The multifocal effects of climate change on marine organisms have received considerable concern and attention from the scientific community over the past few decades (Brierley and Kingsford, 2009; Heuer and Grosell, 2014; Sumby et al., 2021). Marine organisms are affected both by the increasing temperatures of the ocean and also by CO₂-induced ocean acidification. The oceans have been acting as a CO₂ sink, absorbing nearly 30% of atmospheric CO₂ that has been rising for the past few centuries (Doney et al., 2009). Elevated aqueous CO₂ thus results in a set of chemical reactions in the seawater that reduce pH, carbonate ion (CO₃ ^2-) concentrations, and saturation states of calcium carbonate minerals (CaCO₃). Since the Industrial Revolution, ocean pH has declined by approximately 0.1, and is projected to decline an additional 0.1 – 0.4 units by the end of the century (Garcia-Soto et al., 2021).

Fishes exposed to elevated environmental CO₂ can experience acute hypercapnia and acidosis followed by an influx in plasma HCO₃ ^- to either compensate for or neutralize the acidosis (Heuer and Grosell, 2014). The consequences of fish being in a chronic state of balancing has yet to be elucidated. However, studies collectively show that CO₂ exposure affects sensory performance in fishes, e.g. altering taste preferences (Rong et al., 2020), impairing olfaction (Williams et al., 2019; Porteus et al., 2021), slowing retinal function (Chung et al., 2014), and reducing auditory sensitivity (Radford et al., 2021).

Previous studies suggest that neurosensory cues strongly influence fish behavior, interspecific interactions, and even habitat selection (Nagelkerken et al., 2019). This is especially true of olfactory and auditory cues that are used by larval fishes for orientation as well as locating and settling on quality nondegraded habitat (Kaplan and Mooney, 2016; Gordon et al., 2018; Bilodeau and Hay, 2022; Hu et al., 2022). Furthermore, sound can play a critical role in predator-prey interactions, territorial defense, and reproductive behavior (Looby et al., 2022). Auditory impairment could be especially detrimental for soniferous fishes, such as croakers and drums (Sciaenidae) or Arctic cod (Gadidae; Boreogadus saida) that vocalize to locate and aggregate with conspecifics as an intrinsic part of their reproductive behavior (Ramcharitar et al., 2006; Horodysky et al., 2008; Lindseth and Lobel, 2018; Riera et al., 2018). Therefore, the disruption of sensory functions by abiotic environmental and/or anthropogenic change could alter life histories both in single species contexts and across species interactions, resulting in potentially large-scale ecological and ecosystem consequences (Horodysky et al., 2022).

The specific mechanisms through which CO₂-induced ocean acidification hinders auditory performance remain inferred but as yet undemonstrated. Currently, a leading mechanistic hypothesis for observed CO₂-induced auditory deficits centers on the hypertrophy and/or asymmetric growth of otolithic auditory end organs (Bignami et al., 2013; Holmberg et al., 2019) in fish reared under high levels of CO₂ exposure that are consistent with projected future levels of ocean acidification (Radford et al., 2021). However, significant reduction in auditory performance has also been observed in fish exposed to elevated CO₂ conditions in the absence of changes in otolith morphology (Horodysky, in preparation). This suggests strongly that morphology alone may not be the sole or proximate mechanism hindering auditory performance, and that underlying neurological mechanisms are likely (sensu Nilsson et al., 2012). One feasible way to elucidate the molecular underpinnings of this process is to investigate environmentally-induced epigenetic disruption of gene expression associated with neurosensory performance (Gemenetzi and Lotery, 2014; Zhao et al., 2019).

MicroRNAs (miRNA) are epigenetic regulators (Yao et al., 2019), which alter gene expression by promoting messenger RNA (mRNA) degradation or translational inhibition. They consist of short non-coding RNA sequences and their expression levels may be influenced by environmental surroundings, especially environmental stressors (Bonin et al., 2019). The interaction of environment – miRNA – gene expression makes miRNAs integral components of an organism’s ability to adapt to environmental changes. While the effects of some abiotic environmental stressors (e.g. temperature and salinity) on miRNA expression in fishes have been documented (Bizuayehu et al., 2015), the effect of CO₂ –induced ocean acidification on miRNA expression remains relatively unexplored.

The purpose of this preliminary study was to determine if exposure to elevated aqueous CO₂ affects miRNA expression in whole brain tissue of Arctic cod (Boreogadus saida) juveniles. This study focused on miRNAs with potential mRNA targets associated with the auditory system either directly or indirectly.

