RNA editing generates mRNA isoforms with distinct stabilities that may expand the thermal tolerance of mRNA and proteins in Mytilus species

Abstract

Ectothermic organisms may expand their thermal tolerance by producing multiple protein isoforms with differing thermal sensitivities. While such isoforms commonly originate from allelic variation at a single locus (allozymes) or from gene duplication that gives rise to paralogs with distinct thermal responses, this study investigated mRNA editing as an alternative, post-transcriptional mechanism for generating mRNA variants. Cytosolic malate dehydrogenase (cMDH) was examined in foot tissue of two congeners of the marine mussel genus Mytilus, which occupy different thermal environments. Multiple editing events were detected within the mRNA coding region in both species. Editing sites were species-specific, with no shared positions identified. In M. coruscus, editing occurred at 117, 123, 135, 190, 195, 204, 279, and 444, while in M. galloprovincialis, editing was detected at 216 and 597. Each species exhibited multiple edited mRNA variants, and these isoforms were associated with differential protein expression. These findings suggest that mRNA editing may contribute an additional layer of molecular variation. The generation of diverse mRNA isoforms from a single DNA coding sequence may enhance enzymatic flexibility across temperature ranges, supporting eurythermal physiological performance and mitigating thermal stress. Moreover, the presence of multiple edited transcripts within individual organisms raises important caveats about the limitations of approaches that deduce amino acid sequences or estimate adaptive variation solely from genomic data.

Keywords: RNA editing; Temperature adaptation; mRNA isoform; mRNA stability.

Figure 1

RNA editing in cytosolic malate dehydrogenase (cMDH) transcripts from two mesophilic mussel species Mytilus coruscus (Mc) and M. galloprovincialis (Mg) (n=8 individuals per species). A: Percentages of editing sites in cMDH mRNAs. ^*: P<0.05. B: RNA editing level for each editing position.

Figure 2

Percentages of cytosolic malate dehydrogenase (cMDH) mRNA isoforms generated by RNA editing in the coding sequence region Mytilus coruscus (Mc) and M. galloprovincialis (Mg) (n=8). Color-coded bars on the right represent different mRNA isoforms shown in the left panel.

Figure 3

Changes in codon adaptation index (CAI) of cytosolic malate dehydrogenase (cMDH) before and after RNA editing Mytilus coruscus (Mc) and M. galloprovincialis (Mg) (n=8). Calculations before and after editing are based on gDNA and cDNA sequences, respectively. ^*: P<0.05. Maximum value of the upper whisker represents the 1^st quartile (Q1)+1.5×interquartile range (IQR), and minimum value of the lower whisker represents Q1–1.5×IQR (McGill et al., 1978).

Figure 4

Pattern of ΔG_fold among RNA editing-derived isoforms of cytosolic malate dehydrogenase (cMDH) mRNA in two mussel species across different temperatures Mytilus coruscus (Mc) and M. galloprovincialis (Mg) (n=8). A: Frequency distribution of ΔG_fold values for edited cMDH mRNA variants (cDNA) at different temperatures. Thin vertical lines represent frequency (percentage) of observed ΔG_fold values. B: Coefficient of variation of ΔG_fold for each species before (gDNA) and after (cDNA) editing across different temperatures. ^*: P<0.05. Shaded areas represent Gaussian kernel-smoothed trends.

Figure 5

Relationship between ΔG_fold and thermal sensitivities of edited cMDH mRNA isoforms in two mussel species Mytilus coruscus (Mc) and M. galloprovincialis (Mg). Estimated ranges of lethal temperature (LT₅₀) are indicated in the panels. LT₅₀ values were inferred based on previously established regression from 22 marine mollusks (Liao et al., 2021) (y=–92.01x–189.81). Calculations of ΔG_fold were performed with default parameters at a temperature of 37°C.

Figure 6

Characteristics of cMDH mRNA secondary structures before and after RNA editing in two mussel species Mytilus coruscus (Mc) and M. galloprovincialis (Mg). A: Differences in secondary structures between mRNA isoforms before (gDNA) and after (cDNA) editing. Sensitivity scores were calculated based on edited structures (cDNA) compared to original structures (gDNA), with higher sensitivities indicating higher structural differences. B: Comparison of base pairs and internal loops in different mRNA isoforms before (gDNA) and after (cDNA) editing. C: Correlations between the number of internal loops and hairpins and ΔG_fold before (gDNA) and after (cDNA) editing. ns: Not significant; ^*: P<0.05. Maximum value of upper whisker represents the 1^st quartile (Q1)+1.5×interquartile range (IQR), and minimum value of lower whisker represents Q1–1.5×IQR (McGill et al., 1978).