i.e. with a pathlength of 10 mm. Each coral fragment was held in the cuvette with the side containing most polyps facing the light source and the skeleton exposed area (when applicable) in front of the detecting photodiode (Fig. 1c). This guaranteed a controlled illumination of the sample.
In the photosynthetic electron transport chain of peridinin chloroplast-harboring dinoflagellates cyt c 6 is the electron donor to PSI instead of Plastocyanin (PC) 44 . For this reason, light-induced absorption changes of P700 were directly measured at 705 nm 45 by applying μs detecting flashes during short intervals of dark pulses (as described in 46 ). Continuous actinic light at different intensities (25, 85, 210, 380, 670 and 1400 μmol photons m −2 s −1 ) was provided during 5 s (3 min for field coral specimens) by a LED array (640 nm). When coral fragments were exposed to continuous light, a stable negative change in absorbance at 705 nm (due to the formation of oxidized P700) 47 was obtained and recorded as Ps (Fig. 1f). A strong saturating pulse of light (9000 μmol photons m −2 s −1 ) was applied to obtain a more negative absorbance signal corresponding to maximum photooxidation of P700 as Pm' , while after a dark period of 1 s absorbance was recorded as Po. Maximum absorbance change (Pm) was estimated by adding the potent photosystem II inhibitor DCMU (3-(3,4-dichlorophenyl)-1,1-dimethylurea) at 20 μM (Fig. 1f). Photosystem I quantum yield was calculated as Y(I) = (Pm′ − Ps)/(Pm − Po), quantum yield of non-photochemical energy dissipation due to donor side limitation as Y(ND) = (Ps − Po)/ (Pm − Po), and the quantum yield of non-photochemical energy dissipation due to acceptor side limitation as Y(NA) = (Pm − Pm′)/(Pm − Po) 48 . The relative electron transfer rate through PSI (rETR-PSI) was calculated as the product of Y(I) and PPFD.
Absorbance changes of some pigments in the green region of the visible spectrum occur in response to light-induced changes of the trans-thylakoid electric field. This so-called electrochromic shift (ECS) is a useful way to determine photosynthetic parameters such as stoichiometry of active PSI and PSII in photosynthetic eukaryotes 49 , including Symbiodiniaceae 37 . ECS spectra were determined by measuring light-absorption changes in response to 2 ms continuous illumination, each 10 nm using light band-pass filters from 480 to 600 nm, as previously described 50 .
The ratio between active PSII and PSI centers can be quantified because the ECS signal responds linearly to the intensity of the electric field 37 . ECS signal at 554 nm was followed by applying a saturating single-turnover flash of 5 ns provided by a Nd:YAG Laser (Minilite II, Continuum) and its amplitude was considered as PSI + PSII reaction centers contribution. PSII inhibitors DCMU (20 µM, from a stock solution dissolved in ethanol) and hydroxylamine (1 mM, from a stock solution dissolved in distilled water) were added simultaneously and incubated in darkness during 10 min, another single-turnover flash was given and the resulting ECS amplitude accounted for PSI contribution. Ten consecutive repetitions, separated by 5 s in darkness, were averaged per measurement.
Oxygen exchange. Oxygen consumption (respiration) or evolution (photosynthesis) were measured with an Oxygraph + System (Hansatech, UK). Oxygen concentration calibration was carried out at two points with 0.2 µm-filtered F/2 medium either saturated in oxygen after vigorous shaking, or depleted in oxygen by addition of sodium dithionite (Sigma-Aldrich).
Coral fragments were held suspended into a DW1 oximeter chamber (Hansatech, UK) by an inert piece of acetate sheet. This enabled the adequate stirring of the medium by the magnetic stir bar (Fig. 1b). Each fragment was dark acclimated for at least 15 min with the chamber opened before closing it to monitor oxygen consumption in the dark for 5 min. After this dark period, illumination was provided by the Chlorophyll a fluorescence imaging system, so during fluorescence recording oxygen concentration was simultaneously acquired (Fig. 1e). www.nature.com/scientificreports/ 1180 µm; Sigma-Aldrich) for 2 min at 30 Hz with a TissueLyser II (QIAGEN). Pigment extracts were centrifuged at 16,000 g for 10 min to remove debris, and chlorophyll a and c 2 concentrations were calculated according to 51 from absorption measurements at 632 and 665 nm. Tissue-containing surface of coral fragments was calculated from conventional real color pictures of each fragment by using the software ImageJ (version 1.51j8).
