Detection of a chemical cue from the host seaweed Laurencia dendroidea by the associated mollusc Aplysia brasiliana

Chemical cues from sessile hosts can attract mobile and associated organisms and they are also impotant to maintain associations and overall biodiversity, but the identity and molecular structures of these chemicals have been little explored in the marine environment. Secondary metabolites are recognized as possible chemical mediators in the association between species of Laurencia and Aplysia, but the identity of the compounds has not been established. Here, for the first time, we experimentally verify that the sesquiterpene (+)-elatol, a compound produced by the red seaweed Laurencia dendroidea, is a chemical cue attracting the associated sea hare Aplysia brasiliana. In addition to revealing the nature of the chemical mediation between these two species, we provide evidence of a chemical cue that allows young individuals of A. brasiliana to live in association with L. dendroidea. This study highlights the importance of chemical cues in Laurencia-Aplysia association.


Introduction
In the marine environment, chemical cues or waterborne infochemicals act as mediators in different intra-and inter-specific ecological interactions between organisms [1], such as chemical defense [2], settlement and recruitment of larvae [3], feeding stimuli [4], foraging behavior [5], selection and localization of prey [6], among others. Marine organisms produce a wide variety of molecules that mediate these ecological interactions [1,7]. Despite their broad and crucial roles in population structure, community organization and ecosystem function in marine systems, the omnipresence of these chemical cues and their impacts are still inadequately recognized [1]. However, scientific efforts have been increasing to elucidate the mechanisms of chemical communication between organisms in marine environments [8].
Red seaweeds of the genus Laurencia are well established as rich producers of secondary metabolites [9]. Since investigations of this taxon began more than 60 years ago, more than a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 Laurencia-conditioned seawater (LCW) and Laurencia-exuded metabolites in seawater extract (SEM) For the laboratory detection of the L. dendroidea exuded metabolites in seawater, a proportion of 24 g of algae mass to 1 L of artificial seawater ("Sea salts" SIGMA, USA, in distilled water) was used for incubation in fifteen Erlenmeyer flasks of 3 liters each. They were kept for 12 hours in a BOD chamber at 20˚C, under constant ilumination and constant aeration. Then, the algae were taken out from the flasks and the conditioned water was passed through a 0.45 μm filter Millipore system to yield the Laurencia-conditioned seawater (LCW). Part of this LCW was used immediately in Y-maze tests (20 L), and the remaining LCW (10 L) was used to identify the metabolites exuded by L. dendroidea. Artificial seawater (without the presence of L. dendroidea) was used as a control for chemical analyses and as a control stimulus in the Y-maze tests. The portion of the LCW separated for chemical analysis was passed through a glass column filled with DIAON HP20 resin (Sigma-Aldrich). Then the resin was washed with ultrapure water to eliminate salts and the metabolites retained in the resin were eluted with 600 mL of methanol (Tedia, HPLC), that was evaporated to yield the Laurencia-exuded metabolites in seawater extract (SEM). The same procedure was used to produce a representative control from artificial seawater. Both the SEM and the control extract from artificial seawater were analyzed by GC/MS.
The chemical structures of the isolated compounds were established by NMR of 1 H and 13 C, correlation spectroscopy (COSY), heteronuclear multiple-bond correlation (HMBC) and heteronuclear single quantum correlation (HSQC) in a Bruker 1 Avance 600 MHz system using CDCl 3 . Chemical shifts were reported in ppm referenced to solvent signals (CDCl 3 : δ H 7.26, δ C 77.16). Mass spectrometric data were acquired using a VG-Autospec Fisons system in electrospray ionization mode. The specific rotations [α] D were determined using a Polarimeter Perkin-Elmer 1 (model 241), under conditions of 20˚C, sodium D line λ = 589 nm and cell path length 1 dm. All data were compared with the available literature [24][25][26][27][28][29].

