Effects of cold stress on the blood-brain barrier in Plectropomus leopardus Authors: Yilan Guo, Cun Wei, Hui Ding, Peiyu Li, Yurui Gao, Kangning Zhong, Zhenmin Bao, Zhe Qu, Bo Wang, Jingjie Hu Published: 2024 ABSTRACT AbstractBackgroundThe leopard coral grouper (Plectropomus leopardus) is a commercially valuable tropical marine fish species known to be sensitive to low temperatures. A comprehensive understanding of the molecular mechanisms governing its response to acute cold stress is of great importance. However, there is a relative scarcity of fundamental research on low-temperature tolerance in the leopard coral grouper.MethodsIn this study, a cooling and rewarming experiment was conducted on 6-month-old leopard coral groupers. Within 24 h, we decreased the ambient temperature from 25 °C to 13 °C and subsequently allowed it to naturally return to 25 °C. During this process, a comprehensive investigation of serum hormone levels, enzyme activity, and brain transcriptome analysis was performed.ResultsP. leopardus displayed a noticeable adaptive response to the initial temperature decrease by temporarily reducing its life activities. Our transcriptome analysis revealed that the differentially expressed genes (DEGs) were primarily concentrated in crucial pathways including the blood-brain barrier (BBB), inflammatory response, and coagulation cascade. In situ hybridization of claudin 15a (cldn15a), a key gene for BBB maintaining, further confirmed that exposure to low temperatures led to the disruption of the blood-brain barrier and stimulated a pronounced inflammatory reaction within the brain. Upon rewarming, there was a recovery of BBB integrity accompanied by the persistence of inflammation within the brain tissue.ConclusionsOur study reveals the complex interactions between blood-brain barrier function, inflammation, and recovery in P. leopardus during short-term temperature drops and rewarming. These findings provide valuable insights into the physiological responses of this species under cold stress conditions.Supplementary InformationThe online version contains supplementary material available at 10.1186/s12864-024-10943-6.Keywords: Low temperature, Brain, Blood brain barrier, Transcriptome, Claudin Introduction As ectothermic animals, fish experience direct impacts on their body temperature, growth, reproduction, and various life activities from the ambient water temperature. In the context of global climate change, cold and hot extreme weather events are becoming increasingly frequent [1, 2]. In recent decades, marine organism have been increasingly exposed to unsuitable temperature stressors [3, 4]. Natural factors, such as storms, rapid precipitation, abnormal water movements, and extreme weather events associated with La Niña and El Niño, have caused rapid drops in seawater temperature, significantly impacting marine ecosystems [5]. Abnormally low temperatures can affect both wild marine life and aquaculture operations [6], particularly in tropical or subtropical waters [7]. The study of fish temperature tolerance has become topical in recent years, particularly in understanding temperature tolerance variations and delving into the molecular mechanisms involved [8–10]. As ambient temperature changes, fish exhibit adaptive responses at the behavioral, physiological, endocrine, cellular, and molecular levels, with the effects of long-term temperature changes alterations being reflected at the genomic level [11–14]. When vertebrates encounter stressors such as low temperatures, the hypothalamic-pituitary-interrenal (HPI) axis responds [15]. Adrenocorticotropic hormone (ACTH) secreted by the pituitary gland under the influence of the hypothalamus, stimulates the renal interstitium to produce cortisol, aiding in the adjustment of physiological activities to cope with stress [16]. Numerous studies across various species have reported significant changes in hormone levels, antioxidant enzymes, and metabolism-related enzymes under acute cold stress or prolonged exposure to low temperatures. For example, tilapia (Oreochromis niloticus), large yellow croaker (Larimichthys crocea), and juvenile common carp (Cyprinus carpio) have all shown notable variations in these parameters [17–19]. However, there are substantial differences among species. In the case of large yellow croaker, while cortisol and T3 levels increase significantly under cold stress, the plasma T4 level remains unaffected [18]. Conversely, in juvenile common carp maintained at low temperatures for periods ranging from 15 to 45 days, a significant decrease in levels of cortisol, T3, and T4 were shown [19]. The sea raven (Hemitripterus americanus) exhibits a delayed cortisol response following cold exposure [20], whereas cunner (Tautogolabrus adspersus) shows metabolic suppression and relatively lower cortisol levels when acclimated to 0 °C compared to 10 °C [21]. In the subtropical pacu (Piaractus mesopotamicus), no significant changes in cortisol levels or most antioxidant system indicators were observed after 24 h of cold exposure [22]. Therefore, the metabolic responses to cold stress vary among species. If the temperature rapidly decreases and exceeds the tolerance limit, the homeostasis of the fish will be severely disrupted, leading to a series of issues affecting basic functions such as respiration and metabolism, as well as more complex behaviors like predation and locomotion [23–25]. The leopard coral grouper (Plectropomus leopardus), a tropical marine coral reef fish from the Epinephelidae family, has a temperature tolerance range of 16–32 °C, with an optimal temperature of around 26 °C [26]. Moreover, this species has high economic value and broad market prospects due to its beautiful skin color, superior nutrition qualities, and excellent taste. As a valuable and important resource for cage aquaculture and industrialized recirculating aquaculture systems, low temperatures significantly affect P. leopardus survival, growth and limits the extension of this species to high-latitude areas for aquaculture development. In tropical regions where P. leopardus are farmed, semi-open flow-through systems and offshore cage