Bacterial induction of metamorphosis in Hydroides elegans
The serpulid polychaete, Hydroides elegans, is a globally distributed member of the warm water fouling community (Bastida-Zavala and ten Hove 2002; Bastida-Zavala and ten Hove 2003; Sun et al. 2018). Larvae of H. elegans rapidly colonize newly submerged surfaces, where they can reach high densities, before being overgrown by other fouling organisms (Figure 1.2). Examinations of the bacterial communities that induce settlement of H. elegans have demonstrated that monospecific biofilms of taxonomically diverse bacteria, including both Gram-positive and Gram-negative species, isolated from wild biofilms, will induce larvae of H. elegans to settle and metamorphose (Unabia and Hadfield 1999; Lau and Qian 2001; Huang and Hadfield 2003; Hung et al. 2009; Freckelton et al. 2017). However, not all strains of bacteria isolated from the natural habitat of the worm have inductive properties (Unabia and Hadfield 1999; Lau and Qian 2001; Lau et al. 2005; Vijayan and Hadfield 2020). In one study, only eight of 18 isolated strains induced settlement and metamorphosis (Unabia and Hadfield 1999). These inductive strains had varied metabolic capabilities and were from diverse bacterial taxa. Due to these differences, the settlement-inducing cues produced by these bacteria are probably different. Due to its global distribution, it is not surprising that H. elegans evolved to respond to multiple bacterial cues as an adaption for colonizing submerged surfaces in harbors around the world.
Once a correct cue has been detected by a larva, metamorphosis begins within a few minutes and proceeds rapidly. As an early part of the metamorphic cascade in H. elegans, larvae secrete a proteinaceous primary tube that secures them to a surface (Figure 1.2; Carpizo-Ituarte and Hadfield 1998) throughout the initial stages of its metamorphosis. We have evidence that attachment to a biofilm increases the attachment strength of the primary tube significantly (Zardus et al. 2008). That is, the primary-tube cements that are secreted by a larva early in its metamorphosis adhere more firmly to biofilmed surfaces than clean ones (Figure 1.2e; Zardus et al. 2008). An interaction between cements in the primary tube and the biofilm may allow metamorphosing larvae to adhere more tightly to surfaces during this critical shift between a planktonic and a benthic lifestyle.
FIGURE 1.2 Settlement and metamorphosis of Hydroides elegans. (a) Competent larva (scale bar = 50 pm), (b) Juvenile worm having completed metamorphosis and secreted a calcified secondary tube (scale bar = 100 pm), (c) Primary and secondary tubes of H. elegans stained with crystal violet. Primary tube is proteinaceous and is secreted shortly after larva contacts inductive bacteria (scale bar = 100 pm), (d) Dense aggregation of H. elegans (scale bar = 2 cm), (e) Primary tubes of H. elegans settled on a natural biofilm have a higher resistance to detachment that those settled on clean glass when subjected to a turbulent-flow shear of 120 Pa for 4 min. (From Zardus, J. D., Nedved. В. T.. Huang, Y., Tran. C„ and Hadfield. M. G. 2008. Biological Bulletin, 214:91-98.)
Approximately eight hours after the induction of metamorphosis, larvae of H. elegans have completed metamorphosis. During this time, juvenile worms have lost several larva-specific organs (e.g., the prototroch) and begun to concentrate calcium carbonate from the surrounding seawater into a gland that secretes a calcified secondary tube (Figure 1.2b). The primary tube remains attached to the posterior end of the primary tube (Figure 1.2c), and thus, as future research, it will be intriguing to determine how bacteria and tube cements are interacting with each other during both the early and later stages of metamorphosis.
Identification of larval metamorphic cues from biofilm bacteria
A strain of the "'(-proteobacterium, Pseudoalteromonas luteoviolacea, isolated from wild biofilms, was shown to induce metamorphosis of H. elegans in single-strain biofilms (Huang and Hadfield 2003). Random transposon-mediated mutagenesis produced two strains of P. luteoviolacea that were non-inductive and allowed determination of a set of genes that is essential for induction. In-frame deletion mutagenesis of these genes revealed their products to be structural elements derived from phage-tail gene sets termed tailocins. At least four of these genes are required for induction of metamorphosis in H. elegans (Figure 1.3; Huang et al. 2012; Shikuma et al. 2014). However, examination of other bacterial species that induce metamorphosis in larvae of H. elegans revealed that induction cannot be entirely explained by the presence of these tailocins (Freckelton et al. 2017; Vijayan and Hadfield 2020). Freckelton et al. (2017) examined other inductive bacteria (one Gram-negative: Cellulophaga
FIGURE 1.3 Negatively stained transmission electron micrographs of bacterial products involved in induction of metamorphosis of larvae of H. elegans. (a) Outer membrane vesicles (OMVs) from Cellulophaga lytica culture (scale bar = 100 nm); (b) OMVs budding from the outer membrane of C. lytica (scale bar = 200 nm); (c) Tailocin aggregate from Pseudoalteromonas luteoviolacea (scale bar = 200 nm); (d) OMVs and individual tailocins from P. luteoviolacea (scale bar= 100 nm). White arrows indicate OMVs.
lytica, and two Gram-positive: Bacillus aquimaris and Staphylcoccus women) and determined that the cassette of genes encoding phage-tail bacteriocins is not encoded in the genomes of these bacteria. More recently, Vijayan and Hadfield (2020) isolated other inductive Pseudoalteromonads that also lacked tailocins.
