Morphogenesis of Ulva mutabilis (Chlorophyta) induced by Maribacter species (Bacteroidetes, Flavobacteriaceae)

The green macroalga Ulva mutabilis Føyn (Chlorophyta) is not able to develop and differentiate into blade, stem and rhizoid cells under axenic conditions, or when its microbiome is not appropriate. Instead, the alga forms callus-like, slow growing structures with colourless protrusions from the exterior cell wall (Spoerner et al. 2012, Wichard 2015). Early experiments of Provasoli (1958) have already pointed out that treatment of Ulva with antibiotics results in abnormal growth. Further experiments examined the role of isolated bacteria in activating developmental and growth promoting traits in Ulva and revealed species-specific interactions, as no combination of bacteria showed the complete recovery of the normal morphotypes (Provasoli and Pintner 1980, Marshall et al. 2006, Spoerner et al. 2012, Wichard 2015). Two bacterial isolates from non-axenic laboratory cultures of U. mutabilis were later found to induce the complete morphogenesis of U. mutabilis, forming a tripartite community (Spoerner et al. 2012, Wichard 2015). The bacterial strains were originally described as Roseobacter sp. strain MS2 and Cytophaga sp. MS6. The strain MS2 alone induces cell division and elongation of the blades but not of the rhizoid, whereas the strain MS6 promotes rhizoid formation and proper cell wall synthesis without cell-wall protrusions (Figure 1; Spoerner et al. 2012). Overall these morphogenesis-inducing bacteria secreted a bouquet (i.e. MS6- and MS2-factors) of still uncharacterised morphogenesis-inducing factors (=morphogens) into the culture medium of Ulva (Spoerner et al. 2012). However, each of these bacteria can be replaced with partly purified compounds extracted from the corresponding bacterial culture medium (Spoerner et al. 2012, Wichard, unpublished results). Morphogens with similar activity to the compounds released by the MS6-strain were also determined in sterile-filtered lagoon water of the Ria Formosa (Portugal) at sampling sites where Ulva sp. was usually abundant. These results indicated a tight interaction of the MS6-factor-producing bacterium with Ulva. In contrast, the MS2-like bioactivity factor was abundant all over the lagoon regardless of the presence or absence of Ulva species (Grueneberg et al. 2016). In this context further phylogenetic analysis classified the MS6-strain to the genus Maribacter rather than to Cytophaga, while the MS2-strain was reclassified to Roseovarius (Grueneberg et al. 2016). The same study has also shown that cultivable bacteria isolated from the surface of Ulva rigida C. Agardh could not replace the eco-physiological functions of the Maribacter sp. MS6. The uniqueness of the MS6-factors seems to be produced by difficult-to-culture bacteria, which may accidentally get lost in the laboratory (Spoerner et al. 2012, Grueneberg et al. 2016).

Figure 1:

Control experiment for complementary activity of morphogenesis-inducing bacteria (reference strains). Axenic Ulva mutabilis sl G mt(+) germlings (A) were inoculated with either Roseovarius sp. MS2 (B) or Maribacter sp. MS6 (C) or both bacteria (D).

Gametophytes were propagated for gamete production and release under laboratory conditions (Wichard and Oertel 2010). Afterwards axenic gametes were prepared according to Wichard (2015). Purified gametes [mating type (+)] were inoculated with selected bacteria (final concentration OD₆₀₀=1×10⁻⁶) in 10 ml Ulva culture medium and kept in the dark for 24 h to let gametes settle on culture tissue flask. After 14 days growth at 18°C and 90–120 μmol photons s⁻¹ m⁻² for 17 h light and 7 h dark, thallus development of 50–70 germlings derived from triplicate experiments was examined with an inverted Leica DMIL LED microscope (Leica, Solms, Germany) equipped with a digital camera (Nikon, Düsseldorf, Germany). The four morphotypes are colour-coded. Black arrows indicate protrusions from the exterior cell wall. Scale bars=100 μm.

