Transcriptomic analysis of differential host gene expression upon uptake of symbionts: a case study with Symbiodinium and the major bioeroding sponge Cliona varians
- Ana Riesgo†1, 2,
- Kristin Peterson3, 4,
- Crystal Richardson3, 5,
- Tyler Heist3,
- Brian Strehlow3, 6,
- Mark McCauley3, 7,
- Carlos Cotman3,
- Malcolm Hill†3Email author and
- April Hill†3Email author
© Riesgo et al.; licensee BioMed Central Ltd. 2014
Received: 2 August 2013
Accepted: 11 April 2014
Published: 16 May 2014
We have a limited understanding of genomic interactions that occur among partners for many symbioses. One of the most important symbioses in tropical reef habitats involves Symbiodinium. Most work examining Symbiodinium-host interactions involves cnidarian partners. To fully and broadly understand the conditions that permit Symbiodinium to procure intracellular residency, we must explore hosts from different taxa to help uncover universal cellular and genetic strategies for invading and persisting in host cells. Here, we present data from gene expression analyses involving the bioeroding sponge Cliona varians that harbors Clade G Symbiodinium.
Patterns of differential gene expression from distinct symbiont states (“normal”, “reinfected”, and “aposymbiotic”) of the sponge host are presented based on two comparative approaches (transcriptome sequencing and suppressive subtractive hybridization (SSH)). Transcriptomic profiles were different when reinfected tissue was compared to normal and aposymbiotic tissue. We characterized a set of 40 genes drawn from a pool of differentially expressed genes in “reinfected” tissue compared to “aposymbiotic” tissue via SSH. As proof of concept, we determined whether some of the differentially expressed genes identified above could be monitored in sponges grown under ecologically realistic field conditions. We allowed aposymbiotic sponge tissue to become re-populated by natural pools of Symbiodinium in shallow water flats in the Florida Keys, and we analyzed gene expression profiles for two genes found to be increased in expression in “reinfected” tissue in both the transcriptome and via SSH. These experiments highlighted the experimental tractability of C. varians to explore with precision the genetic events that occur upon establishment of the symbiosis. We briefly discuss lab- and field-based experimental approaches that promise to offer insights into the co-opted genetic networks that may modulate uptake and regulation of Symbiondinium populations in hospite.
This work provides a sponge transcriptome, and a database of putative genes and genetic pathways that may be involved in Symbiodinium interactions. The relative patterns of gene expression observed in these experiments will need to be evaluated on a gene-by-gene basis in controlled and natural re-infection experiments. We argue that sponges offer particularly useful characteristics for discerning essential dimensions of the Symbiodinium niche.
It is a truism that most if not all species on the planet serve as habitat for one or more microbial symbiont . These associations can have ecological outcomes that are beneficial (e.g., mutualisms) or deleterious (e.g., parasitisms), and as such are among the most important biological interactions on the planet given that they affect everything from general ecosystem health to human disease. However, our understanding of many major facets of the evolutionary and ecological interactions that occur among partners is limited. New molecular tools and a growing genomic perspective are offering the ability to discern nuanced aspects of host:symbiont interactions while identifying genes and pathways involved in regulating host:symbiont relationships . Here, we employed transcriptomic approaches to elucidate the molecular genetic machinery in operation during re-establishment of an intracellular symbiosis.
The structure and function of coral reefs depends upon trophic interactions that occur between a dinoflagellate symbiont belonging to the diverse lineage referred to as Symbiodinium (Alveolata: Dinoflagellata: Suessioids) and a variety of invertebrate and protistan hosts [3–6]. The algal partners, known colloquially as zooxanthellae, have long been known to be of vital trophic importance to the host ([7–12]). We understand less about the benefits the symbionts receive from the association, though most hypotheses argue that Symbiodinium benefit from intracellular residency by gaining access to nutrients that are limiting outside the host (e.g. [11–13]). The partnership is arguably the most important ecological interaction that occurs in shallow tropical habitats worldwide because Symbiodinium spp. energetically subsidize the entire ecosystem and power calcification processes  that generate the topographic complexity of these systems.
Many Symbiodinium-based symbioses are remarkably sensitive to environmental stressors, notably elevated seawater temperatures (e.g. [15, 16]). Symbionts can be lost from the host through a process known as bleaching, which can have significant deleterious effects on the host . There is growing concern among scientists about what the potential disruption of this important symbiosis means for the future of coral reefs (e.g. [18–20]). In the face of these concerns, it has become apparent that significant gaps exist in our basic comprehension of the natural dynamics of the Symbiodinium:host interaction, and in the degree of cellular and genetic integration among partners. Hosts can recover from mild and even massive losses of their symbiont populations, though mortality rates of the hosts increase under both scenarios, especially the latter . Symbiodinium spp. are also capable of (in fact probably require) existence outside of the host, and Symbiodinium spp. have planktonic, free-living stages that occur even during non-bleaching events (e.g. [22, 23]). Currently, coral reef biologists have a limited capacity to satisfactorily explain the facultative nature of the symbiotic interaction between Symbiodinium and heterotrophic hosts . We do not know how facile/labile the symbiotic association between Symbiodinium spp. and their host partners is, nor what selective landscapes are in place that favor the observed patterns of partner association. Understanding fundamental aspects of symbiont uptake, establishment of intracellular residency, and dynamics behind cellular expulsion will be essential as we attempt to manage the significant environmental changes underway on coral reefs.
As we face warming sea surface temperatures due to human-induced climate change, it has become more pressing to understand the interactions that occur among the partners at the finest molecular genetic levels so that we may better prepare for the ecological realities coral reefs will face. In the broadest terms, we lack a clear understanding of how Symbiodinium navigates a potential host’s cellular and molecular genetic machinery so that digestion, detection and expulsion are avoided; we also lack a clear understanding of what role the host might play in permitting intracellular residency. Recent advances in molecular and genomic approaches have enhanced our understanding of some of the regulatory operations executed between cnidarian hosts and zooxanthella symbionts (e.g. [24–33]). Molecular genetic data has failed to identify “symbiosis-specific” genes that regulate the interaction between partners, but instead has found subtle differences in expression patterns that depend on holobiont context. For example, symbiont cladal identity has been shown to play an important role in transcriptomic profiles . Emphasis has now shifted toward finding the cellular pathways that are modulated such that Symbiodinium maintain their position within the host cell or a particular type of tissue (e.g. [27, 30, 34]).
Given that cnidarians are not the only habitable hosts for Symbiodinium on coral reefs (e.g. [35–38]), we stand to gain insights into nuanced aspects of the entire zooxanthella niche through analysis of non-cnidarian systems (e.g. ). Sponges are ecological important members of many marine ecosystems (e.g. [40, 41]), and their simple body plans affords interesting experimental opportunities [42, 43]. They belong to an ancient metazoan lineage that represents one of the earliest branches of the animal lineage [44, 45]. Sponges use flagellated choanocytes in the choanoderm to propel large volumes of water through an aquiferous system that efficiently remove bacterioplankton and dissolved organic matter while the pinacoderm mediates interaction with the environment .