culture are primarily used [27]. These culture methods are significantly influenced by surface seawater temperature changes [28]. When cold spells occur, water temperatures can drop dramatically within a short period, affecting fish health and leading to economic losses [29]. During artificial breeding process, when the temperature drops below 16 °C, P. leopardus stops feeding and swimming. As the temperature further decreases and persists for several days, even without immediate cold shock and death, the fish will experience a decline in immunity and health, making them susceptible to secondary diseases, such as bacterial infections, which can lead to chronic death. However, understanding of how cold exposure affects the biological and physiological processes of this species is limited. Research on the cold tolerance of bony fish has shed light on the importance of alternative splicing, stress signal transduction, and energy metabolism in responding to cold stress of brain tissue [30–32]. The brain, a crucial part of the CNS, is the control center for all life activities. Brain temperature is an important factor in regulating blood brain barrier (BBB) permeability, alterations in brain water homeostasis, and subsequent structural abnormalities of brain cells in mammals [33]. While hyperthermia has been linked to BBB leakage and brain cell damage [34], limited evidence suggests that hypothermia may induce acute brain dehydration instead of edema and mild BBB disruption along with glial activation in mammals [35]. In common carp, exposure to cold conditions, from 25 °C to 10 °C, resulted in a strong drainage of blood out of the brain, leading to hypoxia and ischemia [36]. However, little is known about the molecular mechanisms underlying these effects. In this study, we aimed to explore the impact of acute temperature fluctuations on the teleost brain by exposing P. leopardus to a gradual decrease from 25 °C to 13°Cover 24 h, followed by natural rewarming back to 25 °C. We monitored the serum hormone levels, enzyme activity and brain transcriptome throughout this temperature change to provide novel insights into how sudden environmental temperature shifts affect the fish brain. Materials and methods Ethics statement This study was carried out with the permission of the College of Marine Life Sciences, the Ocean University of China Institutional Animal Care and Use Committee on 10 October 2018 (Project Identification Code: 20181010). Experimental design and sampling The six-month-old P. leopardus with average body weight (BW) of 19.0 ± 2.1 g were provided by Chenhai Aquatic Co., Ltd., Dongfang, Hainan, China. Fifty P. leopardus were temporarily cultured at 25 °C for 48 h before initiating the temperature decrease in a 100-liter water tank. An aquarium chiller was used to control the water temperature. The culture water temperature gradually decreased from 25 °C to 13 °C in 24 h at a rate of 0.5 °C/h, and then naturally rewarmed to 25 °C in 12 h. Groups were named as 25 °C,19 °C, 13 °C and R-25 °C (rewarmed to 25 °C) group according to the four sampling time points. Before sampling, gently transfer the individuals from groups to a container filled with the prepared solution of 100 mg/L Tricaine Methanesulfonate (MS-222, Sigma-Aldrich, USA) for euthanasia, and continuously observe the status of the fish. Anesthesia was confirmed typically within 2–3 min, continue to maintain for 5 min to ensure complete euthanasia. Blood samples from all groups were collected in triplicate with a sterile syringe from the gill artery and stored at 4 °C for hormone determination. Brains from the 25 °C, 13 °C and R-25 °C groups were excised using sterile scissors and forceps in triplicate, fixed in 4% polyformaldehyde solution, and preserved in 100% ethanol for paraffin embedding, slicing and fluorescence in situ hybridization (FISH). Brains from 3 individuals each in the 25 °C, 13 °C and R-25 °C groups were sampled and stored at -80 °C for RNA extraction and sequencing (Fig. 1). Experimental design. A The leopard coral groupers (n = 50) were cultured in the 100 L tank: temporary cultured in 25 °C for 48 h before cooling at rate of 0.5/h from 25 °C to 13 °C, and naturally rewarmed to 25 °C. A control group (animals at 25 °C before cooling), two low temperature group (animals at 19 °C and 13 °C) and a rewarming group (animals at 25 °C after rewarming). B Blood and brain tissue samples were collected in triplicate from relevant groups for further tests, including ACTH and Cortisol measurement, SOD activity detection, transcriptome sequencing and in situ hybridization experiments Hormone level measurement and enzyme activity assay Enzyme linked immunosorbent assay (ELISA) were used for measuring the level of adrenocorticotropic hormone (ACTH) and cortisol in blood samples (ml003453 and ml003467, Shanghai Enzyme-linked Biotechnology Co., Ltd., Shanghai, China). The blood samples were centrifuged at 2000× g for 20 min to obtaining the serum. The sera were mixed with the antibody and incubated at 37 °C for 60 min in the enzyme-linked strips, and then washed five times using microplate washer (BioTek 50 TS Washer, Agilent, Santa Clara, CA, United States). The chromophoric reagent were added to the strips and incubated at 37 °C for 15 min in the dark. After adding termination solution, the absorption optical density value at 450 nm of each sample was measured using multi-function microplate reader (BioTek Synergy H1, Agilent, Santa Clara, CA, United States). Superoxide Dismutase (SOD) assay kit (WST-1 method) (A001-3-2, Nanjing Jiancheng Bioengineering Institute, Nanjing, China) was used for measuring the activity of SOD in serum per the manufactures’ instruction. In brief, diluted sera were mixed with substrates and incubated at 37 °C for 20 min, and measured the absorption optical density value at 450 nm. RNA extraction and reverse transcription Total RNA was extracted from the brain tissues using TRIzol reagent (Invitrogen, Carlsbad, CA, United States). Total RNA quantity was measured using a UV spectrophotometer (NanoDrop 2000, ThermoFisher Scientific, Waltham, MA, USA). RNA integrity was assessed using 1% agarose gel