Periodic heat waves-induced neuronal etiology in the elderly is mediated by gut-liver-brain axis: a transcriptome profiling approach

Heat stress exposure in intermittent heat waves and subsequent exposure during war theaters pose a clinical challenge that can lead to multi-organ dysfunction and long-term complications in the elderly. Using an aged mouse model and high-throughput sequencing, this study investigated the molecular dynamics of the liver-brain connection during heat stress exposure. Distinctive gene expression patterns induced by periodic heat stress emerged in both brain and liver tissues. An altered transcriptome profile showed heat stress-induced altered acute phase response pathways, causing neural, hepatic, and systemic inflammation and impaired synaptic plasticity. Results also demonstrated that proinflammatory molecules such as S100B, IL-17, IL-33, and neurological disease signaling pathways were upregulated, while protective pathways like aryl hydrocarbon receptor signaling were downregulated. In parallel, Rantes, IRF7, NOD1/2, TREM1, and hepatic injury signaling pathways were upregulated. Furthermore, current research identified Orosomucoid 2 (ORM2) in the liver as one of the mediators of the liver-brain axis due to heat exposure. In conclusion, the transcriptome profiling in elderly heat-stressed mice revealed a coordinated network of liver-brain axis pathways with increased hepatic ORM2 secretion, possibly due to gut inflammation and dysbiosis. The above secretion of ORM2 may impact the brain through a leaky blood–brain barrier, thus emphasizing intricate multi-organ crosstalk.


Heat stress-induced transcriptional alterations in both the brain and liver entailed discernible shifts in the expression patterns of pivotal genes
Upon exposure to heat stress, the cerebral tissues of the heat stress group manifested substantial alterations in the gene expression profile as illustrated in the volcano plot compared to the control (Fig. 2A).A total of 99 genes were significantly upregulated, while 86 were notably downregulated in response to the exposure of heat stress (Fig. 2A, P FDR < 0.05, Log 2 fold change > 1).The heat map, derived from transcriptomic profiling of brain lysates, portrayed the expression changes of genes such as Rpl41, Fopnl, Rarb, Slirp, Hmgn2, Kl, Arc, Lrp1, Npas4, Mast4, among others (Fig. 2B).Similarly, in hepatic tissues, heat stress resulted in a significant upregulation of 91 genes and downregulation of 63 genes, as indicated in the corresponding volcano plot (Fig. 3A, P FDR < 0.05, Log 2 fold change > 1).The pertinent heat map revealed alterations in the expression profile of genes including Mt1, Mt2, Ccl5, Irf7, Pim3, Foxg1, Creld2, Acnat1, Pdia4, and others within the liver tissues when subjected to heat stress (Fig. 3B).

Heat stress-mediated gene expression changes were extrapolated in contributing pathways integral to organ-level pathological manifestations
The pools of differentially expressed genes (DEGs), derived from comprehensive transcriptomic analyses, served as the foundation for predicting enriched biological pathways in the brain, liver, and shared DEGs indicative of the liver-brain axis under the influence of heat stress (Fig. 4).Within the brain, prominently upregulated pathways encompassed inflammatory S100B family signaling, IL-33 signaling associated with microglial activation, degradation pathways for dopamine, nicotine, acetone, bupropion, and melatonin, Th2 signaling, xenobiotic metabolism CAR signaling, IL-17 signaling, DHCR24 signaling, G-protein coupled receptor signaling, pathogeninduced cytokine storm signaling, and acute phase response signaling, among others.Conversely, downregulated pathways included aryl hydrocarbon receptor signaling (AhR), IL-12 signaling, IL-1 mediated inhibition of RXR signaling, inhibition of matrix metalloprotease signaling, extrinsic prothrombin activation signaling, and several other pathways significant for inflammatory pathology (Fig. 4A; P FDR < 0.05, the threshold at -log(p value) ~ 1.3, Log 2 fold change > 1).In the liver, tissues exposed to heat stress, the upregulated pathways comprised of IL-17 signaling, NOD1/2 signaling associated with liver injury, TREM1 signaling of hepatic inflammation, tumor microenvironment signaling, Th1 signaling, and eNOS signaling.Meanwhile, downregulated pathways included G-protein coupled receptor signaling, cyclic AMP (cAMP) mediated signaling, retinoic acid receptor signaling (RAR), SNARE signaling, calcium, and opioid signaling, among others (Fig. 4B; P FDR < 0.05, the threshold at -log(p value) ~ 1.3, Log 2 fold change > 1).The predicted pathways of the liver-brain axis exhibited upregulation in chemokine signaling, Th1 and Th2 signaling, neuroinflammation signaling, synaptic long-term depression signaling, cytokine storm signaling, apelin adipokine signaling, and more.Conversely, downregulated pathways included glutamate signaling and dopamine-DARPP32 feedback in cAMP signaling (Fig. 4C; P FDR < 0.05, the threshold at -log(p value) ~ 1.3, Log 2 fold change > 1).

