Mass spectrometry imaging reveals new biological roles for choline esters and Tyrian purple precursors in muricid molluscs

Despite significant advances in chemical ecology, the biodistribution, temporal changes and ecological function of most marine secondary metabolites remain unknown. One such example is the association between choline esters and Tyrian purple precursors in muricid molluscs. Mass spectrometry imaging (MSI) on nano-structured surfaces has emerged as a sophisticated platform for spatial analysis of low molecular mass metabolites in heterogeneous tissues, ideal for low abundant secondary metabolites. Here we applied desorption-ionisation on porous silicon (DIOS) to examine in situ changes in biodistribution over the reproductive cycle. DIOS-MSI showed muscle-relaxing choline ester murexine to co-localise with tyrindoxyl sulfate in the biosynthetic hypobranchial glands. But during egg-laying, murexine was transferred to the capsule gland, and then to the egg capsules, where chemical ripening resulted in Tyrian purple formation. Murexine was found to tranquilise the larvae and may relax the reproductive tract. This study shows that DIOS-MSI is a powerful tool that can provide new insights into marine chemo-ecology.

explore the biological roles of muricid choline esters and brominted indoles using DIOS-MSI in the context of egg laying and maternal investment, established spatial patterns based on previous histochemical staining 37 were used as a basis to track changes in secondary metabolite distribution at different phases of the reproductive cycle. The reliable performance of pSi in detecting the range of secondary metabolites was validated using organic extracts prepared from the same population of breeding individuals. Extracts were analysed by liquid chromatography mass spectrometry (LC-MS/MS) according to established procedures in our laboratory 35,36,38 ( Supplementary Figs 1 & 2), and used as a comparison for those detected in DIOS-MS and MSI (Figs 2 and 3). The capacity for pSi to desorb and ionise brominated indoles from the tissue samples was confirmed by the detection of isotopic patterns in MS/MS analysis consistent with the crude extracts analysed by LC-MS ( Supplementary Figs 2 and 3) and with purified compounds and synthetic 6 bromoisatin and 6,6′-dibromoindigo spotted directly onto the pSi surface 39 . DIOS-MS takes advantage of the van der Waals interactions and hydrogen bonding between the analyte containing the secondary metabolites and the silanised pSi surface. Attractive properties combined with high porosity (~600 m 2 cm −3 ) allow pSi to selectively extract small molecules in a sponge-like manner 40 . Consistent with our previous studies 25,39 , brominated indoles across a broad range of polarities in the low mass region from m/z 256-421 ( Fig. 2 and Supplementary Fig. 2), showed affinity to the pSi surface.
Stamping allowed low mass metabolite transfer from cryostat sections of the tissue to the pSi (Fig. 2), whilst the hydrophobic pSi surface facilitated the subsequent removal of residual tissue by pipetting with water, with minimal delocalization of the secondary metabolites. The removal of larger biological material reduces spectral complexity by eliminating fragments from highly abundant larger compounds like membrane lipids, which can suppress low abundant target signals. The isotopic clusters for mono-and      Fig. 3, Supplementary Fig. 3), in the first spatial analysis of this well studied 34 secondary metabolite. The presence of murexine was confirmed by extraction from the remaining tissue of a reproductively active female and was structurally elucidated using LC-MS/MS, 1 H-NMR and 13 C-NMR fingerprinting (summarised in Supplementary Table 1), and confirmed by comparison to choline and choline ester salts of tyrindoxyl sulfate previously reported from D. orbita and other muricidae molluscs 32,41 . Apart from major ion m/z 224, murexine showed major fragment ions in LC-MS/MS at m/z 165 and 121 ( Supplementary Fig. 3). These same fragment ions were detectable in DIOS-MS spectra co-localised with the major ion m/z 224 (165 and 121; Supplementary Fig. 4), confirming the detection of murexine using post source decay, and more broadly demonstrating the accuracy of DIOS-MS in the low mass range, whilst eliminating matrix suppression and non-target spectral "noise".