electrophoresis and a Bioanalyzer. The first-strand cDNA was synthesized by All-In-One 5x RT MasterMix kit (abm, Canada), which incubated the reaction mixture containing 0.1 µg/µl total RNA at 37 °C for 15 min and then at 60 °C for 10 min, stored at -20℃. Sequencing and bioinformatics analysis The sequencing libraries were prepared using the Illumina VAHTS® Universal V8 RNA-seq Library Prep Kit (Vazyme NR605, Nanjing, China). Sequencing was performed on the Illumina NovaSeq platform with paired-end 150 bp reads to generate raw data. Clean data was obtained through quality control by using Trimmomatic v0.39. The clean reads of each sample were mapped to the reference genome of P. leopardus (NCBI: GCF_008729295.1) (YSFRI_Pleo_2.0) using HISAT2 v2.2.1 generating alignment file in BAM format. The featureCounts tool was used to extract the read position information from the BAM file, count and reads, and obtain read counts for each gene [37]. TPM (Transcripts Per Million) value was derived from the count values and utilized for correlation and principal component analysis (PCA). Differential gene expression analysis was performed using the DESeq2 package with criteria set at |log2FoldChange| ≥ 2 and padj ≤ 0.05. A Venn diagram was constructed using the gplots package. Heatmaps of differentially expressed genes (DEGs) were generated by ggplot2 and pheatmap package. Time series gene expression data were clustered using the Mfuzz software package. Gene Ontology (GO) annotation and KEGG enrichment analysis of DEGs were conducted using the OmicShare tools available at www.omicshare.com/tools. Volcano plots was plotted by Bioinformatics online platform for data analysis and visualization at https://www.bioinformatics.com.cn. Quantitative real-time polymerase chain reaction In order to validate transcriptomic data, four target genes were selected for qRT-PCR analysis. qRT-PCR assays were performed in a total of 9 samples (25 °C,13 °C and R-25 °C groups). Intron-spanning gene-specific primers were designed by on online primer-blast tool of NCBI (Table. S1). cDNA from the study was serially diluted to construct standard curves, confirming the amplification efficiency of the primers used in the brain tissue samples of this study. The amplification efficiencies ranged from greater than 90% to less than 110%. Detailed information of primers can be found in Table. S1. The cDNA mixture was diluted 10 times with RNase-free water and used as template for qRT-PCR. qRT-PCR assay was performed on the LightCycler 480 (Roche, Forrentrasse, Switzerland). The reaction mix included 10 µl of BlasTaq 2× qPCR MasterMix reagent (abm, Canada), 1 µl of cDNA, 0.5 µL of each primer (10 µM), and 8 µL of nuclease-free H2O. Thermocycling conditions consisted of enzyme activation at 95℃ for 3 min, followed by 40 cycles of denaturation at 95℃ for 15 s and extension 1 min at 60℃. The relative expressions of detected genes were calculated by 2−ΔΔCt comparative Ct method, using the b2m gene as the reference gene. Fluorescence in situ hybridization (FISH) The primers of hybridization probes were designed using the tool of primer-blast in NCBI (Tm around 60°C, GC% around 50%, Table 1), and the promoter sequences of SP6 and T7 polymerase were added to the 5’ end of relevant primers. The RNA probes were transcribed in vitro by T7 and SP6 polymerase and labeled with DIG or Biotin (Roche, Mannheim, Germany). The RNA probes were purified and heated to single chain before used. The brain samples embedded in paraffin were cut into 6 μm slices and put on the glass slices. After dewaxing and rehydration, the brain slices were digested with 2 µg/mL proteinase and fixed with polyformaldehyde. The brain tissue sections underwent hybridization with 1 ng/µL probe at 56 °C for 17 h after a 3-hour pre-hybridization step, followed by washing away any unhybridized probe using saline sodium citrate (SSC, 0.15 mol/L NaCl; 15 mmol/L Sodium Citrate) solution. The brain slices were incubated with Streptavidin − Cy3™ and anti-digoxigenin-fluorescein (from Streptomyces avidinii) diluted in 200 times volume blocking solution for 1 h at 37 °C in dark. The brain cell nuclei were stained with DAPI diluted 1:200 in PBS. Finally, the brain slices were imaged using Olympus FV3000 confocal microscope (Olympus, Japan) with the FV31S-SW-Sfoftware. Table 1Primer sequences of FISHPrimerSequences (5’–3’)FISH tjp1 FATTTAGGTGACACTATAGGTCCCGCAAAAATGAAGCAGAATRTAATACGACTCACTATAGGGGTCGTCCCTCTGAACAACCA cldn15a FATTTAGGTGACACTATAGGACTTTACATCGGCTGGTGCRTAATACGACTCACTATAGGGACATGCAGTGCTGTCCGTTTNote: The underlined sequences indicate the binding sites of SP6 and T7 RNA polymerase, respectively Primer sequences of FISH Note: The underlined sequences indicate the binding sites of SP6 and T7 RNA polymerase, respectively Statistical analysis All data were analyzed using the IBM SPSS statistics version 27.0 software (SPSS Inc., Chicago, IL, USA), and differences were analyzed by one-way analysis of variance (ANOVA). The results are presented as the mean ± standard error of the mean, with a p-value < 0.05 considered significant. When significant differences were found, Duncan’s multiple range test was used to determine differences between experimental groups. Results Hormone and enzyme respond to cold stress To investigate the hormonal response of P. leopardus to cold stress, the level of ACTH and cortisol were quantified using ELISA kits (Fig. 2). Compared with the 25 °C (control) group, the levels of serum ACTH and cortisol showed an increasing trend in 19 °C, 13 °C and R-25 °C groups. Despite no significant difference in SOD activity during the cooling process, a notable rise in SOD activity was observed upon returning the water temperature to 25 °C (Fig. 2). Hormone and enzyme detection results in serum of P. leopardus. The abscissa represents the group, 25 °C (animals at 25 °C before cooling), 19 °C and 13 °C (animals at 19 °C and 13 °C), R- 25 °C (animals at 25 °C after rewarming). The vertical axis represents the unit of hormone level and enzyme activity. The values are expressed as mean ± SEM. n = 3. Significant differences are shown when P < 0.05 Transcriptional changes in brain induced