Expression fold changes of the top genes observed in the brain and liver unveiled potential validatory mediators of the liver-brain axis upon exposure to heat stress
The transcriptomics data was further screened to identify key genes that could mediate the effects of heat stress.This was done by examining the gene list showing differential expression due to heat stress.The changes in gene expression were quantified using log 2 fold changes, which were then converted to expression fold changes.Several genes were found to be upregulated in the brain, including Cd4, Rpl, Ptprv, Adora2a, Gpr6, and Tac1, among others.Conversely, genes such as Eps8l2, F5, Aqp1, Npas4, Slc4a5, Pecam1, and others were downregulated (Fig. 8A; p < 0.001).When comparing the liver data from the control and heat stress groups, a significant set of genes were identified that could potentially act as mediators.Some of these mediators could potentially reach the brain via the systemic route and exert effects on the neuronal circuits.The upregulated genes included acute phase reactants like Orm2, chemokines such as Ccl5, Cxcl9, and other crucial non-mediators like Cib3, Usp2, Socs3, Irf7, etc.On the other hand, genes like Derl3, Gale, Acnat2, Manf, and others were found to be downregulated (Fig. 8B; p < 0.001).www.nature.com/scientificreports/

Heat stress augmented the secretion of hepatic ORM2 and its abundance in the hippocampus along with the frontal cortex, likely traversing through leaky Blood-Brain-Barrier (BBB)
Immunohistochemical analysis of liver slices showed a marked augmentation in Orosomucoid 2 (ORM2) immunoreactivity, signifying an escalated release of this protein at the tissue level within the liver of the heat stressed group (Fig. 9A,B; p < 0.001).Previous research has established a linkage between ORM2 release from the liver and its involvement in intestinal inflammation and gut dysbiosis 30 .Consequently, IL-1β immunoreactivity was assessed as an indicator of intestinal inflammation within the experimental mouse cohorts.IL-1β immunoreactivity exhibited a substantial increase in the intestinal sections from the heat stress group compared to the control (Fig. 9C,D; p = 0.008).Notably, heat stress-induced gut dysbiosis has been acknowledged as a contributing factor to intestinal inflammation 31,32 .Therefore, an investigation was performed into the correlation between gut dysbiosis and intestinal inflammation concerning the release of ORM2 in the liver.The Firmicutes to Bacteroidetes ratio has long been recognized as a marker of gut dysbiosis in the context of hepatic inflammation and diseases [33][34][35] .Hence, the whole metagenome sequencing of fecal samples from the control and heat stress mouse groups were performed.The data elucidated a significant increase in the Firmicutes to Bacteroidetes ratio in the heat stress group compared to the control (Fig. 9E; p < 0.001).Given the previously established correlation between ORM2 abundance and gut dysbiosis, a correlation plot was constructed, revealing a significant positive correlation between the Firmicutes to Bacteroidetes ratio and ORM2 immunoreactivity in liver tissues of www.nature.com/scientificreports/ the heat stress mouse group (Fig. 9E; p = 0.014, r = 0.813).Subsequently, the integrity of the blood-brain barrier (BBB) was examined in brain sections of the respective mouse groups.Immunofluorescence analysis showed BBB disruption, quantified in terms of Claudin 5 (Red) and CD31 (Green) colocalization events (Yellow) counterstained with DAPI for nuclear imaging (Fig. 9G; p < 0.001).Colocalization was markedly diminished in the heat stress group compared to the control, suggesting a potential disruption of the BBB (Fig. 9H; p = 0.0085).ORM2 protein secretion primarily emanated from hepatocytes 36,37 .Past investigations have delineated the role of ORM2 in neuroinflammation and ischemic stroke events 38,39 .Consequently, the abundance of ORM2 was evaluated in brain sections, revealing a significant increase in ORM2 immunoreactivity in the hippocampus and frontal cortex regions of the brain in the heat stress group (Fig. 9A,B; p < 0.001).Intriguingly, the brain transcriptome analysis disclosed no ORM2 mRNA fold changes in the heat stress group when compared to the control.Despite ORM2 being predominantly secreted by the liver, its heightened abundance in the brain suggests increased secretion from hepatocytes during heat stress, possibly reaching brain tissues through a leaky blood-brain barrier via the systemic route.

Heat stress-induced transcriptomic alterations together with acute phase response ensued neural, hepatic, and systemic inflammation further debilitating synaptic plasticity
Immunohistochemical analysis conducted on brain sections aimed to explore the pleiotropic cytokine IL-6, the synaptic plasticity marker Brain-Derived Neurotrophic Factor (BDNF), and the microglial activation marker Translocator Protein (TSPO) (Fig. 10A).Within the heat stress group, a noteworthy augmentation in IL-6 immunoreactivity was discerned (Fig. 10A,B; p = 0.0048), while the immunoreactivity of BDNF exhibited a significant decrease (Fig. 10A,C; p = 0.0031).Additionally, as assessed by TSPO immunoreactivity, microglial activation was significantly heightened in the heat stress group compared to the control (Fig. 10A,D; p = 0.0051).The systemic level of the proinflammatory cytokine IL-1β was quantified via ELISA using serum samples from the control and heat stress mice groups, revealing a substantial elevation in the systemic IL-1β level within the heat stress group (Fig. 10E; p = 0.001).Furthermore, a significant increase in the systemic level of IL-6 was also evident in the heat stress group in comparison to controls (Fig. 10F; p < 0.001).Immunohistochemistry conducted on liver slices from the control and heat stress mouse groups showed a marked escalation in the immunoreactivities of the Kupffer cell activation marker CD68 (Fig. 10G,H; p = 0.0105) and the proinflammatory cytokine IL-1β (Fig. 10G,I; p < 0.001) in the heat stress group.