Biodistribution of secondary metabolites across the reproductive cycle. A major change in the biodistribution of Muricidae secondary metabolites is clearly apparent across the reproductive cycle of female D. orbita, with murexine, in particular, moving from the hypobranchial glands to the capsule glands during the breeding season (Figs 2 and 3). Furthermore, there is a substantial increase in the intensity of murexine detected in egg laying females, in comparison to the pre-and post-breeding stages (Figs 2 and 3). This suggests that the choline ester murexine may play a fundamental role in the reproduction of Muricidae molluscs. A role in reproduction has been previously suggested for the brominated indole precursors of Tyrian purple 31,36 , whereas the choline esters have been suggested to play a role in the feeding activities of these predatory snails 31,34 and/or simply assumed to stabilise the indoxyl sulfate precursors as a salt for storage within the hypobranchial gland tissue 28,32 . Here DIOS-MSI has provided novel information that suggests an up-regulation of choline esters for reproduction, that may in fact be far more important for Neogastropods than previously anticipated.
The co-localisation of murexine with the brominated indoles, tyrindoxyl sulfate and Tyrian purple in the medial region of hypobranchial glands (Fig. 3) was confirmed by correlation analysis of DIOS-MSI for the pre-reproductive females (R-squared 0.883 ± 0.1; Fig. 4a) and post-reproductive females (R-squared 0.713 ± 0.27; Fig. 4c). However, during the spawning (egg deposition) stage of the reproductive cycle, there was a substantial increase in the intensity (Fig. 3) and spatial detection of murexine in the capsule gland (Fig. 2), but not in the brominated indoles. Consequently, there was no longer a correlation between these secondary metabolites in egg laying females (R-squared 0.009; Fig. 4b). The increase in murexine in the capsule gland during egg deposition appears to be at the expense of the medial hypobranchial gland ( Fig. 2 and Supplementary Fig. 5). This implies that availability of murexine for feeding activities 31,34 is reduced during spawning. Few studies have investigated the energetic cost of maternal provisioning for non-primary metabolites 1 , and MSI may provide an attractive method for tracking mother to offspring transfer.
Maternal provisioning and chemical changes during encapsulated larval development. DIOS-MSI on the egg capsules ( Fig. 5) confirmed that murexine is transferred from the female capsule gland into the egg capsules of D. orbita, which is the first report of murexine in Neogastropod egg capsules. DIOS-MSI detected murexine only in the intracapsular fluid surrounding the early stage embryos in freshly laid egg capsules, where it is co-localised with tyrindoxyl sulfate and some minor intermediate brominated indoles and end products (Fig. 5b). In the capsules with late stage veliger larvae, water soluble murexine was no longer detectable. Tyrindoxyl sulfate also decreased in intensity and distribution during encapsulated larval development (Fig. 5, Supplementary Fig. 7). Conversely, insoluble Tyrian purple increased in the late stage capsules but was barely present at the early stage. This provides evidence for a unique form of chemical ripening in the egg capsules 19 involving two distinct classes of secondary metabolites specific to the Muricidae family of marine molluscs. The apparent high abundance of murexine relative to tyrindoxyl sulfate in breeding snails implies a naturally selected role for this compound during reproduction and early stage development, whereas tyrindoxyl sulfate could be acting as a controlled release counter ion to murexine within the egg capsules and hypobranchial gland. In this case, the formation of antimicrobial brominated indole intermediates during intracapsular development 35 could be a secondary function fortuitously acquired over the course of Muricidae evolution.  Biological role for murexine. The purpose of murexine within the female capsule gland is open to interpretation. However, since this compound is a potent muscle relaxant 34 , it may help relax the reproductive tract during egg deposition. Murexine has been suggested as a ligand for the muscle type nicotinic-acetylcholine receptor 34 (nAChR) based on the pharmacological actions of murexine in vivo within vertebrates 34,42,43 . The in vivo paralytic effect of murexine has the greatest similarity to suxamethonium, a muscle type nAChR agonist 34 . The choline esters are expected to become biologically active only when not complexed with tyrindoxyl sulfate 41 , as generally the quaternary ammonium ion of choline esters contribute to receptor binding and are required for the tranquilising activity for a range of other choline ester structures and acetylcholine 44,45 . The detection of murexine in the absence of brominated indoles indicates active muscle relaxing effects in the capsule gland environment.
