Abstract

Introduction

   Extreme heat events caused by occupational exposure and heat waves are
   becoming more common. However, the molecular changes underlying the
   response to heat exposure in humans remain to be elucidated.

Methods

   This study used longitudinal multi-omics profiling to assess the impact
   of acute heat exposure (50°C for 30 min) in 24 subjects from a mine
   rescue team. Intravenous blood samples were collected before acute heat
   exposure (baseline) and at 5 min, 30 min, 1 h, and 24 h after acute
   heat exposure (recovery). In-depth multi-omics profiling was performed
   on each sample, including plasma proteomics (untargeted) and
   metabolomics (untargeted).

Results

   After data curation and annotation, the final dataset contained 2,473
   analytes, including 478 proteins and 1995 metabolites. Time-series
   analysis unveiled an orchestrated molecular choreography of changes
   involving the immune response, coagulation, acid–base balance,
   oxidative stress, cytoskeleton, and energy metabolism. Further analysis
   through protein–protein interactions and network analysis revealed
   potential regulators of acute heat exposure. Moreover, novel
   blood-based analytes that predicted change in cardiopulmonary function
   after acute heat exposure were identified.

Conclusion

   This study provided a comprehensive investigation of the dynamic
   molecular changes that underlie the complex physiological processes
   that occur in human males who undergo heat exposure. Our findings will
   help health impact assessment of extreme high temperature and inspire
   future mechanistic and clinical studies.

   Keywords: acute heat exposure, proteomics, metabolomics, molecular
   choreography, health

1. Introduction

   Prolonged exposure to extreme high temperatures in various industries,
   including steelmaking, tile production, mining, and rescue operations,
   can result in harmful health impact ([35]1). Heat exposure can lead to
   heat-related illnesses encompassing a spectrum of syndromes like heat
   edema, muscle cramps, heat exhaustion and heatstroke ([36]2, [37]3).
   Notably, numerous studies have highlighted moisture and heat as the
   most perceived adverse factors in deep underground work environments
   ([38]4–6). For instance, Hunt et al. found that 79% of underground mine
   workers across three mine sites in northern Australia experienced
   symptoms of heat illness ([39]7). Additionally, an analysis of hazards
   faced by mine rescuers revealed that the challenging mine environment
   led to body overheating and heatstroke, which accounted for 26% of
   total fatalities ([40]8). Furthermore, several studies have
   demonstrated that the extreme heat environment places significant
   physiological and thermoregulatory stresses on firefighters, leading to
   stress, overexertion, and even death ([41]9–11). In addition to
   occupational exposure, the general population is also vulnerable to
   extreme high temperature exposure due to the effects of climate change,
   including global warming ([42]12). In 2019, heat waves in China
   resulted in approximately 26,800 deaths ([43]13). Consequently, these
   studies have sparked significant concerns regarding the potential
   health impacts of extreme high temperatures on populations that are
   regularly exposed to heat.

   The previous studies primarily concentrated on detailing the phenotypic
   and physiological reactions linked to heat exposure. Within the deep
   mine, miners sustained a core body temperature of 38°C throughout most
   of the day, leading to a notable rise in heart rate and severe
   dehydration ([44]14, [45]15). Similarly, firefighters working in hot
   environments experienced elevated core body temperature, skin
   temperature, fluid intake, and heart rate ([46]16–18). Another study
   showed that prolonged heat overload in firefighters working in hot
   environments increased arterial stiffness and vasodilation ([47]19).
   These findings highlight the importance of assessing tissue and organ
   damage and anticipating potential long-term complications after heat
   exposure. It is crucial to note that delayed monitoring of core
   temperature or other physiological symptoms may lead to unexpected
   instances of severe heat-related conditions, such as heatstroke.