In the work presented here, we took advantage of a suite of molecular tools to explore aspects of the intracellular symbiosis that exists between the Caribbean bioeroding sponge Cliona varians and its Clade G Symbiodinium symbionts. Sponge: zooxanthella symbioses are especially important given that Symbiodinium are predominantly associated with the bioeroding sponges that dissolve calcareous structures (e.g. ), which is a growing concern given CO2-driven changes in the pH of seawater . Non-cnidarian systems also offer some empirical and comparative advantages over cnidarian hosts (e.g., the ability to create intracellular associations in hosts that have no evolutionary history of symbiotic associations with Symbiodinium, the ability to compare genetic expression profiles in congeneric species that differ in their ability to form symbioses with Symbiodinium (e.g., C. delitrix versus C. varians), and the ability to produce aposymbiotic cell aggregates (e.g. ) that can then be exposed to Symbiodinium under precisely controlled conditions). In this context, we present C. varians as a useful tool to better understand the Symbiodinium niche sensu lato as well as to achieve a high level of resolution of genetic regulation in sponge:Symbiodinium and all intracellular associations.
Results and discussion
Creation of “aposymbiotic” and “reinfected” tissue
Transcriptome characterization: de novo assembly, BLAST, and functional annotation
We sequenced transcriptomes from “normal”, “reinfected”, and “aposymbiotic” sponges. Each pool of RNA used for subsequent sequencing of the three tissue types was derived from at least three different sponge samples, but these were pooled into a single batch for each symbiont state prior to next generation sequencing. Thus, the sequences we present below come from non-replicated sequence runs (see Methods section). This caveat becomes important when interpreting the putative differences we observed. We recognize a preferable approach would be to sequence several distinct and independent samples from each symbiont state. However, this was a pilot study to determine the feasibility of using C. varians to study Symbiodinium symbioses, and used several approaches (e.g., transcriptomics, suppressive subtractive hybridization (see below)) to assess molecular genetic regulation. At the time we sequenced the transcriptomes, costs associated with sequencing multiple replicates were prohibitive. Furthermore, best practices associated with RNASeq experiments were just being developed (e.g. ). Nonetheless, the success we achieved in obtaining high quality sequences indicated that the database we present below will be a useful resource for the community as future studies attempt to discern significant differences observed at various stages of the establishment and maintenance of Symbiodinium symbioses.
de novo assembly data from the RNA-Seq experiments involving the three symbiont treatments “normal,” “reinfected,” and “aposymbiotic”
N reads BT
GC content (%)
Sequence duplication (%)
N reads trimmed
Avg. L AT
N bases (Mb)
Avg. L Contigs
Max contig L
Reference (pooled data)
For each tissue treatment, most contig sequences with hits returned a BLAST hit against the metazoan database, followed by the bacterial database, and then the protozoan database, with very few contigs obtaining hits against more than one database (Figure 2B). The “normal” treatment obtained more BLAST hits than the other two treatments, whereas the “reinfected” treatment returned the fewest BLAST hits (Figure 2B). This difference in patterns of BLAST hits could be due to differences in sequence read numbers obtained for the different treatments (i.e., 21 M trimmed reads in control vs 9 M reads in reinfected; Additional file 2: Table S1), which could represent experimental error (i.e., technical variation). Alternatively, this pattern could point to an actual molecular genetic response to the onset of symbiosis in the form of global- or chromatin-level gene regulation (e.g. [24, 25]). For example, symbiont-induced, host-gene suppression may be a feature of the initiation of host:symbiont interactions . Further data are necessary to test this hypothesis.
Differential expression analysis
Interesting Gene Ontologies are also revealed when comparing genes expressed at higher levels in “normal” compared to “aposymbiotic” tissue (Figure 5 bottom) including members of the TNF family (e.g., TNF receptor-associated factor 3-like), which are important in immune responses (e.g., “acute-phase response” Figure 5 bottom). Other interesting genes included deleted in malignant brain tumor and niemann pick c1 (Additional file 4: Table S4; Figure 5). These genes are discussed further below. It was intriguing that some of the genes that appear at higher frequency in “normal” tissue compared to “aposymbiotic” tissue (Additional file 4: Table S4 and Figure 5) are involved in “cell adhesion” (e.g., collagen alpha-1(I) chain, basement membrane-specific heparan sulfate proteoglycan core protein-like, and focal adhesion like fibronectin, which is an ECM component that acts as the integrin ligand ). This may relate to the movement and re-organization of Symbiodinium-bearing cells in mature symbiont populations.
The treemaps provided unique insights into some of the patterns observed in our comparison of expression profiles in the different tissue types. The two panels that describe increased levels of expression in “aposymbiotic” tissue (Figure 5 top; Figure 6 bottom) showed very similar patterns in GO assignments. The top three categories for each of these comparisons were “cell cycle,” “tRNA aminoacylation for protein translation,” and “response to bacterium” (Figure 5 top; Figure 6 bottom). Some of the remaining categories were also identical (“carbohydrate catabolism” and “cellular process”). The situation was different for the other two comparisons that involved higher levels of gene expression in the presence of Symbiodinium (Figure 5 bottom; Figure 6 top). The differences in GO assignments here point to the possibility that different cellular processes are operating in a mature symbioses (“normal” tissue) compared to an association that is at an earlier stage of re-establishing Symbiodinium populations (“reinfecting” tissue). For example, “regulation of cell growth” was the predominant GO signature of genes that showed higher expression in “reinfecting” vs. “aposymbiotic” tissue. This broad category presents a suite of genes that would be worthy of future work to ascertain their importance in the development of a stable Symbiodinium symbiosis.
We found interesting patterns in global gene expression patterns among “normal”, “aposymbiotic”, and “reinfected” tissue treatments (Additional file 5: Figure S3). While the significant differences observed using the DESeq analysis described above are interesting, it is important to recognize that subtle differences in gene expression profiles that do not rise to the level of statistical significance estimated with a methodology like the one implemented in DESeq may still play important biological roles in regulating the interaction between partners in this symbiosis. Thus, closer inspection of specific GO categories provides important perspectives on the interplay that may occur between partners in this sponge: algal association. However, high throughput sequencing generates a large and complicated suite of genes and gene networks to consider, thus it is necessary to reduce the complexity of the dataset and identify testable hypotheses for future experiments. Therefore, we examined pathways that might relate to a recent hypothesis that posits that Symbiodinium spp. may mimic the phagosome by releasing materials at a rate and of a quality that would be expected from digesting prey thus securing their intracellular position . This “arrested phagosome hypothesis (APH)” offers a subtly different perspective on the cellular machinations in operation when Symbiodinium take up residency in host cells. If Symbiodinium spp. use their photosynthetic capabilities to maintain residence within the intracellular habitat (but are “parasitic” in other aspects of their life history), then we may expect different types of genetic expression profiles than if the host is somehow controlling the association (e.g. [32, 62]). It is clear, however, that our non-replicated transcriptomes must be interpreted cautiously as trends we observed may not represent statistically significant differences.
In addition to the positive energetic benefits gained by hosts from their symbionts, Symbiodinium partners might also increase physiological stress on their hosts (e.g. ). It is also possible that by inoculating dark-acclimated aposymbiotic C. varians with a large dose of symbionts, and placing them under lighted conditions, we stressed the sponges involved in the reinfection experiments. Thus, we assessed generalized stress responses in “reinfected” compared to “aposymbiotic” tissue. We identified 6 genes involved in response to stress (GO:0006950) that were at least two-fold more common in “reinfected” compared to “aposymbiotic” tissue (Figure 8B; Additional file 6: Table S2). Using that same GO category, we identified 8 genes that were at least two-fold more common in “aposymbiotic” compared to “reinfected” tissue (Figure 8B; Additional file 6: Table S2).