Maribacter is a bacterial genus within the Flavobacteriaceae comprising 20 documented species at present. Maribacter are gram-negative, rod shaped cells, which produce non-diffusible yellow to orange pigments. The metabolism is aerobic or facultatively anaerobic (Nedashkovskaya et al. 2004a, 2010, Barbeyron et al. 2008, Lo et al. 2013, Weerawongwiwat et al. 2013, Hu et al. 2015, Jackson et al. 2015). They were found in various marine sources and climatic zones (Cho et al. 2008, Zhang et al. 2009, Tang et al. 2015) and could be isolated from the water column (Yoon et al. 2005, Barbeyron et al. 2008) and the sea sediment (Nedashkovskaya et al. 2004a, Cho et al. 2008). Maribacter strains are often associated with the surface of macroalgae such as Ulva fenestra Postels & Ruprecht (Nedashkovskaya et al. 2004a, 2010, Weerawongwiwat et al. 2013), U. mutabilis (Spoerner et al. 2012) and Polysiphonia (Nedashkovskaya et al. 2007), where they harbour various enzyme activities involved in degradation of, for example, algal polysaccharides (Bakunina et al. 2012).

In this study, we tested whether the Maribacter genus, and possibly closely related genera, could phenocopy Maribacter sp. MS6 in our tripartite model system (Table 1). To have morphogenesis-inducing strains available for bioassays in the long term, cultivable type strains of the cluster Maribacter/Arenibacter/Muricauda and more distantly related flavobacterial type strains were purchased from the German Collection of Microorganisms and Cell Cultures (DSMZ). Twelve different type strains were tested for their MS6-like bioactivity on the morphogenesis of U. mutabilis (Figures 2 and 3). As the same bacteria (MS2- and MS6-strain) induce in combination the predisposed morphotype (Spoerner et al. 2012), either wildtype or “slender”, the simply organised ribbon-shaped developmental mutant “slender”, with a shorter developmental cycle, was preferred for the bioassays.

Figure 2:

Bioassay screening for morphogenetic activity among selected Maribacter strains (rows).

Two-week old germlings are shown. Maribacter strains were tested with axenic Ulva gametes alone (left column: A–F) and in combination with Roseovarius sp. MS2 (middle column: G–L) or with Maribacter sp. MS6 (right column: M–R). Yellow framing (F) highlights an axenic-like development and morphotype with protrusions from the exterior cell wall (black arrow). Purple framing (L) shows cell divisions and blade formation with malformed cell walls (black arrow) indicating an MS2-like morphotype. Red framing (A–E, M–R) shows longitudinal growth and normal cell wall formation similar to the MS6-like morphotype. If the tested strain harboured an MS6-like bioactivity and was inoculated with Roseovarius sp. MS2, the complete morphogenesis was observed (green framing, G–K). Scale bars=100 μm.

Figure 3:

Bioassay screening for morphogenetic activity among selected Flavobacteriaceae (rows).

Two-week old germlings are shown. Selected strains were tested with axenic Ulva gametes alone (left column: A–F) and in combination with Roseovarius sp. MS2 (middle column: G–L) or with Maribacter sp. MS6 (right column: M–R). Black arrows indicate protrusions from the exterior cell wall. Different colours of frames indicate different morphotypes, as explained in Figure 2 and shown for the control experiments in Figure 1. Scale bars=100 μm.

Experiments were started with axenic gametes prepared and treated under strictly controlled conditions (Spoerner et al. 2012). Control experiments have proven that U. mutabilis grown in the absence of symbiotic bacteria developed into callus-like structures with no cell differentiation and a disturbed cell wall synthesis (Figure 1A). By adding Roseovarius sp. MS2 to axenic cultures, cell division leads to blade formation, but cell wall protrusions are still visible (Figure 1B). Under the influence of Maribacter sp., axenic gametes of the “slender” mutant develop into minute short rows of degenerated blade cells with normal cell walls and rhizoid formation (Figure 1C). Maribacter sp. MS6 in combination with Roseovarius sp. MS2 results in the morphogenesis being completely restored (Figure 1D). Using the “slender” mutant of U. mutabilis the bioassay was thus used to determine the bacterial ability which induced proper cell wall formation and completed the intrinsic morphogenesis in combination with Roseovarius sp. MS2 (Figures 2 and 3).

Cross-testing experiments with bacteria from U. mutabilis on Ulva linza L. have already shown that the combination of MS2- and MS6-strains can also induce the morphogenesis of U. linza. However, bacterial morphogens might have an Ulva-species-specific component, for example, to control the accurate formation of algal holdfasts (Vesty et al. 2015).

The Maribacter strains tested in this study were equally able to induce the morphogenesis of U. mutabilis (slender) regardless of their origin. Indeed, Maribacter chungangensis (Weerawongwiwat et al. 2013), Maribacter stanieri (Nedashkovskaya et al. 2010) and Maribacter ulvicola (Nedashkovskaya et al. 2004a), isolated from green seaweed, and Maribacter arcticus (Cho et al. 2008) and Maribacter sedimenticola (Nedashkovskaya et al. 2004a), isolated from marine sediment, all displayed similar morphogenetic activities (Figure 2A–E). Importantly, in combination with Roseovarius sp. MS2, U. mutabilis cell division was enhanced and the complete morphotype was developed (Figure 2G–K). The only exception, Maribacter polysiphoniae (Nedashkovskaya et al. 2007), which had no effect on the morphogenesis of U. mutabilis, was isolated from the red alga Polysiphonia japonica (Figure 2F, L). It is thus tempting to assume that mutualistic interactions have evolved between Ulva and specific Maribacter strains releasing the MS6-factor or factors similar to the MS6-bioactivity. It is noteworthy that the morphogenetic activity of the MS6-factor from Maribacter sp. MS6, induces the morphogenesis of different Ulva strains and species in the same way (Spoerner et al. 2012, Vesty et al. 2015), but it remains to be proven whether the newly-identified bioactive Maribacter strains have a widespread activity on morphogenesis within the order Ulvales. Interestingly, a general bioactivity was suggested for thallusin, the first isolated morphogen, retrieved from an epiphytic bacterium associated with Monostromaoxyspermum (Matsuo et al. 2005).

In order to assess the distribution of morphogenetic activities similar to Maribacter sp. in the Flavobacteriaceae, a broader screening of different type strains isolated from Ulva spp., seawater samples and from a mangrove sediment sample was performed (Figure 3). Muricauda zhangzhouensis (Yang et al. 2013), Pseudozobellia thermophila (Nedashkovskaya et al. 2009) and Arenibacter palladensis (Nedashkovskaya et al. 2006) showed the same morphogenetic activity on U. mutabilis as the Maribacter strains (Figure 3D–F). However, Algibacter lectus (Nedashkovskaya et al. 2004c) and Ulvibacter litoralis (Nedashkovskaya et al. 2004b), isolated from U. fenestra, and Polaribacter dokdonensis (Yoon et al. 2006), isolated from a seawater sample, had no effect on U.mutabilis (Figure 3A–C). Algae inoculated with these strains developed into calli with colourless protrusions from the exterior cell wall (Figure 1A). Moreover, a detailed analysis of the degree of formation of cell wall protrusions (Figure 4) as a result of a lack of MS6-morphogens confirmed the microscopic observations of the development of Ulva juveniles (Figures 2 and 3). All active strains were able to suppress the formation of cell wall protrusions of the inspected germlings in the tested population of Ulva. However, in case of M. ulvicola, half of the investigated individuals still possessed malformed cell walls (Figure 4). This indicates that the concentration of the morphogen released by the bacterium was close to the threshold concentration that is required to induce a proper cell wall formation of all individuals in the bioassay.

Figure 4:

Relative morphogenetic activity of the tested Flavobacteriaceae.

Algae with normal cell wall formation (i.e. without any protrusions) were counted after 14 days of co-cultivation with following bacterial type strains: Maribacter chungangensis, Maribacter arcticus, Maribacter sedimenticola, Maribacter stanieri, Maribacter ulvicola, Maribacter polysiphoniae, Algibacter lectus, Ulvibacter litoralis, Polaribacter dokdonensis, Pseudozobellia thermophila, Arenibacter palladensis, Muricauda zhangzhouensis and the isolated strains Roseovarius sp. MS2 (formerly Roseobacter sp. MS2) and Maribacter sp. MS6 (formerly Cytophaga sp. MS6). Error bars represent standard deviation (n=50–70 individual algae). Roseovarius sp. MS2 and Maribacter sp. MS6 were used as control strains for comparison.

Phylogenetic analysis revealed that the morphogenesis-inducing genera Maribacter, Muricauda, Pseudozobellia and Arenibacter are more closely related to each other than to the other genera tested within the Flavobacteriaceae that were not active (Figure 5). In fact, the inactive M. polysiphoniae displayed somewhat moderate levels of 16S rRNA gene resemblance (94%–96%) with most Maribacter type strains, being placed next to the Maribacter clade in our phylogenetic assessment (Figure 5). However, more exploratory studies of the Flavobacteriaceae and its 100 genera are necessary for a better understanding of the phylogenetic breadth of bacteria releasing MS6-factors. Attempts to isolate and cultivate Flavobacteriaceae species with alternative and specific procedures (Hahnke and Harder 2013) will increase the number of cultivable Flavobacteriaceae available in culture collections and better reflect the distribution and abundance of this family in various marine habitats. In this context, research on specific nutrient requirements holds promise for broadening our bioassay-based surveys of Flavobacteriaceae in the future.

Figure 5:

Phylogenetic tree of strains of Flavobacteriaceae tested for morphogenetic activities. Maximum Likelihood (ML) phylogenetic inference of 16S rRNA gene sequences of the genus Maribacter and related genera in the family Flavobacteriaceae (phylum Bacteroidetes). Maribacter sp. strain MS6 (bold) and all type strains found to be MS6 substitutes in Ulva morphogenesis bioassays are marked in green. Test type strains which did not elicit the MS6 morphotype are marked in dark-red. For taxonomic robustness, the tree was constructed only with high-quality, ≥1300 bp 16S rRNA gene sequences, mostly from type strains. The gene sequence of strain MBIC04683 (349 bp) was added to the tree a posteriori using the ARB parsimony function. Numbers at tree nodes are bootstrap values calculated in ML analysis, and values ≥70% are shown. The unrooted tree is drawn to scale, and the scale bar represents the number of nucleotide substitutions per site (Costa et al. 2013, Keller-Costa et al. 2014). Arrows indicate strains which were identified as morphogenesis-inducing bacteria in previous studies (Matsuo et al. 2003, Spoerner et al. 2012).

Matsuo et al. (2003) have isolated about 1000 bacteria from algae and sponges. Among these isolates, only eight, mainly retrieved from the green algae Ulva and Monostroma, showed morphogenesis-inducing activity on M. oxyspermum. The main activity of these isolates was attributed to the first identified morphogen, thallusin, which was isolated from the epiphytic bacterium YM2-23 (deposition No. MBIC 04683) associated with M. oxyspermum. The compound was also effective in Ulva (Matsuo et al. 2005). Moreover, strain UP7 (deposition No. MBIC 01484) induced growth acceleration and the development of axenic spores of Ulva pertusa Kjellman, Ulva conglobata Kjellman and Ulva intestinalis L. (formerly Enteromorpha intestinalis) (Matsuo et al. 2003). Strain UP7 could even complement the bacteria-induced morphogenesis of U. pertusa (Matsuo et al. 2003). However in this context, it is noteworthy that specimens of the applied axenic cultures of U. pertusa have already looked similar to the phenotype of an axenic culture inoculated with Roseovarius sp. MS2 in U. mutabilis in our current study. Overall, both strains, YM2-23 (MBIC 04683) and UP7 (MBIC 01484), studied by Matsuo et al. (2003) grouped well together with the tested type specimens showing the typical MS6-like bioactivity in this study (Figure 5, highlighted by the arrows).

In summary, we have established a bioassay to screen for complementary morphogenetic activities of bacteria that control normal development of U. mutabilis. As morphogens are biologically active at very low concentrations in Ulva, it is challenging to identify and elucidate their structures. The current study provides resources necessary to screen for shared metabolites between bioactive strains and thus paves the way for the identification of the chemical nature of morphogenesis-inducing factors. In addition, future comparative genomics of morphogen-producing and non-producing Maribacter strains might support the identification of biopathways involved in the biosynthesis of morphogens.