Discussion
This study is the first report of a molecular investigation using genomic approaches of heat stress pathophysiology, focusing on the concerted actions of brain and liver.Further, the present study focused on the molecular pathways of the gut-liver-brain axis, linking intestinal pathology and gut dysbiosis.Employing multiorgan transcriptomic profiling, it identified common heat stress-affected genes in the brain and liver.Rigorous bioinformatics analysis unveiled significant transcriptome changes and identified enriched pathways.The study uniquely identified liver-derived mediators connecting the gut-brain-liver axis and intricately correlated with heat stress pathophysiology, marking an effort to unravel organ interplay during heat stress, advancing molecular insights into the pathophysiology of adverse heat events.Top differentially expressed genes, like Rpl41, upregulated in the brain under heat stress are linked to ATF4 phosphorylation, oxidative stress, and neuronal degeneration, notably in Alzheimer's and Parkinson's diseases 40 .Another gene, Rarb was also found to be upregulated in heat-stress brain tissues, and gain-of-function mutations of Rarb are associated with intellectual disability and progressive motor impairment, indicating a potential role of Rarb in brain function and development 41 .Hmgn2 is upregulated in cerebral tissue upon heat stress and plays a specific role in the brain, maintaining chromatin accessibility during corticogenesis, crucial for adult neurogenesis and synaptic plasticity 42 .Heat stress in our study was observed to downregulate Npas4, a brainexpressed transcription factor crucial for memory and inhibitory synapse formation, with deficiency linked to cognitive deficits and implications in neuroprotection against stress, offering potential therapeutic insights for conditions like epilepsy, Alzheimer's 43,44 .Furthermore, Npas4 significantly downregulated in brain tissues from the heat stress group was reported previously to serve in the maintenance of the blood-brain barrier integrity, and its deficiency compromises neuroprotection [45][46][47] which is supportive of this study's observations of the leaky blood-brain barrier and deteriorating synaptic plasticity in the heat stress exposed mice.The enriched pathways observed in the brain under heat stress strongly highlighted the activation of S100B signaling, a central nervous system calcium-binding protein concentrated in astrocytes, that serves as a reliable biomarker of neural distress and is implicated in neuropathology.In brain inflammation, the above molecule acts as both a pro-inflammatory cytokine and a Damage-Associated Molecular Pattern (DAMP), displaying concentration-dependent effects on neurons, ranging from neurotrophic support to inducing apoptosis 48,49 .Another pathway predicted in brain tissues of the heat stress group was the activation of Interleukin-17 (IL-17) signaling, which is associated with disrupting the blood-brain barrier, fostering neuroinflammation, and contributing to tissue damage in neuroinflammatory conditions of ischemic brain injury.Its impact on various neural cell types, including microglia, endothelial cells, astrocytes, and neurons, triggers inflammatory responses and poses a risk of tissue damage 50 .Interleukin-33 (IL-33) was also observed to have enriched in brain lysates of the heat stress group known to exhibit roles in brain inflammation.This may act as a pro-inflammatory mediator that can activate cells such as astrocytes and microglia, potentially leading to sustained inflammation, neuronal damage, and cognitive impairment in neuroinflammatory pathology 51,52 .Additionally, downregulation of the Interleukin-12 (IL-12) signaling has been observed in heat stress-induced neuronal pathways; IL-12 is associated with inducing neuroprotection and shaping the trophic factor milieu within the inflamed central nervous system, with neurons being the dominant source of trophic factors 53  www.nature.com/scientificreports/ 4 which is novel in depicting neuronal distress while other proteins of the same family were previously correlated with synaptic plasticity factors under various neurological disease pathology 54,55 .Serpine 3, was inhibited in heat-stressed brain tissues and is known to hasten the neuronal apoptosis and more severe impairments of spatial learning and memory following a hippocampal injury 56 .Observations from the transcriptomic profiling and downstream bioinformatic analyses of brain tissues comparing control and the heat stress groups are in concordance with the functional validations indicating neuroinflammation (IL-6 release, microglial activation, TSPO) and cognitive decline (decrease in BDNF) marked in the heat stress group.Upregulated genes in the liver due to heat stress comprised of Ccl5, or Rantes (Regulated upon Activation, Normal T Cell Expressed and Secreted), a well-known hepatic chemokine influencing immune cell recruitment in liver disease.It activates T cell proliferation and cytokine production, contributing to inflammatory responses, hepatic fibrosis, and early stages of metabolic dysfunction-associated steatotic liver disease (MASLD) by inducing steatosis through the Ccr5 receptor 57,58 .Creld2 or Cysteine-rich with EGF-like domains 2 was also upregulated in the liver lysates of the heat stress group and is reported to be crucial for liver metabolism homeostasis, particularly in the unfolded protein response (UPR).It also acts as an ER-stress inducible gene, enhancing protein folding, and its deficiency is associated with dysregulated UPR, and hepatic steatosis, and impacts energy expenditure, emphasizing its key role in liver health and function 59 .Irf7 (Interferon Regulatory Factor 7) was upregulated, which is also crucial for IL-1β release and concomitant inflammatory response in the liver 60 .Pathway enrichment study identified the activation of NOD1/NOD2 in liver tissues of the heat stress group which triggers proinflammatory and type I interferon responses, contributing to liver injury.These receptors play a role in recognizing bacterial pathogens, inducing liver inflammation, and have associations with innate immune activation in metabolic diseases, with studies suggesting that blocking NOD1 can be protective against tissue injury 61,62 .TREM-1 (Triggering Receptor Expressed on Myeloid Cells 1) signaling was also activated in heat-stressed liver tissues and is implicated in liver inflammation and fibrosis by promoting hepatic inflammation, activating stellate cells, and serving as a master regulator of Kupffer cell activation and linked to inflammation, lipid accumulation, fibrosis, and tumor progression in liver-related diseases 63 .Interleukin-17 (IL-17) signaling activation was notable in the liver transcriptome of the heat stress group, which is linked to liver inflammation and fibrosis, with elevated levels observed in liver diseases, inducing immune cell infiltration, damage, inflammation, and fibrosis.Targeting the IL-17 axis is considered a potential therapeutic approach for treating liver-related conditions 64,65 .Interactome mapping in the liver lysates from the heat stress group showed an array of significant interactions compared to the controls.Projected upregulations were observed in the adipokine LRG1 (Leucine-rich alpha-2-glycoprotein 1) associated with liver diseases, notably in obesity-induced hepatosteatosis and insulin resistance.Its endocrine action, including suppressing fatty acid β-oxidation and promoting de novo lipogenesis, contributes to hepatosteatosis.At the same time, LRG1 also influences Kupfer cell activation in the liver and enhances TGF-β1 signaling, suggesting a significant role in liver metabolism and diseases associated with obesity 66,67 .The Krüppel-like factor 4 (KLF4) transcription factor has been associated with various physiological and pathological processes in the liver and was also found to be upregulated in the liver interactome of the heat stress group is known to augment TGF-β1 expression and liver inflammation, indicating a potential role in modulating the fibrotic response in organ systems including liver fibrosis 68 .The observations from the transcriptomic profiling and downstream bioinformatic analyses of the liver tissues comparing control and the heat stress groups are congruent with the functional studies conducted.Validations revealed a correlative hepatic inflammation (Kupfer cell activation, CD68 immunoreactivity) and IL-1β release, suggesting the onset of hepatic inflammation.