Symbiodiniaceae genotyping. Small fragments of each species (n = 3) were removed from independent colonies and immediately processed or preserved in RNAlater (Invitrogen). Fresh or preserved fragments were washed with filtered artificial sea water and coral tissues were extracted by airbrushing the fragments into sterile PBS. DNA from the microalgal fraction was then extracted by using a DNeasy Plant mini kit (QIAGEN) according to the manufacturer's instructions, with slight modifications. The PCR amplification of the ITS2 gene marker was performed using primers pair SYM_VAR_5.8S2 and SYM_VAR_REV as detailed in 52 . The amplicons library was constructed using the Nextera XT kit (Illumina, Inc., San Diego, CA, USA) and sequencing was carried out on the Illumina MiSeq platform with 2 × 250 bp read configuration. The ITS2-type profiles were obtained by using the SymPortal analytical framework 53 .

Results
Whole S. pistillata coral colonies containing S. microadriaticum exhibited a homogeneous pigmentation (as illustrated in Fig. 2a). Chlorophyll a fluorescence imaging indicated a maximum quantum yield of photosystem II (Fv/ Fm) of 0.53 ± 0.03 (Fig. 2b,d). Small half-cylinder shape fragments were cut from these colonies. They were covered with about 35 mm 2 of coral tissue with an average total chlorophyll content (Chl a + c 2 ) of 30.5 ± 7.6 µg cm −2 , a Chl a/c 2 ratio of 2.39 ± 0.11 and 61 ± 11 polyps cm −2 (n = 16). These values remained stable up to 2 days after cutting and no visual sign of tissue necrosis was observed during this period (Fig. 2c). Although there was a positive correlation (R 2 = 0.41, p = 0.007) between Fv/Fm and fragment thickness (from 1.5 to 5 mm), the slope of the linear regression is very small so that the difference between a 2 mm fragment and a 5 mm fragment is less than 6% (Fig. 2e). In addition, the average of Fv/Fm values of coral fragments is not significantly different to the Fv/Fm value of coral colonies (one-way ANOVA, F(3,20) = 2.14, P = 0.13) (Fig. 2d). The fast polyphasic chlorophyll a fluorescence rise (i.e. OJIP curves) was obtained from freshly cut coral fragments and from 1-or 2-days old fragments (Fig. 2f) and no changes in the relative amplitude or presence of phases were observed.
To further test the good health of coral fragments, we determined respiratory and photosynthetic oxygen exchange rates from 3 mm thick fragments. Dark oxygen consumption (Rd) was on average 13.9 ± 2.1 nmol O 2 cm −2 min −1 over the three days and remained stable (Fig. 3a, inset). Net oxygen evolution as a function of the light intensity (Fig. 3a) was obtained simultaneously with fluorescence-based calculation of relative electron transport rate of PSII (rETR-PSII) (Fig. 3b). Both parameters showed a similar light dependency as indicated by the fact that rETR-PSII linearly correlated to oxygen evolution rates (R 2 = 0.98, p = 0.002) (Fig. 3c). Moreover, non-photochemical quenching (NPQ) capacity also remained stable (Fig. 3d).
Altogether, these data indicated that the global photosynthetic and respiratory activities of the coral fragments remained stable after cutting. In addition, most of the coral fragments showed the same potential to regenerate the tissue and the coral skeleton, as already shown for fragments of similar size 54 .