Suppressive subtractive hybridization
Results from suppressive subtractive hybridization experiments
Insert size (bp)
3-hydroxybutyrate dehydrogenase type 2
degradation of ketone bodies
metabolic process, catalytic activity
Actin-related protein 2/3 complex subunit
cell locomotion & phagocytosis
cytoskeleton, protein binding
AP-2 complex subunit beta
membrane, transport, protein binding
ATPase, H+ transporting, lysosomal, V0 subunit
ion transport, transport
Ca2+-triggered coelenterazine-binding protein 2
calcium ion binding
calcium ion binding
Calcium-binding protein p22; Calcineurin
CHK1 checkpoint-like protein
kinase activity in mitosis
protein kinase activity, nucleotide binding
Creatine kinase U-type, mitochondrial
energy production and transport
nucleotide binding, transferase activity
protein folding, hydrolase activity
Cell death induction
cell motility & maintenance
cellular protein metabolic process,
Deleted in malignant brain tumors 1 protein-like; Scavenger receptor cysteine-rich type protein
removal of foreign substances
mitochondrial glycine cleavage
Dynein heavy chain
cellular transport & maintenance
biological process, transferase activity
Ephrin type-B receptor 1; Protein tyrosine kinase
nucleotide binding, transferase activity
358 - 721
9.00E-86 - 1.00E-18
innate immune recognition
G-protein gamma subunit
Gamma-interferon-inducible lysosomal thiol reductase like
catalytic activity, biological process
cellular nitrogen compound metabolic process
Heat shock protein 70
protein folding & stress protection
response to stress
389 - 564
8.00E-24 - 2.00E-04
calcium absorption & metabolism
lipid metabolism & calcium absorption
cytoplasm, ion binding
MafB chain A
transcription, cell death
Neurogenic locus notch protein homolog
Nuclear pore complex Nup50
intracellular protein transport
carbohydrate metabolic process
Proteasome subunit alpha
processing of MHC class I peptides
cellular nitrogen compound metabolic process
Proteasome subunit beta
intracellular protein degradation
cellular protein metabolic process, gene expression
Ribonuclease K-like; Salivary secreted ribonuclease
degredation & protection
223 - 383
3.00E-37 - 2.00E-04
translation, cellular protein metabolic processes, gene expression
RNA polymerase-associated protein LEO1
protein binding, transcription
Selenoprotein Jb; J1a crystallin
regulation of metabolism
Serum response factor
cytoskeleton, signal transduction
Sulfide quinone reductase
Tubulin alpha chain
Tyrosine 3-monooxygenase/tryptophan 5-monooxygenase activation protein
phosphoserine-binding for signal transduction
cytoplasm, protein targeting
Vacuolar sorting protein; sortilin-related receptor
neuropeptide receptor activity & protein binding
von Willebrand factor A domain-containing protein-5a
intracellular ligand interactions
WAS protein family homolog 1
nucleation promoting factor on endosomal surface
Our results add to the growing perspectives on molecular genetic integration between hosts and symbionts in Symbiodinium-based associations. This is, however, the first that provides insights into the genetic pathways that appear to be important in poriferan: Symbiodinium partnerships. Our results indicate that hosts, regardless of taxonomic origin, engage similar cellular and genetic processes in response to intracellular zooxanthella-residency [25–31, 33, 64]. High-throughput sequencing offers opportunities to generate massive datasets, and we found that comparing the transcriptomic data with results generated through suppressive subtractive hybridization provided an interesting mechanism to validate a portion of our non-replicated RNASeq data. The RNA-Seq experiments and cross-validation with an independent methodology (e.g., SSH) provide confidence that we have identified some appropriate candidate genes for future work focused on detailing precise genetic regulation of symbiont and host interactions. However, any differences observed in the present study should be treated cautiously since they come from transcriptomes that were not replicated within treatments. One of our goals was to demonstrate the importance of integrating ecologically-relevant scenarios with insights gained through acquisition of lab-based gene expression data. Sponges may be exceptionally useful systems of study in this context. Specifically, the temporal variability seen in expression dynamics under natural conditions in the field highlight how nuanced the interaction between the host and symbiont is likely to be, and how much work remains to uncover detailed perspectives on the associations.
Through this and related work, it appears possible to identify some common pathways that Symbiodinium may co-opt to gain entry and to procure residency in a variety of potential hosts. Nonetheless, clear explanatory hypotheses are needed so that we can better understand, and prepare for, changes in the symbiosis that are likely with the rapid shifts in temperature and sea-water chemistry that will accompany global climate change [13, 65, 66]. We also require more detailed knowledge of the interaction between symbiotic partners. We argue that sponge: Symbiodinium associations add important perspectives on the Symbiodinium niche, which will foster greater understanding in other host environments.
Creation of aposymbiotic and reinfected sponges
Cliona varians forma varians were collected from shallow (≈1 m) flats just south of the Mote Tropical Research Laboratory in Summerland Key, FL (24.658, −81.452). All collections performed in the Florida Keys for this study were obtained with all appropriate and relevant permits and licenses. In accordance with policies established by the Florida Keys National Marine Sanctuary, we collected sponges under permit FKNMS-20070094-A1 and under a Florida recreational resident saltwater fishing license issued from Florida Fish and Wildlife Conservation Commission. Sponges were transported to shallow raceways where the Symbiodinium-dense pinacodermal region was removed with a sharp razorblade (Additional file 1: Figure S1). The Symbiodinium devoid choanosome explants were placed in a lightproof container (≈60 L total volume) to heal for several months where they received fresh seawater from an underground aquifer, which is unlikely to contain free-living Symbiodinium, at a rate of approximately 2 L min−1. Small explants (≈6-8 cm3) of the “aposymbiotic” sponges were then exposed to Symbiodinium that had been freshly isolated from C. varians forma varians (Figure 1). After 5 days, signs of reinfection were visible to the naked eye (Figure 1). At this point, tissue was harvested, placed in 1.5 ml tubes, flash-frozen in liquid nitrogen, and immediately stored at −80°C until mRNA was extracted.
For transcriptomic analysis, mRNA was isolated directly from the tissue samples (three biological replicates of each tissue type were pooled within the same tube) using the Micro-FastTrack 2.0 mRNA isolation kit (Invitrogen), and mRNA samples from the replicates were pooled. Quantity and quality (purity and integrity) of mRNA were assessed by three different methods. We measured the absorbance at different wavelengths using a NanoDrop ND-1000 UV spectrophotometer (Thermo Fisher Scientific, Wilmington, Massachusetts, USA). Quantity of mRNA was also assessed with the fluorometric quantitation performed by the QubiT® Fluorometer (Invitrogen, California, USA). Also, capillary electrophoresis in an RNA Pico 6000 chip was performed using an Agilent Bioanalyzer 2100 System with the “mRNA pico Series II” assay (Agilent Technologies, California, USA). Integrity of mRNA was estimated by the electropherogram profile and lack of rRNA contamination (based on rRNA peaks for 18S and 28S rRNA given by the Bioanalyzer software). We used the TruSeq RNA Sample Prep Kit (Illumina, Inc.) to prepare the three different library samples of C. varians using 135.8 ng of mRNA for the normal tissue, 665 ng for the aposymbiotic tissue, and 743.5 ng for the reinfected tissue following the manufacturer’s instructions with minor modifications. Fragmentation was performed on mRNA for 1.5 min, and fragments of 350 bp were targeted through size selection on excised gel bands of 2% agarose. The three samples were multiplexed using Index 4 for the normal tissue, 6 for the aposymbiotic tissue, and 12 for the reinfected tissue from the TruSeq RNA Sample Prep Kit.
