in

Physiological and molecular responses of lobe coral indicate nearshore adaptations to anthropogenic stressors

Physiological responses

Small fragments from five source colonies from the two experimental sites (N- and O-sites) were used to conduct a reciprocal transplant experiment in Maunalua Bay, Hawaii (Fig. 1). The results revealed clear physiological response differences between the two populations. The transplantation resulted in a significant reduction in the average tissue layer thickness (TLT) in only one treatment: O-corals transplanted to N-site (O → N) (Tukey-HSD, P-adj < 0.001, Fig. 2A, SI.1A). Contrary to the expectation of coral lipid content being reduced by environmental stress, the total tissue lipid content in O-corals showed a marginally significant increase when transplanted to N-site (Tukey-HSD, P = 0.059, SI.1A), while N-corals showed very little change in their lipid content between the sites (Fig. 2B). Two-way ANOVA showed no population or transplant-site effects, but a significant interaction between the two factors (P = 0.034). Comparing lipid contents pre- and post-experiment revealed that lipid contents of O-corals increased significantly at both sites, while those of N-corals did not (Fig. 2C).

Figure 1

Experimental location and design. (A) A map of study locations, (B) a diagram of the reciprocal transplant experimental design, and (C) temperatures and light intensity profiles during the reciprocal transplant experiment of N- and O-sites. N-site is located in the inner bay where it receives direct run-off, while O-site is located at the end of the bay, exposed to offshore currents with its environmental conditions closer to those of the offshore, despite its proximity to the shoreline (for detailed water quality profiles, see 8,10).Gray circles in (B) reflect the three colonies used for proteomic analysis. The map was generated using maptools (v. 1.0-1) package in R60.

Full size image
Figure 2

Physiological results of the 30-day reciprocal transplant experiment of P. lobata from N- and O-sites. (A) Tissue layer thickness, (B) tissue lipid content, and (C) short-term growth rate (starting weight adjusted to 6 g), (D) tissue lipid content before and after the experiment. Different letters indicate a statistically significant difference. Error bars denote standard error.

Full size image

The short-term growth rates, assessed using a common-garden experiment, revealed a significantly higher average growth rate in N-corals than O-corals. N-corals grew on average 5.57% (0.035 g) of their initial weight over 11 weeks, while O-corals grew 2.57% (0.011 g) (Fig. 2D, Wilcoxon rank sum test, P = 1.53 × 10–8). The experiment was conducted at the Kewalo Marine Laboratory (KML), University of Hawaii at Manoa using an outdoor flow-through seawater tank. During the experimental period, the maximum tank water temperature was 29.95 °C and the minimum temperature was 26.39 °C, with an average of 27.9 °C. The shade cover was placed over the tank during the experimental period, which resulted in a maximum light intensity of 44,089.2 lx (1014 PPFD), with a daytime average of 2553.4 lx (58.7 PPFD).

Proteomic responses

Transplantation effects on proteomic profiles

A label-free shotgun proteomics approach was used to study the dynamic protein-level response to transplantation. Combining all proteomic results, a total of 3635 proteins were identified with high confidence (FDR < 0.01) when correlated to a customized P. lobata protein database predicted from the transcriptome 14. An average of 2,777 proteins were identified across all biological replicates in each treatment. A total of 1977 proteins were shared across all four treatments, which likely represents the basic homeostatic functions of corals. Proteins unique to a treatment ranged from 105 to 233 (Table S1, Fig. S1A).

To evaluate the differences revealed in the protein-level responses among the treatments, normalized spectral abundance factor (NSAF) values for each protein were visualized using non-metric multidimensional scaling (NMDS). NMDS analysis revealed a clear separation of the protein abundance profiles between the two populations regardless of the transplant sites (ANOSIM, R = 0.7074, P = 0.004, Fig. S2). Transplantation had a greater effect on the final protein profile of O-corals, while protein abundances of N-corals were relatively similar between the transplant sites.

Response differences in transplant to Nearshore site (N → N vs. O → N)

A total of 3,290 distinct coral proteins were identified at N-site: 2365 (72%) were shared between N → N and O → N corals, with 402 proteins unique to N → N corals, and 523 unique to O → N corals (Fig. S1B). Gene Ontology (GO) enrichment analysis (CompGO)15, identified 19 enriched GO terms specific to N → N corals and 42 terms specific to O → N corals from the three GO categories (Biological Process [BP], Molecular Function [MF], and Cellular Component [CC]) (SI.2A). Enriched terms in N → N corals included peptidase inhibitor activity, oxidoreduction coenzyme metabolic process, lyase activity, and regulation of protein polymerization. The enriched terms for O → N corals included detoxification, antioxidant activity, lipid oxidation, intracellular protein transport, tricarboxylic acid (TCA) cycle, and purine-containing compound metabolic process.