We then tested the effect of two classical inhibitors on respiration: potassium cyanide (KCN), a potent inhibitor of mitochondrial cytochrome c oxidase complex (Complex IV), and salicylhydroxamic acid (SHAM), an inhibitor of mitochondrial alternative oxidase (AOX). Addition of 1 or 2 mM KCN reduced Rd to 50% after five minutes (Fig. 4a). Addition of 2 mM SHAM further decreased the remaining cyanide-insensitive oxygen consumption to 26% of total Rd (Fig. 4b), and higher concentrations of SHAM (4 or 6 mM) did not suppress this residual oxygen consumption (data not shown). In contrast, addition of SHAM up to 2 mM alone did not have an impact on Rd during the 20 min of exposure to the inhibitor (Fig. 4a).
We then tested the potential of coral fragments to monitor light-induced absorption changes of PSI primary electron donor (P700) in the far-red region of the visible spectrum (705 nm) (Fig. 1f). These absorption changes corresponding to P700 photo-oxidation are fully consistent with previous observations in plant leaves or algae 48 and allowed us to calculate PSI quantum yield [Y(I)] and both PSI donor and acceptor side limitations [Y(ND) and Y(NA), respectively] at different light intensities (Fig. 5a). In addition, relative electron transport rates of PSI (rETR-PSI) were calculated from Y(I) values and a typical photosynthesis-light curve was obtained (Fig. 5b). We also observed a positive correlation (R 2 = 0.40, p = 0.01) between absorbance changes upon a saturating pulse of light in presence of DCMU (Pm values) and chlorophyll a content per coral fragment (Fig. 5c). In contrast, thickness of coral fragments did not correlate (R 2 = 0.06, p = 0.36) to the maximum amplitude of P700 absorbance change (Pm) (Fig. 5d). This indicates that the coral skeleton did not interfere significantly with our measurements.
We next quantified the fraction of active PSI and PSII centers in coral fragments by monitoring the electrochromic shift signal (ECS), a technique previously used in various algal species, including Symbiodiniaceae 37,49 . The overall shape of the ECS spectrum of S. pistillata was similar to the one of its isolated symbiotic algae (S. microadriaticum), with a maximum and a minimum observed at 510 nm and around 560 nm, respectively (Fig. 6a). The ratio between maximum (510 nm) and minimum (560 nm) values was smaller in coral fragments than in isolated algae. We then assessed the amplitude of the fast ECS (< 1 ms) at 554 nm in response to a saturating single turnover flash (Fig. 6b). Under these conditions, the amplitude of the ECS is proportional to the amount of active PSI + PSII (reviewed in 49 ). Addition of Hydroxylamine (HA) and DCMU, two PSII inhibitors, halved the amplitude of the ECS both in coral fragments of S. pistillata and isolated S. microadriaticum (Fig. 6b), indicating that PSI to PSII ratio is about 1 in both cases.
The four different coral species collected from Palau harbored the same Symbiodiniaceae genus (Durusdinium sp), which is a dominant algal genotype of inshore corals in this area 55   www.nature.com/scientificreports/ obtained (Fig. 6c, d). It was not possible to test PSI:PSII stoichiometry in these coral species because a saturating flash laser was not available at that time of analysis. Spectroscopic-based photosynthesis activity measurements allowed us to compare different parameters obtained from small coral fragments. In S. pistillata, rETR-PSII linearly correlated (R 2 = 0.999, p = 0.0005) to rETR-PSI up to the minimum saturating light intensity (around 200 µmol photons m −2 s −1 ), after which PSI activity continued increasing compared to PSII (Fig. 7a). Corals from the field revealed different responses when PSI and PSII activities were compared (Fig. 7b). A linear response was obtained from the two branching corals A. formosa (R 2 = 0.95, p = 0.004) and P. damicornis (R 2 = 0.96, p = 0.003) along with the different light intensities tested, but the laminar corals P. cactus and P. speciosa exhibited a disrupted relationship when light intensity was higher than 200 µmol photons m −2 s −1 .

Discussion
Coral survival to fragmentation is a well-known and broadly documented phenomenon (e.g. 56 ). In the case of S. pistillata, a robust scleractinian coral species, small nubbins survive after fragmentation as well 54 and have been used in skeleton calcification observations 57 . In this work, we demonstrate that thin coral fragments (around 25 mm 2 ) can be used to study various bioenergetics traits which so far have not been assessed in reef-building coral species.