We obtained an average of ~ 45 million raw reads per sample ( Supplementary Table S1 ). From these, 35,668 mature miRNA sequences were obtained from Arctic cod brain tissues. The vast majority (34,818 sequences) were unannotated, and 870 sequences were identified from known sources in miRbase (annotated; Table 2 ).

This is not surprising due to the lack of an annotated genome for the Arctic cod, which also does not have mature miRNA sequences in miRbase. However, the majority (76%) of the identified mature miRNA of interest showed substantial homogeneity (containing an E-value of ≤ 0.006) when aligned to the mature sequence of the reference organism. Considering all of the data, approximately 70% of annotated mature miRNAs identified overlapped between control and treatment samples (normalized expression values for each mature sequence are available in Supplementary Table S2 ).

Our dataset was further reduced by removing miRNAs that exhibited a normalized expression value level below 175 in either treatment. Among the total annotated mature miRNA sequences, 17 miRNAs of interest were identified by our phenotypic association analysis conducted in miRBase and additional literature. These miRNAs were either directly or indirectly associated with hearing loss in a model organism ( Table 3 ). However, among the total miRNAs of interest, 12 miRNAs were significantly differentially expressed (DE) between the control and low pH treatments ( Figure 1 ).

Inter-individual variance was observed in the expression levels of the 17 miRNA in the two pH treatments ( Figure 1 ). However, mir-135b, mir-144-3p, mir-140-3p, mir-218b, mir-460, and mir-9-1 were significantly DE across all three individuals within the pH 7.2 treatment. Mir-135b had the highest fold change between treatments (2.51; Table 3 ).

MiRNAs are integral components of organismal stress responses, but little is known of their role in maintaining neurosensory function under elevated CO₂ conditions. Using Arctic cod whole brain tissues, our study uncovered the significant differential expression of 12 miRNAs following exposure to acidified conditions. Our analytical focus on highly conserved miRNAs with known neurosensory roles in model organisms revealed several miRNAs of interest for future functional studies and charts a path forward for miRNA investigations of organisms without an annotated genome.

The select group of 17 miRNAs identified in this study have been directly or indirectly associated with hair cell damage and hearing loss in model organisms, and 12 of those were significantly DE. Cod exposed to ocean acidification conditions demonstrated significant upregulation of mir-101a and mir-218b, which can be demonstrative of chronic stress and indirectly associated with sudden sensory neural hearing loss (Masuda and Kanzaki, 2013). Significant DEs of mir-140-3p, mir-210, mir-24-1, mir-181c-3p, and mir-9-1 are directly associated with sudden sensory neural hearing loss in model organisms ( Table 3 ). We caution that mir-181c-3p is not a highly conserved miRNA, so we were unable to acquire the E-value or the percentage of query coverage; this miRNA needs validation in future experiments. Perhaps most interesting was the significant upregulation of mir-144-3p under low pH conditions. The upregulation of this miRNA has been correlated with GABAergic dysfunction (following Nilsson et al., 2012), which is hypothesized to play a role in CO₂-induced sensory impairment and is further associated with neurotoxicity and neural injury.

Strikingly, mir-135b showed the highest fold increase (2.51) in fish exposed to low pH. When upregulated, this miRNA helps protect against neuroinflammation and oxidative stress (Schell et al., 2022); the latter stressors have been correlated with hair cell degradation and hearing loss in zebrafish, mice, and humans (Huang et al., 2000; Canlon et al., 2013; Chen et al., 2022). Interestingly, mir-135b is hypothesized to play a role in inner ear development, maintenance, and protection, suggesting its importance in early life history. Yet this miRNA has also been highlighted in studies of elderly human subjects with and without hearing loss, suggesting the importance of its role as a protectant (Sekine et al., 2017). It is therefore possible that environmental exposure to high CO₂/low pH induces both neural tissue and hair cell damage in fishes, and that the upregulation of mir-135b evidences a strong compensatory physiological protective response.

There is ample evidence of auditory recovery in fishes following exposure to stressors (Scholik and Yan, 2001; Amoser and Ladich, 2003; Breitzler et al., 2020). Smith et al. (2004) demonstrated partial recovery of hair cells 8 days after acoustic trauma in goldfish (Carrasius auratus), with full recovery of auditory function in 14 days. Collectively, this raises an interesting question of whether environmentally-induced dysregulation of miRNAs associated with auditory performance would impede the potential for hair cell regeneration and auditory recovery in fishes. There remain interesting open questions about the deleterious effects of intensity and duration of CO₂ exposure on neurosensory system change and resulting impacts on recovery times from environmental stress. Concerns about auditory performance and recovery in marine organisms are magnified given the rapid increases in anthrophony and other marine noise over the past century as global oceans become louder and auditory cuescapes more complex (Duarte et al., 2021; Horodysky et al., 2022).