Quantitative analysis on the protein level identified 138 proteins significantly more abundant (referred to as the abundant-proteins) in N → N corals and 276 in O → N corals (Fig. 3A, Table S1). GO analysis identified a total of 35 terms from the abundant-proteins in O → N corals (SI.2A), while none met the statistical cutoff in N → N corals. The enriched terms in O → N corals indicated that the abundant-proteins were dominated by those associated with amino acid metabolic process, oxidation–reduction process, and vesicle membrane/coat (SI.2A). Abundant proteins identified in O → N corals further suggested more activities in metabolic and regulatory pathways, including detoxification and glutathione pathways (i.e. antioxidant activity) (Fig. S3). The three most abundant proteins (with annotation) in N → N corals were associated with immune function (hemicentin-2: m.9723, glycoprotein 340: m.13233, and lectin MAL homologue: m.12716), while those in O → N corals were involved with detoxification function (arsenite methyltransferase:m.16246, E3 ubiquitin-protein ligase TRIM71:m.2139, and glutathione-S-transferase:m.16994).

Figure 3

Protein abundance comparison between N- and O-corals. (A) At N-site, and (B) at O-site. Fold differences and Z-statistic values (Z-score) were used to generate volcano plots. Colored points indicate proteins with |fold|> 0.5, |Z-score|> 2 at FDR = 0.01. Proteins associated with key GO terms were colored in different colors, and the top 10 abundant proteins in each population are annotated. The bottom bars indicate the total numbers of significantly abundant proteins for each population.

Full size image

Response difference in transplant to the offshore site (N → O vs. O → O)

A total of 3236 distinct coral proteins were identified at O-site: 2217 (68.5%) were shared between the two populations, 656 unique to N → O corals, and 363 to O → O corals (Fig. S1C). GO analysis identified 35 enriched terms specific to N → O, which involved amino acid biosynthetic process, ATP metabolic process, TCA cycles, fatty acid oxidation, and monosaccharide metabolic process. There were 15 specific GO terms in O → O corals, including nucleotide monophosphate biosynthetic process, intracellular protein transport, vesicle organization, and GTP binding (SI.2B).

Quantitative analysis on protein abundances indicated a total of 665 proteins to be significantly differentially abundant at O-site: N → O corals had 155 abundant-proteins, and O → O corals had 510 abundant-proteins (Fig. 3B). GO analysis resulted in identifying 39 enriched terms from abundant proteins in O → O corals, while only one met the cutoff in N → O corals (SI.2B). Although the number of abundant-proteins and enriched terms identified in O → O corals were relatively high, the enriched terms predominantly consisted of cellular functions related to protein translation; organonitrogen biosynthetic process and organic acid metabolic process, both leading to single child terms for BP, CC, and MF (tRNA aminoacylation for protein translation, cytosolic large ribosomal subunit, and tRNA aminoacyl ligase activity). The enriched term in N → O corals was a non-specific term of ‘extracellular region’, indicating that despite the higher number of abundant-proteins, the main functional difference between N → O and O → O corals was an enhanced protein translation activity in O → O corals.

Response comparisons to cross transplantation

Effects of cross transplantation yielded a more diverse proteomic stress-response in O-corals as they moved nearshore than N-corals as they were moved offshore (Fig. S2). The total number of abundant-proteins between the sites was much higher for O-corals (440, O → N vs. O → O) than N-corals (135, N → N vs. N → O) (Table S1), and the number of unique GO terms identified between the sites was also higher in O-corals (69, SI.2C) than in N-corals (46, SI.2D). The number of overlapping proteins between the sites was lower in O-corals than in N-corals (70% vs. 79%), and log-fold changes of all identified proteins between the sites were significantly larger for O-corals than N-corals (Wilcoxon Rank-Sum test, P = 6.02 × 10–9), all emphasizing the larger metabolic reshuffling needed to respond to cross transplantation in O-corals. GO enrichment analysis indicated that N-corals responded to transplantation to O-site with increased abundance of proteins involved in amino acid biosynthesis, fatty acid beta oxidation, TCA cycle, chitin catabolism, coenzyme biosynthesis and translational initiation. O-corals responded to transplantation to N-site by increasing the abundance of proteins associated with detoxification, antioxidant activity, protein complex subunit organization, and multiple metabolic processes (amino acid, fatty acid, ATP, monosaccharide, and carbohydrate derivative) (SI.2E). The shared responses between the cross-transplanted corals (N → O and O → N corals) included increased proteins involved in fatty-acid beta oxidation, TCA cycle, carbohydrate derivative catabolic process, pyridoxal phosphate binding, and ‘oxidoreductase activity acting on the CH-CH group of donors with flavin as acceptor’, likely representing the effects of transplantation to a non-native environment.