The concentration of the cDNA libraries was measured with the QubiT® dsDNA High Sensitivity (HS) Assay Kit using the QubiT® Fluoremeter (Invitrogen, Carlsbad, California, USA). The quality of the library and size selection were checked using the “HS DNA assay” in a DNA chip for Agilent Bioanalyzer 2100 (Agilent Technologies, California, USA). cDNA libraries were considered successful when the final concentration was higher than 1 ng μl−1 and the bioanalyzer profile was optimal . We obtained 1.065 μg of cDNA for the normal tissue, 0.45 ng for the aposymbiotic tissue, and 0.09 ng for the reinfected tissue. The libraries were brought to 10 nM prior to sequencing. Next-generation sequencing was performed using the platform Illumina HiSeq (Illumina, Inc., San Diego, California, USA) at the FAS Center for Systems Biology at Harvard University. The normal and aposymbiotic treatments were run together with another invertebrate library in one lane, and the reinfected treatment was run in another lane with two more invertebrate libraries. Paired-end reads were run to 101 bp.
Transcriptome assembly and annotation
Trimming analyses for the raw reads of each independent transcriptome dataset were done with CLC Genomics Workbench 5.1 (CLC bio, Aarhus, Denmark). Initial trimming was performed using 0.5 as the limit of the quality score (based on Phred quality scores), and resulting quality of the trimmed reads was visualized with FastQC (http://www.bioinformatics.bbsrc.ac.uk/projects/fastqc/). After this, only those terminal bases with a Phred quality score under 30 were trimmed (where a Phred score of 30 corresponds to a probability of 1 in 1,000 of incorrect base calling), which produced sequences of unequal size. High-quality reads were re-screened to check for presence of adapter or primer sequences using FastQC, and if present, they were removed using with CLC Genomics Workbench 5.1.
Four de novo assemblies were performed using CLC Genomics Workbench 5.1: three separate assemblies containing the raw reads of each treatment, and another one pooling all raw reads (called “reference”). Global alignments for the de novo assemblies were always done using the following default parameters: mismatch cost = 2; insertion cost = 3; deletion cost = 3; length fraction = 0.5; similarity = 0.8; and randomly assigning the non-specific matches. Best k-mer length was estimated by the software. The best assembly for each treatment was selected using an adaptation of the optimality criteria for de novo assembly with 454 data .
From the “reference” transcriptomic dataset, contigs shorter than 300 bp were removed (assuming that shorter contigs would retrieve very few results during blast searches). For the remaining contigs we performed BLAST searches against a database of selected proteins from the nr NCBI database (containing Metazoa, Bacteria, Fungi, Virus, and Protozoa, including Symbiodinium spp.). Since sponges host a wide variety of symbiotic organisms within their tissues, mainly bacteria and protozoans, that cannot be completely removed prior to cDNA construction, we performed separate BLAST searches against three different individual databases containing proteins of Metazoa, Protozoa, and Bacteria, to estimate the amount of contigs belonging to either symbionts or the sponge. Such searches were performed for the “normal”, “aposymbiotic”, and “reinfected” transcriptome datasets and the contigs showing hits against two or all the databases were counted. All BLAST searches were conducted with BLAST v2.2.23+  using an e-value cut-off of 1e-5. With the resulting file, we then used Blast2GO v2.5.0 (Conesa et al. ) to retrieve the Gene Ontology (GO) terms and their parents associated with the top BLAST hit for each sequence. For the metazoan hits, we performed a Fisher’s exact test with multiple test correction by Benjamini–Hochberg false discovery rate (FDR) to analyze the differential GO term enrichment (P > 0.05) in each treatment.
RNAseq analysis and differential expression
Only contigs of 1000 bp or longer from the “reference” transcriptome were used as a mapping reference for the evaluation of expression values because they were, in the majority of cases, assigned BLAST and GO term annotations. Quality trimmed reads from each of the three treatments were mapped against the “reference” dataset with CLC Genomics Workbench 5.1 as a short read aligner. The total number of unambiguously mapped reads (i.e., “unique genes”) of each treatment compared to the “reference” transcriptome was exported as a table to use as count data in further analyses. Differential expression values were computed with the DESeq package  in Bioconductor in R. We performed three different comparisons to find genes up-regulated in each treatment: “normal” versus “aposymbiotic”, “aposymbiotic” versus “reinfected”, and “normal” versus “reinfected”. We first estimated the effective library size, and then estimated the data’s dispersion and mean to identify differentially expressed genes. Due to the lack of non-pooled biological replicates for transcriptome sequences, we instructed the program to ignore the condition (i.e., “treatment”) labels and estimated variance by treating all samples as if they were replicates of the same condition. This approach follows that outlined in Anders . Comparisons were accepted to be significant at an FDR adjusted value of 0.01. Only significant values were plotted as a heatmap using the R heatmap.2 function from the R ‘gplots’ library. We used the default hclust hierarchical clustering algorithm to cluster the rows. Finally, the affiliation of differentially expressed contigs to either Metazoa, Bacteria, Protozoa, Fungi, and Virus was obtained from BLAST results of the “reference” transcriptome.
We performed two enrichment analyses for the differentially expressed genes for which we obtained significant p-values and were also able to find associated GO terms (obtained in the annotation with Blast2GO of de novo assembled “reference” transcriptome). The enrichment analyses were performed for this set of differentially expressed genes using all three possible comparisons (“normal”, “aposymbiotic”, “reinfected”) by testing the up-regulated genes in one treatment against up-regulated genes in the other treatment. Enriched GO-terms were then slimmed in REVIGO and treemaps were produced (following ). We also conducted overall comparisons of the expression profiles of the three C. varians treatments. In addition, for overall comparisons of the expression profiles of the three treatments of C. varians, heat maps were obtained with CLC Genomics Workbench 5.1 by mapping the raw reads of each treatment dataset against the total “reference” contig list (292,182 contigs). Contigs of the “reference” dataset whose size exceeded 1000 bp (N = 15,636 sequences) were represented in detail to ensure full length. Expression was measured in RPKM (Reads Per Kilobase of exon model per Million mapped reads). Since no reference genome is available for C. varians, exons were not annotated for the analysis, and in turn, the assembled contigs were assigned a complete exon. We generated another heat map that included genes that had differences between RPKM among treatments of 2 (N = 13,773 sequences). All these analyses were performed without replication, and thus the results should be taken as a preliminary assessment of the gene expression profile of the tissues under the treatments.
Assuming that the transcriptome dataset “reference” contained most of the sponge genes present in the genome, we also estimated the ortholog hit ratio (OHR) as defined by O'Neil et al. . The OHR describes the percentage of an ortholog “found” in a contig by dividing the number of non-gap characters in the query hit by the length of the subject using a script provided by Ewen-Campen et al. . The workflow used to analyze all our transcriptomic data was provided by Riesgo et al. .