The role of murexine within the egg capsules as a natural tranquiliser during larval development is also plausible. To test this, we purified murexine at 0.22 mM (50 mg/L) and confirmed that it has a temporary tranquilising effect on late stage intracapsular larvae lasting over 60 minutes ( Fig. 6a and Supplementary Video 1). Developing larvae shells are delicate and oxygen is limited during muricid intracapsular development 46 , despite this period being a metabolically active period. D. orbita reproductive females could be provisioning offspring with intracapsular tranquilisers to ensure they develop ready for life in the open sea. Indeed nearly 100% hatching success rate has been reported for the thousands of delicate larvae contained within each egg capsule of D. orbita 47 . Natural tranquilisers have been previously reported in molluscan egg masses, specifically in the common squid Loligo vulgaris, which incorporates an unidentified natural tranquiliser in perivitalline fluid within egg cases 48 , preventing premature hatching. Although the bioactive compound in squid perivitalline fluid does not appear to have been identified to date, cephalopods do not possess a hypobranchial gland and there are no reports of choline esters in the venom glands, ink or chemical messengers involved in the reproduction of cephalopods, despite extensive studies 34,49,50 . Therefore, the new biological role proposed here for murexine, appears to be an interesting example of functional convergent evolution between the gastropods and cephalopods. These two classes of marine molluscs have independantly evolved the deposition of benthic masses, as well as chemical sedatives to protect the encapsulated larvae.
Overall, this study has confirmed the usefulness of MSI for providing new insights into marine chemical ecology. DIOS -MSI has been used to map the distribution of two distinct classes of compounds over the adult reproductive and encapsulated larval stage in a muricid mollusc, revealing a significant upregulation of muscle relaxing choline esters in the capsule gland of females during egg deposition, and defining a role for this compound in the egg masses during the earlier stages of intracapsular development. The co-localisation of murexine with tyrindoxyl sulfate in both the hypobranchial glands of adults and in the early stage egg capsules confirms the intrinsic link between these secondary metabolites and provides evidence for a novel chemical ripening system that is likely to play a fundamental role in the encapsulated development of Muricidae larvae. Prior to the start of the breeding season, adult D. orbita were collected from rocky intertidal shores on Southern metropolitan coast in South Australia and housed in recirculating aquarium systems at Flinders University. The breeding population was maintained at temperate marine conditions (18 °C and Scientific RepoRts | 5:13408 | DOi: 10.1038/srep13408 35 psu seawater) and fed ad libitum on a diet of bivalves. Conditioned boulders were provided for egg capsule deposition. Pre-reproductive adult females (n = 3) were selected for mass spectrometry imaging (MSI) to analyze secondary metabolite distribution 30 days prior to the standard breeding time (early September). During the breeding period, a female (n = 1), observed in the process of egg capsule deposition was selected for MSI. Post-reproductive females (n = 3), two weeks after egg capsule deposition, were selected to image the post-reproductive tissue. Duplicate (n = 2) egg capsules deposited from reproductive females were selected immediately after deposition to assess early stage embryos and maternal capsule contents by MSI. The remaining capsules were maintained for full intra-capsular development of 35 days, after which two (n = 2) capsules with actively swimming larvae were selected to assess the veliger larvae and late stage capsule contents by MSI.

Methods
Tissue preparation for mass spectrometry imaging (MSI). Selected adult specimens were prepared by cracking open the shell with a vice at the junction between the primary body whorl and spire. The soft body was then removed by cutting the columnar muscle. Soft tissue was rinsed in MilliQ water to reduce residual salt. Female hypobranchial glands and pallial gonoduct, including the egg capsule glands, were removed by incision along the connective mantle tissue between the ctenidium and the branchial hypobranchial gland, along the posterior gonoduct and digestive gland. The hypobranchial gland and pallial gonoduct were left connected and were placed in 5 mL polypropylene cryo-vials (Sarstedt, Nümbrecht, Germany) and snap frozen in liquid nitrogen. Frozen tissue samples were protected from light and stored at − 80 °C until required.
