Discussion
Strong transcriptional responses to SCTLD exposure in both coral species
Exposure to stony coral tissue loss disease in these experiments caused evident transcriptional responses in both coral species. In general, we observed a depression of host metabolic pathways and a corresponding increase in expression of stress response genes and nuclear processes (Figure S2). This is unsurprising given consistency in transcriptomic responses to myriad stressors, a phenomenon characterized as a conserved stress response among diverse coral taxa (G. Dixon et al., 2020; Wang et al., 2009). While there was evidence of a strong transcriptional response to disease exposure in host corals, there was little variation in the transcriptional response of symbionts in both experiments. This may be due to lower alignment rates to the symbiont transcriptomes relative to the coral hosts (6.5 ± 0.3% forCladocopium and 15.6 ± 0.8% for Durusdinium versus 61.2 ± 0.7% for M. cavernosa and 60.3 ± 1.6% for O. faveolata , respectively), or comparatively poor annotations in symbiont references. Therefore, we have focused primarily on interpretation of the coral host datasets to minimize speculation with the symbiont datasets.
Through comparative analyses between coral species in the transmission experiment, we found consistent overlap in transcriptional responses following exposure to SCTLD. There were 681 DEGs and 21 GO terms shared between disease-exposed M. cavernosa and O. faveolata(Figures 2; 3), which may be an underestimation of orthologous genes between species, since some orthogroups had different gene-level annotations. The majority of DEGs (~64%) and all GO terms were expressed or enriched in the same direction in response to SCTLD exposure, suggesting that there are many conserved pathways among coral species functioning in the same roles. This relationship has been corroborated in a recent study by MacKnight and colleagues (2022), which examined transcriptional responses of seven coral species (including our study species) to white plague disease. They reported expression of immunity and cytoskeletal arrangement gene pathways across all species following disease exposure, with variation in intracellular protein trafficking pathways related to differences between disease susceptibility among and within species. Similarly, we observed higher enrichment of protein trafficking GO and KOG terms in M. cavernosa , which is less susceptible to SCTLD than O. faveolata(Figure 3; Figure S2) (Meiling et al., 2021; NOAA, 2018).
Conversely, a minority of DEGs showed inverse relationships between species following exposure to SCTLD (Figure 2). Despite this, functional enrichment of GO and KOG terms were largely similar between species (Figure 3; Figure S2), potentially indicating modulation of conserved gene pathways between M. cavernosa and O. faveolata (i.e., different species using the same gene pathways in different ways). This modulation may be indicative of plasticity related to disease susceptibility (e.g., MacKnight et al., 2022), and warrants additional investigation with other species affected by SCTLD.
Traylor-Knowles and colleagues (2021) identified differentially expressed gene processes implicated in coral immunity following exposure to SCTLD. In this study, we observed 26 DEGs in SCTLD-exposed M. cavernosa associated with these same immune processes, and 19 DEGs in O. faveolata . These included two apoptosis regulation pathways that were differentially expressed inM. cavernosa only, which are implicated in cell death and necrosis (D’Arcy, 2019; Fuess et al., 2017). Extracellular matrix genes (four forM. cavernosa and three for O. faveolata ) were generally downregulated in both species, and are hypothesized to be related to wound healing and the prevention of lesion progression (Traylor-Knowles et al., 2021; Young et al., 2020). Four transforming growth factor-beta (TGF-β) DEGs were downregulated in M. cavernosa only, which are likely part of the coral’s immune response (Fuess et al., 2020; Traylor-Knowles et al., 2021). Two NF-𝜅B activation genes for each species were upregulated (though one M. cavernosa DEG was downregulated), where this transcription factor is well-known for its role in immune activation and potentially in regulation of the coral-algal symbiosis (Voolstra et al., 2009; Williams et al., 2018). This is perhaps related to the hypothesis that SCTLD may be a virus targeting Symbiodiniaceae (Veglia et al., 2022; Work et al., 2021), but this observation requires further study.
Four peroxidases were upregulated in M. cavernosa , but ten DEGs had variable expression levels in O. faveolata . Peroxidases have been shown to be a component of the innate immune response to disease and thermal stress (Mydlarz & Harvell, 2007; Palmer, 2018). Protein tyrosine kinase (PTK) genes, which promote inflammatory responses to pathogens in corals (Fuess et al., 2016), had variable expression levels (nine in M. cavernosa and three inO. faveolata ), corroborating that cytokine production may play a role in the response to SCTLD exposure (Traylor-Knowles et al., 2021). Finally, one WD-repeat gene per species was downregulated, which is likely involved in apoptotic and immune responses (Aranda et al., 2011). While these genes do not provide the complete picture of coral’s transcriptomic responses to SCTLD, their differential expression in both studies provides evidence of their importance in the immune response to this disease and may be useful in the identification of biomarkers of disease exposure in wild corals.
