As shown in previous studies, anthropogenic CO₂ emissions are increasing and cause lots of environmental changes, such as global warming; however, this change is also threatening the ocean. Several pieces of evidence pointed out that ocean acidification would impact marine life, especially in the early developmental stages. To understand the long-term and transgenerational effects of marine fish, we applied a marine model animal, Oryzias melastigma, to a CO₂-induced acidified condition (pH 7.6). This study focused on the larvae's physiological acclimation and how they face acidified stress through gene expression changes. Furthermore, we checked the epigenetic modifications, like DNA methylation, to know if there were any parental effects under this stress.

1. Breeding conditions:

   We used the pH controller (E-M03, Macro, Taiwan) to monitor the pH value in each tank and pumped the extra CO₂ into the seawater to maintain a low pH condition (pH 7.6). The fertilized eggs from the control tank (WT, pH 8.1) were transferred immediately to acidified seawater until mature (F0), then swapped for the next generation (F1) and kept for the second generation (F2).

   The seawater system was kept in a similar condition (salinity 31–33, temperature 28±1 °C, 10 h dark/14 h light cycle) at Marine Research Station (the Institute of Cellular and Organismic Biology, Academia Sinica, Taiwan). During the exposure period, the water conditions were also recorded, like pH, temperature(T, °C), salinity (S, ‰), total alkalinity (TA,μM) and the total dissolved inorganic carbon (DIC,μM) and pCO₂ was calculated by CO2SYS module.

   The pH and temperature (°C) were measured using a pH meter (pH 3310, WTW, Germany) equipped with a pH electrode (SenTix® 41, WTW, Germany).

   TA (μM) was measured following a published method (Sarazin et al. 1999).

   Ref: Sarazin, G., Michard, G., & Prevot, F. (1999). A rapid and accurate spectroscopic method for alkalinity measurements in sea water samples. Water Research, 33(1), 290-294. doi: 10.1016/S0043-1354(98)00168-7

   DIC (μM) was measured using an analyzer (AS-C3, Apollo SciTech, Newark, DE, USA) with quantification via a non-dispersive infrared CO2/H2O analyzer (LI-7000, LI-COR, Lincoln, NE, USA).

2. Larve growth patterns: 

   We recorded the growth patterns from 2 days post-fertilization (dpf) to 7 dpf in each group. Measured the parameters (eye diameter, eye distance, body width, body height, and total length) under a microscope on a randomly selected individual every day.

3. Oxygen consumption and ammonium excretion: 

   For the oxygen taking and ammonium excretion, 60 eggs were used for measurement as a simple point at 5 dpf. The eggs were gently transferred to the airtight glass respiration chambers with filtered seawater, and then the oxygen concentrations were monitored in real-time by an OXY-4 mini multichannel fiber optic oxygen transmitter (PreSens) within one hour. After testing, water samples were collected from chambers and then measured the ammonium concentration following the published method (Holmes et.al., 1999).

   Ref: Holmes, R. M., Aminot, A., Kérouel, R., Hooker, B. A., & Peterson, B. J. (1999). A simple and precise method for measuring ammonium in marine and freshwater ecosystems. Canadian Journal of Fisheries and Aquatic Sciences, 56(10), 1801-1808. doi: 10.1139/f99-128

4. Heartbeat frequency: 

   A randomly chosen individual was gently placed under a microscope, and their heartbeat was recorded. Those videos were analyzed with Fiji to obtain a clear diagram. Further analyze the frequency of the heartbeat by changing the area of the heart.

5. RNA-seq profile: 

   All RNA samples were extracted using the RNeasy Plus Universal Mini Kit (QIAGEN, Germany) in accordance with the manufacturer's instructions. Genomic (Taipei, Taiwan) performed the quality test using the BioAnalyzer2100 and Agilent RNA 6000 Nano Kit. The passed samples were sequenced using the NovaSeq 6000 (Illumina) 150 bp paired-end sequencing libraries for RNA-seq, giving 6 Gb data per sample, which was likewise processed by Genomics (Taipei, Taiwan). The reads were sufficiently trimmed and mapped to the Oryzias melastigma genome (ASM292280v2, NCBI) using the CLC genomic workbench (v. 20.0, CLC Bio-Qiagen, Denmark). The RNA-seq data were further analyzed by iDEP.96 (integrated Differential Expression Pathway analysis) to identify the significantly different Gene Ontology (GO) Molecular function pathway across groups. Based on the results, we selected some genes and then normalized reads count using Z-score to create a heatmap.

   Ref: Ge, S. X., Son, E. W., & Yao, R. (2018). iDEP: an integrated web application for differential expression and pathway analysis of RNA-Seq data. BMC bioinformatics, 19, 1-24. doi: 10.1186/s12859-018-2486-6

6. Gene expression: 

   To examine the gene expression of acid-base regulators, we synthesize the cDNA through SuperScript IV Reverse Transcriptase (Invitrogen, USA) and then measured the gene expression by a quantitative real-time PCR (qRT-PCR) with the Roche LightCycler 480 system (Roche Applied Science, Germany). The standard curves for each target gene were proven to be linear, with ribosomal protein L7 (rpl7) of O. melastigma serving as a reference gene.

7. Epigenetic analysis: 

   The gDNA of 5 dpf larvae was purified using the Wizard® Genomic DNA Purification Kit (Promega, United States). Sonication was used to fragment the gDNA before proceeding with the MeDIP (Methylated DNA Immunoprecipitation) study, which followed the manufacturer's technique for the EpiQuikTM Tissue Methylated DNA Immunoprecipitation Kit (Epigentek, United States). After anti-5-methylcytosine immunoprecipitation, the immunoprecipitated (IP) DNA was collected and specific primers determined the promoter’s methylation levels. To confirm the dynamic DNA methylation pattern, we did the bisulfide conversion by EZ DNA MethylationTM Kit (ZYMO Research, United States) and amplified the AE1a’s promoter with bisulfite-treated DNA-specific primer by PCR. The PCR products were sequenced by Genomics (Taipei, Taiwan) to determine unmethylated and methylated CpG sites, which were predicted by MethPrimer.