Egg capsules were retrieved from the substrate by an incision underneath the basal membrane of the capsule wall, to maintain capsule integrity, rinsed in MilliQ filtered water and snap frozen in liquid nitrogen within 5 mL cryo-vials for storage at − 80 °C until required.

MSI pSi substrate fabrication, oxidation and functionalisation.
Monocrystalline (0.008-0.02 Ωcm) antimony doped n-type Si (100) wafers (Silicon Quest International, CA, USA) were cut, methanol sonicated for cleaning and dried prior to substrate fabrication by light-assisted anodic etching 23 . Photopatterned pSi arrays were secured in a custom built Teflon cell in contact with a gold foil anode (Space Products International, CA, USA), with platinum wire (0.5 mm, 99.9%; Aldrich, WA, USA) shaped into a ring acting as a cathode. The teflon cell was then filled with an electrolyte solution of 1:1 hydrofluoric acid (HF): ethanol. The submerged Si surfaces were illuminated using a fiber optic light source passing through a set of two aspheric lenses, f = 80 mm (OptoSigma, CA, USA) for collimation. A 20 mA constant current was then applied across the cell for 2 min via a source meter program, constructed in LabView 6.1 to operate a 2425 current source meter (Keithley, Ohio, USA). Fabricated pSi were washed several times with methanol prior to being dried under nitrogen gas.
Freshly etched pSi were ozone-oxidised with a flow rate of 3.25 g/hr using an Ozone-Generator 500 (Fischer, Germany). After oxidation, pSi were subjected to a second pore broadening etch with 5% HF/ H 2 O for 30 s. The double etched pSi surfaces were ozone oxidised as above. Etched hydroxy-terminated pSi surfaces were then silanized using 80 μ L of neat silane (F5PhPr) for 15 min at 90 °C. Silanized pSi arrays were washed with methanol, dried under nitrogen gas and stored in a dessicator until required.
Tissue sectioning and imprinting for DIOS-MS. Hypobranchial glands (with connected pallial gonoduct) were mounted on cryo-section specimen holders and fixed into place with a minimal amount of embedding medium (Optimum Cutting Temperature Compounds (OCT); Tissue-Tek), on the base away from the target tissue. The hypobranchial gland with connected pallial gonoduct was serially transverse cryo-sectioned (Leica 1800 Cryostat, Leica Microsystems) until the mid-region of the medial hypobranchial gland was exposed. Sections to be imaged contained both medial hypobranchial tissue and attached capsule gland tissue. A 15 μ m thick cryo-section was placed on a glass slide for optical imaging using light microscopy (Zeiss Axio Imager Compound Microscope and Axio Imaging software). The serial 15 μ m thick cryo-section was placed on a functionalised pSi chip for tissue imprinting and kept for 30 minutes at room temperature in a desiccator to promote tissue analyte-surface interaction and extract small molecules by affinity. Imprinted pSi chips were digitally scanned using a conventional desktop scanner (Epson V700 Photo Scanner). Prior to MSI the residual tissue on the pSi surface was removed by immersion in MilliQ at 70 °C for 10 minutes. Tissue removal was aided with a gentle stream of hot water from a pipette, dried and rinsed twice in fresh MilliQ water.
Egg capsules were mounted on cryo-section specimen holders on an anterior-posterior axis and fixed in place with a minimal amount of OCT. Capsules were serially sectioned down to half width, removed from the specimen holder and stamped onto pSi chips for 30 minutes in a desiccator for capsule analyte-surface interaction at room temperature. Imprinted pSi chips were digitally scanned (Epson V700 Photoscanner) and remaining tissue was removed as above.

DIOS-MS and MSI.
