Microbial diversity of biofilm communities in microniches associated with the didemnid ascidian Lissoclinum patella

Abstract

We assessed the microbial diversity and microenvironmental niche characteristics in the didemnid ascidian Lissoclinum patella using 16S rRNA gene sequencing, microsensor and imaging techniques. L. patella harbors three distinct microbial communities spatially separated by few millimeters of tunic tissue: (i) a biofilm on its upper surface exposed to high irradiance and O(2) levels, (ii) a cloacal cavity dominated by the prochlorophyte Prochloron spp. characterized by strong depletion of visible light and a dynamic chemical microenvironment ranging from hyperoxia in light to anoxia in darkness and (iii) a biofilm covering the underside of the animal, where light is depleted of visible wavelengths and enriched in near-infrared radiation (NIR). Variable chlorophyll fluorescence imaging demonstrated photosynthetic activity, and hyperspectral imaging revealed a diversity of photopigments in all microhabitats. Amplicon sequencing revealed the dominance of cyanobacteria in all three layers. Sequences representing the chlorophyll d containing cyanobacterium Acaryochloris marina and anoxygenic phototrophs were abundant on the underside of the ascidian in shallow waters but declined in deeper waters. This depth dependency was supported by a negative correlation between A. marina abundance and collection depth, explained by the increased attenuation of NIR as a function of water depth. The combination of microenvironmental analysis and fine-scale sampling techniques used in this investigation gives valuable first insights into the distribution, abundance and diversity of bacterial communities associated with tropical ascidians. In particular, we show that microenvironments and microbial diversity can vary significantly over scales of a few millimeters in such habitats; which is information easily lost by bulk sampling.

Figure 1

The tropical didemnid ascidian L. patella in its natural habitat. (a) Inner coral reef crest at low tide on Heron Island, QLD, Australia. (b) Specimen of L. patella found nested within dead and living coral branches. (c) Deep-green specimen of L. patella, where the green coloration originates from its obligate symbiont Prochloron spp. that resides in the cloacal cavities of the ascidian. (d) Cross-section of L. patella. Note the thick reddish biofilm covering the underside of the tunic and the green Prochloron cells in the cloacal cavity of the ascidian. ‘Surface', ‘cloacal cavity' and ‘underside' denote the sites were samples were taken for subsequent analysis in this study.

Figure 2

(a) Typical biofilm found on the underside of the tropical ascidian L. patella. (b) Cross-section of the same ascidian showing filamentous cyanobacteria (FCy) on the surface, part of the cloacal cavity containing green Prochloron cells (Pro) and the animal zooid. (c) Color-coded composite images of hyperspectral image stacks taken from the biofilm displayed in a, red quantifies the absorption at 710 nm (specific for Chl d), whereas green quantifies absorption at 675 nm (specific for Chl a). (d) Hyperspectral composite images color coded to quantify absorption at 560 nm (specific for phycoerythrin); images were taken from the same biofilm as displayed in b. The numbers in both c and d denote specific areas of interest, exhibiting the reflectance spectra shown in e and f.

Figure 3

Microenvironments and photosynthetic activity in L. patella and its associated biofilm communities. (a) Cross-section of L. patella. Colored dots indicate the different layers of the animal, where measurements where performed. Red denotes measurements taken just below the surface, green just below the cloacal cavity, and blue just below the underside of the ascidian. (b) Oxygen concentration gradients measured through an intact specimen of L. patella in darkness or under an irradiance of 1350 μmol photons m⁻² s⁻¹. The upper surface of the tunic and the boundary toward the cloacal cavity are indicated by the two horizontal lines. (c) Spectral scalar irradiance (in percent (%) of incident irradiance at the tunic surface) as measured with a fiber-optic microprobe just below the three different compartments within L. patella. Color coding is the same as in a. (d) Photosynthetic activity measured as the photosystem II-related rETR versus irradiance in the three different biofilms.

Figure 4

Heatmap of 16S rRNA gene sequences obtained from bacterial biofilms associated with the underside, the surface or the cloacal cavity (=Prochloron) of L. patella, respectively. Independent biological replicates of the sampled biofilms are labeled sequentially, that is, ‘Surface.deep.A–C' indicating three replicates. The three sampling depths are indicated after the name of the biofilm and denote the ‘deep' (2.5–3.5 m), ‘intermediate' (1.5–2.5 m) and ‘shallow' site (0.3–1 m). Only OTUs with a sum of >100 assigned sequences across all samples were used for further log₁₀+1 transformation. Taxonomy was assigned to OTU representatives using best Blast hits against the Greengenes database. Duplicate OTUs were numbered sequentially for clarification. OTU clustering is shown along the y axis; dendrogram distances are based upon relative abundances within the data matrix and not on phylogenetic relationships. The top dendrogram is based on Euclidean distances and represents clustering into four distinct clusters (1–4) according to relative abundances within the data matrix. Cluster I (light blue) contains all samples originating from the underside of the ascidian, and cluster II (green) contains all samples taken from the cloacal cavity (=Prochloron), except one. Cluster III (yellow) is nested within cluster II and contains three surface samples. Cluster IV (red) contains the remaining surface samples and the last sample taken from the cloacal cavity.

Figure 5

Principal component analysis of jackknifed-weighted UniFrac distances. Colors indicate the three different microhabitats sampled in L. patella; green triangles=underside; blue circles=surface; red squares=cloacal cavity.

Figure 6

Percent of phototrophic bacteria-related OTUs versus sampling depth. The sum of several OTUs taxonomically assigned to either (a) Rhodospirillaceae, (b) A. marina, (c) Rhodobacteraceae or (d) Chloracidobacteria were calculated for each sampling site and depth. Two (intermediate depth) or three (shallow and deep site) biological replicates are displayed in the graph as individual points. The relative percentage of sequences was calculated for the biofilm sample and correlated with the depth at which samples were taken. Correlation coefficients were determined by linear regression.