While several promising miRNAs associated with audition were detected in our study, our inferences were admittedly limited by a high degree of individual variation across small sample sizes. The individual variation in miRNA expression may be due to considerable intraspecific variation in auditory performance and/or differential resilience to environmental change (Ladich and Fay, 2013; Schunter et al., 2016; Horodysky, in preparation); individual trait diversity must be understood to predict ecological resilience to climate change (Ward et al., 2016). In fact, we suggest such individual variance in miRNA expression could actually be harnessed in future studies to provide insight into differential adaptation to environmental change. Some individuals may be more resilient to certain environmental stressors, with adaptive potential for positive transgenerational outcomes; conversely, transgenerational effects of ocean acidification may have adverse epigenetic consequences for poorly adapted lineages, reducing fitness of offspring (Schunter et al., 2019).

Other caveats of our approach bear consideration, yet we collectively believe that this technique provides an exciting new tool for studies of physiological performance of fisheries resources in the Anthropocene. The use of whole brain in our miRNA extractions may have contributed to within-treatment variance, and a finer dissection of brain structures could reduce potential for cross-contamination of tissue surrounding or encapsulated in the brain and adjacent otoliths (e.g., blood and endolymph). Whole brain samples may dilute region-specific concentrations of specialized miRNAs that are highly DE in very defined locations of the brain. Finer-scale brain-tissue expression profiles can also reveal more specific miRNA roles in the auditory process, while simultaneously identifying other neurosensory processes and potential impairments. To that end, we observed DE of miRNAs associated with olfactory (mi-140 and mir-214), reproductive (mir-29), and visual (retinal degeneration; mir-183, mir-182, mir-96, mir-7, mir-124) impairment. However, DE in the whole brain alone may not be sufficiently indicative to characterize visual or reproductive impairment; we strongly recommend also assaying additional tissues from relevant organ systems (i.e. retinas and/or gonads, in this example) in the future. Although beyond the scope of the present Brief Research Report, transcriptomics would allow researchers to connect the regulome with functional consequences, allowing far more explicit prediction/identification of novel miRNAs and their target genes that may be affected by CO₂-induced ocean acidification. Such an approach requires a genome assembly for the study species, which does not exist at present for B. saida; we suggest that this is a worthwhile undertaking to advance mechanistic understanding. Finally, and perhaps most excitingly, this miRNA approach can efficiently survey a study system or exposure protocol to generate mechanistic hypotheses across a myriad of physiological systems. Such an approach may be ideal to guide future experiments and extends well to studies of stressors resulting from climate change, ocean acidification, pollutants, and other human activities.

In this study, we identified 12 significantly DE miRNAs between low and high CO₂ exposures, which are either directly or indirectly associated with reduced auditory performance in Arctic cod. We suggest that CO₂ exposure induces miRNA dysregulation that can lead to or exacerbate hair cell damage, hinder recovery, and promote neuroinflammation; conjointly these may mechanistically drive or magnify the observed reductions in auditory performance in fish hearing studies. In species with observed otolith hypertrophy or asymmetry following exposure to ocean acidification, CO₂-induced miRNA dysregulation may be synergistic in causing observed auditory deficits. But more importantly, this neurological mechanism may be the proximal cause of auditory deficits in species with unaltered otolith morphology following CO₂ exposure. We contend that many neurological processes in fishes may be generally affected by elevated CO₂, however, the exact mechanistic underpinnings, as well as the extent and duration of this impairment may be species and/or life stage specific, warranting careful investigation. Collectively, we conclude that miRNA sequencing provides a promising approach to guide future physiological and behavioral studies of the effects of abiotic and anthropogenic change on organismal sensory structure and function, particularly when paired with physiological performance assays and miRNA functional studies.

The data presented in the study are deposited in the NIH SRA repository, BioProject accession PRJNA1007783; sample accession #s: SRX21439117-SRX21439122 and are accessible at: https://www.ncbi.nlm.nih.gov/sra/PRJNA1007783.

The animal study was approved by Hampton University Institutional Animal Care and Use Committee (IACUC). The study was conducted in accordance with the local legislation and institutional requirements.

CS analyzed the data, created figures, and wrote the manuscript. CB funded microRNA extraction and analysis, assisted with data analysis, and edited the manuscript. CM conducted fish husbandry, collected seawater carbonate chemistry samples, and edited the manuscript. TH funded and oversaw all fish husbandry components and CO₂ exposures, and edited the manuscript. AH planned experiments and logistics, selected and dissected fish, funded microRNA analysis, and assisted in manuscript writing and editing. All authors contributed to the article and approved the submitted version.