Figure 9.
Heat stress augmented the secretion of hepatic ORM2 and its abundance in the hippocampus along with the frontal cortex, traversing through leaky Blood-Brain-Barrier (BBB).(A) ORM2 immunoreactivity was shown by immunohistochemistry in liver sections from Control and Heat Stress mouse groups; Images were taken at 20X magnification and displayed with a scale of 200 μm.(B) ORM2 immunoreactivity in both groups was measured as arbitrary light units from six separate microscopic fields and plotted along the ordinate.All p values were derived through an unpaired t-test, with the significance level set at p < 0.05.(C) IL-1β immunoreactivity was shown by immunohistochemistry in small intestinal sections from Control and Heat Stress mouse groups; Images were taken at 20X magnification and displayed with a scale of 100 μm.(D) IL-1β immunoreactivity in both groups was measured as arbitrary light units from six separate microscopic fields and plotted along the ordinate.All p values were derived through an unpaired t-test, with the significance level set at p < 0.05.(E) The bar graph was plotted as the difference in ratio value of the relative abundance of the gut microbial phyla Firmicutes and Bacteroidetes compared between control and heat stress groups.The p value was derived through an unpaired t-test, with the significance level set at p < 0.05.(F) A correlation analysis was conducted between the gut dysbiosis marker Firmicutes/Bacteroidetes ratio and the abundance of ORM2 in liver tissues.The correlation plot includes Pearson's linear regression depicted in red, along with a 95% confidence interval.(G) Claudin5 (red) and CD31 (green) were dual labeled through immunofluorescence staining with colocalization (yellow) observed at 60X (oil) magnification.The images were presented with a scale of 20 μm in the brain sections from both the Control and Heat Stress mouse groups.(H) Claudin5-CD31 colocalization (yellow) events in both groups were quantified as arbitrary fluorescent units from six distinct microscopic fields and plotted along the ordinate.All p values were derived through an unpaired t-test, with the significance level set at p < 0.05.(I) ORM2 immunoreactivity was shown by immunohistochemistry in brain sections (hippocampus and frontal cortex) from Control and Heat Stress mouse groups; Images were taken at 10X, 40X magnification displayed with the scale of 200 μm.(J) ORM2 immunoreactivity in both groups was measured as arbitrary light units from six separate microscopic fields and plotted along the ordinate.All p values were derived through an unpaired t-test, with the significance level set at p < 0.05.www.nature.com/scientificreports/ The thorough bioinformatic analyses further revealed the liver-brain connection in heat stress based on the commonly differentiated genes in both organs.The liver-brain axis is a vital communication network influencing human health, connecting the gut, liver, and brain.It plays a role in various disease conditions involving endocrine, humoral, metabolic, and immune pathways 69 .The impact of climate change on neurological diseases and gastrointestinal health, including the liver-brain axis, has been highlighted, with potential implications for metabolic liver diseases and associated neuropathologies 70 .The increased activity in chemokine signaling, Th1 and Th2 signaling, neuroinflammation signaling, synaptic long-term depression signaling, adipokine signaling, etc. were observed in the enriched pathways of the Liver-Brain axis upon exposure to heat stress.An array of cytokines such as CCL3, CCL4, CCL5, IL-35, IL-12, etc., different acute phase responders like ORM2, ORM3, etc., and other intricate non-mediators were identified.Among these, the ORM2, a liver-produced hepatokine, Vol:.( 1234567890 www.nature.com/scientificreports/plays a crucial role in maintaining hepatic and systemic lipid balance and is implicated in regulating de novo lipogenesis 71 .ORM2 levels are associated with MASLD, while studies suggest its potential as a therapeutic target for liver diseases, particularly those related to lipid metabolism and steatohepatitis 36 .Additionally, ORM2 has shown promise as a biomarker for liver diseases, contributing to diagnostics and monitoring, emphasizing its significance in liver health and potential for therapeutic interventions 72 .Following the literature, our observations also revealed a significant increase in the production of hepatic ORM2 in the liver tissues of the heat stress group, compared with the control group, indicating heat stress elevated ORM2 production in the liver.Existing literature described the link between ORM2 release from the liver and its involvement with intestinal inflammation and gut dysbiosis 30 .Our results revealed the intestinal inflammation caused due to exposure to heat stress and concomitant events of gut dysbiosis measured through the alteration of Firmicutes to Bacteroidetes ratio.The Firmicutes to Bacteroidetes ratio is a gold-standard marker of gut dysbiosis 33,34 .Heat stress is well characterized in inducing gut dysbiosis and has been proven robust in causing intestinal inflammation 31 .Our results indicated a positive correlation plotted with gut dysbiosis and hepatic ORM2 production.Therefore, our data successfully delineated the compelling correlation between intestinal inflammation, gut dysbiosis, and hepatic ORM2 release in the context of heat stress-induced pathophysiology.This tended to justify the cause of augmented ORM2 production from the liver under heat stress.Also, systemic markers of inflammation, such as the levels of IL-6 evident in our data sets, pointed to the acute phase response in heat stress, which is consistent with existing literature 73,74 .Past studies have described the role of ORM2 in neuroinflammation and ischemic stroke events, and the existence of astrocyte-produced ORM2 was also deciphered 38,39 .Our results revealed a significant rise in ORM2 abundance in the brain's hippocampus and frontal cortex regions upon exposure to heat stress.However, the transcriptomic profiling of the brain tissues showed no ORM2 mRNA increase in the heat stress group compared to the control.Therefore, it may be predicted that the augmented ORM2 level in the brain tissues of the heat stress group may not be autocrine in nature that arose from astrocytes.Conversely, the ORM2 possibly reached the brain tissues predominantly secreted from the hepatocytes in the liver, and the increased abundance in the brain was probably due to its increased production in the liver and affected the neuronal tissues via an endocrine route following a leaky blood-brain barrier, though the above pathway may be hypothetical at this point.While this study delved into the functional aspects of heat stress pathophysiology, it primarily adopted an observational approach, relying on transcriptomic analyses, microbial sequencing, bioinformatic mining, and prediction models.The research, centered on mechanistic validations and pathological evaluations, is confined to identified genes and pathways.Systemic-level and multifactorial measurements including thorough mechanistic insights of the liver-brain mediators are beyond the study's scope.The authors propose subsequent deep investigations into ORM2 mechanisms, connecting the liver-brain axis in heat stress, using tissue-specific knockout models to confirm ORM2's source and eliminate astrocytic ORM2 interference in neuronal etiologies.In conclusion, the study was foundational in employing high-throughput sequencing, explored organ-specific heat stress pathologies, the possible involvement of the gut-liver-brain axis, and molecular interplays, especially in aged mice.The research identified mediators of gut-liver-brain communication, offering implications for developing prognostic and therapeutic markers, crucial for alleviating heat stress pathophysiology, particularly in the aged population.

