*4.1. Biosynthesis of* D. orbita *Brominated Indoles*

Basic gaps in our understanding of the gene and protein machinery that underlie Tyrian purple biosynthesis allow for new and exciting discoveries on biohalogenation and methane thiol incorporation into secondary metabolites. Tyrian purple is thought to be synthesized from dietary derived tryptophan in the Muricidae [4,43]. Tryptophan has been detected in the hypobranchial secretory cells of several Muricidae species [14,15]. It is particularly prevalent in the rectum of *D. orbita*, which is embedded in the hypobranchial gland [15,18]. Although it remains unclear how this amino acid is specifically converted into tyrindoxyl sulfate in Muricidae, several biosynthetic enzymes are likely to be involved (Table 3). Tryptophanase is typically involved in converting tryptophan to indole in bacteria, which can then be converted to indoxyl sulphate by a mono- or di-oxygenase enzyme system [95]. The specific enzymes involved in adding methane thiol groups onto the indole ring are unknown, but may involve some sulfur transferase and reductase enzymes (Table 3). Further investigation of these enzymes could uncover novel mechanisms for biotransformation in secondary metabolism.

**Table 3.** Biosynthetic enzymes proposed to be involved in the production of Tyrian purple precursors. The order of enzyme reactions generating the bromo and methylthio derivatives is not known.


Specific incorporation of bromine into the 6-position of the indole ring is an unusual feature found in several bioactive marine indoles [96]. Since bromination more ready occurs in the 4 or 7 position, this strongly implies enzymatic bromination during the biosynthesis of tryindoxyl sulfate. Several regiospecific halogenases have been previously identified from bacteria, which are highly substrate specific for tryptophan [97,98]. However, the tryptophan-halogenases reported to date all appear to utilise chlorine over bromine. Jannun and Coe [99,100] reported bromoperoxidase activity in homogenates from hypobranchial glands of *Murex trunculus* and recent histochemical studies by Westley have confirmed the bromoperoxidase activity in *D. orbita* hypobranchial gland tissue [14,18]. A range of bromoperoxidase enzymes have been previously identified from marine algae, bacteria and fungi [101,102], but these do not generally appear to be substrate or regiospecific in their brominating activity. In a preliminary attempt to identify the bromoperoxidase gene from *D. orbita*, Laffy [21] developed primers from consensus sequence regions after multiple sequence alignment of 11 bromoperoxidases available on genebank (4 algal and 3 bacteria). No PCR products were amplified with these primers, despite successful positive controls. This indicates the muricid enzyme shares low sequence conservation at these primer sites or may be a distinct type of brominating enzyme with specificity for 6-bromination of tryptophan/indole for Tyrian purple biosynthesis. Bromination of indole derivatives has been shown to increase their biological activity [80,103] and the identification of novel halogenation strategies will facilitate alternative mechanisms for generating halogenated biologically active molecules for drug development [97].

The conversion of tyrindoxyl sulfate salt to tyrindoxyl and ultimately Tyrian purple requires an aryl sulfatase enzyme [8]. Histochemical studies have confirmed the release of aryl sulfatase on the epithelium of the hypobranchial gland of *D. orbita* [14,18]. Preliminary analysis of the transcriptome from *D. orbita* hypobranchial gland was successful in detecting the aryl sulfatase gene [5,21,50] and full length sequencing has confirmed the molluscan origin of this enzyme [21]. No other biosynthetic genes were identified in this mollusc transcriptome library, although there is good support for a primary role of the hypobranchial gland in protein synthesis, post translational modification and transport [5, 21,25,50]. A large number of unidentified sequences were also present in the hyporbanchial gland transcriptome, suggesting possible novel genes, although the suppressive subtractive hybridization technique used only produces short reads, which may have reduced the chance of successful matches to conserved areas of the open reading frames. Nevertheless, there remains a good possibility for the discovery of novel biosynthetic enzymes from *D. orbita*.

#### *4.2. Biodistribution of the Secondary Metabolites in* D. orbita

Knowledge of the anatomical distribution of natural products is essential for understanding the biosynthesis process and optimal methods for extraction. On a basic level, different tissues can be dissected and extracted to determine which produce and/or store the secondary metabolites. This approach was applied to establish the distribution of Tyrian purple pigments and precursor compounds in the male and female reproductive organs of D. orbita [13]. These compounds were found throughout the female pallial gonoduct [13], with significant quantities in the capsule gland, which lies adjacent to the hypobranchial glands, thus providing evidence for maternal investment of these compounds in the egg masses of D. orbita [4,12]. Despite the production of more oxidized compounds in the male prostate gland, relative to the female gonoduct [13], the presence of significant quantities of these brominated compounds in the males suggests that these compounds are not exclusively produced for defense of the egg masses and likely play some role in the adult life history.