by cooling and rewarming To understand how the brain responds to cooling and rewarming processes, transcriptome sequencing was performed on nine brain tissue samples from the following temperature groups: 25 °C, 13 °C, and R-25 °C. After quality control, a total of 105.3 gigabytes of clean data was obtained, averaging 21.9 million clean reads per sample. The transcriptome clean data has been uploaded to the NCBI sequence Read Archive (SRA) with the accession number PRJNA1105558. The clean reads were mapped to the reference genome, achieving an average mapping percentage of 93.01%, and annotated a total of 26,109 genes. TPM (Transcripts Per Million) values were derived from the count values and used for principal component analysis (PCA). The PCA results indicated significant differences among the three temperature groups, with the most significant difference was observed between the 25 °C and R-25 °C groups (Fig. 3A). Using DESeq2 for pairwise comparisons, we identified a total of 1,028 differentially expressed genes (DEGs) based on the criteria of |log2FoldChange| ≥ 2 and padj ≤ 0.05. Analysis of differentially expressed genes in brain transcriptome of P. leopardus. A principal component analysis plot of 25 °C (animals at 25 °C before cooling), 13 °C (animals at 13 °C) and R- 25 °C (animals at 25 °C after rewarming). B Venn diagram of the distribution of the DEGs of groups. C Statistical table for the number of differentially expressed genes (log2FoldChange > = 2, P < = 0.05) pairwise comparison between groups. D The heatmap of the differentially expressed genes. E Soft clustering plots of DEGs in the time series: Each line in graph represents the expression profile of an individual gene In the 13°C_vs_25°C, there were 451 DEGs, comprising 297 upregulated and 154 downregulated genes. Similarly, in the R-25°C_vs_13°C, we found 277 DEGs, with 116 upregulated and 161 downregulated genes. The R-25°C_vs_25°C exhibited 610 DEGs, including 359 upregulated and 251 downregulated genes. Observation of the Venn diagram reveals that there are more DEGs between the 13°C_vs_25°C and R-25°C_vs_25°C groups, and the number of DEGs between the temperature-influenced 13 °C and R-25 °C groups is relatively low (Fig. 3B, C, D). To gain a better understanding of DEG expression trends during cooling and rewarming processes, we conducted the time-series gene expression analysis, which yielded four distinct clusters (Fig. 3E). Cluster 1: Genes in this cluster exhibited a consistent decrease in expression as the cooling process progressed and maintained this reduced expression until after the rewarming phase. Cluster 2: Genes in this cluster displayed a continuous increase in expression as the cooling process progressed and continued to rise until the rewarming phase. Cluster 3: Genes in this cluster initially showed an increase in expression, followed by a subsequent decrease in expression levels. Cluster 4: Genes in this cluster exhibited an increasing trend in expression levels as the rewarming process progressed. The function analysis of DEGs in brain To visualize the differential expression between groups in our transcriptome data, we generated enhanced volcano plots (Fig. 4A-C), where marked genes indicate significant changes in gene expression associated with temperature variations. To gain a deeper insight into the functions of DEGs, we conducted Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) enrichment analyses. Functional enrichment analysis of differentially expressed genes in brain transcriptome of P. leopardus.A-C Enhanced volcano plots with annotated partial genes that enriched in significant pathways. The x-axis represents log2 fold change (FC) in expression levels, the y-axis shows -log10(p-value) to indicate statistical significance. D-F KEGG bubble plot illustrating the top 10 pathway enrichment analysis of DEGs. The x-axis indicates the enrichment score, reflecting the degree of enrichment in each pathway. The y-axis lists the KEGG pathways, and the size of the bubbles corresponds to the number of genes mapped to each pathway. The color gradient indicates the statistical significance (-log10(p-value)) of the enrichment The most enriched GO level 2 terms in all three groups were ‘cellular process,’ ‘binding,’ and ‘membrane.’ Detailed information on the GO level 2 terms of each group is provided in Fig.S1 Furthermore, the top significant GO pathways, as displayed in dot plot (Fig.S2), varied across the different group comparisons. For the 13°C_vs_25°C group: The most significant GO pathways were ‘complement activation (GO:0006956),’ ‘creatine kinase activity (GO:0004111),’ and ‘extracellular space (GO:0005615).’ For the R-25°C_vs_13°C group: The top enriched GO terms were ‘negative regulation of cellular protein metabolic process (GO:0032269),’ ‘negative regulation of protein metabolic process (GO:0051248),’ and ‘negative regulation of protein modification process (GO:0031400).’ For the R-25°C_vs_25°C group: The most significant GO pathways included ‘complement activation (GO:0006956),’ ‘humoral immune response (GO:0006959),’ and ‘protein activation cascade (GO:0072376).’ In comparing the 13 °C group with the 25 °C group, we identified several significant KEGG pathways that can be broadly categorized into three main aspects: the complement system, the blood-brain barrier (BBB), and inflammation-related pathways (as illustrated in Fig. 4D). Specifically, pathways such as leukocyte transendothelial migration, cell adhesion molecules, and tight junctions are associated with BBB function. Additionally, the TNF signaling pathway and IL-17 signaling pathway are linked to inflammation and immune responses. Related genes include cebpd, ccl20, socs3, c8g, fosl1a. To provide further insights, we have compiled information on nine DEGs of high interest related to blood-brain barrier function and inflammation (Table 2). The heatmap illustrating the expression patterns of these nine genes is shown in Fig.S3. The blood-brain barrier is characterized by tight junctions (TJs) located at the membrane of adjacent endothelial cells. These TJs involve integral membrane proteins such as claudin, polarization, and junction adhesion molecules (JAMs), and various cytoplasmic accessory proteins [38]. Table 2Selected DEGs information associated with BBB and inflammationDescriptionGene idGene nameClusterFunction in BBBReferencesJunctional adhesion molecules (JAMs) jam2a junctional adhesion molecule 2a1JAMs participate in intercellular adhesion, polarization, and migration. The knockout of JAM-A enhancing inflammatory activity.