Gene expression analysis
Total RNA was extracted from dry frozen coral samples (size ~0.5 - 1.5cm) with RiboZolTM RNA Extraction Reagent (VWR Life Science) followed by DNase treatment (DNaseI, Zymo Research) and phenol-chloroform extraction. We performed the DNase treatment in 50 μl with 1U of DNase I at 37°C for 15 min and – to increase the efficiency – with additional 1U for 25 min. 1μg of RNA was reversely transcribed with High-Capacity cDNA reverse transcription kit (ThermoFisher Scientific) with RiboLock RNase inhibitor (ThermoFisher Scientific). Reverse transcription quantitative PCR (RT-qPCR) reactions were ran in 12 μl with PowerUpTM SYBRTM Green Master Mix (Applied Biosystems) for 40 cycles. Four gene expression profiles –pActin, pEF-1a (elongation factor 1), pCalm (calmodulin) andpAHC (AdenosylHomoCysteinase) - were compared in five different treatments and pEF-1a showed the highest expression stability upon heat stress and thus chosen as the reference gene. All primer sequences are listed in Table S1.
Expression of target genes was calculated relatively to the control treatment at time 0 with ΔΔCt method.
\begin{equation} \text{ΔΔ}\text{Ct}=2^{-\left[\left(\text{Ct}_{target,\ treated}-\text{Ct}_{ref,\ treated}\right)-(\text{Ct}_{target,\ control}-\text{Ct}_{ref,\ control}\right]}\nonumber \\ \end{equation}
ΔΔCt calculation method returns gene expression relative to control coral at time 0 without reflecting RNA levels at other timepoints. Moreover, we analyzed mostly genes involved in cell signaling whose gene expression can change very promptly after the signaling occurs and the change can last for only a very short period of time. Thus, the gene expression levels are similar between treatments for most of the time, with only a few differing timepoints. For these reasons, we excluded time as a variable from our statistical analysis and analyzed expression rates individually for each timepoint using One-way ANOVA with Tukey’s post hoc test. We sampled corals at different times only to increase the chance of capturing gene expression changes, not to compare the entire gene expression patterns between different treatments. Including time as a variable decreases the probability that a gene expression difference will be detected.
To compare gene expression profiles between experimentally preconditioned (PC) and naturally acclimatized (WT) corals during heat stress, linear regression model with Pearson correlation was calculated for mean single gene expression and mean gene expression ratios between PC and WT.