Histochemical techniques for proposed biosynthetic constituents can further aid in establishing the primary metabolic origin of natural products and sites of active biosynthesis [14]. Histomorphological properties of biosynthetic tissues may also reveal regulatory mechanisms, modes of transport, storage and secretion, while histological examination can reveal the presence of potential symbionts (see Section 4.3). The hypobranchial glands of D. orbita show remarkable complexity, with seven distinct types of secretory cells located on the epithelial cell surface [15]. At least two cell types appear to be specifically associated with Tyrian purple synthesis. A subepithelial vascular sinus occurs between the hypobranchial gland and gonoduct, surrounding the rectum and rectal gland [15]. However, there appears to be no direct anatomical mechanism for the transfer of precursors to the gonoduct, suggesting that the compounds are independently synthesized in the reproductive organs. This is supported by the presence of bromoperoxidase and aryl sulfatase activity in the female egg capsule gland [14,18].

Histochemical examination of the biosynthetic enzyme activity and precursors in the hypobranchial glands of *D. orbita* by Westley [18] has further revealed that tyrindoxyl sulfate is biosynthesized through the post-translational bromination of dietary-derived tryptophan, within two discrete sites by two distinct modes. Regulated synthesis occurs on the surface of the lateral hypobranchial epithelium, while the subepithelial vascular sinus of the medial hypobranchial gland appears to constitutively synthesize these compounds. Aryl sulfatase is stored in adjacent supportive cells and exocytosis onto the epithelium surface appears to be regulated [18]. The distinct distribution and regulated activity of aryl sulfatase and bromoperoxidase implies *D. orbita* has evolved the capacity to control the release of bioactive indoles and choline esters. This histological evidence provides further support for a naturally selected role of these secondary metabolites in the life history of the mollusc.

More recently, mass spectrometry imaging (MSI) using desorption/Ionization on porous silicon (DIOS) and nanostructured initiator mass spectrometry (NIMS) was applied to examine the biodistribution of secondary metabolites in *D. orbita* tissues [52]. MSI of biological tissues is becoming a popular tool for biodistributional studies of proteins and pharmaceuticals. However, standard Matrix Assisted Laser Desorption/Ionization Mass Spectrometry (MALDI MS) MSI is challenging for secondary metabolites with low molecular weight due to intense matrix signals, interfering with the detection of signals from the less abundant target compounds. Due to the broad range of polarities in the brominated indoles, *D. orbita* hypobranchial gland chemistry proved to be a good model system for "proof of principle" of a new technique involving direct tissue stamping onto porous silicon and NALDI targets [52]. Ongoing research using this technique is providing interesting insights into the distribution of choline esters and changes the secondary metabolite profile over the reproductive season [104]. Mass spectrometry imaging could also be applied to examine the biodistribution of the bioactive compounds in preclinical trials, as previously done using MALDI with pharmaceutical compounds in rodent models [105].

#### *4.3. Microbial Symbionts*

Tyrian purple, a uniquely marine metabolite, is the brominated derivative of the blue dye indigo, derived from plants in the genus *Isatis* and a range of bacteria [106,107]. This appears to be an interesting case of convergent evolution, although the potential role of bacteria in the production of Tyrian purple precursors is yet to be ruled out. To date it has been assumed that muricid molluscs themselves are responsible for the biosynthesis of Tyrian purple [4]. However, over the last decade there has been increasing recognition for the key role of microbial symbionts in the biosynthesis of marine natural products [108]. The rectal gland, which is embedded in the hypobranchial gland of *D. orbita*, contains an abundant supply of the tryptophan precursor and also appears to be associated with bromoperoxidase activity [14,18]. Bacteria have been observed within specialized invaginations of the rectal gland in the muricid *Nucella lapillus* [109]. The positive identification of biosynthetic bacterial symbionts involved in Tyrian purple precursor production would present a paradigm shift, providing new options for large scale sustainable production of these bioactive metabolites and valuable pigments.

Preliminary attempts to culture the bacteria from *Dicathais orbita* using standard techniques have isolated only one species from the hypobranchial gland and three from the rectal gland, compared to 35 from nonbiosynthetic tissues [30]. The sole bacterium isolated from both of these biosynthetic organs was positive for indole production, suggesting a possible role in Tyrian purple synthesis, although further chemical analysis of the culture supernatant is required. It is also possible that the diversity of bacteria in these biosynthetic organs has been underestimated due to specific environmental requirements for growth. The high concentration of mercaptans, such as dimethyl disulfide, in the hypobranchial gland is likely to create a reducing environment [13]. Furthermore, the production of Tyrian purple precursors in culture must require sufficient bromine ion availability. Therefore, a range of novel culture conditions may be required to facilitate the growth and secondary metabolism of Muricidae symbionts. Considering that by far the majority of microorganisms can not be easily cultured [110], the application of culture techniques alone may not be sufficient to identify the diversity of microbial symbionts in *D. orbita*. Metagenomic-based approaches have provided evidence of a microbial origin for several metabolites produced by marine invertebrates [108] and have been successfully applied to the identification of indigo producing bacterial strains in soil [107].