GC/MS analysis
Chemical profiles were obtained on a GC/MS Agilent 7890B, MS7000C (using an automatic injector CombiPAL with the control software MassHunter) in the electron impact mode (70 eV). Chromatographic separation was achieved with a HP-5MS UL capillary column (15 m x 0.25 mm x 0.25 μm). Helium was used as the carrier gas (flow rate of 1 mL min -1 ). The oven temperature was 60˚C for the first 2.5 min, then increased at a rate of 10˚C.min -1 to 280˚C and held at the latter temperature for 6 min, with a total run time of 30.5 min. Injector and ion source temperatures were 300 and 240˚C. Samples were injected in splitless mode. The detection was performed in the full scan mode, using a mass range of 60-400 m/z. All samples were solubilized in dichloromethane and filtered through a PTFE Syringe filter (pore width of 0.45μm; diameter of 13mm) before the injections.
The chemical profiles of all the extracts (whole algal extract, SEM and control) were obtained by GC/MS. The isolated compounds (1-5) and purified standards (6-15) (  available in our laboratory were also analyzed by GC/MS to construct a L. dendroidea metabolites database. Compounds 6-15 were obtained as described by the literature [30][31][32][33]. Identification of metabolites exuded into seawater by L. dendroidea (SEM) was performed by comparing the peaks of the SEM chromatogram with its respective control (artificial seawater extract). A baseline alignment of each chromatogram was performed prior to comparison and the time-shift among peaks was corrected by the Correlation Warping Algorithm (COW), as described by Nielsen and collaborators (1998) [34], using the software COW (http://www2. biocentrum.dtu.dk/mycology/analysis/cow). Compounds present in the algal extract, and SEM extract were identified based on their mass spectra data, retention time, and comparison with pure compounds available in our laboratory.
The compound identified from the SEM extract (elatol) was quantified using external standardization by means of an analytical curve calculated from four different concentrations of purified elatol in dichloromethane (10 to 10 −2 μg mL -1 ). Detection and quantification limits were determined according to United States Environmental Protection Agency provisions; we obtained an analytical curve in a concentration range almost 5 times the noise of the baseline. Then, the concentrations of three replicates of the standard were determined from this curve, and its standard deviation was calculated. The relative standard deviation (RSD) or coefficient of variation (CV) was 10%.

Y-maze experiments
Y-maze behavioural bioassays were performed to determine whether the specimens of A. brasiliana would be attracted by L. dendroidea compounds. The maze used in this study consisted of a Y-shaped plastic tube connected to two 10 L aquariums, with seawater flowing by gravity from the aquariums through each branch of the Y-maze (Fig 2). One of the aquariums contained artificial seawater with the tested stimuli (treatment), while the other was filled with pure artificial seawater, i.e. the control (see below). Water flowed from the two aquariums (treatment and control) through the Y-maze and exited at the base of the conjoined branch. Soluble coloring was used to verify the water plume and flow regulators were used to standardize the flux in the control and treatment aquaria before each experiment. The artificial seawater was used at 26˚C and a salinity of 35. In the intervals between each assay, the entire Y-maze was washed with distilled water and rinsed with seawater and control and stimulus arms of the Y-maze were switched. The seawater flow in each branch of the Y-maze was controlled by flow controllers. Individuals of A. brasiliana were initially positioned at the base of the Y-maze at the start of each bioassay.