Imprinted pSi chips were mounted onto a customised MTP 384 ground steel target plate (Bruker-Daltronics GmbH, Bremen, Germany), secured with conductive carbon tape, and loaded into an Autoflex III TOF/TOF mass spectrometer (Bruker-Daltronics) equipped with a SmartBeam 200 Hz laser. Scanned tissue images, on pSi substrate prior to tissue removal, were loaded into FlexImaging 2.1 (build 25) and aligned with the steel target plate containing the pSi sample based Scientific RepoRts | 5:13408 | DOi: 10.1038/srep13408 on three teaching points. A small margin around each tissue area was allowed for capillary movement of fluids out of the tissue. FlexImaging 2.1 distribution maps were used to control FlexControl 3.3 (build 85) during image acquisition.
Quadratic external calibration of the TOF tube was performed prior to each acquisition using α -cyano-4-hydroxycinnamic acid (CHCA) adducts together with bradykinin (1-7) and angiotensin II, spotted on a free area of the pSi substrate. Specifically, monoisotopic peaks for the calibration range included: Samples were run in reflectron positive mode in the 20-1000 Da range, with a spatial resolution of 100 μ m and medium laser focus, corresponding to a 50 μ m diameter. Mass spectrometer settings: Ion source 1-19.00 kV, ion source 2-16.80 kV, lens− 8.25 kV, reflector 1-21.00 kV, reflector 2-9.40 kV. Data sets were further processed by baseline subtraction (1) and normalised threshold settings (0.5).
Ion peak distribution maps were generated in FlexImaging 2.1 and mass spectra were analysed in FlexAnalysis 3.3. Resulting masses associated with known compounds were assigned colours for visual localisation with reference to histological sections.
To calculate co-localisation of mass spectra across all tissue samples from DIOS-MS imaging, data files were imported into SCiLS Lab (Bremen, Germany) and run through an unsupervised cluster analysis of spatially similar m/z distribution patterns providing a summed spectra of peaks that co-localise within tissue regions. Correlation analysis identified where brominated indole distribution patterns overlapped with murexine based on the reproductive cycle. The peak distribution correlation of representative samples were plotted with corresponding peak intensity for each m/z value.

Extraction and elucidation of brominated indoles.
Fresh hypobranchial glands (6.58 g) were solvent extracted 35  Phenomenex, Lane Cove, NSW, Australia) according to the previously reported elution gradient 36 of acetonitrile in water with 1% formic acid at a flow rate of 300 μ L/min with parallel UV/Vis photo-diode array (PDA) detection at 300 and 600 nm. Electrospray ionisation (ESI-MS) facilitated the identification of brominated indole structures and data were analysed using the Masslynx 4.1 data system (Waters). Retention times were standardised using 40 μ M synthetic 5-bromoisatin (Sigma-Aldrich, technical grade) in acetonitrile, a structural isomer of the Tyrian purple precursor 6-bromoisatin. The identification of brominated indoles was based on peak retention time, expected mass and isotopic clusters for the monoand dibrominated compounds within mass spectra, with reference to previously published structures for this species 26,32,35,51 , listed in Table S1.
Extraction and elucidation of murexine structure. The remaining frozen tissue (1.3 g) from the reproductive female, after MSI tissue collection, was used for extraction of reproductive female-derived murexine to ensure specificity of detection in DIOS-MSI. Murexine was extracted three times in 30 mL volumes of acetone and pooled. The extract was vacuum filtered through a PTFE membrane filter (pore size 0.2 μ m), evaporated to dryness, and washed three times with 10 mL of diethyl ether (Ajax Finechem, AR grade) to remove fats 18 . Total extract was then taken up in 5 mL of acetonitrile (ACN) for thin layer chromatography (TLC) detection 18 , LC-MS and NMR analysis.
TLC was used for the initial detection of murexine, based on previous reports of choline ester compounds in members of the Muricidae 34 . Approximately 5 μ L of extract was collected into glass capillary tubes (32 mm, 0.6 mm i.d., BLU-TIP ® ), spotted on aluminum-backed silica TLC plates (F60, Merck) and separated using an n-butanol-ethanol-acetic acid-water (8:2:1:3) gradient for the mobile phase. Visualisation of the choline ester spots was achieved using Dragondorff reagent (Fluka, Sigma-Aldrich Chemie GmbH) according to the retention times previously reported 34 .