We compared the expression values to identify contigs that had either the highest or lowest occurrence in the “reinfected” tissue compared to the “aposymbiotic” tissue types. These values were reported as fold increase. It is important to note that we cannot assign significance to these differences – they are meant to demonstrate how candidate genes might be first identified. To narrow the large universe of genes that could possibly be examined, we focused our attention on GO terms that may be associated with pathways related to recently proposed hypotheses . We truncated our analysis to genes that showed a 2-fold or higher difference between reinfected and aposymbiotic tissue.
Suppressive subtractive hybridization
Suppressive subtractive hybridization (SSH) was performed using the Clontech PCR-Select cDNA Subtraction Kit®, following the manufacturer’s protocol. Poly A + mRNA was isolated from three biological replicates of C. varians aposymbiotic and reinfected tissue using the Micro-FastTrack 2.0 mRNA isolation kit (Invitrogen) and pooled before cDNA synthesis of RNA, which was performed using the Super Smart cDNA Synthesis Kit (Clontech). cDNA from reinfected tissue was used as the tester and cDNA from aposymbiotic tissue was used as the driver for the forward subtraction reactions. PCR products generated from the subtracted library, representing mRNAs putatively over-expressed in reinfected tissue, were sub-cloned into the TOPO TA cloning vector using OneShot TOP10 competent cells (Invitrogen) and plasmids were prepared using the QIAprep Spin Miniprep kit (Qiagen). Sequencing of 173 individual clones from the subtracted library was performed on an ABI 3130 × L Genetic Analyzer at Virginia Commonwealth University’s sequencing facility. Sequences were searched using the blastx and tblastx algorithms in the Genbank database. To validate that a subset of the identified genes were differentially expressed, RNA was isolated from aposymbiotic and reinfected C. varians using the RNeasy® Mini Kit (Qiagen), limiting genomic DNA contamination through an additional on-column DNase I treatment. cDNA was synthesized from equal amounts of sponge mRNA (125–200 ng/μl) using Superscript III reverse transcriptase (Invitrogen) and oligodT primer. In some cases, RT-PCR was conducted followed by gel electrophoresis to allow visual inspection of differential gene expression. In other cases, SYBR Green (Invitrogen) chemistry and Chromo4 (BioRad) were used to obtain relative levels of expression by qRT-PCR. Expression levels were normalized to the housekeeping gene Ef1a that qRT-PCR showed to be consistently expressed at high levels in both sets of tissues. For all qRT-PCR experiments, duplicates were performed from master mixes, and in most cases each experiment was repeated twice. Threshold values for Ct calculation were manually selected for all samples by placing the threshold line at the intersection where the signal intensities of the fluorescence traces surpassed background levels and began to increase (i.e., the linear portion of the curve). Both data and standard graphs were considered when establishing the position of the threshold line to optimize efficiency. Reaction efficiencies were recorded as efficiency per well in the linear range of the Ct and two points above. Standard curves, using plasmid dilutions of known quantities as templates, were generated for each gene in each qPCR experiment. Efficiency-corrected Ct values were compared to these curves (based on log of standard DNA concentration vs. Ct value for each sample) to calculate relative concentrations of samples using Opticon Monitor software (BioRad). The relative concentration values of duplicates were averaged and experimental averages were normalized to Ef1a values.
We used the SSH library as a partial validation of the gene expression values observed in the transcriptomic analysis. We first used the BLAST algorithm to search the transcriptome for transcripts matching our SSH clones. We used BLAST to verify GO terms and thus gene identity for contigs identified as having significant overlap with the SSH clone. In two cases, none of the contigs recovered the gene identified in SSH (i.e., cyplasin, a ribosomal protein). For the other 54 genes, we could verify contigs that aligned with our SSH gene. In some cases more than one contig aligned with the SSH clone so we examined the expression levels for each contig aligning with our SSH clone.
Experimental analysis of gene expression profiles
A natural reinfection experiment was conducted in the flats south of Mote Tropical Research Laboratory (24.6605, −81.4551). Forty-two aposymbiotic C. varians explants were transplanted from their lightproof container into shallow water (>1 m). Explants were secured to a sheet of fiberglass window with monofilament. The window screen with sponges was situated on top of the substratum in an area populated with several potential Symbiodinium donors (e.g., Porites divericata, Siderastrea radians, Cassiopea xamanchana, and Cliona varians).
Preliminary experiments indicated that populations of intracellular Symbiodinium began to appear in aposymbiotic C. varians transplants after approximately six days in the field. Thus, from May to July, 2012, we sampled 3 explants from the field at 0, 2, 4, 5, 6, 7, 8, 9, 10, 12, 14, 16, 18, and 48 days post transplantation. The explants were transported to the lab within 30 min of collection where they were immediately processed for subsequent work. One section of each explant was immediately placed in RNAlater RNA Stabilization Reagent from QIAGEN (for gene expression analyses), and stored overnight at 4°C. The next morning, RNAlater was drained from the tube, and the tissue was frozen and stored at −80°C. A second section was snap frozen for DNA isolation. Another two sub-sections were taken from each explant and fixed in either a 4% paraformaldehyde: 2.5% gluteraldehyde solution (for electron and light microscopy work) or a 3.7% formaldehyde solution (for zooxanthella cell counts). Tissues were stored at 4°C, and after 24 h, the gluteraldehyde-containing samples were transferred to filter sterilized seawater and stored at 4°C until embedding, sectioning and visualization. Three randomly chosen C. varians individuals were sampled from the flats to serve as controls and were processed in the same manner described above. Differential expression of two genes (nup50 and vacuolar sorting protein) as a function of time post-transplantation was assessed by qRT-PCR as described above, however, expression values are plotted relative to time 0 after normalization to the housekeeping gene Ef1a. We selected nup50 and vacuolar sorting protein because they showed strong levels of up-regulation in reinfecting tissue, and thus represented robust candidates to demonstrate that this empirical approach would be a useful tool to test gene expression hypotheses generated by the transcriptome and SSH databases.
Paraformaldehyde: gluteraldehyde-fixed samples were embedded in OCT™ medium, frozen in liquid nitrogen, and sectioned with a Leica CM1850 cryostat at a thickness of 10 μm. Sections were stained with SYBR® green (1 μg μl−1) in 80% glycerol, and imaged using a Hamamatsu ORCA-ER camera attached to an Olympus BX61 microscope with a DG4 fluorescent lamp. Symbiodinium were visualized with a TX-RED filter (936 ms exposure) while SYBR green-stained nuclei could be distinguished using the FITC filter (1302 ms exposure). Symbiodinium-depth within sponge tissue was determined by stitching together successive images starting at the pinacoderm and moving deeper into the choanosome with Adobe Photoshop. Algal cells were counted in triplicate along microtransects (5 μm by 10 μm) that ran 3 cm into the choanosome. Total Symbiodinium cell counts were performed with formaldehyde-fixed samples. A block of known dimensions was cut from the pinacoderm into the choanosome. The tissue was ground with a mortar and pestle and the resultant slurry was suspended in 5 ml of filter-sterilized seawater. Symbiodinium cell concentrations were measured with a 0.1 mm deep Bright-line® hemacytometer. Five independent samples were taken from the suspension to calculate average zooxanthellae densities (cells mm−3 sponge tissue). DNA was isolated from frozen samples using a modified CTAB protocol and used in PCR reactions to amplify 23S rDNA . PCR products were gel purified (Qiagen) before being sent to VCU’s DNA sequencing facility. Using BLAST, sequences were compared to NCBI’s nucleotide collection database to determine identity.
