Abstract Background The wild stocks of Pinctada maxima pearl oysters found off the coast of northern Australia are of critical importance for the sustainability of Australia’s pearling industry. Locations inhabited by pearl oysters often have oil and gas reserves in the seafloor below and are therefore potentially subjected to seismic exploration surveys. The present study assessed the impact of a simulated commercial seismic survey on the transcriptome of pearl oysters. Animals were placed at seven distances (-1000, 0, 300, 500, 1000, 2000, and 6000 m) from the first of six operational seismic source sail lines. Vessel control groups were collected before the seismic survey started and exposed groups were collected after completion of six operational seismic sail lines (operated at varying distances over a four-day period). Samples from these groups were taken immediately and at 1, 3, and 6 months post-exposure. RNA-seq was used to identify candidate genes and pathways impacted by the seismic noise in pearl oyster mantle tissues. The quantified transcripts were compared using DESeq2 and pathway enrichment analysis was conducted using KEGG pathway, identifying differentially expressed genes and pathways associated with the seismic activity. Results The study revealed the highest gene expression and pathway dysregulation after four days of exposure and a month post-exposure. However, this dysregulation diminished after three months, with only oysters at -1000 and 0 m displaying differential gene expression and pathway disruption six months post-exposure. Stress-induced responses were evident and impacted energy production, transcription, translation, and protein synthesis. Conclusion Seismic activity impacted the gene expression and pathways of pearl oysters at distances up to 2000 m from the source after four days of exposure, and at distances up to 1000 m from the source one-month post-exposure. At three- and six-months post-exposure, gene and pathway dysregulations were mostly observed in oysters located closest to the seismic source at 0 and − 1000 m. Overall, our results suggest that oysters successfully activated stress responses to mitigate damage and maintain cellular homeostasis and growth in response to seismic noise exposure. Supplementary Information The online version contains supplementary material available at 10.1186/s12864-024-11091-7. Keywords: Transcriptome, Differential gene expression, Aquatic noise, Acoustic stress, Cellular stress, Enriched pathways, Oysters Introduction The Pinctada maxima pearling industry is one of Australia’s most valuable and iconic fisheries, creating significant economic and employment opportunities across Northern Australia. Pinctada maxima produce Australian South Sea pearls, which are the most highly regarded pearls in the world. The Australian cultured pearling industry was established in the 1950s, utilising pearl oysters collected from the wild stocks of 80 Mile Beach, Western Australia, which is the “only remaining significant wild-stock fishery for pearl oysters in the world” [[46]1]. Today, the Australian pearling industry is still underpinned by the collection of wild pearl oysters from 80 Mile Beach, which are then seeded with a nucleus and mantle allograft to initiate cultured pearl formation. In addition to their economic significance, Pinctada maxima are ecologically important as ecosystem engineers. They contribute to habitat complexity and biodiversity by providing hard substrate for other marine organisms and play a crucial role in nutrient cycling and water filtration. The locations in which pearl oysters naturally occur in northern Australia (Western Australia and Northern Territory) are also key prospecting areas for offshore oil and gas exploration. Offshore oil and gas is a large industry in northwest Australia with multiple projects in various stages of development, production and exploration [[47]2]. One of the main disturbances to the marine environment associated with oil and gas exploration and production includes noise pollution from seismic surveys [[48]2]. Seismic surveys are used to acquire data about the geological formations below the earth’s surface and produce maps showing potential locations of gas deposits. In the marine environment, seismic surveys are conducted using strings of airguns towed behind a vessel in parallel paths to form an array of airguns that can be fired simultaneously to produce a desired acoustic beam. Airguns produce intense, low-frequency impulsive sounds that travel through the water to reach the ocean floor. Some of the sound energy is reflected by features within the seafloor and travels back to the sea surface, where hydrophone streamers record it. Three variables characterise the sound related to seismic surveys: (1) the sound pressure, which acts in all directions and can be described by its magnitude and its temporal and frequency characteristics; (2) the particle motion, characterised by the oscillation back and forth in a specific direction and can be defined by its magnitude, direction of the motion, and temporal and frequency characteristics [[49]3]; (3) the ground motion, which is the movement of the seafloor in response to the acoustic signal. As anthropogenic activities in the marine environment increase, understanding the impact of noise such as seismic surveys on marine animals, particularly on ecosystem engineers like pearl oysters, is crucial. Low-frequency noise generated by seismic surveys can propagate over long distances and harm fish and marine mammals [[50]3, [51]4]. To date, the effect of seismic surveys has mostly been investigated on large mammals, such as whales and, to a lesser extent, on finfish [[52]5]. The impact on invertebrates has been poorly studied and is yet to be fully understood [[53]3, [54]4, [55]6]. Among marine invertebrates, most of the studies have focused on crabs, lobsters and squid [[56]6–[57]17], and only a few studies have investigated the effect of seismic energy on molluscs [[58]4, [59]5]. As such, there is currently a lack of dose-response data for stress-related effects in bivalve molluscs exposed to this type of noise [[60]3]. Only six publications have studied the impact of a seismic survey on molluscs in the field [[61]18–[62]23] and four of them were conducted on the same species: the scallop Pecten fumatus [[63]18–[64]21]. Three of those publications did not detect short- or long-term effect on the mortality, condition and adductor muscle strength of the scallops Pecten fumatus [[65]19–[66]21], whereas a single study reported an increase in mortality, changes in the behaviour and a compromised capacity for maintaining homeostasis and immunity [[67]18]. One study investigated the impact of a seismic survey on the clam Paphia aurea and observed an increase in biochemical parameters such as hydrocortisone, glucose and lactate levels but no change in clam density [[68]22]. The other studies available in the literature were performed in a laboratory setup and some have provided evidence that the wavelength used in seismic surveys can affect key functions in marine molluscs [[69]24, [70]25]. Indeed, exposure of mussels (Mytilus galloprovincialis) to low-frequency acoustic signals in a laboratory environment resulted in elevated values of biochemical stress parameters [[71]24]. The mussel, Mytilus edulis, closed its valves in response to sinusoidal vibratory signals, with the greatest sensitivity in the frequency range of 10 to 210 Hz [[72]25]. In response to pulsed sounds, Ruditapes phillipinarum reduced their surface relocation activity, moved to a position above the sediment-water interface, and closed their valves [[73]26]. Only two studies have investigated the impact of seismic noise on the transcriptome of an invertebrate. Both examined the snow crab Chionoecetes opilio [[74]27, [75]28]. Overall, comparing peer-reviewed studies is challenging because of the absence of a standardised