Decoding cellular dialogues between sponges, bacteria, and phages

Lara Schmittmann, Martin T. ]ahn, Lucfa Pita, and Ute Hentschel


The evolution of multicellularity has not only enabled the specialization of eukaryote cell types, but also provided stable confined habitats for microbes to engage in symbiotic associations with metazoans. Animal-microbe interactions presented new challenges, such as self/nonself recognition, but also new opportunities that have shaped the evolution and diversification of holobionts. Sponges (Porifera), as one of the most basal animals, provide a fundamental resource to decipher key mechanisms of animal-microbe interactions with implications for our understanding of the interactions between more complex invertebrates/vertebrates and microbes. Significant progress has been made in our comprehension of the metabolic interactions and (meta-)genomic repertoires of sponge symbioses and the reader is kindly referred to several excellent reviews in the field (Taylor et al. 2007; Webster et al. 2016; Pita et al. 2018b). In this chapter, we summarize the current knowledge on cellular dialogues within sponge holobionts by taking a close look at the different players and interactions that make sponges one of the most diverse and successful marine animal groups.

Sponges have a fossil record dating back to ~600 Mya (Yin et al. 2015) and around 9,000 extant species have been described (Van Soest et al. 2012). Despite their high taxonomic diversity, all sponges possess a sessile, filter-feeding lifestyle (exceptions: carnivorous sponges and the pelagic larval phase). Sponges continuously pump water and consume large amounts of microbial cells as well as dissolved organic matter (DOM) (De Goeij et al. 2013). Specialized flagellated cells (choanocytes) capture particles from the surrounding water and transfer them into the sponge interior, the sponge extracellular matrix (mesohyl). Once inside the mesohyl, the particles are digested by phagocytotically active, amoeboid cells (archaeocytes). While seawater bacteria constitute one of the main sponge food sources, the mesohyl also harbors dense bacterial symbiotic communities within the sponge host (Thomas et al. 2016; Moitinho-Silva et al. 2017b).

Comprehensive knowledge on sponge microbiome diversity and functions has been gained from 16S rRNA gene sequencing and high-throughput sequencing technologies such as metagenomics, metatranscriptomics, and single-cell genomics (Horn et al. 2016; Thomas et al. 2016; Moitinho-Silva et al. 2017a; Podell et al. 2019). An extremely high diversity of sponge-associated bacteria has been discovered w'ith members of >60 bacterial phyla so far (Thomas et al. 2016; Moitinho-Silva et al. 2017b). Although a fraction of sponge-associated microbes also occurs in the surrounding environment, each sponge species maintains a specific and stable microbial community (Thomas et al. 2016). Notably, the sponge-specific symbionts are adapted to live within the sponge habitat (Siegl et al. 2011; Jahn et al. 2016), making sponges a refuge for novel biodiversity. The symbionts are either maintained by vertical transmission from adults to offspring (Sharp et al. 2007; Schmitt et al. 2008; Webster et al. 2010; Sipkema et al. 2015; Bjork et al. 2018; Russell 2019) or are acquired horizontally from the seaw'ater (Bjork et al. 2018). Based on the composition and abundance of microbes in their tissues, sponges can be defined in high and low microbial abundance (HMA/LMA) sponges (Gloeckner et al. 2014). Those two lifestyles can be differentiated based on microbial microscopy and taxonomy (Gloeckner et al. 2014; Moitinho-Silva et al. 2017c), as well as on sponge physiology and pumping rates (Weisz et al. 2008).

In contrast to most other animals, microbes in sponges mainly occur extracellularly, in close vicinity to sponge cells (but note exceptions w'here bacteria are enclosed in bacteriocytes e.g. Burgsdorf et al. 2019; Tianero et al. 2019). Bacterial cell densities can reach up to 109 cells per cm3 of sponge tissue and outnumber sponge cell abundance by orders of magnitude (Taylor et al. 2007). Thus, these morphological basal animals constitute one of the most complex holobionts w ith several types of sponge cells and a large diversity of microbial symbiont lineages coexisting in the same matrix. In this chapter, we focus on three main types of interactions: (i) the dialogue between sponge cells and bacteria, (ii) the dialogue between bacterial cells, and (iii) the tripartite interaction between sponge cells, bacteria, and bacteriophages (Figure 4.1). In the first section, we w ill discuss the current knowledge on host mechanisms for microbial recognition as well as microbial features to promote tolerance. In the second section, we will present recent literature on bacteria-bacteria interactions in the context of

Schematic presentation of the cellular interactions on the platform that the sponge mesohyl provides (central circle)

FIGURE 4.1 Schematic presentation of the cellular interactions on the platform that the sponge mesohyl provides (central circle): sponge and bacterial cells (upper panel); bacterial cells (right panel); sponge and bacterial cells with bacteriophages, (left panel; adapted from Jahn, M. T. et al. 2019. Cell Host Microbe 26: 542-550.e5.) LPS, lipopolysaccharides: PPG. peptidoglycan; AHL. N-acyl homoserine lactones; QQ/QSI. quorum sensing/quorum sensing inhibition. (The figure was created with the visualization tool

quorum sensing/quenching. In the third section, we will present a recent discovery on how bacteriophages can foster sponge-bacteria symbiosis. Finally, we will highlight emerging topics in sponge-microbe research.