Recent histological and genetic studies have also revealed the presence of ciliate protozoans within the hypobranchial glands of *D. orbita* [21,50]. These ciliates are most likely feeding on bacteria on the epithelial surfaces and interstitial spaces. At present, it is unclear whether these ciliates are pathogens, symbionts or just facultative opportunists. The ciliates do not seem to be directly involved in the production of Tyrian purple based on a lack of histological correlation in the location of the ciliates [21], compared to the biosynthetic enzymes and precursor compounds [18]. However, the abundance of the ciliates does increase towards the reproductive season [21], which correlates with an increase in biosynthetic activity and indole precursor storage prior to spawning [4,12,18,84]. This suggests that the brominated indole precursors could be involved in regulating the activity and/or abundance of ciliates in *D. orbita*. The secondary metabolites from *D. orbita* have not yet been tested for anti-protozoan activity, however a number of other indoles are known to possess anti-parasitic activity [96].

## *4.4. Sustainable Supply*

Tyrian purple is the world's most expensive colorant (1 g = 2439.50 EUR) [111], and is currently extracted from *Purpura lapillus* (10,000 adult snails for 1 g) and South American Muricidae considered at risk from over fishing [89]. The bioactive properties of the brominated indole precursors and the potential for nutraceutical development from the bioactive extracts, provides a further incentive for large-scale sustainable supply. Ecological and life history studies on *D. orbita* [24,31,33,37–39,41,112] contribute to our ability to effectively monitor the population size and recruitment potential of this species. In fact *D. orbita* has been used as a model species for estimating population size [113,114] and for monitoring TBT pollution in the Australian marine environment [32,115,116]. However, as top invertebrate predators, Muricidae molluscs are susceptible to population crashes and the persistence of imposex in some populations further increases their susceptibility to over harvest.

Some progress has been made towards the larval culture [41] and sea-based polyculture of *D. orbita* on abalone farms [40]. However, it has not yet been possible to close the life cycle of this species due to the long planktotrophic (feeding) larval stage and lack of known cues for settlement and metamorphosis [41,117]. Nevertheless, progress has been made towards understanding the growth rates and dietary preferences of the juvenile snails [44,118]. Furthermore, Noble *et al*. [112] have established that it is possible to obtain the bioactive indole precursors from a mucus secretion of *D. orbita*, which offers the potential for non-lethal harvest.

Although generally not suitable for nutraceuticals, chemical synthesis of bioactive metabolites is generally the preferred option for pharmaceutical supply [119]. This can be efficiently achieved for 6-bromoisatin [20,80] and the choline esters [16,62]. These well known molecules can not be patented, but nevertheless provide interesting leads for the chemical synthesis of a range of structural analogues [16,72,79,80,103], thus permitting the assessment of structure activity relationships. Some bioactive marine metabolites are too difficult or expensive to chemically synthesize and previous attempts to chemically synthesis the anticancer precursor of Tyrian purple, tyrindoleninone, have been unsuccessful [16,20]. This is partly due to nonspecific bromination favoring the 5 or 7 position on the indole ring, thus generating low yields for 6-bromoindole derivatives. However, a greater problem occurs in relation to the addition of a methane thiol group at position 2, due to uncontrollable rapid oxidation to 6-bromoisatin. Consequently, tyrindoleninone is not optimal for pharmaceutical development, and holds better potential for human health applications if incorporated into nutraceutical extracts.

The identification of biosynthetic bacteria, enzymes and gene clusters involved in Tyrian purple production could have important implications for application in sustainable production of *D. orbita* brominated indole derivatives, as well as the bioengineering of novel compounds through recombinant expression. Identification of bacterial symbionts that can produce tyrindoxyl sulfate would facilitate the large scale sustainable production of bioactive brominated indoles and Tyrian purple, assuming these bacteria can be cultured. Over the last decade, there have been increasingly frequent reports of gene clusters or gene cassettes for the biosynthesis of marine natural products [108]. Identification of the full gene cluster associated with tyrindoxyl sulphate biosynthesis in Muricidae would open up the potential for recombinant expression of the entire pathway in an heterologous host. This could also facilitate the rational engineering of new metabolites using combinations of enzymes from distinct biosynthetic pathways, which is an important goal for future drug development [98].