Availability of supporting data
Transcriptomic sequences were deposited in the NCBI Sequence Read Archive. The experiment accession numbers for the raw reads deposit is as follows: “normal”: SRX333053, “aposymbiotic”: SRX333054, and “reinfected”: SRX333055. The Bioproject accession number for the whole project is: PRJNA214560, and the Biosample accession number is: SAMN02304131.
The authors would like to thank Sarah Friday, Andrew Massaro, Samuel Hill, and Blake Ramsby for their help with aspects of the field- and lab-work described in this paper. We also thank the Mote Tropical Research Laboratory on Summerland Key, FL for logistical support (especially Erich Bartels for helping us locate appropriate sites). Three anonymous reviewers provided helpful comments. This work was supported by a Juan de la Cierva contract to AR and the US National Science Foundation (grant numbers 0647119, 0829763) to MH and AH.
- Douglas AE: The symbiotic habit. 2010, Princeton: Princeton University PressGoogle Scholar
- Medina M, Sachs JL: Symbiont genomics, our new tangled bank. Genomics. 2010, 95: 129-137. 10.1016/j.ygeno.2009.12.004.PubMedView ArticleGoogle Scholar
- Muscatine L, Porter JW: Reef corals - mutualistic symbioses adapted to nutrient-poor environments. Bioscience. 1977, 27: 454-460. 10.2307/1297526.View ArticleGoogle Scholar
- Trench RK: Dinoflagellates in non-parasitic symbioses. The biology of dinoflagellates. Edited by: Taylor FJR. 1987, Oxford: Blackwell, 530-570.Google Scholar
- Veron JEN: Corals in space and time: the biogeography and evolution of the Scleractinia. 1995, Ithaca, NY: Cornell University PressGoogle Scholar
- Coffroth M, Santos S: Genetic diversity of symbiotic dinoflagellates in the genus Symbiodinium. Protist. 2005, 156: 19-34. 10.1016/j.protis.2005.02.004.PubMedView ArticleGoogle Scholar
- Boschma H: On the feeding reactions and digestion in the coral polyp Astrangia danae, with notes on its symbionts with zooxanthellae. Biol Bull. 1925, 49: 407-439. 10.2307/1536652.View ArticleGoogle Scholar
- Kawaguchi S: On the physiology of reef corals. VII. The zooxanthella of the reef corals is Gymnodinium sp. Dinoflagellata its culture in vitro. Palao Trop Biol Stn Stud. 1944, 2: 675-679.Google Scholar
- Stat M, Carter D, Hoegh-Guldberg O: The evolutionary history of Symbiodinium and scleractinian hosts - symbiosis, diversity, and the effect of climate change. Perspect Plant Ecol. 2006, 8: 23-43. 10.1016/j.ppees.2006.04.001.View ArticleGoogle Scholar
- Weisz J, Massaro A, Ramsby B, Hill M: Zooxanthellar symbionts shape host sponge trophic status through translocation of carbon. Biol Bull. 2010, 219: 189-197.PubMedGoogle Scholar
- Stambler N: Marine microralgae/cyanobacteria -invertebrate symbiosis, trading energy for strategic material. All flesh is grass: plant-animal interrelationships. Edited by: Seckbach J, Dubinsky Z. 2011, 383-414. 16Google Scholar
- Stambler N: Zooxanthellae: The yellow symbionts inside animals. Coral Reefs: An Ecosystem in Transition. Edited by: Dubinsky Z, Stambler N. 2011, New York (NY): Springer, 87-106.View ArticleGoogle Scholar
- Hill MS, Hill AL: The arrested phagosome and magnesium inhibition hypothesis: novel perspectives on Symbiodinium symbioses. Biol Rev. 2012, 87: 804-821. 10.1111/j.1469-185X.2012.00223.x.View ArticleGoogle Scholar
- Colombo-Pallotta MF, Rodríguez-Román A, Iglesias-Prieto R: Calcification in bleached and unbleached Montastrea faveolata: evaluating the role of oxygen and glycerol. Coral Reefs. 2010, 29: 899-907. 10.1007/s00338-010-0638-x.View ArticleGoogle Scholar
- Oliver TA, Palumbi SR: Do fluctuating temperature environments elevate coral thermal tolerance?. Coral Reefs. 2011, 30: 429-440. 10.1007/s00338-011-0721-y.View ArticleGoogle Scholar
- Pandolfi JM, Connolly SR, Marshall DJ, Cohen AL: Projecting coral reef futures under global warming and ocean acidification. Science. 2011, 333: 418-422. 10.1126/science.1204794.PubMedView ArticleGoogle Scholar
- McClanahan T, Weil E, Cortés J, Baird AH, Ateweberhan M: Consequences of coral bleaching for sessile reef organisms. Ecological studies: Coral bleaching: patterns, processes, causes and consequences. Edited by: van Oppen MJH, Lough JM. 2009, Berlin Heidelberg: Springer, 121-138.Google Scholar
- Brandt ME, McManus JW: Disease incidence is related to bleaching extent. Ecology. 2009, 90: 2859-2867. 10.1890/08-0445.1.PubMedView ArticleGoogle Scholar
- Hughes TP, Baird AH, Bellwood DR, Card M, Connolly SR, Folke C, Grosberg R, Hoegh-Guldberg O, Jackson JBC, Kleypas J, Lough JM, Marshall P, Nyström M, Palumbi SR, Pandolfi JM, Rosen B, Roughgarden J: Climate change, human impacts, and the resilience of coral reefs. Science. 2003, 301: 929-933. 10.1126/science.1085046.PubMedView ArticleGoogle Scholar
- Wilkinson C: Status of coral reefs of the world: 2008. 2008, Townsville, Australia: Global coral reef monitoring network and reef and rainforest research centreGoogle Scholar
- Maynard JAM, Turner PJ, Anthony KRN, Baird AH, Berkelmans R, Eakin CM, Johnson J, Marshall PA, Packer GR, Rea A, Willis BL: ReefTemp: an interactive monitoring system for coral bleaching using high-resolution SST and improved stress predictors. Geophys Res Lett. 2008, 35: L0560-View ArticleGoogle Scholar
- Jeong HJ, Du Yoo Y, Kang NS, Lim AS, Seong KA, Lee SY, Lee MJ, Lee KH, Kim HS, Shin W, Nam SW, Yih W, Lee K: Heterotrophic feeding as a newly identified survival strategy of the dinoflagellate Symbiodinium. Proc Natl Acad Sci U S A. 2012, 109: 12604-12609. 10.1073/pnas.1204302109.PubMed CentralPubMedView ArticleGoogle Scholar
- Takabayashi M, Adams LM, Pochon X, Gates RD: Genetic diversity of free-living Symbiodinium in surface water and sediment of Hawaii and Florida. Coral Reefs. 2012, 31: 157-167. 10.1007/s00338-011-0832-5.View ArticleGoogle Scholar