The chlorophyll a fluorescence-based photosynthetic parameter Fv/Fm, which accounts for the maximum quantum yield of PSII, has been used to track damages on the photosynthetic capacity and health in photosynthetic organisms, including corals (e.g. 58 ). On the other hand, OJIP chlorophyll fluorescence rise has been used to reveal alterations of the PSII reaction center in Symbiodiniaceae cultures 59 and changes in the redox state of PSII acceptor side in corals 60 . The fact that we did not observe any significant modification neither of the Fv/Fm value nor in OJIP chlorophyll a fluorescence kinetics up to 48 h after fragmentation in small coral fragments (Fig. 2), indicates the absence of structural or functional disturbance in the photosynthetic electron transfer chain. Moreover, our coral fragments exhibited a photoprotection capacity (NPQ; Fig. 3d) comparable to values registered from branches of S. pistillata collected from shallow waters in the Red Sea 61 and from colonies maintained in aquarium conditions for several years 31 . Respiration measured in our coral fragments (Rd = 13.9 ± 2.1 nmol O 2 cm −2 min −1 ; Fig. 3a) agreed with previous studies on 2 cm long nubbins or 10 to 15 cm long branches from S. Rd and Pgmax values can be used to roughly calculate an extrapolated daily photosynthesis to respiration ratio (Pgmax/Rd*2) 63 . In our thin coral fragments of S. pistillata this ratio is 1.23 ± 0.16, a value in good agreement with S. pistillata branches (10 to 15 cm long) collected in summer from the Gulf of Aqaba (1.4) 56 , or from nubbins (2 to 3 cm long) prepared from S. pistillata coral colonies grown in aquarium conditions (1.1 to 1.5) 63 . Similarities between our results and data from the literature show that values measured on small thin fragments give reliable information on energetic relationships as those measured from bigger pieces of corals. Alternative electron transport pathways in mitochondria and chloroplasts are described as regulators of the redox poise in the main electron transfer chains, and more generally are known for their influence on the energetic state of the cell 64 . By measuring oxygen consumption rates in S. pistillata in presence of inhibitors (Fig. 4), we evidenced, for the first time, the presence of a cyanide-insensitive but SHAM-sensitive mitochondrial alternative oxidase (AOX) activity in coral. In several algal species, AOX contributes to dissipate excess of redox equivalents produced in the chloroplast during photosynthetic activity (i.e. NAD(P)H) (e.g. 66 ). The capacity of the AOX pathway measured in S. pistillata was 24% of Rd and is in the range as values previously reported for different species of Symbiodiniaceae grown in culture 20,65 . The observed lack of effect of the AOX inhibitor (i.e. SHAM) addition to fully respiring coral fragments demonstrates a high capacity of mitochondrial respiration through Complex IV. However, further studies are needed to determine the extent of both the cnidarian and photosymbiont's AOX contribution to whole holobiont respiration and to elucidate the role of AOX under control and stress conditions. Besides, in several species of Symbiodiniaceae in culture, a high light-dependent oxygen uptake was found to occur at the acceptor side of PSI through a Mehler-type reaction 37,38 . As illustrated in Fig. 3, the linear relationship between rETR-PSII and oxygen evolution observed in S. pistillata suggests that the photosynthetic electron transfer rate in high light does not depend on an alternative electron pathway to oxygen reduction as it does in free living cultures (e.g. 37,67 ). This result may be explained by a lower CO 2 limitation for symbionts in hospite www.nature.com/scientificreports/ compared to those isolated in culture 26,37 , but does not exclude that a different strategy operates under high light intensities. In this respect, PSI-dependent CEF is another alternative photosynthetic pathway that has been suggested to play an important role in the energetic balance in Symbiodiniaceae 23,36 . The most recommended way to assess CEF in vivo is to compare PSI and PSII activities 68 (Fig. 7). Despite the fact that calculation of absolute electron transport was not assessed 69 in the present study, our results indicate the occurrence of CEF in high light in S. pistillata and in two