Methods
The temperature and humidity-controlled rodent chamber (Model No# RIS26SD) was obtained from Powers Scientific Inc. (Doylestown, PA, USA).Primary antibodies against Cluster of Differentiation 68 (CD68), Translocator Protein (TSPO)/ Peripheral Benzodiazepine Receptor (PBR), and secondary antibodies conjugated with horseradish peroxidase (HRP) were sourced from Abcam (Cambridge, MA, USA).The primary antibody against Orosomucoid 2 (ORM2) was purchased from Proteintech Group Inc (Rosemont, IL, USA).Primary antibodies Figure 10.Heat stress-induced transcriptomic alterations together with acute phase response ensued in neural, hepatic, and systemic inflammation debilitating synaptic plasticity.(A) IL-6, BDNF, and TSPO immunoreactivities were shown by immunohistochemistry in the brain (frontal cortex) sections from Control and Heat Stress mouse groups; Images were taken at 40X magnification displayed with the scale of 50 μm.(B) IL-6 immunoreactivity in both groups was measured as arbitrary light units from six separate microscopic fields and plotted along the ordinate.All p values were derived through an unpaired t-test, with the significance level set at p < 0.05.(C) BDNF immunoreactivity in both groups was measured as arbitrary light units from six separate microscopic fields and plotted along the ordinate.All p values were derived through an unpaired t-test, with the significance level set at p < 0.05.(D) TSPO immunoreactivity in both groups was measured as arbitrary light units from six separate microscopic fields and plotted along the ordinate.All p values were derived through an unpaired t-test, with the significance level set at p < 0.05.(E) Serum IL-1β level in pg/mL was graphed in violin plots from Control and Heat Stress mouse serum samples (n = 3).All p values were derived through an unpaired t-test, with the significance level set at p < 0.05.(F) Serum IL-6 level in pg/mL was graphed in violin plots from Control and Heat Stress mouse serum samples (n = 3).All p values were derived through an unpaired t-test, with the significance level set at p < 0.05.(G) CD68 and IL-1β immunoreactivities were shown by immunohistochemistry in liver sections from Control and Heat Stress mouse groups; Images were taken at 20X magnification and displayed with a scale of 100 μm.(H) CD68 immunoreactivity in both groups was measured as arbitrary light units from six separate microscopic fields and plotted along the ordinate.All p values were derived through an unpaired t-test, with the significance level set at p < 0.05.(I) IL-1β immunoreactivity in both groups was measured as arbitrary light units from six separate microscopic fields and plotted along the ordinate.All p values were derived through an unpaired t-test, with the significance level set at p < 0.05.www.nature.com/scientificreports/