methodology for conducting seismic impact studies on aquatic animals. Studies have used different seismic sources, which generated different sounds, different designs for the experiment, as well as different metrics to assess the responses of the animal. In addition, there is a lack of knowledge about the effect of low-frequency seismic air gun signals on pearl oysters. Part 1 of the present study revealed no impact on the mortality or pearl quality [[76]23]. Considering the potential for negative impact of real-world exposure levels on lobster and scallops [[77]4, [78]16, [79]18, [80]29], and to evaluate whether the pearling and the oil and gas industries can co-exist in northern Australia, the present study assessed the impact of a commercial seismic survey on the transcriptomes of pearl oysters P. maxima. Materials and methods Experimental design Samples were collected in conjunction with samples for other studies examining the effects of a simulated real-world seismic survey experiment on the pearl oyster Pinctada maxima mortality and pearl quality [[81]23, [82]30]. The heights of P. maxima ranged from 104 to 108 mm and length varied between 102 and 172 mm. McCauley et al. [[83]30] detailed the information regarding the acoustic properties of the seismic source exposure experiment including the maximum and cumulative levels of acoustic pressure such as root mean squared level, sound exposure level (SEL), particle motion, and ground motion, whereas [[84]23] outlined the experimental design, locations and utilised the seismic exposure metrics measured by [[85]30] to assess any changes in mortality and pearl production by P. maxima post exposure. The experimental design is summarised in the paragraph below, however, more details are available in [[86]30] and [[87]23]. Briefly, wild adult P. maxima pearl oysters were hand-collected offshore from 80 Mile Beach in August 2018 and translocated to the experimental site on the seafloor in bottom culture panels at a depth of 15–25 m (tide dependent) with eight oysters per panel, following standard industry practices. In total, 1,390 panels were split across all the rows of the experiment, with the number of panels in each row determined by the sampling design (see Tables S2 and S4 from [[88]23] for full details). Panels were placed across seven parallel rows at the following closest point of approach (CPA) distances: −1000, 0, 300, 500, 1000, 2000, and 6000 m, with 0 m being the reference line directly underneath a seismic sail line run by BGPs RV Explorer during the first passages as described in Fig. [89]1 and S1 by [[90]23]. The closest point of approach is the shortest distance between a point of interest and a line transited by the vessel (see Fig. S1 of [[91]23]). The translocated oysters were allowed to acclimatise for one month before being exposed to the seismic survey. In September 2018, the 3D seismic source survey was carried out using a 2600 cubic inch air gun array [[92]30]. Pearl oysters at the experimental site were exposed to seismic energy similar to that experienced by marine fauna during a commercial seismic survey [[93]30]. At the start of the experiment, the vessel followed a 20 km line commencing at the 0 m row without activating its seismic equipment; this was the vessel control line. Oysters subjected to the vessel control were then retrieved from the seafloor and transported to a pearl farm in Roebuck Bay where the panels containing oysters were suspended on farming longlines in the water column. The farm was at a distance sufficient not to receive any energy from the seismic survey. Fig. 1. [94]Fig. 1 [95]Open in a new tab Total number of differentially expressed genes (DEGs) for all sampling times between the vessel control (Day 1) and oysters exposed over six sailing lines to the seismic energy (Day 5) across 7 distances: −1000 m, 0 m, 300 m, 500 m, 1000 m, 2000 m, and 6000 m Following the translocation of the control oysters, the vessel conducted six operational sail lines moving from the 0 m row towards the − 1000 m oyster groups at distances of 0, −500, −1000, −1500, −2000, and − 2500 m from the 0 m reference line, with each line being separated in time by either 12–24 h. Randomly selected groups of panels were retrieved from the seafloor and transported to the pearl farm at Roebuck Bay after the first, second, fourth, and sixth operational lines, which resulted in 35 groups of oysters, corresponding to the seven distances or lines multiplied by the five exposure/ retrieval times (vessel control (VC) or Day 1, Day 2, Day 3, Day 4, and Day 5). Oyster panels from each of the 35 groups were retrieved for destructive laboratory analyses directly from the exposure site in September 2018 (after 4 days of exposure). The remaining pearl oysters were translocated to the pearl farm, where they were maintained under standard farming conditions, before further panels from each treatment group were destructively sampled in October 2018 (1 month post-exposure), December 2018 (3 months post-exposure) and March 2019 (6 months post-exposure). Collection of samples Upon collection from the pearl farm, the oysters were transported approximately 2 km to the laboratory in batches of four panels (32 oysters). The panels of pearl oysters were cleaned of epibionts using a brush, then hosed with freshwater. Pearl oysters initially close tightly when emersed, so the animals’ tissues were not exposed to freshwater. The panel identity (numbered tag) and the number of missing and dead pearl oysters were recorded [[96]23]. Panels were processed randomly and without knowledge of specific treatments by the researchers involved to prevent sampling bias (see [[97]23] Supplemental Material for tagging protocols). Following initial processing, ten individual pearl oysters were placed in an opening basket to allow the animals to relax and undertake a natural air-gaping response. Once the pearl oyster had opened, a plastic wedge was carefully inserted between the shell valves to keep them open. Then, the pearl oyster was sacrificed for sampling by severing the adductor muscle using a scalpel blade. Sections of mantle tissue (> 100 mg) were aseptically excised from the same location in each oyster, placed in a cryotube and snap frozen in liquid nitrogen for transcriptomics analyses. Transcriptome analysis Samples processed included all seven distances or lines for two exposures only: vessel controls or Day 1 (VC) and Day 5, which were the oysters that received the higher number of exposures to the seismic energy. Samples of three to four oysters were pooled, which resulted in three biological replicates per treatment. The frozen mantle tissues were removed from the tubes and placed inside a petri dish on dry ice. Thin slices of frozen tissue were excised with a sterile scalpel blade, and quickly weighed to prevent the defrosting of the tissue. One hundred mg of tissue was returned to the dry ice on a sterile petri dish and chopped finely while frozen. The samples were then transferred to 2 mL tubes containing ceramic beads (MPBio, lysing matrix-D tubes, 116913050-CF) and 500 µL of ice-cold TRIzol reagent (Invitrogen, Carlsbad, USA). All samples were kept at 4 °C until RNA extraction commenced. RNA extraction The samples were bead beaten three times at 3000 beats per minute for 2.5 min with 30-second intervals in pre-cooled (at −20 °C) 96 sample adapters using the Qiagen Tissue Lyser II. After each bead beating session, the tubes were briefly centrifuged in a pre-cooled (4⁰C) centrifuge to reduce foaming. An additional 200 µL of TRIzol reagent was added thereafter, and the samples were vortexed and incubated for 10 min at room temperature. The samples were transferred to new, clear, low-bind 2 mL tubes and centrifuged for 5 min at maximum speed to remove any debris. The supernatant was then transferred to a new tube, and an equal volume of ice-cold absolute ethanol was added. The remainder of the extraction was completed using the Direct-zol RNA MiniPrep Kit (Zymo Research, Irvine, CA, USA) according to the protocol provided, with the exception of an additional 700 µL RNA wash step and 2 min dry spin before elution. On column DNase I, treatment was also executed using the DNase I provided and kit directions. The RNA was finally eluted in 60 µL nuclease-free water. After extraction, the concentration of the RNA was measured using the Qubit 3.0 Fluorometer (Invitrogen, Carlsbad, USA) with the Qubit RNA Assay Kit (Invitrogen, Carlsbad, United States). Next, the purity of the RNA was determined using a NanoDrop 1000 (Thermo Fisher Scientific, Waltham, USA), while the RNA integrity was evaluated using the LabChip GX Touch 24 (PerkinElmer, Waltham, United States), 24 DNA5K/RNA/CZE Chip with HT RNA Reagents Standard Sensitivity reagents and protocol. The remaining RNA was transferred to 0.5 mL GenTegra-RNA safe screw cap tubes (GenTegra, Pleasanton, CA, USA) and stored at −80 °C. Library preparation, sequencing and data analysis The RNA samples were shipped on ice and submitted to NovoGene (Singapore/Hong Kong) for mRNA library preparation and sequencing. The mRNA from the oyster mantle samples was enriched using oligo(dT) beads, and the rRNA was removed using the Ribo-Zero kit. The mRNA was randomly fragmented upon the addition of the fragmentation buffer. Next, cDNA was synthesised using random hexamer primers followed by second-strand synthesis (Illumina, San Diego, CA, USA). After terminal repair, A-tailing and sequencing adaptor ligation, the double-stranded cDNA libraries were completed through size selection and PCR enrichment according to the Illumina protocol (Novogene sequencing report). The prepared libraries were pooled and sequenced on the NovaSeq platform. A total of 1202.6 GB of raw sequencing data were generated. Quality filtering of the raw sequencing data was carried out to remove reads containing adapters, poly-Ns > 10%, and low-quality reads with a Q-score ≤ 5. An index of the P. maxima reference genome was built, and paired-end RNASeq reads were aligned to the reference genome using HISAT2 v2.2.1 [[98]31]. The P. maxima genome is not publicly available. The genome was sequenced and assembled in FRDC-funded project 2018 − 198 led by The University of Sunshine Coast, Australia. The genome was assembled in collaboration with Dovetail Genomics. A Dovetail whole-genome shotgun strategy was used to sequence and assemble the P. maxima genome from PacBio long-read data and completed using long-range Chicago and Hi-C library data scaffolded with the HiRise software. The final ~ 1.2 Gb assembly consists of 1,116 scaffolds with a N50 of 87,994,994, which includes 14 pseudomolecules greater than 40 Mb in length, corresponding to the chromosome number for P. maxima. The completeness of the genome assembly was assessed with BUSCO using the Eukaryota (odb10) database. The gene prediction contained 94.72% of the highly conserved orthologs (92.55% complete and single-copy) in the eukaryotic lineage.” The transcripts were quantified using StringTie v2.1.5 [[99]32]. DESeq2 R package (2_1.6.3) [[100]33] was used to infer and compare differential expression across distances and sampling events. Benjamin and Hochberg’s approach [[101]34] was used to calculate the adjusted p (padj) values to control the false discovery rate and differentially expressed genes (DEGs) with an adjusted p ≤ 0.05 and log2FoldChange ≥ 0.5 were retained. The intersection of differentially expressed genes between the distances and times was analysed and visualised using Intervene [[102]35]. The annotation of genes expressed was deferred from the P. maxima genome annotation. Pannzer [[103]36], EggNOG-mapper v2 [[104]37], InterProScan v5.54_87.0 and UniProtKB were employed to improve the functional annotation of the DEGs. Functional annotation is the process of attaching biological information to sequences of genes, i.e., providing the gene’s names and functions. We hypothesise that the DEGs capture not only the response of the oysters to the seismic activity, but also the response to any other biotic and abiotic factors present at that site/ distance or during the collection and transportation of the panels. DEGs that were shared between different distances were considered to be associated with seismic activity. Pathway and process enrichment analysis was carried out using KEGG Pathway. All genes in the genome were used as the enrichment background. Terms with a p-value < 0.01, calculated based on the accumulative hypergeometric distribution [[105]38], a minimum count of 3, and an enrichment factor > 1.5 were clustered based on their membership similarities. Kappa scores [[106]39] were used as the similarity metric upon performing hierarchical clustering on the enriched terms, and sub-trees with a similarity of > 0.3 are considered a cluster, which is represented by the statistically most significant term. The Benjamini-Hochberg procedure was used to calculate the q-score to account for multiple testings [[107]40]. All lists were finally combined, and the best p-value was selected where more than one term from the lists as the final p-value, as described by [[108]41]. Annotation classes were then grouped into broader Biological Processes to promote comparative analysis. KOBAS-i/ KOBAS v3 [[109]42] was then used to carry out further detailed KEGG pathway enrichment analyses using the closely related Pacific oyster (C. gigas) as a reference organism. Enriched terms with a corrected p-value < 0.05 were selected. Similarly to the DEGs, this study focused on identifying pathways that were shared among exposed oysters collected from several distances. Results Transcriptome sequencing and assembly RNA-seq libraries were generated for 183 samples, each sample representing a pool of 3 to 4 pearl oyster mantles. The samples processed related to two levels of seismic exposure: vessel control (VC or Day 1) and exposure over six sail lines (Day 5) from seven distances and were sampled at four different sampling times: September, October, and December 2018, and March 2019. The RNA extractions yielded an average of 208 mg of RNA per sample with average ratios of 2.05 and 2.11 for 260/280 nm and 260/230 nm, respectively. After quality filtering, an average of 6.5 × 10^9 raw reads per sequenced sample was obtained and was used for transcriptome assembly. RNA seq libraries aligned with an average percentage of 71% ranging from 48 to 79% (Table S1S). Differentially expressed genes (DEGs) A total of 10,414 differentially expressed genes (DEGs) between the vessel control and the exposed oysters were recorded and varied with the distance from the seismic source. Oysters at 6000 and − 1000 m displayed the least differentially expressed genes, with 256 and 458 DEGs, respectively (Fig. [110]1). In contrast, the largest number of DEGs was recorded from oysters located at 2000 m with 3,896 DEGs, followed by 0 (1,614), 1000 (1535), 300 (1226) and 500 m (1001) (Fig. [111]1). Overall, approximately 80% of all the DEGs were recorded in September and October 2018 (Fig. [112]2). A gradual decrease in DEGs was observed over time from September 2018 to March 2019 at 500 and 2000 m. An increase in DEGs was observed from September to October 2018 followed by a decrease from October to December 2018 and a slight increase between December 2018 and March 2019 at 0, 300 and 1000 m (Fig. [113]2). Furthermore, no noticeable trend was observed over time at −1000 and 6000 m, which were the distances showing the lowest number of DEGs. Fig. 2. [114]Fig. 2 [115]Open in a new tab Total number of differentially expressed genes (DEGs) at every sampling time between the vessel control (Day 1) and oysters exposed to six operational seismic sail lines (Day 5). * No data was available in March 2019 for the 1000 m distance The number of up- and down-regulated DEGs are shown in Table [116]1 and varied according to the distance. An increase in the number of up-regulated DEGs was observed between September and October 2018 at 0, 300, 1000 and 6000 m, whereas a decrease was noticed at −1000, 500, and 2000 m (Table [117]1). Then, between October and December 2018, the number of up-and down-regulated DEGs decreased for all distances except for the up-regulated DEGs for 6000 m. Of 10,414 DEGs, only 3,012 DEGs could be annotated. Table 1. Comparison of the number of up- and down-regulated DEGs between vessel controls (day 1) and oysters exposed over six sailing lines to the seismic energy (day 5) across all distances and sampling times Distance Regulation September 2018 October 2018 December 2018 March 2019 Total −1000 m Up 84 43 28 69 224 0 m Up 203 223 50 231 707 300 m Up 92 325 146 71 634 500 m Up 195 136 59 60 450 1000 m Up 30 436 152 ^a 618 2000 m Up 1492 387 107 17 2003 6000 m Up 7 23 49 53 132 Total Up 2103 1573 591 501 4768 −1000 m Down 105 62 44 42 253 0 m Down 265 535 43 210 1053 300 m Down 58 572 33 154 817 500 m Down 275 213 80 41 609 1000 m Down 25 778 87 a 890 2000 m Down 1536 308 39 10 1893 6000 m Down 12 43 44 32 131 Total Down 2276 2511 370 489 5646 [118]Open in a new tab ^aNo data was available for the 1000 m distance in March 201 Annotation and intersection of DEGs Overall, most of the DEGs were unique to each distance and sampling time, and the majority of DEGs recorded at each distance at one sampling time were not shared with the other sampling times (Fig. [119]3). In September 2018, the highest number of shared genes between two distances was observed between 0 and 2000 m with 237 shared DEGs, followed by 101 DEGs shared between 2000 and 500 m, and 55 DEGs shared between 2000 and − 1000 m (Fig. [120]3). Oysters at 0 and − 1000 m, which received the highest SEL (Fig. [121]1), shared two down-regulated DEGs: TIMP 2 ([122]ANN28417) and a hypothetical protein FSP39_015959 ([123]ANN17612) also found in Pinctada imbricata that might be a rootletin-like protein (Fig. [124]3). The latter was also down-regulated in exposed oysters at 0 m in October 2018. The distances 0, 300 and 500 m shared only one down-regulated DEG, i.e. MYH2 or Myosin heavy chain-2 C striated muscle ([125]ANN38325, Fig. [126]3). Fig. 3. [127]Fig. 3 [128]Open in a new tab Upset plots of the DEGs shared between the vessel control (Day 1) and oysters exposed over six sailing lines to the seismic energy (Day 5) across distances at each sampling time One month after the exposure (October 2018), distances 0, 300, and 1000 m shared 154 DEGs (Fig. [129]3). There were also 144 common DEGs between 1000 and 300 m, 137 DEGs between 1000 and 0 m and 82 DEGs between 300 m and 0 m. Exposed oysters sampled at 0 m and − 1000 m, which received most of the energy, shared four DEGs, including two down-regulated genes: LIPI ([130]ANN00995) and a hypothetical protein FSP39_022102 ([131]ANN37254) found in P. imbricata similar to solute carrier family 4, sodium bicarbonate cotransporter, member 8 and two up-regulated genes: GSTCD ([132]ANN05328), hypothetical protein FSP39_009367 ([133]ANN26599) found in P. imbricata similar to SLX1A gene. One common gene that could not be annotated was shared between the 0 m, 300 m, and 500 m group oysters, while exposed oysters at −1000 m, 0 m, and 300 m, shared DEG PHKG2 ([134]ANN23636). Furthermore, exposed oysters at −1000 m, 0 m, 300 m, and 500 m all differentially expressed the Hlf gene ([135]ANN20733, Fig. [136]3). In addition, Klf11 ([137]ANN01904) was differentially expressed by seismic exposed oysters from all distances except for 6000 m. In December 2023, a significant reduction in the number of DEGs was observed (Table [138]1; Fig. [139]3); 13 DEGs were shared between 500 and 1000 m and 12 DEGs were common to 0 and 1000 m (Fig. [140]3). Exposed oysters at 0 and − 1000 m down-regulated two genes: Alk ([141]ANN21601) and REST ([142]ANN30031), while those at −1000 and 300 m strongly up-regulated the EEF1A1 gene ([143]ANN16547). In March 2019, the number of DEGs was similar to December 2018 (Table [144]1). The distances 0 and 300 m shared 46 DEGs, and 0 and − 1000 m shared 6 DEGs, including two down-regulated genes identified as putative ankyrin repeat protein ([145]ANN23144), an uncharacterised gene ([146]ANN37111), and two up-regulated genes: Tyrosinase-like protein 1 ([147]ANN33133), and an uncharacterised gene ([148]ANN31371). TMPRSS9 ([149]ANN07098) and an uncharacterised gene ([150]ANN18294) were also differentially expressed (Fig. [151]3). Exposed oysters at −1000 m, 0 m, and 300 m shared 3 DEGs: Hlf ([152]ANN20733) and an unidentified gene ([153]ANN35332) were up-regulated at −1000 m and 0 m and down-regulated at 300 m; HR4 ([154]ANN15777) was down-regulated at −1000 and 0 m, and up-regulated at 300 m (Fig. [155]3). Among all the shared genes, EEF1A1 appeared to display the highest log2 fold change. Indeed, it was strongly down-regulated in September 2018 at 0 m with a log2 fold change of −28.30. Then, EEF1A1 was strongly up-regulated in subsequent samplings in October 2018 at 2000 m (log2FoldChange = 26.24), in December 2018 at −1000 m (log2 fold change = 27.68) and 300 m (log2 fold change = 27.79), in March 2024 at 500 (log2 fold change = 27.61). The DEG HR4 was present at all sampling times and multiple distances (n = 12), followed by Klf11 (n = 11), Hlf (n = 7), EEF1A1 (n = 5), Alk (n = 5), PHKG2 (n = 5), TMPRSS9 (n = 5), and an uncharacterised gene ([156]ANN35332) (n = 5). MYH2 was only present in September and October 2018 (n = 4) at the most exposed distances and was the only gene that did not experience a change in its regulation across different distances and sampling times. In September 2018, − 1000, 0, and 2000 m were the distances where the exposed oysters presented the highest number of genes with a log2 fold change over 10. Overall, the highest log2 fold change was observed for the genes Rpl35a (−30.00 in September 2018, [157]ANN13368) and EEF1A1 (28.30 in September 2018, [158]ANN16547). The majority of genes that were strongly differentially expressed (log2 fold change > 10) could not be annotated. However, strongly differentially expressed genes that could be annotated coded for proteins involved in transcription and translation, energy metabolism, cellular processes, cell growth, muscular activity, hormones, and immunity and included: RPL35A ([159]ANN13368), EEF1A1 ([160]ANN16547), LANCL2 ([161]ANN18713), CRT9 ([162]ANN06803), BACE1 ([163]ANN26629), Cox6b1 ([164]ANN03357), RNASEH1 ([165]ANN05658), EMC6 ([166]ANN22058), MAL ([167]ANN21771), COL9A1 ([168]ANN14062), ACTR8 ([169]ANN30165), PIN1 ([170]ANN16524), AOC1 ([171]ANN31577), VARS1 ([172]ANN17506), TYRL ([173]ANN38093), DDP4 ([174]ANN31977), ARF1 ([175]ANN04978), SH3BGRL3 ([176]ANN33725), RAB5B ([177]ANN38247), GSK3B ([178]ANN26499), AQP9 ([179]ANN37490), GRHPR ([180]ANN33431), TPM1 ([181]ANN08876), MAPK1 ([182]ANN24590), CTR9 ([183]ANN06803), GMFG ([184]ANN18394), WBP2 ([185]ANN03176). KEGG pathway enrichment analysis Pathway enrichment analysis identified enriched metabolic pathways for the annotated DEGs and revealed significantly altered pathways in the transcriptome. All DEGs were mapped to KEGG pathways and Table [186]2 shows the number of significant pathways using a p-corrected threshold value of 0.05. Table 2. Number of KEGG pathways significantly enriched between vessel controls (day 1) and oysters exposed over six sailing lines to the seismic energy (day 4) across all distances and sampling times based on a p-corrected value below 0.05 Distance −1000 m 0 m 300 m 500 m 1000 m 2000 m 6000 m Total September 2018 2 0 2 7 1 33 0 45 October 2018 0 24 26 14 24 7 0 95 December 2018 0 2 2 2 4 2 0 12 March 2019 4 6 1 4 n/a 1 0 16 Total 6 32 31 27 29 43 0 168 [187]Open in a new tab In September 2018, oysters located at 2000 m exhibited the highest number of impacted pathways with 33, followed by 500 m with seven impacted pathways. In line with the DEGs analyses, the number of impacted pathways increased in October, with 26 pathways impacted at 300 m, 24 at 0 m and 1000 m, and 14 at 500 m (Table [188]2). In December, the number of impacted pathways was considerably reduced, with a maximum of four pathways impacted at 1000 m. This number slightly increased in March 2019 with six pathways impacted at 0 m and four at −1000 and 500 m (Table [189]2). Oysters at 6000 m did not show any significantly impacted pathways at any sampling time. In September 2018, pathways involved in metabolism were the most impacted across all distances (Fig. [190]4). Indeed, the porphyrin and chlorophyll pathway was impacted at −1000, 300, and 2000 m and various amino acid metabolism pathways were impacted across most of the distances. The taurine and hypotaurin pathway was impacted at −1000 m; so were the arginine and proline pathways at 500 and 2000 m and the alanine, aspartate and glutamate pathways at 500 and 1000 m (Fig. [191]4). Genetic information and processing appear to be significant with the mRNA pathway at 300 and 2000 m and the protein processing in endoplasmic reticulum pathway at 500 and 2000 m were also impacted. The highest enrichment ratios were observed at −1000 and 2000 m. At 2000 m, many more pathways involved in metabolism (n = 10), genetic information and processing (n = 11), environmental information processing (n = 4), and cellular processes (n = 4) were significantly enriched (Fig. [192]4). Fig. 4. [193]Fig. 4 [194]Open in a new tab Enriched KEGG pathways of the DEGs between the vessel control (Day 1) and oysters exposed over six sailing lines to the seismic energy (Day 5) across distances in September 2018. The bar plot displays enriched terms with each bar representing an enriched function. Bar length represents the enrichment ratio = input gene number/ background gene number. Bar terms with similar functions are clustered by colour. Only the significant pathways using the corrected p value < 0.05 are represented. The enrichment analysis was conducted using KOBAS v3. No pathway was significantly enriched for the distances 0 m and 6000 m We recorded the highest number of enriched pathways and enrichment ratios in October 2018. The category metabolism was again the most important. Overall, amino acid metabolism was affected at all distances except for 6000 m (Fig. [195]5). This included many amino acid pathways such as the valine, leucine and isoleucine degradation pathway common at 0, 300, and 1000 m, the arginine and proline metabolism pathway shared between 500 and 1000 m, the histidine metabolism pathway shared between 0 and 500 m, and the lysine degradation metabolism pathway common at 0 and 300 m (Fig. [196]5). Carbohydrate metabolism pathways were dysregulated including the citrate cycle at 300 and 1000 m; the fructose and mannose metabolism at 0, 300 and 1000 m as well as the glycolysis/ gluconeogenesis at 0, 300, 500, and 1000 m; the galactose metabolism at 0 and 500 m; the pentose phosphate at 0, 300, and 2000 m (Fig. [197]5). The lipid metabolism was also disturbed with the fatty acid degradation dysregulated at 0, 300, 500, and 1000 m as well as the glycerolipid metabolism at 0, 300, and 1000 m. In the metabolism of cofactors and vitamins category, the folate biosynthesis was affected at 0 and 300 m. The Genetic Information processing category was also significantly affected by the seismic energy including the transcription with spliceosome at 0, 300, and 1000 m; the translation with the mRNA surveillance and RNA transport pathways affected at 0, 300, and 1000 m; the ribosome at 300 and 2000 m. Protein processing in the endoplasmic reticulum was also dysregulated at 0, 300, and 1000 m. Among the Environmental information processing category, the signal transduction pathway was impacted with mTOR signalling pathways dysregulated at 300 and 1000 m, so was the ECM receptor interaction at 0, 300, and 1000 m in the sub-category signalling molecules and interaction (Fig. [198]5). Fig. 5. [199]Fig. 5 [200]Open in a new tab Enriched KEGG pathways of the DEGs between the vessel control (Day 1) and oysters exposed over six sailing lines to the seismic energy (Day 5) across distances in October 2018. The bar plot displays enriched terms with each bar representing an enriched function. Bar length represents the enrichment ratio = input gene number/ background gene number. Bar terms with similar functions are clustered by colour. Only the significant pathways using the corrected p value < 0.05 are represented. The enrichment analysis was conducted using KOBAS v3. No pathway was significantly enriched at -1000 m and 6000 m A very low number of pathways was dysregulated in December 2018, with again the category metabolism being the most prevalent, including amino acid, glycan, and carbohydrate metabolism (Fig. [201]6). Only two pathways were shared between different distances: the other types of o-glycan biosynthesis pathway was enriched at 500 and 1000 m as well as the taurine and hypotaurine metabolism at 300 and 500 m. 1000 m was the only distance where the seismic energy impacted the spliceosome (transcription) (Fig. [202]6). Fig. 6. [203]Fig. 6 [204]Open in a new tab Enriched KEGG pathways of the DEGs between the vessel control (Day 1) and oysters exposed over six sailing lines to the seismic energy (Day 5) across distances in December 2018. The bar plot displays enriched terms with each bar representing an enriched function. Bar length represents the enrichment ratio = input gene number/ background gene number. Bar terms with similar functions are clustered by colour. Only the significant pathways using the corrected p value < 0.05 are represented. The enrichment analysis was conducted using KOBAS v3. No pathway was significantly enriched at -1000 m and 6000 m In March 2019, the distance 0 m had the highest number of pathways impacted, with six pathways of the 12 pathways that were enriched, especially pathways related to the biosynthesis of amino acid, transcription and translation. Then, the group at −1000 m followed with four impacted pathways related to the metabolism of amino acids and carbon metabolism as well as the distance 500 m with four pathways related to carbohydrate metabolism, metabolism, carbon metabolism and folding, sorting and degradation. The distance 300 m had only 1 enriched pathway: protein processing in endoplasmic reticulum. There was no impacted pathway for the other distances: 1000 m, 2000 m, and 6000 m. Pathways belonging to genetic information processing were the most impacted in March 2019 across most distances, such as 0, 300, and 500 m (Fig. [205]7). Indeed, at 0 m, the transcription with the spliceosome, the translation with the ribosome and mRNA surveillance pathways were enriched. Exposed oysters at 300 and 2000 m shared the protein processing in the endoplasmic reticulum pathway. The metabolism of amino acid was impacted at −1000 and 0 m; so was the carbon metabolism at −1000 and 500 m (Fig. [206]7). Fig. 7. [207]Fig. 7 [208]Open in a new tab Enriched KEGG pathways of the DEGs between the vessel control (Day 1) and oysters exposed over six sailing lines to the seismic energy (Day 5) across distances in March 2019. The bar plot displays enriched terms with each bar representing an enriched function. Bar length represents the enrichment ratio = input gene number/ background gene number. Bar terms with similar functions are clustered by colour. Only the significant pathways using the corrected p value < 0.05 are represented. The enrichment analysis was conducted using KOBAS v3. No pathway was significantly enriched at 1000 m and 6000 m Discussion Our study identified numerous differentially expressed genes (DEGs) and enriched pathways immediately and at 1, 3, and 6 months post-exposure. Seismic noise exposure induced cellular stress, notably evident one-month post-exposure, with effects diminishing over time. Oysters from groups positioned at 0 and − 1000 m continued to exhibit DEGs and enriched pathways six months post-exposure. To ensure optimal cellular growth and metabolic processes, cells possess intricate mechanisms to adjust their gene expression and modulate their metabolic pathways in response to internal and external stimuli. During adverse conditions, these regulatory mechanisms help to facilitate adaptation, maintain homeostasis, alleviate impairments and restore normal physiological functions [[209]43]. Our study, consistent with others, highlights the potential impact of seismic-induced stress on translation factors involved in mRNA translation initiation or elongation [[210]43], and underscores the overlap with specific pathways affected by other stress in molluscs. Similar to our findings where the greatest response was reported one month after exposure [[211]27, [212]28], observed that the snow crab Chionoecetes opilio crab’s response peaked at 3 weeks post-exposure and subsided by 6 weeks post-exposure. Interestingly, our study did not identify a single gene consistently differentially expressed across all distances. This complexity could be attributed to the ecological and environmental variations inherent to each site, despite similar ocean floor compositions and habitats [[213]23, [214]30]. Hence, we hypothesise that the expression of P. maxima genes was influenced by both biotic and abiotic variations and exposure to seismic energy. Hall et al. (2021) also noted similar challenges as DEGs identified in one survey were not consistently expressed in another. To minimise external variations that could influence stress responses, Hall et al. (2023) conducted tank-based noise exposure experiments to identify biomarkers associated with stress and immune response and validated these measures in field studies. This underscores the good practice of combining lab- and field-based experiments. This is something that could not be achieved in the present study as we did not have the facilities to do so. Additionally, our sampling after four days of exposure may have missed early responses occurring between initial exposure to the seismic signals and the sampling on Day 5, as acclimatisation to seismic energy and noise disturbance likely occurred over this period [[215]44]. Similar observations were reported in coral reef fish, where physiological responses to noise decreased with repeated exposures [[216]45]. Over the duration of the experiment and sample collection, the lowest number of DEGs was observed at 6000 m, potentially due to less noise disturbances than the other sites. The second lowest number of DEGs was observed at −1000 m, which received some of the highest sound exposure level (SEL). It was also the only distance where SEL increased during the experiment as the vessel conducted the sail lines every 500 m from the 0 m site toward the − 1000 m site. With the highest number of DEGs among all distances in September 2018 with other distances, it is possible that oysters at 2000 m experienced the greatest response due to the biotic and abiotic conditions intrinsic to that site. Such a synergistic effect was noted as a potential reason for increased mortality on Day 3 oyster groups at that site, in part 1 of this study [[217]23]. In the subsequent discussion, we will specifically focus on the responses shared among oysters located the closest to the source. The identified DEGs and KEGG pathways affected by the noise play critical roles in various cellular functions, including translation, transcription, muscle contraction, melanin, metabolism of amino acids, lipids and glycogen, protein synthesis, cell growth, cell migration, inflammation, apoptosis, and nervous system. September 2018, after four days of subsequent seismic exposures After four days of subsequent exposures, DEGs and enriched pathways revealed that the seismic noise disturbed the osmoregulation and affected critical cellular processes. Oysters located closest to the seismic source (mainly 0 and − 1000 m) down-regulated the genes TIMP2, MYH2, and an uncharacterised gene coding for a rootletin-like protein. Tissue Inhibitor of Metalloproteinases 2 (TIMP 2) codes for a protein that inhibits matrix metalloproteases and degrades the extracellular matrix, however, it is also involved in cell differentiation, growth, migration, angiogenesis, apoptosis, and tissue homeostasis [[218]46, [219]47]. The downregulation of TIMP2 at 0 and − 1000 m suggests the seismic noise disrupted these cellular processes and aligns with previous findings [[220]18], where seismic exposure caused persistent osmoregulation disturbance in the haemolymph of the scallop Pecten fumatus. In addition, Myosin Heavy Chain 2 (MYH2) appeared downregulated by the seismic noise at 0, 300, and 500 m in September 2018 and at 0 m in October 2018. MYH2 codes for an actine-based motor protein, essential for muscle growth, development, and contraction, indicating that the seismic noise may have disrupted these functions in the mantle of P. maxima. The seismic noise also putatively reduced the expression of a Rootletin-like protein at 0 and − 1000 m, which is crucial for maintaining the integrity and stability of ciliated roots and axoneme length [[221]48, [222]49]. Its dysregulation may affect the integrity of ciliated cells from the marginal mantle [[223]50]. The seismic noise may have primarily impacted pathways related to amino acid metabolism, particularly those involving porphyrin and chlorophyll at-1000, 300, and 2000 m, arginine and proline at 500 and 2000 m, and taurine and hypotaurine at −1000 m. Porphyrins serve vital functions in oxygen and carbon dioxide transport, biocatalysis [[224]51], and contribute to shell pigmentation in oysters [[225]52]. Proline acts as a molecular chaperone, regulating protein conformational changes [[226]53], stabilising proteins and enzymes, preserving membrane integrity, osmosis, and detoxifying ROS [[227]53, [228]54]. This pathway is primarily mobilised under stress, as seen in molluscs facing long-term hypoxia, heat waves, salinity changes, or exposure to nanoplastics [[229]53–[230]58]. Proline metabolism also triggers downstream effects like cell cycle arrest, autophagy, and apoptosis [[231]54]. Taurine is a key intracellular osmolyte, aiding cell volume adjustment [[232]59–[233]61], and it also supports mitochondrial function, acts as an antioxidant, and minimises ROS production [[234]62]. It was affected by hyper- and hypo-salinity stress in oysters, temperature increase in clams, abalone, and mussels, and by contaminants and hypoxia in abalone [[235]59, [236]63–[237]66]. Regarding genetic information and processing, the seismic noise potentially impacted the mRNA surveillance pathway at 300 and 2000 m and protein processing in endoplasmic reticulum pathway at 500 and 2000 m. External stress can damage RNAs, resulting in non-functional proteins [[238]67] but the mRNA surveillance pathway resolves translation issues [[239]43]. This pathway’s impact at multiple sampling times and distances suggests seismic noise caused RNA damage requiring repair. The protein processing in the endoplasmic reticulum (ER) pathway, responsible for protein folding, sorting, and degradation acts as a quality-control system and improves cell survival under stress conditions in plants [[240]68]. In the ER, proteins fold with the aid of chaperones and are transported to the Golgi complex if correctly folded, or retained and degraded if misfolded. Under external stress, the accumulation of misfolded proteins can lead to ER stress and potentially trigger apoptosis. Various stressors, such as cold in fish, viral stress in carp, heat in sea urchins and abalone, and thermal stress in clams, can affect this pathway [[241]69–[242]72]. Overall, the disruption of those genes and pathways suggests potential disturbances in the mantle homeostasis. October 2018, one month after seismic exposure One month post-exposure, oysters at 0, 300, and 1000 m showed the greatest responses to seismic energy with the highest number of DEGs and enriched pathways. Oysters closest to the source (0 and − 1000 m) up-regulated GSTCD and an uncharacterised gene related to genome stability, while down-regulating LIPI and an uncharacterised gene linked to homeostasis. DEGs included PHKG2 (glycogen metabolism), Hlf (transcription factor), and Klf11 (transcription factor). Dysregulation was observed in pathways generating energy from amino acids, carbohydrates, and lipids, along with pathways synthesising proteins and repairing transcription and translation errors, as detailed below. Glutathione-S-transferase C Domain (GSTCD) is commonly used as a biomarker in toxicology studies in bivalves [[243]73, [244]74]. GST plays a crucial role in protecting tissues from oxidative damage [[245]75], and was identified as a yellow pigment-related gene involved in the melanin pathway in P. margaritifera [[246]76]. Given the oyster mantle’s role in shell formation [[247]77] and colouration, the observed over-expression of GST in exposed oysters may indicate disruptions in shell and mantle pigmentation, as well as oxidative stress. Exposed oysters showed downregulation of the LIPI (lipase I) gene, consistent with findings from other molluscan studies examining the effects of stressors on lipid metabolism. For instance, decreased lipase activity was reported under hypoxia in pearl oysters [[248]78]; exposure to microplastic in the giant clam and its symbiont [[249]79]; and seawater acidification in mussels [[250]80]. The down-regulation of LIPI and PHKG2 genes suggests a potential disruption of lipid and carbohydrate metabolisms in exposed oysters, consistent with pathway analysis results. Phorphorylase Kinase Gamma 2 subunits(PHKG2) are integral to the glycogenolytic pathway and serve as a multimeric regulatory enzyme [[251]81]. Other studies on oysters have observed reduced glycogen levels in polluted sites [[252]82] and disturbances in glycogen metabolism under increased pCO2 [[253]83]. Hlf or Hepatic Leukemia factor is a transcription factor belonging to the proline and acid-rich protein family, known to activate transcription and linked to cellular circadian rhythms [[254]84]. Also, previously unreported in molluscs, Hlf regulates hematopoietic stem cells in humans [[255]85] and may serve vital functions in P. maxima, if conserved. Its varied expression across different distances and times in our experiment suggests its significance in oyster response to seismic noise. The KLF Transcription Factor 11 (KLF11) encodes a zinc finger transcription factor critical in humans [[256]86], implicated in cell growth inhibition, apoptosis, and regulation of inflammation, proliferation and differentiation [[257]87–[258]90]. It also binds to promoters of genes involved in various metabolic pathways, such as cholesterol, prostaglandin, collagen, etc. [[259]90, [260]91]. Its detection at 11 combinations of distances and times in our experiment potentially highlights its importance in the response to seismic noise. The dysregulation of HLF and KLF11 revealed a putative impact of the seismic activity on the transcription of P. maxima. The KEGG pathway analysis revealed metabolic impacts across all distances except 6000 m. These effects included disruptions in carbohydrate, lipid, amino acid, and vitamin metabolism, as well as processes related to transcription, translation, protein folding, sorting, degradation, and signaling. Pathways such as arginine and proline metabolism, mRNA surveillance, and protein processing in the endoplasmic reticulum were also shared with oysters collected in September 2018. Oysters at 0 and 300 m exhibited the most extensive pathway dysregulation, including disruptions in fructose and mannose metabolism, glucogenesis, pentose phosphate pathway, fatty acid degradation, glycerolipid metabolism, lysine degradation, valine, leucine, and isoleucine degradation, folate biosynthesis, spliceosome, mRNA surveillance, RNA transport, protein processing in the endoplasmic reticulum, mTOR signaling, and ECM-receptor interaction. Dysregulated amino acid pathways included arginine and proline metabolism at 500 and 1000 m, histidine metabolism at 0 and 500 m, lysine degradation at 0 and 500 m, and valine, leucine, and isoleucine degradation at 0, 300, and 1000 m. The disturbance in histidine, crucial for protein synthesis and enzyme activity, aligns with previous findings in P. maxima and C. gigas under heat and salinity stress [[261]56, [262]92]. Furthermore, lysine degradation, essential for protein synthesis, was dysregulated and was shown in other studies to be affected by stressors such as contaminants and nanoplastics [[263]58, [264]66]. The disruption of lysine and fatty acid metabolism pathways underscores the potential broader impact of seismic-induced stress on lipid metabolism [[265]93]. Our results indicate that exposure to seismic noise disrupted energy regulation, including amino acids, lipids, and carbohydrates pathways between 0 and 1000 m. Specifically, the valine, leucine, and isoleucine degradation pathway, crucial for energy production as it produces acetyl-CoA, which contributes to glucose production was disrupted [[266]94]. It was activated in other organisms under stress: in zebrafish exposed to carbon nanodots [[267]95], and in pearl oysters with growth retardation [[268]96]. The glycolysis/ gluconeogenesis pathway, enriched at multiple distances serves as a primary energy source and was shown to be impacted by stress in pearl oysters during nucleus implantation [[269]97]. Metabolic fluxes between glycolysis and the pentose phosphate pathway regulate cellular energy needs [[270]98]. The pentose phosphate pathway produces energy and nucleotide precursors, as described in oyster larvae [[271]99]. It was enriched at multiple distances, and is crucial under environmental stress such as cold stress [[272]100]. The fructose and mannose metabolism pathway, impacted at multiple distances, may aim to maintain cellular redox balance by scavenging reactive oxygen species generated during stress [[273]101, [274]102]. This pathway was also dysregulated by a nanoplastic exposure in oysters [[275]58]. Our findings reveal disruption in energy metabolism not only in amino acid and carbohydrate pathways but also in lipid metabolism. Fatty acid degradation and glycerolipid metabolism pathways, enriched at multiple distances, serve as a major energy source in oysters [[276]99] and also play crucial roles in signaling processes regulating various biological functions [[277]103]. Glycerolipid metabolism also regulates inflammatory and immune responses [[278]104]. Lastly, our results suggested that seismic exposure also impacted genetic information and processing, with enriched pathways observed in transcription, translation, folding, sorting, and degradation, as well as signaling molecules and interaction at three distances: 0 m, 300 m, and 1000 m. Similar to September 2018, the mRNA surveillance pathway and ER protein processing were enriched, so was the spliceosome pathway. Spliceosomes play a crucial role in generating transcript variability and proteome diversity [[279]105], and may have helped pearl oysters cope with seismic stress. This pathway was also enriched in fish facing cold stress [[280]105, [281]106]. The RNA transport pathway, fundamental for gene expression was also enriched. Furthermore, the ECM-receptor interaction pathway, mediating critical cellular processes such as adhesion, migration, differentiation, proliferation, and apoptosis, was enriched in our study and in pearl oysters under hypoxic stress [[282]107] and after mantle grafting [[283]108]. These findings, therefore suggest that seismic noise may have caused dysfunctions in metabolism and transcription in P. maxima. December 2018, three months after seismic exposure Three months after the exposure, the number of differentially expressed genes and enriched pathways decreased. Oysters at 0 and − 1000 m displayed dysregulation of genes involved in the nervous system, along with enriched pathways related to amino acid and glycan metabolism. Specifically, the genes Alk and REST were down-regulated, while EEF1A showed strong up-regulation. Alk (Anaplastic Lymphoma Receptor Tyrosine Kinase) encodes protein kinase receptors critical for brain development and the function of the nervous system [[284]109]. REST (RE1 Silencing Transcription Factor) is a transcriptional repressor that suppresses neuronal genes in non-neuronal tissues and regulates neurogenesis [[285]110, [286]111]. Notably, this is the first identification of REST and Alk in a mollusc model, previously believed absent in invertebrates [[287]112]. EEF1A (Eukaryotic Translation Elongation Factor 1 Alpha) encodes an isoform of the alpha subunit of the elongation factor-1 complex, responsible for delivering aminoacyl tRNAs to the ribosome’s A site [[288]113, [289]114]. Beyond its role in protein synthesis, EEF1A1 has multiple functions in regulating cell growth and proliferation, cytoskeleton organisation, axon repair, apoptosis, and protecting against ER stress [[290]113, [291]115–[292]121]. In our study, EEF1A emerged as the most upregulated gene, showing significant changes at multiple distances and times. These findings align with previous reports indicating that translation, such as elongation factors, can be affected by various stressors [[293]43]. Similar to September 2018, seismic activity affected the taurine and hypotaurine pathway at 300 and 500 m. Enrichment of the spliceosome pathway was observed at 1000 m only. Overall, the limited impact on pathways suggests that energy metabolism has largely recovered three months after exposure. However, the differential expression of the three genes ALK, REST, and EEF1A related to the nervous system may indicate a lingering effect of seismic noise on the P. maxima nervous system after the initial exposure. March 2019, six months after seismic exposure Six months after exposure, only the oysters closest to the seismic source at 0 and − 1000 m shared DEGs, and to a lesser extent 300 m. KEGG pathway analysis showed enriched pathways in oysters at 0, 500 and − 1000 m. Among the DEGs, a putative-ankyrin repeat protein gene was down-regulated, while a Tyrosinase-like protein-1 was up-regulated. Additionally, TMPRSS9, Hlf, and HR4 revealed differential expression. Ankyrins are membrane adaptor molecules associated with the cytoskeleton network and membrane proteins [[294]122]. They are crucial for cell motility, activation, and proliferation [[295]123]. In humans, decreased expression is linked to abnormal eNOS signaling and reduced cell proliferation [[296]124]. In oysters, they play pivotal roles in NF-KB pathways and structural organisation like cytoskeleton rearrangement, and maintaining pressure balance during salinity stress [[297]122]. Seismic noise may disrupt these functions, akin to cytoskeleton remodelling observed in oysters under heat stress [[298]125]. Tyrosinase-like protein 1 is a key enzyme in melanogenesis [[299]126] and plays essential roles in shell growth, and pigmentation [[300]127, [301]128]. Highly expressed in the mantle edge of pearl oysters, it likely contributes to shell formation [[302]128] and nacre matrix cross-linking [[303]129]. Similar to our findings, its expression decreases in response to hypoxia in pearl oysters [[304]53]. The transmembrane serine protease 9 (TMPRSS9) is a membrane-bound enzyme involved in proteolysis, organelle development, homeostasis, and regulation of cell surface phenomena in humans [[305]130–[306]132]. While not yet described in molluscs, its presence suggests potential similar impacts of seismic noise. Hormone receptor 4 (HR4), part of the nuclear receptors-6 subfamily, plays pivotal roles in various signalling and metabolic pathways. In insects, it regulates chitin synthesis, epidermal cuticle formation [[307]133], larval development and metamorphosis [[308]134] while in mammals, its ortholog, Germ Cell Nuclear Factor, acts as a transcriptional repressor [[309]135]. Differentially expressed across several distances and times in our study, HR4 emerges as a potentially significant gene. Six months after the exposure, the seismic noise affected vital functions such as transcription., cell proliferation, development, and differentiation, as well as proteolysis, melanisation, haematopoiesis, and amino acid metabolism. At 0 m, mRNA surveillance and spliceosome pathways were enriched, indicating seismic activity’s potential impact on transcription and translation. Moreover, the ribosome pathway, crucial for protein synthesis, was affected, consistent with findings in pearl oysters stressed by an allograft [[310]97]. This suggests that seismic surveys may have had an impact six months after the exposure in oysters closest to the noise source. Conclusion In summary, seismic noise impacted genes expression in P. maxima, especially from exposure time up to one month post-exposure, with varying effects at different distances, except for 6000 m. The impact appeared the greatest closer to the seismic source at CPAs of 0 and − 1000 m. Other distances such as 2000 m presented the highest level of dysregulation with the higher number of DEGs straight after the exposure in September 2018, probably due to other contributing biotic and abiotic factors at that site. As observed here, responses to stressors typically involve alterations in energy consumption and production, modulation of transcription and translation, and synthesis of stress response proteins [[311]43]. Following four days of subsequent exposures, amino acid metabolism, translation, and protein processing in the endoplasmic reticulum were notably affected up to the distance 2000 m, along with the down-regulation of genes related to muscle contraction and cilia cell stability. This suggests that seismic noise-induced stress may lead to DNA and RNA aberrations and the production of non-functional proteins, as evidenced by enriched mRNA surveillance and ER protein processing pathways. At one-month post-exposure, disturbances in mantle energy metabolism were mostly observed at distances up to 1000 m, with dysregulated pathways involved in energy regulation using amino acids, lipids, and carbohydrates. Enriched pathways indicated impacts on transcription, translation, and protein synthesis. By three months, the impact on gene expression was significantly reduced across all distances, with fewer differentially expressed genes at 0 and − 1000 m. Six months post-exposure, the metabolism and genetic information processing pathways remained impacted in oysters, mainly at 0 and − 1000 m. Despite the impact on genes expression, no mortality was attributable to the seismic survey in the oysters from Day 5, used in the present experiment [[312]23]. However, mortality and lower pearl quality were reported in seeded oysters after 2 days of seismic exposure at 0 and 2000 m, with no data for the other distances. As discussed in part 1 of the present study, the mortality was probably due to other factors during the experiment such as biotic and abiotic conditions at the experimental site, the farm site or during themanipulation and transportation of the panels [[313]23]. Overall, our findings suggest that oysters activated stress responses to seismic energy exposure, focusing on damage repair, maintaining cellular homeostasis, and supporting growth. This aligns with some studies indicating a minimal impact of noise on crab behaviour, physiology, and catch rate [[314]11, [315]12, [316]14, [317]27] and differs from other studies on scallops and rock lobsters that reported increased mortality rates and impairment of cellular and immune functions [[318]18, [319]136]. Analysis of physiological data from the present experiment is underway to further investigate the effect of seismic activity on pearl oysters. Supplementary Information [320]Supplementary Material 1.^ (42.4KB, xlsx) Acknowledgements