[39, 40]Claudin (CLDN) cldn15a claudin 15a3CLDNs form a dimer in the extracellular loop by binding to CLDN molecules in adjacent endothelial cells, thereby forming the backbone of TJ.[41] cldn7bl claudin 7bl cldni claudin iCell adhesion molecules itga9 integrin, alpha 92Integrin regulates pericyte recruitment to support BBB maintenance and repair following ischemic stroke.[42] svep1 sushi, von Willebrand factor type A, EGF and pentraxin domain containing 13SVEP1 regulates vasoconstriction and influences monocyte to macrophage differentiation.[43, 44] sell L-selectin-like3L-selectin plays a role in inflammation processes which levels are higher in patients with ischemic stroke.[45] tiam2a TIAM Rac1 associated GEF 2a1Tiam1 are involved in neurite formation through the activation of Rac1 and successive cytoskeletal reorganization of neuronal cells during development.[46, 47]Hypoxia inducible factor-1 (HIF-1) hif1al hypoxia inducible factor 1 subunit alpha, like4HIF-1 drives hypoxia-induced vascular permeability in vivo, increased levels of proangiogenic cytokines.[48, 49]Note: Cluster refers to the results of time series analysis, and the trend can be seen in Fig. 3D Selected DEGs information associated with BBB and inflammation Note: Cluster refers to the results of time series analysis, and the trend can be seen in Fig. 3D In Table 2, we present gene information related to JAMs and claudin, including junctional adhesion molecule 2a (jam2a), claudin 15a (cldn15a), claudin 7bl (cldn7bl), and claudin I (cldni). Additionally, we include genes related to vascular permeability and the brain’s inflammatory response, such as integrin, alpha 9 (itga9), sushi, von Willebrand factor type A, EGF and pentraxin domain containing 1 (svep1), L-selectin-like (sell), TIAM Rac1 associated GEF 2a (tiam2a), hypoxia inducible factor 1 subunit alpha, like (hif1al). The information includes details about the gene’ associated proteins, their temporal expression clusters (refer to Fig. 3E), and their functions within the blood-brain barrier. These gene expression changes suggest the possibility of blood-brain barrier dysfunction at the mRNA level in the brain of P. leopardus under low-temperature stress. In the R-25°C_vs_13°C comparison, we identified only nine significant KEGG pathways, which are related to energy metabolism, inflammatory reactions, and platelet activation (as shown in Fig. 4E). Notably, the significant pathways in the R-25°C_vs_25°C group predominantly encompassed complement and coagulation cascades, inflammatory reactions, and responses to fluid shear stress (as depicted in Fig. 4F). Some genes in these significantly enriched pathways are indicated in the enhanced volcano plots. The enrichment analysis of DEGs using GO and KEGG revealed distinct biological processes in the brain of P. leopardus in response to temperature changes. When the temperature dropped to 13 °C, significant changes were observed in several key pathways. These included coagulation dysfunction, blood-brain barrier dysfunction, and inflammatory reactions. These findings suggest that the brain experiences disruptions in these critical processes in response to cold stress. Upon rewarming, a different pattern emerged. Comparing the R-25 °C group with both the 13 °C and 25 °C groups, the brain tissue displayed metabolic adjustments, platelet activation, and immune responses. This indicates that as the temperature returns to normal, the brain undergoes metabolic changes, activates platelets, and mounts immune responses to adapt to the changing environmental conditions. In summary, the enrichment analysis of DEGs demonstrates temperature-induced alterations in brain function, encompassing coagulation, barrier integrity, inflammation, metabolism, platelet activation, and immune responses. Validation of gene expression profiles To determine the reliability of the transcriptomic data and further validate the gene expression changes within key pathways, qRT-PCR analysis on DEGs (cldn15a, cebpd, chac1, cyp1a1, Jam2a, pfkfb3, fosl1a, hist1h1d, rorb1) enriched in the pathways of interest and reference gene b2m and FISH control gene tjp1 was performed. The expression changes of these gene in P. leopardus brain obtained through transcriptome analysis and qPCR are consistent, as shown in Fig. 5A and Fig. S4. The cldn15a, cebpd and cyp1a1 gene had the highest expression and significant differences in the brain of 13 °C group, followed by R-25 °C group and control 25 °C group. The cldn15a, cebpd, and cyp1a1 genes are involved in cell junctions, immunity, and metabolism functions, respectively. The expression of chac1 gene which involved in stress response significantly increased after rewarming. By focusing on these genes, we aimed to shed light on the intricate molecular mechanisms governing BBB integrity within the brain. Quantitative PCR and FISH results. A Relative expression level of cldn15a, cebpd, chac1, cyp1a1 and tjp1 in 13 °C, 25 °C and R-25 °C groups. B Co-localization of cldn15a and tjp1 antisense probe in brain at 25 °C, 13 °C and R-25 °C groups. a, e and i Nucleus was counterstained with DAPI. b, f and j The fluorescent signal of cldn15a antisense probe. c, g and k The fluorescent signal of tjp1 antisense probe. d, h and l Merged images. Scale bars = 50 μm. Arrows indicate capillaries In addition, to better visualize how genes related to the BBB change their expression patterns in the brain, we employed fluorescence in situ hybridization (FISH) to focus on a specific gene: claudin 15a (cldn15a). Meanwhile, we selected Tight Junction Protein 1 (tjp1) gene as a control gene. The stability of the gene expression was validated through qRT-PCR, and the results are shown in Fig. 5A. The protein Zona occludens-1encoded by tjp1 gene