RNA isolation
Samples of frozen tissue were cut and weighed to obtain ≤ 25 mg.The tissue was placed into a 1.7 mL microfuge tube with 500uL Trizol (Life Technologies) and kept on ice.The tissue was homogenized with a disperser (IKA) with three cycles of 20 s of dispersing (speed 6-liver; speed 3-brain) followed by 10 s of rest.The tube was kept in an ice water bath during the homogenization.After the homogenization, the tube was centrifuged for 12,000×g for 1 min.The supernatant was removed to a new 1.7 mL microfuge tube without disturbing cell debris that may have been pelleted.The remainder of the protocol followed the instructions for the Zymo Research Direct-zol RNA miniprep kit with some minor edits.An equal volume of ethanol (90-100%) was added to the supernatant and briefly vortexed.The mixture was transferred to the Zymo Spin IICR column in a collection tube and centrifuged for 30 s at 12,000×g.The flow-through was discarded and 400uL of Direct-zol RNA Wash Buffer was added to the column.The column was centrifuged for 30 s at 12,000×g.A mixture of DNase digestion buffer (75uL) and DNase I (5uL) was added to the column and incubated at room temperature for 15 min.After the DNase I incubation, 400uL of Direct-zol RNA PreWash was added to the column, followed by centrifugation for 30 s at 12,000×g.The flow-through was discarded and the PreWash step was repeated.The flow through was discarded and 700uL of Direct-zol RNA Wash Buffer was added to the column and centrifuged for 2 min at 12,000×g.The spin column was removed to a fresh collection tube and centrifuged again for 1 min at 12,000×g.
To elute the RNA, 30uL of DNase/RNase-free water was added directly to the column and incubated for 5 min at room temperature.The RNA was eluted by centrifugation for 1 min at 16,000×g into a new 1.7 mL lo-bind RNase-free microtube.The RNA was stored at − 80 °C.

RNA sequencing
Total RNA was monitored for quality control using the Agilent Bioanalyzer Nano RNA chip and Nanodrop absorbance ratios for 260/280 nm and 260/230 nm.Library construction was performed according to the Illumina TruSeq mRNA stranded protocol.The input quantity for total RNA was ~ 550 ng and mRNA was enriched using oligo dT magnetic beads.The enriched mRNA was chemically fragmented for three minutes.First-strand synthesis used random primers and reverse transcriptase to make cDNA.After second-strand synthesis, the ds cDNA was cleaned using AMPure XP beads, and the cDNA was end-repaired and then the 3' ends were adenylated.Illumina unique dual indexed adapters were ligated on the ends and the adapter-ligated fragments were enriched by nine cycles of PCR.The resulting libraries were validated by qPCR and sized by Agilent Bioanalyzer DNA high-sensitivity chip.The concentrations for the libraries were normalized and multiplexed together.The multiplexed libraries were sequenced using paired-end 100 cycles chemistry on the NovaSeq 6000.We targeted 20 M reads per sample.

RNA sequence analysis
The quality of the sequencing was first assessed using the fastQC tool (v0.11.9) on the paired-end reads.Raw reads were then illumina adapter and quality trimmed and filtered by a length of 20 bases using trimmomatic (v0.39).Trimmed reads were aligned to indexed mouse reference genome mm10 using short read aligner hisat2 (v2.2.1) and gene expression levels were quantified using featureCount from Subread (v2.0.1) for all annotated genes and samples.Differential gene expression analysis was performed using R package DESeq2(v1.22.2) implementing a negative binomial model of the gene count with an FDR (false discovery rate) cut-off of 0.05.PCA plots and heatmaps were generated using DESeq2(v1.22.2).

Pathway and network analysis
The differentially expressed gene lists (DEG) were imported into the Ingenuity Pathway Analysis tool (IPA: v094302991) using default parameters with a log fold change cutoff of less than − 1.0 and greater than 1.0 for pathway enrichment analysis.Statistically significant canonical pathways were identified and filtered for pathways of interest to create a customized chart.Of the 25 networks generated by IPA, those of interest were selected and exported.Cytoscape (v3.10.0) was used to generate protein-protein interaction networks.The DEG lists were filtered for downregulated and upregulated genes using log fold change cutoff of less than − 1.0 and greater than 1.0 and p value less than 0.05 and imported into Cytoscape.The STRING protein database (v11.5)generated networks based on the gene list.The generated networks were filtered for the largest subnetwork.The final network was generated, and edges were colored by combining log fold change and adjusted p values.

Animal model
The study used pathogen-free, adult, male, 24 months old C57BL/

Analysis of gut microbiome
At the conclusion of the experimental timeline, fecal pellets from the control and heat stress mouse groups were subjected to DNA isolation and purification using the ZymoBIOMICS Miniprep kit.Following this, DNA libraries were prepared, and QC was performed using the Qubit dsDNA HS assay from ThermoFisher (Waltham, MA, USA).Whole metagenome sequencing (WMS) was performed on the Illumina NovaseqX high-throughput sequencer, generating 12 million reads per sample conducted by CosmosID (Rockville, MD, USA).The unassembled sequencing reads for each sample were then analyzed using the vendor-optimized bioinformatics platform and pipelines.The bar graph was generated using the ratio value from the observed relative abundance of the bacterial phyla Firmicutes and Bacteroidetes from respective mouse groups.

Experimental model used
To study the underlying pathological changes in exposure to heat stress, an aging mouse model was used.Aged C57BL6/J (24 months old) mice were exposed to heat stress at 40 ± 0.5 °C with a relative humidity of 60% ± 5% for three hours/day for 15 days mimicking a periodic heat wave exposure.The total number of animals in each group (n = 3) was determined based on statistical power calculations to ensure adequate statistical analysis.The mice were randomly allocated to their respective cages following a randomization procedure.At 24 months and 16 days of age, all mice were euthanized, and serum and liver tissues were collected for further analysis.Brain and liver tissues were fixed in 10% neutral buffered formalin after euthanization to prepare them for sectioning and subsequent processing.