Effects of antibiotic treatments on coral transcriptional patterns
Treatment of corals with amoxicillin appears to result in a ‘normalization’ of transcriptional pathways associated with the response to SCTLD. When comparing expression profiles across all genes, treated corals appeared more similar to apparently-healthy controls than they did their original diseased state (Figure 4). Comparisons of diseased versus healthy samples in both transmission and intervention experiments revealed mostly similar patterns, with downregulation of metabolic pathways and upregulation of stress response, immune, and nuclear processes (Figure S4). The majority of shared DEGs between diseased individuals in both experiments were expressed in the same direction (~64%), however, examination of treated versus diseased colonies showed an inverse relationship of expression patterns across the majority of shared DEGs (~76%, Figure 5). This implies a reversal of the same transcriptomic mechanisms involved in the immune response to SCTLD following treatment with amoxicillin, and provides evidence that disease intervention may be beneficial to the coral beyond removal of potential pathogens and co-occurring opportunistic microbes that may be affecting host transcription. As studies evaluating the effectiveness of antibiotic treatments on diseased corals often focus on visible observations (i.e., lesion progression and/or quiescence; Forrester et al., 2022; Neely et al., 2020; Shilling et al., 2021; Walker et al., 2021) or shifts in microbial communities (Sweet et al., 2011, 2014) as a means of assessing treatment success, this hypothesis requires further testing to elucidate the impacts of antibiotic treatment on other members of the holobiont including the coral host. For example, antibiotics have additional immune-modulating and anti-inflammatory impacts on humans beyond alteration of associated microbial communities (Pradhan et al., 2016), which may prove beneficial to corals affected by SCTLD.
Examination of the SCTLD-associated immune responses described by Traylor-Knowles and colleagues (2021) in our datasets also corroborated the reversal of transcriptional patterns following amoxicillin treatment. Eleven immune-related pathways were differentially expressed in diseased versus healthy corals, with fourteen in treated versus diseased corals. While overlap was minimal between treatment comparisons (four identical DEGs), expression patterns were often inversely related among immune classes. Four extracellular matrix genes were upregulated in treated colonies compared to downregulated in diseased samples from the transmission experiment (no DEGs were observed for diseased versus healthy colonies in the intervention experiment). Two NF-𝜅B activation DEGs were found for each treatment comparison and were upregulated in diseased versus downregulated in treated colonies. Three PTK genes were upregulated in diseased colonies while two matching genes (and one additional) were downregulated following treatment. TGF-β and WD-repeat DEGs were also inversely-expressed between diseased and treated colonies, indicative of a shift in immune responses following antibiotic treatment.
Evaluation of antibiotic treatment as an SCTLD intervention
Amoxicillin is effective at slowing or halting SCTLD lesion progression on individual colonies, including in a field setting (Forrester et al., 2022; Neely et al., 2020; Shilling et al., 2021; Walker et al., 2021). There are multiple limitations of antibiotic effectiveness on broad-scale disease intervention efforts, including that this approach often requires multiple re-treatments, continuous monitoring, and is labor-intensive (Neely et al., 2020; Walker et al., 2021). The time investment is especially important when considering reef systems with substantially more corals than southeast Florida, such as the Flower Garden Banks that recently experienced its first potential signs of a SCTLD outbreak (Johnston et al., 2023). While application of amoxicillin shows a high rate of success (~95%) in halting individual lesions, it does not necessarily prevent the formation of new lesions on a treated colony through time (e.g., Shilling et al., 2021). Finally, it is currently unknown what impacts antibiotic application may have on diverse coral reef ecosystems, notably potential antibiotic resistance in microbial communities (Griffin et al., 2020; Liu et al., 2020). Antibiotic treatment may additionally have negative impacts on healthy coral microbiomes, potentially increasing susceptibility of corals to other stressors such as elevated temperatures (Connelly et al., 2022).
Alternative treatment approaches are therefore warranted for further investigation in the mitigation of SCTLD, as well as future coral disease outbreaks. In particular, probiotic treatments are gaining research interest due to their potential benefits to coral survival following disease exposure (Peixoto et al., 2021), thermal stress (Doering et al., 2021; Santoro et al., 2021), and exposure to pollutants (Silva et al., 2021). Probiotic treatments are in development for SCTLD (Deutsch et al., 2022), and field trials are necessary to assess treatment effectiveness in a reef environment (Peixoto et al., 2021). Phage therapy also shows promise for the treatment of bacterial pathogens that affect corals (Jacquemot et al., 2018; Teplitski & Ritchie, 2009; Thurber et al., 2020), but many of these approaches are not yet fully operational.
While treatment of individual coral colonies with antibiotics is not intended to be a long-term solution to curbing the spread of SCTLD, it may represent a feasible method of mitigating impacts on high-value conservation and restoration targets, particularly ecologically-important, rare, and reproductively-viable members of the population. This is of particular importance while disease diagnostics are still being developed for SCTLD and other diseases. Disease diagnostics, when operational and scalable, may instead allow broad mitigation of pathogen transport via disease vectors and sources, potentially eliminating the need for colony-level disease intervention. Continued examination of gene mechanisms and functions will also improve understanding of coral immune responses (Traylor-Knowles et al., 2022), facilitating targeted treatment approaches in future disease outbreaks. Disease response efforts focusing on identification and diagnosis, mitigation of spread, and treatment of affected individuals require a holistic understanding of coral immunity and resilience at the individual level in order to maximize conservation and restoration success at the population level.