Reciprocal transplantation of the heterotrophic coral Tubastraea coccinea (Scleractinia: Dendrophylliidae) between distinct habitats did not alter its venom toxin composition

Abstract Tubastraea coccinea is an azooxanthellate coral species recorded in the Indian and Atlantic oceans and is presently widespread in the southwestern Atlantic with an alien status for Brazil. T. coccinea outcompete other native coral species by using a varied repertoire of biological traits. For example, T. coccinea has evolved potent venom capable of immobilizing and digesting zooplankton prey. Diversification and modification of venom toxins can provide potential adaptive benefits to individual fitness, yet acquired alteration of venom composition in cnidarians is poorly understood as the adaptive flexibility affecting toxin composition in these ancient lineages has been largely ignored. We used quantitative high‐throughput proteomics to detect changes in toxin expression in clonal fragments of specimens collected and interchanged from two environmentally distinct and geographically separate study sites. Unexpectedly, despite global changes in protein expression, there were no changes in the composition and abundance of toxins from coral fragments recovered from either site, and following clonal transplantation between sites. There were also no apparent changes to the cnidome (cnidae) and gross skeletal or soft tissue morphologies of the specimens. These results suggest that the conserved toxin complexity of T. coccinea co‐evolved with innovation of the venom delivery system, and its morphological development and phenotypic expression are not modulated by habitat pressures over short periods of time. The adaptive response of the venom trait to specific predatory regimes, however, necessitates further consideration.


| INTRODUC TI ON
The major and most lethal components of animal venoms are protein and peptide toxins. The toxin compositions of venom are attributed to serve in both offensive and defensive functions to facilitate prey capture and provide protection from predation (Casewell, Wüster, Vonk, Harrison, & Fry, 2013). Animals must adjust to divergent and changing conditions in their biotic and abiotic environment, which offer opportunities or pose challenges for feeding and defense.
Adapting toxin composition to accommodate these changes has been extensively documented in prominent venomous bilaterians.
For example, variability in snake venom toxins has been associated with geographical distribution (Alape-Girón et al., 2008) and their ecological conditions (Strickland et al., 2018), but are most widely attributed with their ability to capture, consume, and digest a wide variety of different prey types (Daltry, Wüster, & Thorpe, 1996;Gibbs & Mackessy, 2009). Likewise, geographical variations in venom toxin composition have been documented in some scorpions, spiders, and species of cone snails, which has been linked also to changes in habitat or diet (Abdel-Rahman, Omran, Abdel-Nabi, Ueda, & McVean, 2009;Duda, Chang, Lewis, & Lee, 2009;Pekár, Petráková, Šedo, Korenko, & Zdráhal, 2018). However, data are scarce or nonexistent for geographical or intraspecific variation for the majority of the venomous taxa.
Cnidaria (corals, sea anemones, hydroids, jellyfish, myxozoans, and kin) possess a unique venom delivery system-nematocytes or "stinging cells," that release their toxic payload by discharging a penetrative barb from an intracellular cnida organelle called a nematocyst. Approximately 30 different varieties of nematocyst are known to exist, but individual species usually combine no more than two to six structural types that are collectively known as the organism's cnidome (Östman, 2000). Cnidarians are possibly the earliest diverging venomous animal lineage to deploy venom for both predation and defense (Jaimes-Becerra et al., 2017). Venom production and maintenance are therefore central to cnidarian existence and evolution, and increasingly more is known about the evolutionary history and phyletic distribution of cnidarian toxins. A pattern is emerging which suggests venoms with predominantly cytolytic and neurotoxic activities were established by early cnidarian ancestors followed by lineage-specific recruitment of certain toxin protein families, with cytolysins diversifying prominently in Medusozoa and neurotoxins in Anthozoa (Balasubramanian et al., 2012;Brinkman et al., 2015;Huang et al., 2016;Jaimes-Becerra et al., 2017;Li et al., 2014Li et al., , 2016Macrander, Brugler, & Daly, 2015;Madio, Undheim, & King, 2017;Ponce, Brinkman, Potriquet, & Mulvenna, 2016;Rachamim et al., 2015). When an innovative unsupervised clustering approach was used to compare toxin composition between groups of venomous animals, the results revealed that despite the early divergence and morphological simplicity of cnidarians, their toxin composition was as complex as those of venomous insects, gastropods, and elapid snakes (Jaimes-Becerra et al., 2019).
The orange cup coral Tubastraea coccinea Lesson 1829 (Figure 1) is presently broadly distributed in the Indian and Atlantic oceans.
Corals of the genus Tubastraea are obligate heterotrophs, which lack autotrophic dinoflagellate endosymbionts to provide host photosynthetic nutritional energy, so instead feed predominantly on pelagic zooplankton. T. coccinea is considered an alien and invasive species to the southwestern Atlantic Ocean that has now expanded its habitat range to much of Brazil's southern coastal reefs, which is believed to have been propagated by a single invasion event in the 1990s (reviewed by Miranda, Costa, Lorders, Nunes, & Barros, 2016). T. coccinea is a hermaphroditic coral; therefore, settlement of broadcast planula via sexual reproduction does not occur. Proliferation ensues by the release of brooding larvae (Ayre & Resing, 1986) with high fecundity (de Paula, Pires, & Creed, 2014), or is spread by runners (Vermeij, 2005) and by the fragmentation of colonies, including regeneration from undifferentiated coral tissue which fosters the rapid propagation of clonal offspring (Capel, Migotto, Zilberberg, & Kitahara, 2014;Luz et al., 2018).
Recent proteomic analysis of nematocyst venom isolated from primary tentacles of T. coccinea revealed a complement of 17 likely toxins, mainly with predicted cytolytic or protease inhibitor activities, consistent with consuming a zooplankton diet (Jaimes-Becerra et al., 2019). Given the vastly different trophic and interspecific interactions cnidarians encounter in diverse benthic and pelagic habitats, surprisingly little is known about their venom composition in response to diverse ecological conditions. Here, we use a tandem mass tag (TMT)-based proteomics approach to compare wild clonal populations of T. coccinea from near-shore and offshore habitats to determine whether geographical distribution and environmental factors could influence venom toxin modification in this early diverging metazoan, or whether such alterations are elaborated only by more advanced bilaterian taxa.

| Study sites and sampling design
Clonal populations of T. coccinea from two environmentally distinct and geographically separate study sites were collected from the northern coast of São Paulo State, Brazil, and were analyzed for changes in protein expression at the beginning and end of a six-week

| Isolation and identification of nematocysts
Individual T. coccinea polyps are relatively large, and although the tentacles retract upon freezing, the soft tentacle tissues could easily be dissected using forceps and a scalpel blade. Intact nematocysts were isolated from the dissected tentacles as previously described (Weston et al., 2013) and, after microscopic inspection, were lyophilized. An optical microscope (Nikon ECLIPSE 80i with a 100x objective lens with immersion oil) equipped with a digital camera (Nikon DS-Ri1) for documentation was used for the identification of nematocysts. The abundance of different nematocyst types was recorded as either "very common," "common," or "rare" based on the scheme of Picciani, Pires, and Silva (2011).

| Protein extraction
A volume of 200 µl of lysis buffer (8 M urea, 75 mM NaCl, 50 mM Tris, pH 8.2, plus one tablet each of Roche protease (cOmplete) and phosphatase (PhosStop) inhibitor cocktail per 10 ml of lysis buffer) was added to approximately 10 mg of freeze dried nematocyst tissue. The reconstituted material was disrupted on ice using a Qiagen TissueLyser II operated at 20-30 Hz for 2 min. The extracts were then centrifuged for 5 min at 10,000 × g and 4°C. The supernatants were decanted, and soluble protein concentrations were quantified by Nanodrop (Thermo Scientific) spectrophotometry and then lyophilized.

| Protein preparation and labeling
The lyophilized samples were reconstituted in 50 µl TEAB and 1 µl of each protein extract (equivalent to 25 µg of protein in each sample), protein disulfides were reduced with 10 µl of 8 mM TCEP in 100 mM TEAB and 0.1% (w/v) SDS at 55°C for 1 hr, and then alkylated with 10 µl of 67.5 mM iodoacetamide in 100 mM TEAB and 0.1% (w/v) SDS at room temperature for 30 min in the dark. Trypsin digestion was then performed overnight at 37°C by adding 10 µl of 0.2 mg/ ml trypsin in 100 mM TEAB and 0.1% (v/v) TFA. Following trypsin digestion, each sample was labeled with an isobaric tandem mass tag (TMT) 6-plex reagent set using the protocol supplied with the reagent kit (Lot# SF251226 Thermo Fisher). Once labeled with a unique TMT reagent, the six individual samples were combined to create the TMT6plex analytical sample mixture. This combined sample of TMT labeled peptides was then fractionated into 12 fractions by off-gel electrophoresis using an IPG strip pH 3-10 (Bio-Rad) for 20 kV hr in 1× OFFGEL buffer, pH 3-10. Fractions were solubilized in 50 mM ammonium bicarbonate prior to LC-MS/MS.

| LC-MS/MS tandem mass spectrometry
Chromatographic separations were performed on each fraction using an Ultimate 3000 nano-LC system in line with an Orbitrap Fusion Tribrid mass spectrometer (Thermo Scientific). In brief, peptide samples in 1% (v/v) formic acid were injected onto an Acclaim PepMap C18 nano-trap column (Thermo Scientific). After washing with 0.5% (v/v) acetonitrile and 0.1% (v/v) formic acid, peptides were resolved on a 250 mm × 75 μm Acclaim PepMap C18 reversephase analytical column (Thermo Scientific) over a 150-min organic gradient, using 7 gradient segments (1%-6% solvent B over 1 min, 6%-15% B over 58 min, 15%-32% B over 58 min, 32%-40% B over 5 min, 40%-90% B over 1 min, held at 90% B for 6 min, and then reduced to 1% B over 1 min.) with a flow rate of 300 nl/min. Solvent A was 0.1% (v/v) formic acid, and Solvent B was aqueous 80% (v/v) acetonitrile in 0.1% (v/v) formic acid. Peptides were ionized by nanoelectrospray ionization at 2.2 kV using a stainless-steel emitter with an internal diameter of 30 μm (Thermo Scientific) and a capillary temperature of 250°C. All spectra were acquired using an Orbitrap Fusion Tribrid mass spectrometer controlled by Xcalibur 2.0 software (Thermo Scientific) and operated in data-dependent acquisition mode. FTMS1 spectra were collected at a resolution of 120,000 over a scan range (m/z) of 350-1,550, with an automatic gain control (AGC) target of 400,000 and a maximum injection time of 100 ms.

F I G U R E 1 Gross skeletal and soft tissue morphology of Tubastraea coccinea
The data-dependent mode was set to cycle time with 3 s between master scans. Precursors were filtered according to charge state (to include charge states 2-7), with monoisotopic precursor selection and using an intensity range of 5E3 to 1E20. Previously interrogated precursors were excluded using a dynamic window (40 s ± 10 ppm).
The MS2 precursors were isolated with a quadrupole mass filter set to a width of 1.6m/z. ITMS2 spectra were collected with an AGC target of 5,000, maximum injection time of 50 ms, and HCD collision energy of 35%. with an e-value of <1.0e -5 to entries in the custom toxin database were used as input to a machine learning tool called "ToxClassifier" that excludes proteins with possible nontoxic physiological functions (Gacesa, Barlow, & Long, 2016). Sequences that met the stringent validation process were considered bone fide potential toxins.

| Data analysis
A second analysis was carried out to assign putative biological functions to sequences not identified as potential toxins, by comparison against the annotated transcriptome of Eguchipsammia fistula (Yum et al., 2017), the closest species to T. coccinea for which gene predictions are available. Sequences were assigned an annotation that gave a match against the E. fistula transcriptome with an e-value of <1.0e -5 . The PEAKS Q algorithm was used to measure the abundance of peptides across the TMT6plex dataset where TMT reporter ion signals were above the limit of quantitation (LOQ value). The value for each peptide in a protein was summed into one measurement according to an up(+) or down(−) fold-change calculation. The quantitative results were represented as a clustered heat map (double dendrogram) using Neighbor-joining tree clustering based on the Euclidean distance. MS/MS spectra corresponding to regulated features of interest were manually reviewed to validate the assignment, and these raw spectral data have been deposited via the PRIDE partner repository with the dataset identifier PXD015559.

| RE SULTS
There were no apparent gross morphological differences among the coral fragments collected from the offshore or inshore sites at T0 and T6; and the transplanted Tx fragments recovered at the two experimental sites. The coral skeleton and soft tissues formed typical clumps of pink-red calcareous cups around a single deep red The composition and abundance of cnidae types were also uniform between T0, T6, and Tx offshore and inshore fragments, consistent with expected variation between specimens (Table 1). The most common capsule types were spirocysts, tentacle type holotrichous and tentacle type b-rabdoids; while mesentery type holotrichous and mesentery type b-rabdoids were also noticed, but to a lesser extent. All morphological features are consistent for this species (Picciani et al., 2011).
The composition of recognized toxins extracted from specimens' nematocysts was identical irrespective of site, time of sampling, and transplantation (File S1 PRIDE repository dataset identifier PXD015559  Table 2  fold-changes that were significantly different to the T6 specimen from the offshore to inshore sites. Protein fold-changes in the nontoxin proteome of the Tx fragment relocated from inshore to offshore sites was also affected six weeks after transplantation. While fold-changes in 20/74 (27.0%) of proteins were expressed at identical levels as those of the T6 specimen at the inshore site, 24/74 (32.4%) of proteins shifted to levels of expression equivalent to the native T6 fragments collected at the offshore site. For 30/74 (40.5%) of proteins in this Tx fragment, fold-changes did not resemble the levels of protein expression in either of the local T6 coral fragments from offshore or inshore sites.

| D ISCUSS I ON
This study examined potential differences in venom toxins of T. coccinea colonies inhabiting an inshore and offshore site in southern Brazil. The potential for T. coccinea venom to adapt its toxin composition in response to different habitat environments was also as- resulting in local and differential selection for a greater proportion of mammals over lizard prey items (Strickland et al., 2018). A previous study that compared the toxicological characteristics of cnidarians, however, concurred with our results. In that study (Radwan et al., 2001), the cnidome toxin composition of Cassiopea and Aurelia species collected from very distinct habitats and geographical locations (viz. Bahamas, Chesapeake Bay, and Red Sea) were identical.
The toxicological profiles of the venoms from isolated nematocysts were also alike between locations for each species; nevertheless, the potency of the venom in the Red Sea specimens was much greater (Radwan et al., 2001).
After six weeks, fragments of native and transplanted colonies collected at the sample sites retained identical macro (skeletal and soft tissue features, Figure 2) and micro (cnidome, Note: Putative toxins were annotated by homology of peptide sequences obtained from de novo sequencing of unique peptide MS/MS events with a custom database of known animal venom toxins. Venomous animals and their toxins with closest sequence similarity are given together with accession numbers corresponding to either UniProt or * ZoophyteBase  assignments. Note that the accession numbers in the left-hand column refer to the laboratory numbers used for the proteomics analysis given in File S1 (PRIDE repository dataset identifier PXD015559). a Potential toxins which met the criteria for quantification, data and calculations for which are given in File S1 (PRIDE repository with the dataset identifier PXD015559). ference also in the relative abundance of toxins between the two nematocyst preparations . A pattern is emerging from published data Radwan et al., 2001) and that of this study, whereby cnidarians appear to retain an array of different nematocyst types designed to deliver a single and invariant venom, with both venom and the delivery system having co-evolved independently of geographical or environmental forces. Such may reveal, however, that relative concentrations of certain toxins may vary between nematocyst types  or in species from different habitats, perhaps altering venom potency by varying nematocyst types and densities to allow adaptation concerning to prey preference or predator type, which warrants further investigation.
It may be argued that the six-week study period was not long enough for T. coccinea to adapt change in its venom composition, cnidome, and gross morphology following transplantation between the offshore and inshore sites. T. coccinea is an azooxanthellate and obligate heterotroph coral with an extreme regenerative capacity (Luz et al., 2018) that depends on prey capture to meet its metabolic needs, including its reproductive energy demands. It was thus expected that the venom composition of T. coccinea would be different in colonies sampled at inshore and offshore sites, due to potential F I G U R E 3 Heatmap showing quantified changes of proteins from discharged nematocysts isolated from Tubastraea coccinea specimens sampled offshore, inshore, and those transplanted between the two sites. Fold-change values were considered either a significant reduction (<2) or significant increase (>2) in tag intensity disparities in habitat prey populations. Yet, there were significant fold-changes in the cnida proteomes of specimens collected at each site and those following reciprocal transplantation after six weeks of acclimation. Nevertheless, it should be noted that the natural distributional range of this southwestern Atlantic invasive species has a much more diverse coral fauna and, therefore, complex "within order" substrate competition. Such competition might have shaped T. coccinea venon/toxins, which are potent enough to outcompete the native species from invaded areas.
However, comparatively, it is evident that a six-week period was sufficient for transcription and translation processes to alter expression patterns in the nontoxin proteome of T. coccinea nematocytes, but such adaptation certainly did not extend to the venom pheno- Our research will further test the influence of how diet affects the expressed venome composition, by selective transcription and translation of component toxins, which will improve our understanding on how venom evolved in early cnidarian development.

| CON CLUS IONS
The complexity of toxins in isolated nematocysts from the heterotrophic coral T. coccinea did not vary significantly in genotypically identical colonies taken from offshore and inshore sites, or following reciprocal transplantation of clonal fragments between the two sites. The structural morphology and cnidome toxin composition also did not change during the 6-week study period. These findings suggest that T. coccinea produces a single and relatively invariant venom, and that the venom and its delivery system are conservative and expressed independently of geographical influence and environmental selection in the invaded habitats of the northern São Paulo coast.

ACK N OWLED G M ENTS
We thank Prof Beth Okamura and Dr Alvaro Esteves Migotto for their advice during the early stages of this project and to Dr Walter Paulo.

CO N FLI C T O F I NTE R E S T S
None declared.

AUTH O R CO NTR I B UTI O N S
MVK and PFL designed research; MVK, AJB, EGM, MW, and PFL performed research; MW, GP, ACMa, and ACMo contributed analytical tools; and all authors analyzed data and wrote the paper.

DATA AVA I L A B I L I T Y S TAT E M E N T
The mass spectrometry proteomics data have been deposited