Host–bacteria dialogue

Already in the early 1980s, Wilkinson et al. (1979) reported that sponges can distinguish between seaw'ater bacteria and their symbionts by feeding tritium-labelled bacteria to sponges followed by high-resolution radioautography of sponge tissue. While most cells of the bacterial symbionts passed through the sponge unharmed and were expelled via the exhalant water, the seawater bacteria (Vibrio alginolyticus) were retained by the sponge and were digested. Later, Wehrl et al. (2007) confirmed that feeding rates of sponges on seawater bacteria are higher than on sponge symbionts. How does the sponge differentiate between food bacteria and symbionts?

To address this question in an experimental way, differential gene expression analyses was used to characterize the molecular response of sponges towards microbial elicitors. Two sponges that are representatives of the HMA/LM A dichotomy (Aplysina aerophoba, HMA, Dysidea avara, LMA) were exposed to a cocktail of lipopolysaccharide and peptidoglycan as signals (Pita et al. 2018a). We hypothesized that the different microbial densities in these sponges would affect the host’s responses towards bacterial elicitors. Both species responded to microbial stimuli by increasing the expression of a subset of immune receptors (such as NLRs in D. avara, SRCR and GPCRs in A. aerophoba) and activating kinase cascades likely yielding apoptotic and phagocytotic processes (Figure 4.2). Moreover, the magnitude of the transcriptionally- regulated response (in terms of number of differentially expressed genes) was more complex in A. aerophoba (HMA) than in D. avara (LMA). We propose that the HMA species requires a more fine-tuned regulated response to deal with conflicting signals coming from the microbial stimuli versus those from the symbionts.

Other studies support the role of the sponge immune system in the crosstalk with microbes. The sponge Petrosia ficiformis displayed an increased expression of a gene containing the conserved SRCR domain when living in symbiosis with a cyanobacterium, in comparison to the aposymbiotic status (Steindler et al. 2007). In juvenile Amphimedon queenslandica, bacterial encounter involved regulation of

Differentially expressed genes in one LMA

FIGURE 4.2 Differentially expressed genes in one LMA (Dysidea avara. left side) and one HMA (Aplysina aerophoba. right side) sponge after exposure to bacterial elicitors (LPS and peptidoglycan). (Adapted from Pita et al. 2018a. Sci. Rep. 8: 1-15.) The upregulated immune receptors with characteristic, conserved domains are depicted, as well as additional regulated functions (up- and/or downregulation represented by arrows).

SRCR-containing genes, but the downstream signaling response differed depending on the origin of the bacteria (Yuen 2016). In particular, the transcription factors FoxO and NFnii were upregulated upon exposure to own symbionts, but not to a bacterial fraction from another sponge species (Yuen 2016). Finally, the components of the TLR pathway such as MyD88 were activated in response to microbial signals in different sponge species (Wiens et al. 2005; Yuen 2016).

Sponge immune receptors

The sponge cellular immune response was studied already in the 19th century by Nobel laureate Elias Metchnikoff and colleagues (Metchnikoff 1893). While these first studies were not focused on the sponge’s response to microbes, the observed cell behaviors suggested that two cell types are the key players for mediating the interactions with microbes: choanocytes (representing the first barrier for external microbes) and archeocytes (representing the patrol of the sponge matrix). Despite this promising takeoff, our understanding of sponge cellular immunity is still at the beginning. The publication of the first sponge genome, that of the Great Barrier Reef species Amphimedon queenslandica, brought a complex and expanded repertoire of immune receptors into light (Srivastava et al. 2010). This repertoire included several extracellular (i.e., scavenger receptor cystein-rich, SRCR, domain), membrane-bound (immunoglobulin-like domains), and intracellular (NOD-like receptor, NLR. domains) receptors (Srivastava et al. 2010; reviewed in Hentschel et al. 2012). These gene families were identified at sequence level because the domains (sequence patterns) were arranged in a particular architecture that is conserved from early metazoans to vertebrates. The conserved gene structure suggests a conserved function, which is not always clear. For example. Toll-like receptors (TLRs) are transmembrane receptors characterized by several extracellular leucin-rich repeat (LRR) motifs and an intracellular Toll/interleukin-1 receptor (TIR) domain. In vertebrates, the extracellular LRR motifs recognize the ligand (e.g. bacteria) and transduce the signal via the TIR domain. All components of the signaling cascade induced by TLRs are present in A. queenslandica, but not the conventional TLR. The genome of A. queenslandica contained a TIR domain-containing gene, homolog of the TIR-domain in vertebrates TLRs, which was combined with extracellular Ig domains rather than LRR motifs. Therefore, its role in bacteria recognition remains to be validated.