is a crucial component of tight junctions, and the transcription of this gene showed no significant difference among all groups in this study. It can be clearly seen from the FISH results (Fig. 5B) that the fluorescence signal of cldn15a was strongest in the cerebral capillary cells of the 13 °C groups. Discussion We aimed to investigated the impact of low-temperature stress on the brain and the changes that occur in P. leopardus during a simulated cold event under natural conditions. To achieve this goal, we conducted transcriptome analysis, blood enzyme activity measurement, and FISH experiments to explore the effects of ambient temperature cooling and the regulation after rewarming on the brain tissue of P. leopardus. Our results indicate: (1) P. leopardus responds to acute cold stress by reducing life activities; (2) Low temperature results in blood-brain barrier disruption and inflammatory reaction; (3) The rewarming process partially recovered BBB injury while being accompanied by persistent inflammation. Effects of cold stress on hormone levels and behavioral changes The hormone content results of this low-temperature experiments, simulating natural low-temperature events, show that the changes in hormone levels in the serum of P. leopardus during the acute cooling process were not significant. Hormone levels can be indicative of level of stress experienced by an organism. Cortisol is commonly released during stress in teleost [50, 51]. The main regulator of cortisol secretion is adrenocorticotropic hormone [52]. Evidence from previous cold-treatment experiments on fish indicates that different species exhibit varying biological stress responses to low temperatures. For instance, in large yellow croaker, serum cortisol levels significantly increased when exposed to sublethal temperatures for 0–48 h [18]. In contrast, the sea raven does not immediately display changes in cortisol levels following cold exposure [20]. Similarly, in the subtropical pacu, cortisol levels do not show significant changes even after 24 h of cold exposure [22]. Furthermore, within the same species, the variations in stress hormone levels due to cold treatment can depend on both the intensity and duration of the exposure. For example, the concentrations of epinephrine and cortisol in plasma did not significantly differ between the cold-treated 19 °C group and the control 25 °C group. However, they were significantly higher in the cold-treated 12 °C group, which experienced more severe cold conditions [50]. Similarly, in adult Danio rerio, cortisol levels were higher in the group subjected to lower temperatures for a duration of 6 days [24]. In Atlantic salmon (Salmo salar), stress hormone levels suggested adrenergic control and responsiveness in response to cold temperatures, but these changes were specific to longer exposure (> 3 weeks) rather than acute exposure (8 –1 °C at 1 °C per hour) [53]. Therefore, it is evident that the primary responses in teleost fish governed by the activation of the hypothalamus-pituitary-interrenal gland/head kidney (HPI) axis are influenced not only by the magnitude but also the duration of exposure to stress. During our experiment, we observed the distinct changes in individual behaviors. Initially, the P. leopardus engaged in normal foraging activities at 25 °C; however, as the temperature dropped, they gradually reduced their activity levels and even exhibited signs of fainting and rolling onto sides. Detailed behavioral records are available in Table. S2. When the temperature reached 19 °C, the fish began to congregate, and their swimming and foraging activities decreased. At 16 °C, some fish showed mild surfacing and tilting behaviors. By 13 °C, some exhibited severe surfacing, became unresponsive to external stimuli, and rolled onto their sides. However, these adverse behaviors gradually diminished with rewarming. The changes in behavior observed in this study are similar to those observed in other fish such as Atlantic salmon (Salmo salar) and Silver barb (Barbonymus gonionotus) [54, 55]. Together with our hormonal results, we speculate that the leopard coral grouper minimizes short-term energy expenditure by reducing its activities level in response to rapid cooling. This leads to a gradual transition into a state of low physiological activity, enabling the fish to adapt to the drop in external environmental temperature and ensuring their survival under low-temperature conditions, thereby helping them endure the cold period. When organism is trapped in an unsuitable environment, cell apoptosis or immune regulation can increase the content of reactive oxygen species (ROS), which can attack the mitochondria of cells and cause cell damage [56, 57]. Consequently, when faced with such circumstances, organisms must deploy antioxidant enzymes to counteract the detrimental effects of excessive ROS, striving to maintain a state of relative homeostasis [58]. SOD, an antioxidant enzyme, can promote the decomposition of free radicals, and relieve the damage caused by ROS to cells. For instance, in juvenile tilapia Oreochromis niloticus, the low temperature treatment for 24 h caused a significant increase in SOD activity in comparison to the control group [17, 59]. In this study, there was no significant change of the SOD activity in serum of P. leopardus during cooling process, but the activity was significantly upregulated after rewarming. It can be inferred that after experiencing temperature rewarming, P. leopardus may has produced and accumulated a large amount of ROS. However, our study does not provide evidence to prove whether the elevation of ROS originates from specific tissues or cells, or whether it is due to metabolic changes, cellular damage, or alterations in immune function. Given that the primary source of ROS, the electron transport chain, may be blocked or become less active under low-temperature conditions, the reason why we did not observe an increase in SOD during hypothermia could be due to the inhibition of ROS production [60]. However, it cannot be ruled out that the sampling time or the lag in SOD enzyme activity changes might have contributed to the lack of significant variation observed. Additionally, our study did not obtain data on enzymatic activity in brain tissue. Nevertheless, we emphasize that the oxidative stress conditions generated within the brain tissue may differ from those observed in the serum and hold greater relevance for understanding the direct impact of low temperatures on the brain. Previous research has shown that low temperatures can lead to hypoxia and ischemia in the brain tissue of common carp [36]. This aligns with our observation of P. leopardus exhibiting surfacing behavior under low-temperature conditions (Table. S2), which is generally considered a sign of hypoxia in fish. Erwin H. used functional magnetic resonance imaging fMRI to study brain responses to acute cooling in common carp, observed cerebral blood volume (CBV) and blood oxygenation level-dependent (BOLD) both decreases after the temperature drop [61]. The reduction in CBV is thought to be a protective mechanism that slows the influx of cold blood from the gills into the brain, akin to the vasoconstriction response observed in mammals during hypothermia. Meanwhile, the decrease of BOLD, caused by the increased affinity of hemoglobin for oxygen, and CBV exacerbates brain cells lack oxygen damage [61, 62]. In the transcriptome results of this study, the expression level of hypoxia inducible factor 1 gene (hif1al) and ChaC, cation transport regulator homolog 1 (chac1) increased and was significantly upregulated after rewarming. The chac1 has been identified as a novel pro-apoptotic factor in cells under endoplasmic reticulum (ER) stress [63]. Among its related pathways are cellular responses to stimuli and metapathway biotransformation [64]. This suggests that the brain of P. leopardus may have experienced hypoxia and injury during rewarming processes, consistent with the behavioral changes we observed, where surfacing behavior worsened as the temperature reached its lowest point and became even more pronounced during the rewarming phase. The lack of significant upregulation of hif1al expression in the hypothermia group could be attributed to the fact that the species chose to reduce its activity level to decrease oxygen consumption, making the hypoxic condition less apparent. The worsening surfacing behavior observed in P. leopardus during the initial stages of rewarming could be due to the gradual recovery of cellular metabolic activity without a corresponding timely restoration of cerebral blood flow, resulting in a more severe hypoxic condition. This change is attributed to reoxygenation, a phase where oxygen consumption often surpasses that during the initial normoxic exposure [65]. Additionally, in a study involving goldfish exposed to hypoxia, it was found that reoxygenation doubled the levels of lipid peroxidation in metabolically active brain organ, but not during anoxic exposure [58]. With the reintroduction of oxygen, electrons accumulated during anoxia may intensively “leak” from the electron transport chain, generating a burst of oxyradicals [58]. This surge in radical production prompts the organism to produce antioxidant enzymes to eliminate excess ROS and combat the negative effects caused by these radicals. Thus, we observed enhanced SOD enzyme activity during the rewarming phase. This finding suggests that maintaining homeostasis during the rewarming phase following cold exposure may present a greater challenge due to the complex responses required to counteract the effects of accumulated ROS. Bblood-brain barrier The blood-brain barrier (BBB) plays a pivotal role in meticulously regulating the exchange of substances between the bloodstream and the central nervous system (CNS). This stringent control safeguards the homeostasis of the cerebral microenvironment, thereby maintaining the normal operation of the CNS [66]. BBB dysfunction is reported in many pathological conditions in mammal including hypoxic and ischemic insult, Parkinson’s and Alzheimer’s disease, epilepsy and brain tumors [67–70]. The damage to the integrity of cerebral blood vessels is related to the decrease in blood volume and hypoxia in the brain caused by hypothermia [71]. Once the integrity of the BBB was compromised, permeability increased and the physical structure was damaged. This allowed substances that normally cannot cross the BBB to enter the CNS, leading to neurotoxicity and causing damage that is difficult to repair, thereby affecting normal function [72]. There are investigations suggest that TJs disruption and increased BBB permeability triggered by hypoxia–ischemia in mammals, accompanying astrocytes releasing inflammatory mediators, cerebrovascular endothelial cells up-regulating genes related interleukin, TNF-α, monocyte chemoattractant protein-1, subsequently leads to leukocyte recruitment and transmigrated across the endothelium and the BBB [40, 66, 67, 73, 74]. The nervous function will be affected if nerves are damaged, which may explain why fish was gradually showing weakened reaction and inactive mobility when water temperatures drop. The inner ear of fish that maintains balance relies on the seventh nerve to transmit signals to the brain to maintain movement balance [75]. We observed that P. leopardus exhibits difficulties in maintaining body balance during cooling, often manifesting as tilting sideways, head down, or even belly up. However, it is noteworthy that these phenomena are transient, and the capacity for balance is reinstated as temperatures return to their normal range. In the transcriptome results of this study, the expression levels of junctional adhesion molecules, claudin genes have undergone alterations in the brain. These genes encode structural proteins intimately associated with the BBB, thereby contributing significantly to the preservation of its integrity. Claudins are integral membrane proteins that play a crucial role in forming the tight junction strands between adjacent endothelial cells in capillaries [41, 76]. Among the claudin family, cldn15a is a protein-coding gene and one of at least 26 claudin family members identified in P. leopardus, with a total of 62 annotated genes. Our comprehensive transcriptome analysis