Cellulophaga lytica was revealed to produce inductive cues by an entirely different mechanism: outer membrane vesicles (OMVs; Figure 1.3; Freckelton et al. 2017). OMVs are ubiquitously produced by Gram-negative bacteria and have frequently been associated with bacterial cell signaling (Kulp and Kuehn 2010; Deatherage and Cookson 2012; Lynch and Alegado 2017). OMVs are produced when small sections of the outer membrane first bulge outward and then pinch off from the bacterial cell. Consequently. OMVs can be highly diverse and include not only the outer membrane, with the bioactive molecules peptidoglycan and lipopolysaccharide (LPS), but also any compounds associated with the periplasmic space directly below the outer membrane. The OMVs thus included a list of possible components, including proteins (membrane bound or soluble), virulence factors, nucleic acids, or peptidoglycan (Deatherage and Cookson 2012).
To gain key insight into the molecular structure of the bacterial inducer produced by C. lytica, its OMV fractions were recently subjected to a battery of enzymatic and chemical treatments, and tested for loss of inductive activity using the settlement bioassay (Freckelton et al. 2020). These experiments asked whether OMV induction of metamorphosis was due to the presence of a protein, nucleic acid, or a lipid (Figure 1.4). OMV-induced metamorphosis was impacted only by treatment with lipases, indicating that either a lipid or an intact membrane is required for OMV induction (Freckelton et al. 2020). Assessment of the bacterial sacculus in combination with interrogation of OMVs with lysozyme, and finally isolation and purification of lipopolysaccharide (LPS), has led us to conclude that LPS is the molecule of interest for C. /y//ca-induced metamorphosis of H. elegans (Figure 1.4).
How variability of inductive bacteria and identified settlement cues relate to variable larval settlement and recruitment
The capacity to induce settlement in one invertebrate species by one strain of a bacterium species does not translate to the same inductive activity by all strains of that bacterium. This is not entirely unexpected, as strain-level differences in symbiotic ability have been observed in other classical symbioses as well (Bongrand and Ruby 2019). One possible explanation for strain specificity in bacteria-eukaryotic interactions is suggested by our recent work identifying the cell-surface lipopolysaccharide (LPS) as the relevant bacterial cueing molecule for larvae of H. elegans. The high level of structural variability of this macromolecule is closely tied to bacterial taxonomic lines and growth conditions (Aucken and Pitt 1993; Schletter et al. 1995; Zhang et al. 2006; Raetz et al. 2007; Leker et al. 2017), revealing that LPS may be responsible for induction of metamorphosis in a broad swathe of marine invertebrates. Most importantly, it would explain the selective settlement of different invertebrates into different habitats. If variations in the structure of LPS are responsible for variable inductive activity of different species and strains of bacteria, it would explain why many—but definitely not all—marine Gram-negative bacteria induce settlement and metamorphosis in the larvae of H. elegans. The taxonomic variability of LPS may
FIGURE 1.4 Lipopolysaccharide is the morphogen produced by Cellulophaga lytica.
(a) Outer membrane vesicles (OMVs) and seawater (SW) controls were exposed to either: DNase (20 U); RNase (20 U); trypsin (1 mg/mL) and Proteinase К (50 U) or trypsin and Pronase E, sonicated and then incubated for 2 hrs at 37°C. Prior to settlement testing, proteases were removed by ultracentrifugation (200.000 g, 4°C. 2 hrs). (b) OMVs and SW controls were exposed to lysozyme (20 U), sonicated, and then incubated for 2 hrs at 37°C. (c) Lipopolysaccharide isolated using the hot phenol method induced metamorphosis in the larvae of H. elegans after 24 hrs exposure. In all cases, metamorphosis was counted after 24 hrs exposure with SW as a negative control, a multispecies biofilm as a positive control, and an untreated OMV experimental control (Freckelton et al. 2020).
also explain why not all strains of a bacterium induce larvae of an invertebrate species to settle (Huang et al. 2012), as well as why larvae may settle selectively in some habitats and not others, even though the same bacterial species is present.
Lipopolysaccharide mediates both symbiotic and pre-symbiotic interactions
LPS was traditionally recognized as the prototypical class of Pathogen-Associated Molecular Patterns (PAMP). PAMPs are molecules that trigger the innate immune response of eukaryotes, and thereby notify host organisms of potential infection. With the recognition of increased benign and beneficial roles for host-associated bacteria, it has been argued that this terminology is restrictive (Koropatnick et al. 2004) and is more accurately portrayed by the term Microbe-Associated Molecular Pattern (MAMP). The ubiquity of a molecule such as LPS means that its detection by a host is entirely context dependent. Indeed, since MAMPs encompass a wide variety of molecular classes, many MAMPs are implicated in both pre-symbiotic associations, such as larval settlement (Tebben et al. 2015), and w'ell-established symbiotic relationships, such as the Hawaiian Bobtail squid Euprymna scolopes and the bacterium Vibrio fischeri (Rader et al. 2012). In light of the growing perspective of benign and beneficial interactions between bacteria and eukaryotes, it is clear that examining the pre-symbiotic dealings between eukaryotes and bacteria will increase our understanding of the evolution of both facultative and obligate symbiotic relationships among animals and microbes.