- Rodriguez-Lanetty M, Phillips WS, Weis VM: Transcriptome analysis of a cnidarian – dinoflagellate mutualism reveals complex modulation of host gene expression. BMC Genomics. 2006, 7: 23-10.1186/1471-2164-7-23.PubMed CentralPubMedView ArticleGoogle Scholar
- Rodriguez-Lanetty M, Wood-Charlson EM, Hollingsworth LL, Krupp DA, Weis VM: Temporal and spatial infection dynamics indicate recognition events in the early hours of a dinoflagellate/coral symbiosis. Mar Biol. 2006, 149: 713-719. 10.1007/s00227-006-0272-x.View ArticleGoogle Scholar
- Sunagawa S, Wilson EC, Thaler M, Smith ML, Caruso C, Pringle JR, Weis VM, Medina M, Schwarz JA: Generation and analysis of transcriptomic resources for a model system on the rise: the sea anemone Aiptasia pallida and its dinoflagellate endosymbiont. BMC Genomics. 2009, 10: 258-10.1186/1471-2164-10-258.PubMed CentralPubMedView ArticleGoogle Scholar
- Voolstra CR, Schwarz JA, Schnetzer J, Sunagawa S, Desalvo MK, Szmant AM, Coffroth MA, Medina M: The host transcriptome remains unaltered during the establishment of coral–algal symbioses. Mol Ecol. 2009, 18: 1823-1833. 10.1111/j.1365-294X.2009.04167.x.PubMedView ArticleGoogle Scholar
- De Salvo MK, Sunagawa S, Fisher PL, Voolstra CR, Iglesias-Prieto R, Medina M: Coral host transcriptomic states are correlated with Symbiodinium genotypes. Mol Ecol. 2010, 19: 1174-1186. 10.1111/j.1365-294X.2010.04534.x.View ArticleGoogle Scholar
- Peng S, Wang Y, Wang L, Chen WU, Lu C, Fang L, Chen C: Proteomic analysis of symbiosome membranes in cnidaria-dinoflagellate endosymbiosis. Proteomics. 2010, 10: 1002-1016.PubMedView ArticleGoogle Scholar
- Ganot P, Moya A, Magnone V, Allemand D, Furla P, Sabourault C: Adaptations to endosymbiosis in a cnidarian-dinoflagellate association: differential gene expression and specific gene duplications. PLoS Genet. 2011, 7: e1002187-10.1371/journal.pgen.1002187.PubMed CentralPubMedView ArticleGoogle Scholar
- Levy O, Kaniewska P, Alon S, Eisenberg E, Karako-Lampert S, Bay LK, Reef R, Rodriguez-Lanetty M, Miller DJ, Hoegh-Guldberg O: Complex diel cycles of gene expression in coral-algal symbiosis. Science. 2011, 331: 175-10.1126/science.1196419.PubMedView ArticleGoogle Scholar
- Wooldridge SA: Is the coral-algae symbiosis really ‘mutually beneficial’ for the partners?. Bioessays. 2010, 32: 615-625. 10.1002/bies.200900182.PubMedView ArticleGoogle Scholar
- Meyer E, Weis VM: Study of cnidarian-algal symbiosis in the “Omics” age. Biol Bull. 2012, 223: 44-65.PubMedGoogle Scholar
- Weis VM, Davy SK, Hoegh-Guldberg O, Rodriguez-Lanetty M, Pringe JR: Cell biology in model systems as the key to understanding corals. Trends Ecol Evol. 2008, 23: 369-376. 10.1016/j.tree.2008.03.004.PubMedView ArticleGoogle Scholar
- Schönberg CHL, Loh WKW: Molecular identity of the unique symbiotic dinoflagellates found in the bioeroding demosponge Cliona orientalis. Mar Ecol Prog Ser. 2005, 299: 157-166.View ArticleGoogle Scholar
- Granados C, Camargo C, Zea S, Sanchez JA: Phylogenetic relationships among zooxanthellae (Symbiodinium) associated to excavating sponges (Cliona spp.) reveal an unexpected lineage in the Caribbean. Mol Phylogenet Evol. 2008, 49: 554-560. 10.1016/j.ympev.2008.07.023.PubMedView ArticleGoogle Scholar
- Pochon X, Gates RD: A new Symbiodinium clade (Dinophyceae) from soritid foraminifera in Hawai'i. Mol Phylogenet Evol. 2010, 56: 492-497. 10.1016/j.ympev.2010.03.040.PubMedView ArticleGoogle Scholar
- Hill M, Allenby A, Ramsby B, Schönberg C, Hill A: Symbiodinium diversity among host clionaid sponges from Caribbean and Pacific reefs: evidence of heteroplasmy and putative host-specific symbiont lineages. Mol Phylogenet Evol. 2011, 59: 81-88. 10.1016/j.ympev.2011.01.006.PubMedView ArticleGoogle Scholar
- Hill M, Wilcox T: Unusual mode of symbiont repopulation after bleaching in Anthosigmella varians: acquisition of different zooxanthellae strains. Symbiosis. 1998, 25: 279-289.Google Scholar
- Hill MS: Spongivory on Caribbean reefs releases corals from competition with sponges. Oecologia. 1998, 117: 143-150. 10.1007/s004420050642.View ArticleGoogle Scholar
- Hill MS, Hill AL: Porifera (Sponges). Encyclopedia of Inland Waters, Volume 2. Edited by: Likens GE. 2009, Oxford: Elsevier, 423-432.View ArticleGoogle Scholar
- Rivera A, Hammel J, Haen K, Danka ES, Cieniewicz B, Winters IP, Posfai D, Wörheide G, Lavrov DV, Knight SW, Hill MS, Hill AL: RNA interference in marine and freshwater sponges: actin knockdown in Tethya wilhelma and Ephydatia muelleri by ingested dsRNA expressing bacteria. BMC Biotechnol. 2011, 11: 67-10.1186/1472-6750-11-67.PubMed CentralPubMedView ArticleGoogle Scholar
- Richardson C, Hill M, Runyen-Janecky L, Hill A: Experimental manipulation of sponge: bacterial symbiont community composition with antibiotics: sponge cell aggregates as a unique tool to study animal: microbe symbiosis. FEMS Microbiol Ecol. 2012, 81: 407-418. 10.1111/j.1574-6941.2012.01365.x.PubMedView ArticleGoogle Scholar
- Hill MS, Hill AL, Lopez J, Peterson KJ, Pomponi S, Diaz MC, Thacker RW, Adamska M, Boury-Esnault N, Cárdenas P, Chaves-Fonnegra A, Danka E, De Laine B, Formica D, Hajdu E, Lobo-Hajdu G, Klontz S, Morrow CC, Patel J, Picton B, Pisani D, Pohlmann D, Redmond NE, Reed J, Richie S, Riesgo A, Rubin E, Russell Z, Rützler K, Sperling EA, et al: Reconstruction of family-level phylogenetic relationships within Demospongiae (Porifera) using nuclear encoded housekeeping genes. PLoS One. 2013, 8: e50437-10.1371/journal.pone.0050437.PubMed CentralPubMedView ArticleGoogle Scholar
