Computational protocol: The development of contemporary European sea bass larvae (Dicentrarchus labrax) is not affected by projected ocean acidification scenarios

Similar protocols

Protocol publication

[…] Larvae were homogenized individually, and total RNA was extracted as previously described (Vanderplancke et al. ). Microarray analysis: Transcription profiles of five samples per tanks (i.e., 15 samples per experimental condition, 45 samples in total) were performed using 44 K whole sea bass genome microarrays (Agilent) that contained 38,130 probes, providing high coverage of sea bass transcripts. Double-stranded cDNA was synthesized from 500 ng of total RNA using the Quick Amp Labeling kit, One color (Agilent). Labeling with cyanine 3-CTP, fragmentation of cRNA, hybridization, and washing were performed according to the manufacturer’s instructions (Agilent) at the Laboratoire de Génétique Moléculaire et d’Histocompatibilité (Brest, France). The microarrays were then scanned and the data extracted with the Agilent Feature Extraction Software. qPCR analysis: Total RNA of the 30 samples per condition was reverse-transcribed into cDNA using iScript™ cDNA Synthesis kit (Bio-Rad Laboratories, Hercules, CA, USA). Quantitative real-time polymerase chain reaction (qPCR) was used to specifically analyze seven candidate genes: osteocalcin, which are implied in osteoblast differentiation and mineralization (Lein and Stein ); carbonic anhydrase 2 (CA2) and carbonic anhydrase 4 (CA4), which contribute to CO2 hydration; NA+/H+ exchanger 1 (Slc9a1), NA+/H+ exchanger 3 (Slc9a3), NA+/HCO3 − co-transporter (Slc4a4), and HCO3 − transporters (Slc4a1), which are involved in dynamic adjustments of acid–base balance (reviewed in Heuer and Grosell ). The design of specific primers for each gene was performed using Primer3Plus software based on cDNA sequences available in public databases NCBI (https://www.ncbi.nlm.nih.gov) or SIGENAE (http://www.sigenae.org/) (Table ). The qPCR for each gene was performed using iQ™ SYBR® Green Supermix (Bio-Rad Laboratories, Hercule, USA) as previously described (Vanderplancke et al. ). Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) gene was chosen as the housekeeping gene as it showed minimal variance in its transcript level, both within and between conditions. [...] Data normality and homogeneity of all variances were tested with Kolmogorov–Smirnov and Brown–Forsythe tests, respectively. The macroscopically evident deformities index had to be rank transformed to obtain normality (Quinn and Keough ). Body mass and fork length were then analyzed using mixed model:yijk=μ+ASi+ACj+(AS×AC)ij+Tjk+eijk,where yijk is the phenotypic observation, μ is the sample mean, ASi is the effect of the ith age stage, and ACj is the effect of the jth acidification condition, which were all fitted as fixed effects as well as their interactions; Tjk is the effect of the kth tank nested in jth acidification condition fitted as a random effect, and eijk is the random residual effect. The percentage of larvae survival at 45 dph was analyzed using one-way ANOVA with the acidification condition as factor. The calcification ratio was analyzed using ANCOVAs with the acidification condition fitted as a fixed effect, the tank effect nested in the acidification condition fitted as a random effect, and the fork length as the co-variable. Other variables (i.e., percentage of deformities, enzymatic capacities, enzymatic ratios and genes expression profiles) were analyzed using the mixed model with the acidification condition fitted as a fixed effect and the tank effect nested in the acidification condition fitted as a random effect. The a posteriori Tukey’s honest significant difference (HSD) tests were used for mean comparisons when possible or replaced by Games and Howell tests when variances were not homogenous. Analyses were carried out using Statistica version 7.0 for windows (StatSoft, USA). The microarray expression data were analyzed at the BIOSIT Health Genomic core facility (Rennes, France). Data were normalized using Agilent Genespring GX software, and statistical analysis was performed with FactoMineR and nlme R packages using mixed model with the acidification condition as a fixed effect and the tank as a random effect. The P values were then corrected for multiple testing using Benjamini–Hochberg threshold. A significance level of α = 0.05 was used in all statistical tests. […]

Pipeline specifications

Software tools Agilent Feature Extraction, Primer3, Statistica, GeneSpring GX, FactoMineR
Databases SIGENAE
Applications Miscellaneous, qPCR
Organisms Dicentrarchus labrax
Diseases Musculoskeletal Diseases
Chemicals Carbon Dioxide