Proteome patterns across the four treatments

Comparing enriched GO terms across all treatments (SI.2E) highlighted the unique state of O → N corals; O → N corals had a much higher number of uniquely enriched GO terms (n = 27) compared to those in the rests (4 in O → O, 5 in N → N, and 15 in N → O corals). The most notable difference among the treatments was enrichment of detoxification and antioxidant activity exclusively in O → N corals (Fig. 4). Also, lipid oxidation was highly enriched in O → N corals with four terms associated to this category identified (Fig. 4, SI.2E).

Figure 4

Enriched GO terms uniquely identified to specific treatment groups. Treatment groups are shown in the right column (e.g. N-coral = N-corals at both sites, N-site = N- and O-corals at N-site, CrossT = cross transplantation). The heat-map represents P-values for the associated GO terms. The GO terms are grouped by the parent–child terms with the most parent term in bold (for values, see SI-2E).

Full size image

Examining the relative abundance of individual proteins associated with detoxification (‘detox-proteins’) revealed the following interesting patterns. (1) Distinct sets of proteins were abundant in different treatments, rather than all detox-proteins to be elevated in one treatment, and the direction and magnitude of responses to transplantation were protein specific and varied between populations (Fig. S4A). (2) Two peroxiredoxin (Prx) proteins, Prx-1 (m.6147) and Prx-6 (m.9595), dominated the relative abundance of detox-proteins by having over an order of magnitude higher abundance values, and they were consistently more abundant in N-corals than O-corals (ave. 44%, Kruskal Test, P = 0.004–0.01) (Fig. S4B, SI.1B). (3) Some proteins with the same or similar annotations had contrasting responses between the populations. For example, Prx-4 (m.17739), which belongs to the same subfamily as Prx-1, was significantly more abundant in O-corals at both sites (Fig. S4B, SI.2F,G), while Prx-1 was more abundant in N-corals. Similarly, seven peroxidasin (PXDN) homologs were identified, of which m.17686 was significantly more abundant in O → N corals, while m.9432 was significantly more abundant in N → N corals (Fig. S4B, SI.2F), suggesting that the two populations potentially utilize different class/kind of enzymes as primary proteins in detoxification/antioxidant pathways. Of the seven PXDN homologs, two (m.1440, m.9432) were consistently higher in N-corals, two (m.10928, m.15200) were consistently higher in O-corals, and three (m.12572, m.17686, m.9657) increased abundance at N-site in both corals, but m.12572 and m.17686 being higher in O-corals, while m.9657 higher in N-corals (Fig. S3B).

To ascertain that the proteins with the same annotations are indeed different proteins, sequences of matched peptides were assessed for those that showed contrasting responses. The pairwise comparison of Prx-1 and Prx-4 showed only seven of the total 65 peptides (11%) were identical between the two, revealing that these protein sequences are significantly different and they each have unique peptides that be detected and quantified accurately (SI.1C1). Similarly the majority of PXDN-like proteins identified had no overlapping peptides between the contrasting pairs (0–19%, median = 0, SI.1C2), indicating that corals possess multiple types of PXDN, and N- and O-corals respond to stressors with different sets of PXDN.

In addition to lipid oxidation being significantly enriched in O → N corals, a single term (fatty acid beta-oxidation,) was also enriched in N → O corals, which suggests that cross-transplantation had an effect on lipid oxidation processes. However, the abundances of most proteins associated with lipid oxidation were higher in O-corals than N-corals at both sites (Fig. S4A). Statistically, three proteins (medium-chain sp acyl-CoA:m.22274, very-long-chain sp. acyl-CoA:m.17984, and trifunctional enzyme subunit alpha:m.6724) showed a difference in abundance between the two populations at N-site (Fig. S4C) and one (isovaleryl-CoA dehydrogenase:m.27714) at O-site, all of which were higher in O-corals than N-corals.


Source: Ecology - nature.com

George Shultz PhD ’49, renowned statesman and former professor, dies at 100

Descriptive multi-agent epidemiology via molecular screening on Atlantic salmon farms in the northeast Pacific Ocean