uncovered a noteworthy upregulation of three claudin genes, including cldn15a, cldni, and cldn7b-like, in response to temperature decreases. There have been reports indicating differential DNA methylation of the cldn15 gene in patients with Alzheimer’s disease (AD) [77]. Studies in mice have found that chronic exposure to ambient particulate matter (PM2.5) leads to significant decreased of the cldn15, cldn2, and cldn3 genes in the brain, which disrupts tight junction protein (TJP) function and increases the permeability of the BBB, ultimately resulting in cognitive decline [78]. Compared to mammals, the Claudin gene family exhibits extensive expansion in teleost fish, Fugu rubripes [79]. However, there have been no reports regarding the Claudin gene family or its functions in the P. leopardus. Therefore, further analysis and understanding of this gene family in this species are essential. In the transcriptome results, we also screened for downregulation of the jam2a. Junctional adhesion molecules as a classical cell-cell adhesion molecule is enriched at cell-cell contacts of epithelial cells and endothelial cells [80]. JAMs are involved in varieious biological processes depend on physical and functional interaction with integrins [39]. The redistribution of JAMs between cell-cell junctions in endothelial cells is influenced by pro-inflammatory stimulation, thus making JAMs exposed to leukocyte interactions at the luminal surface [81, 82]. Furthermore, genes linked to processes involving cell adhesion, cell recruitment, vascular permeability, and the pro-inflammatory response, such as itga9, l-selectin, hif1al and tiam2a, have also exhibited modifications in expression level in this transcriptome results. The incomplete blood-brain barrier can cause inflammatory reactions in the brain. In the transcriptome results, we have enriched the expression changes of genes such as cebpd, ccl20, socs3, c8g, fosl1a, which are involved in processes such as the IL-17 signaling pathway and the TNF signaling pathway. Similar findings have also been reported of mice Chronic PM2.5 exposed where IL-1β, IL-6, and TNFα showed obvious upregulation in the brains, Enhanced neuroinflammation [78]. Transcriptomic analyses also indicate that as temperature recovers, the expression levels of specific genes associated with the BBB tend to normalize, akin to those observed in the control group. Collectively, these alterations in gene expression patterns serve as indicative evidence of the impact of cold stress and subsequent rewarming on the blood-brain barrier in the brain of P. leopardus. Drawing from these observations, we posit that the reduced ability of P. leopardus to maintain bodily equilibrium may be ascribed to nerve damage and interruption in nerve signal transmission resulting from cold-induced blood-brain barrier damage. As temperature recovers, the damage is gradually repaired, subsequently restoring their equilibrium capacity. Conclusion In this study, we investigated the impact of environmental temperature fluctuations, including cooling and rewarming, on the brain tissue of Plectropomus. leopardus. Our goal was to clarify the relationship between the adaptability of P. leopardus and the changes in physiological activity during short-term exposure to low temperatures. We observed the life activity decreased and exhibited abnormal behaviors such as surfacing and lose balance during the cooling and rewarming phases. Transcriptome analysis revealed blood-brain barrier dysfunction and inflammation in the brain during cold exposure. Post-rewarming, significantly elevated SOD enzyme activity was detected, along with hypoxia signals and persistent inflammatory signals in the brain. These findings provide valuable insights into the physiological responses of this species when faced with cold stress conditions. Electronic Supplementary Material Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 2 Supplementary Material 3 Supplementary Material 4 Supplementary Material 5 Supplementary Material 6 Acknowledgements This work was supported by the High-performance Computing Platform of YZBSTCACC. Author contributions YLG , CW, BW conceptualization and methodology; YLG, CW, HD, PYL, YRG and KNZ, sampling; CW and HD, investigation; ZMB, resources; YLG, writing—original draft preparation; ZQ, BW and JJH, writing—review and editing. All authors reviewed the manuscript. Funding This research was funded by the Key R&D Project of Hainan Province (ZDYF2024XDNY248 and ZDYF2023XDNY182), the Project of Sanya Yazhouwan Science and Technology City Management Foundation (SKJC-2023-01-004), the Shandong Excellent Young Scientists Fund Program (Overseas, 2023HWYQ-055), Young Taishan Scholar Program of Shandong Province (tsqnz20221112). Data availability The Sequence Read Archive (SRA) has been deposited at GenBank in NCBI, and it will be released upon this study publication. The transcriptome data has been available in the NCBI GenBank repository with the accession number PRJNA1105558, please see the link below for details (https://dataview.ncbi.nlm.nih.gov/object/PRJNA1105558). Declarations Ethics approval and consent to participate This study was approved by the College of Marine Life Sciences, Ocean University of China Institutional Animal Care and Use Committee on 10 October 2018 (Project Identification Code: 20181010). This study complies with ARRIVE guidelines (https://arriveguidelines.org). Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Yilan Guo and Cun Wei contributed equally to this work. Contributor Information Zhe Qu, Email: qz@ouc.edu.cn. Bo Wang, Email: wb@ouc.edu.cn. References Associated Data Supplementary Materials Supplementary Material 1 Supplementary Material 2 Supplementary Material 3 Supplementary Material 4 Supplementary Material 5 Supplementary Material 6 Data Availability Statement The Sequence Read Archive (SRA) has been deposited at GenBank in NCBI, and it will be released upon this study publication. The transcriptome data has been available in the NCBI GenBank repository with the accession number PRJNA1105558, please see the link below for details (https://dataview.ncbi.nlm.nih.gov/object/PRJNA1105558).