- Thacker RW, Hill AL, Hill MS, Redmond NE, Collins AG, Morrow CC, Spicer L, Carmack CA, Zappe ME, Pohlmann D, Hall C, Diaz MC, Bangalore PV: Nearly complete 28S rRNA gene sequences confirm new hypotheses of sponge evolution. Integr Comp Biol. 2013, 53: 373-387. 10.1093/icb/ict071.PubMed CentralPubMedView ArticleGoogle Scholar
- Hill MS: Symbiotic zooxanthellae enhance boring and growth rates of the tropical sponge Anthosigmella varians forma varians. Mar Biol. 1996, 125: 649-654. 10.1007/BF00349246.View ArticleGoogle Scholar
- Andersonn AJ, Gledhill D: Ocean acidification and coral reefs: effects on breakdown, dissolution, and net ecosystem calcification. Annu Rev Mar Sci. 2013, 5: 321-348. 10.1146/annurev-marine-121211-172241.View ArticleGoogle Scholar
- Scalera-Liaci L, Sciscioli M, Lepore E, Gaino E: Symbiotic zooxanthellae in Cinachyra tarentina, a non-boring demosponge. Endocyt Cell Res. 1999, 13: 105-114.Google Scholar
- De Wit P, Pespeni MH, Ladner JT, Barshis DJ, Seneca F, Jaris H, Overgaard Therkildsen N, Morikawa M, Palumbi SR: The simple fool’s guide to population genomics via RNA-Seq: an introduction to high-throughput sequencing data analysis. Mol Ecol Res. 2012, 12: 1058-1067. 10.1111/1755-0998.12003.View ArticleGoogle Scholar
- Riesgo A, Andrade SCS, Sharma PP, Novo M, Pérez-Porro AR, Vahtera V, González VL, Kawauchi GY, Giribet G: Comparative description of ten transcriptomes of newly sequenced invertebrates and efficiency estimation of genomic sampling in non-model taxa. Front Zool. 2012, 9: 1-24. 10.1186/1742-9994-9-1.View ArticleGoogle Scholar
- Pérez‒Porro AR, Navarro‒Gómez D, Uriz MJ, Giribet G: A NGS approach to the encrusting Mediterranean sponge Crella elegans (Porifera, Demospongiae, Poecilosclerida): transcriptome sequencing, characterization and overview of the gene expression along three life cycle stages. Mol Ecol Resour. 2013, 13: 494-509. 10.1111/1755-0998.12085.View ArticleGoogle Scholar
- Novo M, Riesgo A, Fernández-Guerra A, Giribet G: Pheromone evolution, reproductive genes, and comparative transcriptomics in Mediterranean earthworms (Annelida, Oligochaeta, Hormogastridae). Mol Biol Evol. 2013, 30: 1614-1629. 10.1093/molbev/mst074.PubMedView ArticleGoogle Scholar
- Novara D, Gay A, Lacomme C, Shaw J, Ridout C, Douchkov D, Hensel G, Kumlehn J, Schweizer P: HIGS: host-induced gene silencing in the obligate biotrophic fungal pathogen Blumeria graminis. Plant Cell. 2010, 22: 3130-3141. 10.1105/tpc.110.077040.View ArticleGoogle Scholar
- Ewen-Campen B, Shaner N, Panfilio KA, Suzuki Y, Roth S, Extavour CG: The maternal and early embryonic transcriptome of the milkweed bug Oncopeltus fasciatus. BMC Genomics. 2011, 12: 61-10.1186/1471-2164-12-61.PubMed CentralPubMedView ArticleGoogle Scholar
- Anders S, Huber W: Differential expression analysis for sequence count data. Genome Biol. 2010, 11 (10): R106-10.1186/gb-2010-11-10-r106.PubMed CentralPubMedView ArticleGoogle Scholar
- Supek F, Bošnjak M, Škunca N, Šmuc T: REVIGO summarizes and visualizes long lists of gene ontology terms. PLoS One. 2011, 6: e21800-10.1371/journal.pone.0021800.PubMed CentralPubMedView ArticleGoogle Scholar
- Julian D, Statile JL, Wohlgemuth SE, Arp AJ: Enzymatic hydrogen sulfide production in marine invertebrate tissues. Comp Biochem Physiol. 2002, 133: 105-115.View ArticleGoogle Scholar
- Taylor MW, Radax R, Steger D, Wagner M: Sponge-associated microorganisms: evolution, ecology, and biotechnological potential. Microbiol Mol Biol Rev. 2007, 71: 295-347. 10.1128/MMBR.00040-06.PubMed CentralPubMedView ArticleGoogle Scholar
- Luo BH, Carman CV, Springer TA: Structural basis of integrin regulation and signaling. Annu Rev Immunol. 2007, 25: 619-647. 10.1146/annurev.immunol.25.022106.141618.PubMed CentralPubMedView ArticleGoogle Scholar
- Bond JS, Beynon RJ: The astacin family of metalloendopeptidases. Protein Sci. 1995, 4: 1247-1261. 10.1002/pro.5560040701.PubMed CentralPubMedView ArticleGoogle Scholar
- Ben-Shlomo R: The molecular basis of allorecognition in ascidians. Bioessays. 2008, 30: 1048-1051. 10.1002/bies.20848.PubMedView ArticleGoogle Scholar
- Buddemeier RW, Fautin DG: Coral bleaching as an adaptive mechanism - a testable hypothesis. Bioscience. 1993, 43: 320-326. 10.2307/1312064.View ArticleGoogle Scholar
- Cunning R, Baker AC: Excess algal symbionts increase the susceptibility of reef corals to bleaching. Nat Clim Change. 2012, 3: 259-262. 10.1038/nclimate1711.View ArticleGoogle Scholar
- Yuyama I, Watanabe T, Takei Y: Profiling differential gene expression of symbiotic and aposymbiotic corals using a high coverage gene expression profiling (HiCEP) analysis. Mar Biotechnol. 2011, 13: 32-40. 10.1007/s10126-010-9265-3.PubMedView ArticleGoogle Scholar
- Hoegh-Guldberg O: Climate change, coral bleaching and the future of the world's coral reefs. Mar Freshwater Res. 1999, 50: 839-866. 10.1071/MF99078.View ArticleGoogle Scholar
- Hoegh-Guldberg O, Bruno JF: The impact of climate change on the world’s marine ecosystems. Science. 2010, 328: 1523-1528. 10.1126/science.1189930.PubMedView ArticleGoogle Scholar
- Camacho C, Coulouris G, Avagyan V, Ma N, Papadopoulos J, Bealer K, Madden TL: BLAST+: architecture and applications. BMC Bioinforma. 2009, 10: 421-10.1186/1471-2105-10-421.View ArticleGoogle Scholar
- Conesa A, Götz S, Garcia-Gomez JM, Terol J, Talon M, Robles M: Blast2GO: a universal tool for annotation, visualization and analysis in functional genomics research. Bioinformatics. 2005, 21: 3674-3676. 10.1093/bioinformatics/bti610.PubMedView ArticleGoogle Scholar
- O'Neil ST, Dzurisin JDK, Carmichael RD, Lobo NF, Emrich SJ, Hellmann JJ: Population-level transcriptome sequencing of nonmodel organisms Erynnis propertius and Papilio zelicaon. BMC Genomics. 2010, 11: 310-10.1186/1471-2164-11-310.PubMed CentralPubMedView ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited.