Macroalgal Morphogenesis Induced by Waterborne Compounds and Bacteria in Coastal Seawater

Axenic gametes of the marine green macroalga Ulva mutabilis Føyn (Ria Formosa, locus typicus) exhibit abnormal development into slow-growing callus-like colonies with aberrant cell walls. Under laboratory conditions, it was previously demonstrated that all defects in growth and thallus development can be completely abolished when axenic gametes are inoculated with a combination of two specific bacterial strains originally identified as Roseobacter sp. strain MS2 and Cytophaga sp. strain MS6. These bacteria release diffusible morphogenetic compounds (= morphogens), which act similar to cytokinin and auxin. To investigate the ecological relevance of the waterborne bacterial morphogens, seawater samples were collected in the Ria Formosa lagoon (Algarve, Southern Portugal) at 20 sampling sites and tidal pools to assess their morphogenetic effects on the axenic gametes of U. mutabilis. Specifically the survey revealed that sterile-filtered seawater samples can completely recover growth and morphogenesis of U. mutabilis under axenic conditions. Morphogenetic activities of free-living and epiphytic bacteria isolated from the locally very abundant Ulva species (i.e., U. rigida) were screened using a multiwell-based testing system. The most represented genera isolated from U. rigida were Alteromonas, Pseudoalteromonas and Sulfitobacter followed by Psychrobacter and Polaribacter. Several naturally occurring bacterial species could emulate MS2 activity (= induction of cell divisions) regardless of taxonomic affiliation, whereas the MS6 activity (= induction of cell differentiation and cell wall formation) was species-specific and is probably a feature of difficult-to-culture bacteria. Interestingly, isolated bacteroidetes such as Algoriphagus sp. and Polaribacter sp. could individually trigger complete Ulva morphogenesis and thus provide a novel mode of action for bacterial-induced algal development. This study also highlights that the accumulation of algal growth factors in a shallow water body separated from the open ocean by barrier islands might have strong implications to, for example, the wide usage of natural coastal seawater in algal (land based) aquacultures of Ulva.


Metagenomic DNA extraction
Lagoon water and algal surface samples were collected for PCR-DGGE analysis to determine (i) whether epiphytic bacterial communities from tidal pools substantially differ from those of the surrounding pool water and (ii) whether PCR-DGGE bands from the algal fingerprints matched the electrophoretic mobility of bacterial strains previously isolated from Ulva mutabilis (see main text).
In total, six algal samples were taken from various tidal pools for this purpose, namely Ulva rigida (n=2), Fucus vesiculosus (n=1), Blidingia sp. (n=2) and Zostera noltii (n=1). Algae were taken to a laminar flow cabinet and one piece in the range of 1 × 1 to 2 × 2 cm per specimen (about 10 mg fresh weight) were transferred with sterile tweezers into 10 mL of sterile Ulva culture medium (UCM [1]), in which they were incubated for 5 min while shaking gently every minute. The washing step was repeated once with fresh sterile UCM. The washed pieces of the thallus were then swabbed with sterile swabs (Omni Swab, Whatman Bioscience, USA). The swab heads were transferred into a sterile 2 mL polypropylene microcentrifuge tube (Eppendorf, Wesseling-Berzdorf, Germany) and frozen at -80°C until DNA extraction. Seawater samples (50 mL) were passed through a sterile paper filter to remove bigger particles, organisms and debris. The sample was then passed through a sterile filter (0.2 µm, Isopore TM filter, type GTTP, Millipore, Schwalbach, Germany) using a polysulfone syringe filter holder assembly (Nalgene Nunc, USA). The filter was subsequently transferred with sterile tweezers into a sterile 2 mL polypropylene microcentrifuge tube and stored at -80 °C until DNA extraction. Metagenomic DNA was extracted from bacteria collected on filters (50 mL water samples) or on swabs (Ulva 3 surface) with the QIAamp DNA Mini Kit (Qiagen, Hilden, Germany) following the company's protocol for DNA purification from buccal swabs modified with an additional lysis step for gram-positive bacteria. The extracted DNA was finally eluted in 50 µL Tris-Cl / EDTA (10 mmol L -1 / 0.5 mmol L -1 , pH = 9) and stored at -80 °C.

PCR-DGGE bacterial community fingerprinting
Cultivation-independent analysis of bacterial community structure in Ulva surface and lagoon water samples was performed by PCR-DGGE fingerprinting of 16S rRNA gene fragments amplified from metagenomic DNA. To this end, the primer pair 358fGC (cgc ccg ccg cgc gcg gcg ggc ggg gcg ggg gca cgg ggg gcc tac ggg agg cag cag) and 907rM (ccg tca att cmt ttg agt tt) [2,3]

PCR-DGGE
The most abundant Ulva species of the survey in the Ria Formosa, U. rigida, [6] was used along with Fucus vesiculosus, Blidingia sp. and Z. noltii to determine whether epibacterial communities from tidal pools differed from the surrounding bacterioplankton.
To this end, 16SrRNA gene fragments PCR-amplified from epiphytic community metagenomic DNA were profiled by PCR-DGGE and compared with those obtained from water samples in the vicinity of the algae (Fig. S1). Thus, epiphytic and free-living bacteria were directly explored by PCR-DGGE without any bias towards cultivatable bacteria, providing a snapshot of bacterial community structuring in the surveyed samples. By means of multidimensional scaling analysis of the similarity matrix of PCR-DGGE band patterns, differences were observed between epibacterial communities on the macroalgae and freeliving bacteria in their direct surroundings. The PCR-DGGE profiles retrieved from the seawater of tidal pools and the nearby main channel of the lagoon (Fig. S1A, W2-W8) showed high similarity and grouped closely in a principal coordinate analysis (PCO) (Fig.   S1B). In contrast, the PCR-DGGE profiles of swabbed surface bacteria revealed a high diversity at the community level, even from the surfaces of U. rigida collected from the same environment (Fig. S1A, A10, A11), but were clearly separated from the water body profiles illustrated by the PCO analysis and confirmed by an additional supervised discriminant analysis (DA) with two groups (i) water samples (n = 7) and (ii) algal surface isolates (n = 6). The overall misclassification error of the cross-validation of the DA was only 8% indicating the significant differences between both groups. 6 We could not recognize PCR-DGGE bands from either free-living or algal surface communities that unequivocally matched the electrophoretic mobility of bands amplified from isolates retrieved from the surface of lab-cultured U. mutabilis. Therefore, to verify if morphogenesis-inducing bacteria could be isolated from their natural habitat (the Ria Formosa lagoon) and to determine whether they were phylogenetically similar to our MS2 and MS6 strains, we isolated, cultivated, and tested naturally occurring Ulva bacterial epibionts using the "Ulva bioassay array" for morphogenesis assessment (see main text).