laminar corals from the field (P. cactus and P. speciosa). In contrast, we found that light influenced at the same proportion the relative electron transport rate of both photosystems in A. formosa and P. damicornis, two branching coral species, indicating that CEF is neglectable in high light (Fig. 7). It is worth to notice that these last four corals species harbour the same Durusdinium genus. A previous study, based on chlorophyll a fluorescence Serial Irradiation Pulses (SIP) method, indicated the absence of CEF in isolates of Durusdinium genus (referred as Clade D) 36 . Altogether, this highlights that involvement of photosynthetic alternative electron flows may strongly differ between corals harboring Symbiodiniaceae from the same genus. Contrastingly to PSII, the role of PSI on photoacclimation has just started to be revealed in Symbiodiniaceae dinoflagellates 70 . PSI has been the target of very few studies on coral photosynthesis due to technical constraints. An indirect determination made on freshly extracted algae from S. pistillata showed a higher PSI content in coral colonies in symbiosis with Symbiodinium sp. (referred as Clade A) than Cladocopium sp. (referred as Clade C) 61 , suggesting that PSI content can be determinant for coral photosynthesis adaptation to changing light regimes. Other studies, aiming to assess in vivo PSI activity in nubbins from other coral species were performed by following changes in light reflectance of 820 or 830 nm, and correcting the signal at 870 nm with a pulseamplitude-modulated Dual-PAM instrument (Walz, Germany) 39,40 . With this method a correction at 870 nm is commonly implemented in PSI activity determination to deconvolute PC contribution in the far-red region 48 . However, due to the absence of PC in peridinin chloroplasts 44 we consider that this correction is not necessary in photosynthetic corals, and light-induced absorption change at 705 nm can be used (Fig. 1f), as it commonly yields a high negative value in plants 46 and green algae 71 . In the case of Symbiodiniaceae isolates, absorbance change at 705 nm has been used with no major differences in kinetics of P700 photo-oxidation when a correction at 740 nm was applied (personal observation 37,38 ). Assuming that the molar attenuation coefficient of P700 at 705 nm of Symbiodiniaceae is in the range of values determined for P700 in green photosynthetic organisms   35,37 . Despite missing a molar absorption coefficient of P700 for peridinin plastids, our results indicate that spectroscopy-based measurement in thin coral fragments is a reliable and accurate strategy to ascertain PSI activity of coral endosymbionts (e.g. Fig. 5).
As thin coral fragments enabled spectroscopy-based measurements, ECS spectrum could be acquired. The presence of an ECS signal allows a direct measure of the electric field across algal thylakoid membranes 49 . The ECS spectrum measured in S. pistillata showed a small ratio between maximum (510 nm) and minimum (560 nm) ECS values in comparison to Symbiodiniaceae in culture (Fig. 6a 37 ). On the other hand, field coral specimens showed different ECS signal amplitudes at 554 nm compared to 510 nm, despite harbouring the same Symbiodiniaceae genera Durusdinium (Fig. 6c). In general, this difference can be attributed to pigment composition, distribution and orientation along the thylakoid membranes 49 , but we cannot exclude that other effects such as differences in coral tissue or skeleton optical properties contribute to this difference. ECS-based  www.nature.com/scientificreports/ measurements of PSI to PSII active center ratio in coral fragments gave comparable values to those previously reported in Symbiodiniaceae isolates (Fig. 6b 37,38 ).
In conclusion, in vivo monitoring of photosynthesis and respiration, as well as PSI and PSII activities, of S. pistillata coral fragments resulted in robust and reliable information. We thus consider that coral fragments are good spectroscopic objects to obtain reliable information about coral photosynthesis. This approach does not consume large amounts of coral colonies or nubbins allowing to limit the loss of living samples when destructive measures are taken. This work paves the way towards a better understanding of the role of bioenergetics processes in coral symbiosis establishment and persistence.