   Molecular responses induced by extreme high temperatures can provide
   valuable indications of the physiological and metabolic changes that
   occur under such conditions. By monitoring these molecular responses,
   we can gain a better understanding of the body’s adaptive capacity to
   extreme high temperature and assess the potential degree of damage
   before clear clinical signs of thermal injury emerge. For example,
   analysis of blood mononuclear gene expression patterns in individuals
   exposed to extreme heat in a sauna (temperature of 75.7 ± 0.86°C)
   showed rapid changes in gene expression without a significant increase
   in core temperature. The reprogrammed transcriptome was mainly
   inhibitory, as genes involved in protein synthesis, mitochondrial
   bioenergetics, and immune function showing reduced expression ([48]20).
   In another study by Bouchama et al. ([49]21) the whole genome
   transcriptome in peripheral blood mononuclear cells of an adult cohort
   with heatstroke was examined. It was found that in heatstroke, the heat
   shock response was robust but failed to restore homeostasis due to
   proteotoxicity and a reduction in energy production ([50]21). A
   meta-analysis of public gene expression database from human and mouse
   samples identified previously overlooked genes responding to heat, such
   as ABHD3, ZFAND2A, and USPL1 ([51]22). However, knowledge of the
   response to heat exposure is limited to a small range of molecules and
   biological processes. This highlights the critical need of a systematic
   exploration of molecular dynamics in individuals exposed to extreme
   high temperature environments. In this study, we hypothesize that after
   heat exposure, the molecules undergo dynamic changes with distinct
   trends. While some molecules show reversible alterations, others do
   not. These diverse trends partly reflect the body’s adaptive capacity
   and the extent of damage sustained.

   The objective of this study was to perform longitudinal multi-omics
   profiling of blood components from 24 mine rescuers, before and after a
   30 min stay in a simulated extreme high temperature environment (50°C).
   The goal was to understand the short-term tolerance of humans to
   extreme high temperature and characterize the detailed series of events
   that occur in response to acute heat exposure. Our findings revealed an
   orchestrated molecular choreography of changes involving numerous
   molecules and pathways, and demonstrate the potential of multi-omics
   analytes application in monitoring recovery and predict complications
   following heat exposure. This study provides a comprehensive overview
   of the physiological processes that occur in humans who undergo extreme
   high temperature exposure and will inspire novel prevention strategies
   to reduce organ damage and long-term sequelae.

2. Methods

2.1. Study subjects

   In May 2022, 24 participants were enlisted from the National Mine
   Emergency Rescue Furong Team (China) for this study. The study protocol
   received approval from the Ethics Review Committee (IRB number, 1351)
   of West China Hospital of Sichuan University. Prior to engaging in any
   study procedures, all participants provided written informed consent. A
   health questionnaire was used to screen for contraindications and/or
   comorbidities that might have prevented their exposure to extreme high
   temperatures. None of the participants had a history of underlying
   cardiovascular disease, cancer, or other chronic illnesses, and none
   had undergone heat or other stress training within 3 months prior to
   the study commencement.

2.2. Study design

   In a systematic review assessing the impact of occupational heat
   strain, the studies included in the analysis reported a wide range of
   wet-bulb globe temperature (WBGT, 19.3 to 52.0°C) and air temperature
   (21.2–150.0°C, this extreme value was observed in a steel plant
   worksite) ([52]23). Additionally, the rising global temperatures are
   pushing the human to endure extreme heat, with temperatures even
   reaching up to 54°C. As a balance between the intention of extreme high
   temperature exposure investigation and security considerations, we
   specifically chose 50°C as the environmental exposure temperature.
   Regarding the exposure time, a duration of 30 min was used based on the
   previous study that employed passive whole-body heat exposure
   ([53]24–26). The research comprised three phases: baseline, exposure,
   and recovery. The baseline phase spanned from 09:00 am on the day prior
   to exposure until the day of exposure. The exposure phase occurred
   between 09:00 and 09:30 am. The recovery phase extended from 09:30 am
   until 09:30 am the following morning.

2.2.1. Baseline period

   Subjects’ physical condition was assessed on the day preceding
   exposure. Oral temperature was measured using a liquid crystal
   thermometer (Watermark^TM). Grip strength of both hands was assessed
   using a grip dynamometer (EH 101, CAMRY). Objective fatigue was
   measured using performance tasks (BD-V-302A, BD-V-509A, Beda Qingniao).
   Pulmonary function was measured using spirometry. Heart function was
   examined with Doppler echocardiography. Urine specific gravity and pH
   were evaluated with urinalysis. Blood samples were collected at
   07:30 am on the day of exposure.