Immunohistochemistry
The brain and liver tissue sections were deparaffinized according to a standard laboratory protocol, utilizing a solution and steamer from IHC-World (Woodstock, MD, USA) for antigen epitope retrieval.To inhibit endogenous peroxidase activity, a 3% H2O2 solution was applied for 20 min.Subsequently, serum blocking was carried out using 5% goat serum for 1 h.Overnight incubation at 4 °C with primary antibodies (ORM2, CD68, α-SMA, IL-1β, IL-6, TSPO, BDNF), appropriately diluted in a blocking buffer as per the manufacturer's recommended dilutions, were conducted in a humidified chamber.For detection, species-specific biotinylated secondary antibodies and streptavidin-conjugated with horseradish peroxidase were employed, following standard protocols from the manufacturer.3,3-diaminobenzidine (DAB) from Sigma-Aldrich (St. Louis, MO, USA) was utilized as a chromogenic substrate to visualize immunoreactivity.Counterstaining with Mayer's hematoxylin from Sigma-Aldrich (St. Louis, MO, USA) was performed.Throughout the procedure, tissue sections were washed with 1X PBS-T (PBS + 0.05% Tween 20) between steps to eliminate any unbound reagents.Finally, the sections were mounted in Aqua Mount from Lerner Laboratories (Stamford, CT, USA).Image acquisition was carried out using an Olympus BX63 microscope (Olympus, Center Valley, PA, USA), and morphometry analysis was performed using Cellsens Software V2.2 from Olympus (Center Valley, PA, USA).

Immunofluorescence
The formalin-fixed, paraffin-embedded brain tissue sections were deparaffinized following standard instructions.Epitope retrieval of the deparaffinized tissue sections was performed using an epitope retrieval solution and a steamer from IHC World, following the manufacturer's protocol.Primary antibodies for CD31 and Claudin 5 were used at the manufacturer's recommended dilutions and incubated overnight at 4 °C.Species-specific anti-IgG secondary antibodies conjugated with Alexa Fluor 633 or 488 from Invitrogen (Waltham, MA, USA), USA) were used for detection.The tissue sections were mounted using ProLong Gold antifade mounting media with DAPI from Thermofisher Scientific (Grand Island, NY, USA) which helps preserve fluorescence and provides nuclear staining with DAPI.Images of the tissue sections were captured under 60X magnification using an Olympus BX63 microscope (Olympus, Center Valley, PA, USA).Morphometry analysis was performed using Cellsens Software V2.2 (Olympus, Center Valley, PA, USA).

ELISA
Serum levels of IL-1β and IL-6 were estimated in the serum samples collected from the mice groups by using the commercially available mouse IL-1β (Cat# KE10003) and IL-6 (Cat# KE10007) ELISA kits, purchased from ProteinTech Group Inc (Rosemont, IL, USA).

Statistical analyses
Results from transcriptomics were analyzed through a test of significance using the False Discovery Rate (FDR) method providing P FDR Values (P FDR < 0.05).The other results were presented as mean ± standard error of the mean (S.E.M.).Statistical analyses were performed using an unpaired t-test, and significance was determined by the Bonferroni Dunn post hoc correction method using GraphPad Software V10.1.0(San Diego, CA, USA).A p value of less than 0.05 (p < 0.05) was considered statistically significant.

Figure 1 .
Figure 1.Periodic heat stress showed distinctive gene expression patterns in brain and liver tissues of aging mice.(A) Schematic representation of the basic experimental design.(B) Principal component analysis plot of brain tissues comparing the differential gene expressions between the control (red) and heat stress (blue) groups, Principal component 1 (PC1) and 2 (PC2) plotted along abscissa and ordinate respectively (PC1 at 25% and PC2 at 50% variance).(C) Principal component analysis plot of brain tissues comparing the differential gene expressions between the control (red) and heat stress (blue) groups, Principal component 1 (PC1) and 2 (PC2) plotted along abscissa and ordinate respectively (significance was set conventionally with PC1 at 25% and PC2 at 50% variance).

Figure 2 .
Figure 2. Heat stress-induced transcriptional alterations in the brain.(A) Volcano plot obtained from comparing the brain tissues of control and heat stress groups that were analyzed for differential gene expression changes through transcriptomic profiling, -Log 10 P (P FDR ) and Log 2 fold changes were plotted respectively along the ordinate and abscissa with a significance cut off set at P FDR < 0.05, Log 2 fold change > 1.The blue dots represented downregulated, and the red dots represented the upregulated genes.Genes represented in gray dots were laid in the zone of no significance.The total number of upregulated and downregulated genes was indicated with similar color arrowheads.(B) The heatmap generated from the brain tissue volcano plot was delineated with a Log 2 fold change gradient between 1.5 and − 1.5 where the extremities of red and blue signified up-and downregulations respectively.The same blue and red color code was set to represent the intensity of gene expression falling within the given gradient reflecting the Log 2 fold change of individual genes named parallel.

Figure 3 .
Figure3.Heat stress-induced transcriptional alterations in the liver.(A) Volcano plot obtained from comparing the liver tissues of control and heat stress groups that were analyzed for differential gene expression changes through transcriptomic profiling, -Log 10 P (P FDR ) and Log 2 fold changes were plotted respectively along the ordinate and abscissa with a significance cut off set at P FDR < 0.05, Log 2 fold change > 1.The blue dots represented downregulated, and the red dots represented the upregulated genes.Genes represented in gray dots were laid in the zone of no significance.The total number of upregulated and downregulated genes was indicated with similar color arrowheads.(B) The heatmap generated from the liver tissue volcano plot was delineated with a Log 2 fold change gradient between 1.5 and − 1.5 where the extremities of red and blue signified up-and downregulations respectively.The same blue and red color code was set to represent the intensity of gene expression falling within the given gradient reflecting the Log 2 fold change of individual genes named parallel.