Three replicate 10 μ L subsamples of the capsule gland extract were subjected to ultra-performance liquid chromatography (UPLC)-MS, based on a modified LC-MS protocol for detection of acetylcholine 52 . Separation was provided by an Acquity UPLC system (Waters), 10 μ L injection volume, on a reverse-phase column (Atlantis T3, 3 μ m C 18 , 3 × 100 mm i.d., Waters) using 0.5% formic acid (A) and acetonitrile (B) at a flow rate of 0.5 mL/min (gradient of solvent: 0-10 min, 98% A and 2% B), with UV PDA. ESI-MS detected the murexine structure on the Micromass Quatro micro ™ tandem quadrupole mass spectrometer and MS and UV data was acquired using Masslynx 4.1 data system (Waters). To see the structural features of murexine, in ESI, a scan at 20 V was compared to a collision induced dissociation (CID) scan at 35 V cone voltage (positive ion electrospray, 80 to 500 m/z mass range).
The remaining reproductive female tissue extract was used for 1 H-NMR and 13 C-NMR fingerprinting for identification of reproductive female capsule gland extract components. The bulk of contaminants were removed by solid phase extraction column (Prevail ™ C 18 reversed-phase, 4 mL 500 mg packed bed, Grace, Deerfield, IL, USA), resuspended in methanol-d 4 followed by ACN-d 3 in 5 mm (600 MHz) NMR tubes (Bruker-Daltronics). Chemical shifts were recorded on a 600 MHz NMR spectrometer (Bruker Avance, Karlsruhe, Germany) using an inverse multinuclear probe (5 mm) and a triple resonance HCN probe and referenced to residual solvent peaks. Structural confirmation of individual compounds was elucidated using 1 H-NMR and 13 C-NMR chemical shifts and correlation analysis. Chemical shift assignment of murexine was based on similarity to previously reported 1 H-NMR from murexine (in a mixture of 90% ACN-d 3 ) 53 and 1 H-NMR for tyrindoxyl sulfate and the complex in methanol-d 32 . Reproductive derived murexine was validated against a semi-purified acetone extract from adult hypobranchial gland extractions (106 mg).
Bioassay of murexine in hatching stage D. orbita larvae. To assess the effect of murexine on larval motility, fresh hypobranchial glands were extracted (127 g) using acetone. Fresh glands were extracted overnight, with continuous stirring, in 200 mL of acetone and remaining tissue further extracted in 200 mL overnight. The supernatant was collected, pooled, then vacuum filtered through a PTFE membrane filter (0.2 μ m) and evaporated to dryness. The extract was resuspended in 30 mL of MilliQ filtered water. Fats were removed by three washes with an equal portion of diethyl ether. Murexine was separated from the bulk of contaminants using a solid phase extraction column (Prevail ™ C18 reversed-phase, 4 mL 500 mg packed bed, Grace, Deerfield, IL, USA) and a purified fraction of the choline ester was collected by concentration on a normal-phase SPE column (Alltech ® normal-phase silica column, 4 mL 500 mg packed bed) eluted with successive additions of 10, 30, 40, 50, 60, 70, 80, 95, 99% ethanol. Fractions of 10-30 mL were collected and run on TLC for purity. The fractions containing murexine were evaporated to dryness and resuspended in artificial seawater to collect murexine for bioassays.
Late stage intracapsular D. orbita larvae were subjected to semi-purified murexine at a concentration of 50 ppm and motility was recorded over a 60 min motility assay. Replicate (n = 12) assays were run in 24 well cell culture plates using 4 mL of artificial seawater. Seawater was pipetted onto larvae just prior to the addition of murexine to ensure larvae were actively swimming. Murexine in artificial seawater (0.25 mg in 1 mL) was added slowly to larvae (in two amounts) and swimming activity was assessed by short 30 s video recordings at timed intervals 0, 5, 10, 20, 30 and 60 min. Movement by beating cilia across still shots during short videos were scored as swimming.