Six experiments were undertaken with individual A. brasiliana choosing between stimulus vs artificial seawater. Each experiment consisted to introduce one sea hare at the base of the Ymaze and its behavior in terms of choosing between the arms containing either the stimulus or the artificial seawater was observed for up to 15 min. Three responses were considered: (a) positive, when A. brasiliana swam to the end of the stimulus arm; (b) negative, when A. brasiliana swam to the end of the control arm; and (c) no response, when A. brasiliana did not choose either one of the arms within the 15 min time frame. The stimulus tested in each experiment was drive in the follow sequence: 1) Control stimulus (n = 21): both arms of the Y-maze tube were filled only with artificial seawater in order to verify the ability of A. brasiliana individuals to overcome the water flow in the branches of the Y-maze structure; 2) Fresh L. dendroidea (n = 32): one specimen of this seaweed was placed in the stimulus arm of the Y-maze, while a plastic L. dendroidea mimic with equivalent size and volume of this fresh seaweed was placed in the artificial seawater arm (control) to maintain an equivalent flow of water through the tube during the experiment; 3) Laurencia-conditioned seawater (LCW): the stimulus aquarium was filled with LCW while only artificial seawater was used in the control tank. A total of 43 replicates were used; 4) L. dendroidea extract: the effect of compounds obtained from fresh L. dendroidea on A. brasiliana was evaluated using the extract at its obtained concentration (the amount of extract per algae dry mass, 146 mg g -1 ). A piece of filter paper was cut into squares 1.5 × 1.5 cm and each square was equivalently soaked into the extract previously dissolved in dichloromethane (concentration of extract determined as the extract equivalent to the dry weight of alga = dry weight of filter paper). For the control, an equivalent square piece of paper was soaked in dichloromethane without the extract. After evaporating the solvent, each piece of paper of control and treatment was positionated at its respective Y-maze arm. A total of 42 replicates was used; 5) Laurencia-exuded metabolites in seawater extract (SEM): To evaluate the effect of SEM extract (using the obtained concentration of extract per algal dry mass incubated, 5 mg g -1 ) on the Aplysia specimens, the experiments were conducted in the same way as described for LCW above. A total of 30 replicates was used; and 6) Elatol-conditioned water (ECW): the effect of elatol (2), the compound identified in the SEM by GC/MS, on Aplysia specimens was investigated. The amount of treatment added was such that the final concentration was equal to the natural isolated concentration based on seawater volume used for extraction (0.02 mg L -1 ). The compound was solubilized in 35μL DMSO and added to the artificial seawater in the stimulus arm. The same amount of 35μL DMSO was added to the artificial seawater in the control arm. A total of 26 replicates were used in this experiment. In the intervals between each assay, the entire Y-maze was washed and the control and stimulus arms of the Ymaze were switched in order to remove any remaining A. brasiliana scents or previous chemical cues. For each assay a new A. brasiliana specimen was used. In the end of the experiments all the animals were returned to field.
The significance of A. brasiliana behavioral choice was evaluated by a chi-squared test that compared the number of times that A. brasiliana moved to for the three response groups (positive response, negative resposnde and no response) against an equal random distribution. Differences between the groups were deemed statistically significant when p 0.05.

Identification of metabolites in extracts of Laurencia dendroidea by GC/ MS analysis
The L. dendroidea extract, Laurencia-exuded metabolites in seawater extract (SEM) and control extract (extract from artificial seawater) were analyzed by GC/MS. The retention times and mass spectroscopy patterns of the compounds in these extracts were compared to the L. dendroidea metabolites database (Table 1 and Fig 3). Compounds 1-8 were detected in the L. dendroidea extract (Fig 3A), whereas only compound 2 (elatol) was found in the exudate (SEM) (Fig 3B) at a concentration of 0.02 mg L -1 (or 0.0007 mg g -1 by algal wet weight; y = 1E +10x-3E+08; R 2 = 0.99). No metabolites were identified in the control extract (Fig 3C).

Discussion
Our results showed that the mollusc Aplysia brasiliana is able to detect and respond to chemicals produced and exuded by Laurencia dendroidea. This mollusc showed a positive chemotaxis by the presence of live seaweed, its water exudates in conditioned water (LCW) and also to its lipophilic extract (SEM).
Benthic marine invertebrates sense molecules from other organisms and use these compounds in different ecological contexts as communication, navigation and foraging behaviour [36]. These organisms produce and detect a wide range of polar and nonpolar compounds, but Table 1 their identities are poorly understood [7]. Here, the lipophilic compounds exuded by L. dendroidea into seawater were detected by GC/MS analysis. Among them, the halogenated sesquiterpene (+)-elatol (2) was identified and revealed as one of the major compounds. In adition, the analysis of L. dendroidea extract revealed (+)-elatol as the major compound produced. These results indicated that (+)-elatol was exuded by alga into seawater. The nonpolar (+)-elatol (2) is a known terpenoid isolated from Laurencia species that has multiple ecological functions, including anti-herbivory [12] and antifouling activities [37]. In our work, (+)-elatol attracted A. brasiliana in Y-maze experiments indicating its potencial as a chemical mediator in the interaction between Laurencia and Aplysia, its associated fauna.
Chemoreception can be the principal sensory system driving behavior for several marine organisms, including slowmoving molluscs belonging to the genus Aplysia [36]. Specific receptors involved in chemical signaling are able to detect and to bind to different compounds as water-soluble amino acids, organic acids or liposoluble terpenoids [44]. The sea hare Aplysia californica, for example, has a well-developed chemical sense with rhinophores and tentacles that help it to localize odors of its main food, i.e., seaweeds belonging to Plocamium and Laurencia genus [45]. A. brasiliana, which usually swims in order to find food [46], probably use similar sensors for localization. Our results provide evidence that secondary metabolites are important chemical cues mediating the interaction between A. brasiliana and L. dendroidea.
Identification of chemical communication and discrimination between evolved functions (signals) and unintentional releases (cues) are significant challenges in aquatic biology [8]. Chemical signals are intentionally released by the sender to convey information about presence of the signaler or the signaler's environment changing the behavior of the receiver; this change of behavior must be profitable to sender and receiver. In general, signals must be honest and reliable. Already chemical cues are compounds unintentionally released by the sender that any other organism can use as a guide to display a particular behavior, benefiting the receiver exclusively [47] Distinguishing these two modes of information is essential in identifying, properly, a chemical substance that triggers a certain behavior in another individual, and only a closer look at the chemical compounds involved brings clarification about its ecological and evolutionary interactions [48]. Here, for the first time, the combination of experiments allowed to verify that the sea hare A. brasiliana is attracted by the sesquiterpene (+)-elatol as a chemical cue at different concentrations. In the experiments using conditioned water and pure elatol, the concentration of this compound was probably higher than that in natural conditions. On the other hand, in experiments using a unique piece of algal thallus, the exudated concentrations are expected to be similar to natural conditions, and the result obtained in this assay was similar that verified using conditioned water and pure elatol. In all cases, elatol present in the water was sufficient to attract A. brasiliana.
Additionaly, the diurnal production of elatol by L. dendroidea is dynamic and strongly correlated to the primary photosynthetic metabolism [49]. So, it is expected that its surface concentration is variable creating different scenarios for the chemical sinalization and interactions between the host seaweed and its associated fauna including the mollusc A. brasiliana.
Several species of Laurencia contain refractile membrane-bound vesicles (corps en cerise) in the outer cells or cortical layer, which act as locations for the biosynthesis and storage of halogenated secondary metabolites [50], including elatol. Through a transport system, these metabolites reach the surface and provide essential cues as surface-mediated chemicals [51]. A previous study revealed that elatol occurs in very low amounts on the surface of L. dendroidea (0.006 mg g -1 ) when compared to the highest within-thallus quantities (9.89 mg g -1 ) based on algal dry weight [52]. Low concentrations of elatol on the surface of L. dendroidea did not inhibit herbivory by sea urchins, settlement of barnacle larvae, or mussel attachment [51], but is probably enough to attract A. brasiliana.
Compared to adults, juvenile sea hares are known to be subject to high predation pressure [53]. In addition, juvenile sea hares have restricted diets, often eating only one or a few types of red seaweeds [54]. Several small sedentary herbivores (mesograzers such as amphipods, crabs and polychaetes) can significantly decrease their susceptibility to predation by living among and feeding on seaweeds that are avoided by predators [55]. In our bioassays, we only used juvenile A. brasiliana that were found associated with L. dendroidea in the field. Thus, young individuals of A. brasiliana are mesograzers that use L. dendroidea both food and as a chemical source that protects from predators.
In conclusion, our discovery of Aplysia foraging cues reveals that this mollusc uses prey secondary metabolites as foraging cues. The sesquterpene (+)-elatol can facilitate the encounter of A. brasiliana with L. dendroidea, but it is still not known if other compounds are involved in this process. Thus, elatol represents for this alga both an evolved function against natural enemies such as pressure by conusmers [12] and fouling [37], and also a cue that benefits the sea hare A. brasiliana.
Supporting information S1 Table. NMR data and all the MS and NMR spectra of the isolated compounds 1-5 found in L. dendroidea. (DOC)