Multi-marker metabarcoding of coral skeletons reveals a rich microbiome and diverse evolutionary origins of endolithic algae

Abstract

Bacteria, fungi and green algae are common inhabitants of coral skeletons. Their diversity is poorly characterized because they are difficult to identify with microscopy or environmental sequencing, as common metabarcoding markers have low phylogenetic resolution and miss a large portion of the biodiversity. We used a cost-effective protocol and a combination of markers (tufA, 16S rDNA, 18S rDNA and 23S rDNA) to characterize the microbiome of 132 coral skeleton samples. We identified a wide range of prokaryotic and eukaryotic organisms, many never reported in corals before. We additionally investigated the phylogenetic diversity of the green algae-the most abundant eukaryotic member of this community, for which previous literature recognizes only a handful of endolithic species. We found more than 120 taxonomic units (near species level), including six family-level lineages mostly new to science. The results suggest that the existence of lineages with an endolithic lifestyle predates the existence of modern scleractinian corals by ca. 250my, and that this particular niche was independently invaded by over 20 lineages in green algae evolution. These results highlight the potential of the multi-marker approach to assist in species discovery and, when combined with a phylogenetic framework, clarify the evolutionary origins of host-microbiota associations.

Figure 1. The multi-marker metabarcoding approach.

The library preparation (A) consists of a 2-step PCR amplification: the first PCR amplifies the target markers and the second PCR adds the indices and the Illumina adapters (T = tail, P = amplicon-specific primer, P5 = Illumina adapter, FI = forward index, A = amplicon). The amplicons are then purified with magnetic beads, quantified and pooled together to be sequenced in an Illumina’s MiSeq platform. The sequence reads of the 4 markers are teased apart in the analysis pipeline (B) based on primers sequences, and go through a series of quality control steps (including pipelines available in QIIME28), OTU clustering (using UPARSE46), alignment and classification.

Figure 2. Pie charts indicating the relative abundance of sequence reads matching the main taxa assigned with RDP-classifier.

Note that the relative abundances do not always reflect diversity, as indicated by the total and green algal number of Operation Taxonomic Units retrieved with each marker.

Figure 3. Maximum Likelihood tree of green algae (4833 bp alignment) including the OTUs retrieved with the tufA metabarcode.

The green algal families where OTUs were found are indicated in square brackets. The OTUs composing the endolithic clades were given an identifier for reference in future biodiversity screenings using the tufA marker. The ancestral states reconstruction of the endolithic nature is plotted along the tree and indicate the probability of the ancestral lineage being coral-endolithic (black) or not (grey).