Figure 4 .
Figure 4. Heat stress-mediated gene expression changes were extrapolated in contributing pathways integral to organ-level pathological manifestations.(A) Biological pathways were predicted based on the differentially expressed gene list obtained by comparing the brain tissues from control and heat stress groups.Significance level set at P FDR < 0.05, threshold at -log(p value) ~ 1.3, Log 2 fold change > 1.The gradients of blue and red colors represented the up-and downregulations of the pathways' respectively.(B) Biological pathways were predicted based on the differentially expressed gene list obtained by comparing the liver tissues from control and heat stress groups.Significance level set at P FDR < 0.05, threshold at -log(p value) ~ 1.3, Log 2 fold change > 1.The gradients of blue and red colors represented the up-and downregulations of the pathways' respectively.(C) Biological pathways were predicted based on the commonly differentially expressed gene list obtained by comparing both the brain and liver tissues from control and heat stress groups.Significance level set at P FDR < 0.05, threshold at -log(p value) ~ 1.3, Log 2 fold change > 1.The gradients of blue and red colors represented the up-and downregulations of the pathways' respectively.

Figure 5 .
Figure 5. Predicted pathway circuits unveiled a neurological disease network within the brain and an inflammatory control network within the liver tissues following exposure to heat stress.(A) Pathway circuit predicted based on the projected pathways obtained by comparing the brain tissues from control and heat stress groups.Each shape represented different biological kinds of molecules and pathway components and each line connector style represented activation/inhibition status.Detailed legends of prediction, network, and path designer shapes were included in the figure.(B) Pathway circuit predicted based on the projected pathways obtained by comparing the liver tissues from control and heat stress groups.Each shape represented different biological kinds of molecules and pathway components and each line connector style represented activation/ inhibition status.Detailed legends of prediction, network, and path designer shapes were included in the figure.

Figure 6 .
Figure 6.Altered functional interactome emerged in both brain and liver tissues upon exposure to heat stress.(A) Interactome was predicted based on the differentially expressed genes obtained from transcriptomic profiling and then by comparing the brain tissues from control and heat stress groups.Interaction intensities were signified by possible upregulations and downregulations of the respective gene clusters with a Log 2 fold change gradient between − 3.9 and − 4.8 where the extremities of red and blue indicated up-and downregulations respectively.The same blue and red color code was set to mark the genes in clusters falling within the given gradient, gene names were provided in each representing shape.The connector line indicated the possible biological strength and interconnections of the interacting gene products.(B) Interactome was predicted based on the differentially expressed genes obtained from transcriptomic profiling and then by comparing the liver tissues from control and heat stress groups.Interaction intensities were signified by possible upregulations and downregulations of the respective gene clusters with a Log 2 fold change gradient between − 3.9 and − 4.8 where the extremities of red and blue indicated up-and downregulations respectively.The same blue and red color code was set to mark the genes in clusters falling within the given gradient, gene names were provided in each representing shape.The connector line indicated the possible biological strength and interconnections of the interacting gene products.

Figure 7 .
Figure 7.The liver-brain connection in heat stress pathologies was projected through a shared predictive pathway circuit and interactome.(A) Pathway circuit predicted based on the projected pathways obtained by comparing both brain and liver tissues from control and heat stress groups.Each shape represented different biological kinds of molecules and pathway components and each line connector style represented activation/ inhibition status.Detailed legends of prediction, network, and path designer shapes were included in the figure.(B) Interactome was predicted based on the differentially expressed genes obtained from transcriptomic profiling and then by comparing both brain and liver tissues from control and heat stress groups.Interaction intensities were signified by possible upregulations and downregulations of the respective gene clusters with a Log 2 fold change gradient between − 3.9 and − 4.8 where the extremities of red and blue indicated up-and downregulations respectively.The same blue and red color code was set to mark the genes in clusters falling within the given gradient, gene names were provided in each representing shape.The connector line indicated the possible biological strength and interconnections of the interacting gene products.

Figure 8 .
Figure 8. Expression fold changes of the top genes observed in the brain and liver unveiled potential validatory mediators of the liver-brain axis upon exposure to heat stress.(A) Heat Stress-induced relative mRNA expression changes (18S rRNA normalized) of the top genes in brain samples were converted from log2 fold changes based on the transcriptomic data plotted in bar graphs representing each gene (Control representing the baseline, 0).Negative fold changes were indicated in the blue color gradient whereas positive fold changes were indicated in the red gradient.All p values were derived through an unpaired t-test, with the significance level set at p < 0.05.(B) Heat Stress-induced relative mRNA expression changes (18S rRNA normalized) of the top genes in liver samples were converted from log2 fold changes based on the transcriptomic data plotted in bar graphs representing each gene (Control representing the baseline, 0).Negative fold changes were indicated in a blue color gradient whereas positive fold changes were indicated in a red color gradient.All p values were derived through an unpaired t-test, with the significance level set at p < 0.05.

The liver-brain connection in heat stress pathologies was projected through a shared predic- tive pathway circuit and interactome
outlined in the NIH Guide for the Humane Care and Use of Laboratory Animals and ARRIVE GUIDELINES.The study protocol was approved by the institutional review board at the University of South Carolina (Animal Protocol Number: 2488-101501-051220; approval date: 5th October 2021).The animals were handled in compliance with local IACUC (Institutional Animal Care and Use Committee) standards.After the completion of the experimental treatments, all mice were euthanized.
6 J wild-type (WT) mice obtained from the Jackson Laboratories (Bar Harbor, ME, USA).The mean body weight of the animals used was 36.93 ± 1.69 g at day 0 of the experiment.The mice were housed in a temperature-controlled room with a 12-h light/dark cycle and had ad libitum access to food and water.All experimental procedures were conducted following the guidelines Vol.:(0123456789) Scientific Reports | (2024) 14:10555 | https://doi.org/10.1038/s41598-024-60664-9www.nature.com/scientificreports/