A striking feature of the A. queenslandica genome was the high diversification of two other immune receptor families (Hentschel et al. 2012). The NLR family are defined according to the presence of a nucleotide-binding domain combined with a leucine-rich repeat domain (Ting et al. 2008). The genetic animal models Drosophila melanogaster and C. elegans lost this receptor family and, therefore, it was long thought that these receptors had their origin in the teleost. An interesting feature of A. queenslandica NLRs is their enormous diversity: A. queenslandica genome comprises 135 genes, which is in stark contrast to 20 NLR genes in humans (Yuen et al. 2014). Similarly, the family of scavenger receptors cysteine rich (SRCRs) in A. queenslandica is also highly expanded (ca. 300 genes) when compared to vertebrates (e.g. 16 genes in humans) and other invertebrates (Buckley and Rast 2015). All cell types in A. queenslandica adults express genes with SRCR domains, but they are significantly enriched in choanocytes (Sebe-Pedros et al. 2018). The evolutionary forces driving a high diversity of pattern recognition receptors (PRRs) are suggested to be related to the specific recognition of a wide variety of microbial compounds and has been proposed as a mechanism for specificity in sponges and other invertebrates (Schulenburg et al. 2007; Messier-Solek et al. 2010; Buckley and Rast 2015; Degnan 2015).

Since the publication of A. queenslandica genome in 2010, further sponges have been sequenced at genome and transcriptome level. Most of these reference genomes and transcriptomes are incomplete, yet they are adding new knowledge to our understanding on sponge molecular repertoire of immunity. Poriferan TLR/IL-lR-like receptors as well as their downstream signaling cascades were detected in other sponge genomes and transcriptomes (Riesgo et al. 2014; Germer et al. 2017; Pita et al. 2018a). These new data confirmed the complex and expanded repertoire of Poriferan immune receptors, notably NLRs and SRCRs (Germer et al. 2017; Pita et al. 2018a). However, there are also differences among sponge species that may be related to their symbiotic status (HMA or LMA). Ryu et al. (Ryu et al. 2016) detected different enrichment in immune domains depending on symbiont densities within the mesohyl when comparing the genomes of LMA sponges A. queenslandica and Stylissa carteri versus the HMA sponge Xestospongia testudinaria. Along similar lines, we detected 80 bona fide NLRs in the reference transcriptome of the LMA sponge D. a vara: whereas, using the same experimental setup, we found only one bona fide NLR gene in A. aerophoba (HMA) reference transcriptome (Pita et al. 2018a). The reference transcriptome of the HMA sponge Vaceletia sp. contained no NLR (Germer et al. 2017). These distinct signatures in the HMA and LMA genomic repertoires of immune receptors support the different evolutionary trajectories imposed by the symbiosis with microbes.

Apart from the above-mentioned receptors, lectins are also diversified in sponges. This class of soluble or membrane-bound proteins recognizes carbohydrates and mediates cell adherence, self/nonself recognition, and symbiotic relationships (Brown et al. 2018; Dinh et al. 2018). A few dozen lectins from sponges are known so far of which some may aid in bacterial recognition by responding to carbohydrates from gram-positive (peptidoglycan) as well as from gram-negative (lipopolysaccharides) bacteria (reviewed in Garderes et al. 2012). In a growth assay, a lectin from Halichondria panicea stimulated bacterial proliferation of sponge derived bacterial strains (Muller et al. 1981).

Microbe associated molecular patterns (MAMPs)

Immune receptors detect microorganisms via molecules that are present in prokaryotes but absent in eukaryotes, the so-called “pathogen associated molecular patterns” (PAMPs). PAMPs include components of bacterial cell walls and membranes such as peptidoglycans (PPGs) of gram-positive bacteria and lipopolysaccharides (LPS) from gram-negative bacteria. If such PAMPs are recognized by an immune receptor, they will trigger a signaling cascade yielding the elimination of the microbial invader (e.g. via phagocytosis). It was soon recognized that PAMPs are not exclusive for pathogens, but are also present in bacterial symbionts (Koropatnick et al. 2004), leading to the alternative use of the term “microbe associated molecular patterns”, or MAMPs. Therefore, host recognition of microorganisms must be specific enough to yield an appropriate response, which may be either to eliminate or tolerate microorganisms.

We have limited experimental evidence for which symbiont MAMPs may be recognized by poriferan receptors. However, genetic features enriched or depleted in sponge-associated versus free-living bacteria revealed interesting patterns. In this context, it is remarkable that sponge-associated microbes lack flagella (Siegl et al. 2011). Flagellin is known as a powerful immune stimulator that initiated, for example, signal transduction mediated by TLR5 receptors (Hayashi et al. 2001). The absence of flagella (thus, flagellin) could allow microbes to evade host immunity and persist within the sponge holobiont. On a different note, a common sponge symbiont, “Candidatus Synechococcus spongiarum” (Cyanobacteria) presents a modified O-antigen in its LPS, as compared to free-living Synechococcus relatives (Burgsdorf et al. 2015). This modification could represent another mechanism for recognition. Thus, modifications in MAMP structure could help microbes to escape recognition as “nonself’ by the host. Altogether, these studies provide evidence of adaption to symbiosis in both the host and the microbial side.

< Prev   CONTENTS   Source   Next >