2.2.2. Exposure period

   At 09:00 am on the day of exposure, participants were subjected to a
   high-temperature environment of 50°C and 37–40% relative humidity for
   30 min. This was achieved using a mine roadway simulating test device
   designed to replicate fire conditions and aid in the training of the
   National Mine Emergency Rescue Team. Industrial heaters (JH-H150F,
   Cameron, China) were positioned at both ends of the roadway to generate
   the heat. Throughout the acute heat exposure, changes in heart rate
   were monitored using an electrocardiogram patch.

2.2.3. Recovery period

   Immediately after the heat exposure, oral temperature, grip strength,
   objective fatigue, subjective fatigue score, lung function, color
   Doppler echocardiography, urine samples were detected.

2.3. Objective fatigue

   To measure the effects of heat exposure on fatigue, subjects were asked
   to perform two tasks. First, subjects participated in a bimanual
   interaction technique, where they controlled a cursor on a computer
   screen with a joystick using both hands, and the number of
   cursor-trajectory errors were recorded. Next, subjects participated in
   a visual reaction time test where they had to press color buttons as
   soon as corresponding-colored icons appeared on a screen using both
   hands. Reaction times were recorded.

2.4. Cardiopulmonary function tests

   To assess the pulmonary function of subjects, we conducted a specific
   assay which measured Forced Expiratory Volume in One Second (FEV 1),
   Forced Vital Capacity (FVC), Vital Capacity (VC), Maximum Voluntary
   Ventilation (MVV), Maximum Mid-Respiratory Flow (MMF), and Maximum
   Expiratory Flow at 50% of Vital Capacity (V50). To assess the cardiac
   function of subjects, Echocardiograms were performed by experienced
   cardiologists using GE Vivid E9 and E9-S5 probes (1.6–3.2 MHz).
   According to international consensus guidelines, post-exposure images
   were obtained immediately after heat exposure, and imaging was
   satisfactory for all subjects. Echocardiography was analyzed according
   to the American Society of Echocardiography (ASE) guidelines (2015).
   Left ventricular function was quantified with the Simpson method using
   measurements of left ventricular ejection fraction (LVEF), stroke
   output (SV) and fractional shortening (FS). For statistical analysis of
   these parameters, Wilcoxon Signed Rank Test was performed as these
   quantitative data did not adhere to a normal distribution.

2.5. Blood collection and sample preparation

   Following a 12-h overnight fast, blood samples were obtained from a
   vein in the upper forearm at baseline, 5 min, 30 min, and 1 h post-heat
   exposure, as well as the morning after heat exposure. The blood samples
   were collected in purple top vacutainers (BD), chilled on ice, and
   promptly processed. Subsequently, the blood samples were centrifuged at
   3000 rpm for 10 min at 4°C. The top layer of EDTA plasma was removed,
   divided into 8 equal portions, and immediately frozen at −80°C. All
   blood samples were subjected to multi-level molecular profiling. An
   untargeted proteomics (proteins) and metabolomics (metabolites)
   approach was used to analyze the EDTA plasma samples.

2.6. Untargeted proteomics and data processing

   Untargeted proteomics used Data independent acquisition (DIA). After
   the protein extraction from the plasma samples, protein concentration
   was determined using a BCA protein assay kit. Subsequently, 200
   micrograms of each sample was subjected to 120 μL reducing buffer
   (10 mM DTT, 8 M Urea, 100 mM TEAB, pH 8.0). The solution was incubated
   at 60°C for 1 h, and IAA was added to the solution for 40 min at room
   temperature. Then the solutions were centrifuged on the filters at
   12,000 rpm for 20 min at 4°C, followed by extensive washing. The
   purified samples were then subjected to overnight trypsin digestion at
   37°C, resulting in the collection of peptides as filtrate. To
   facilitate further analysis, the peptide mixture was labeled using the
   TMT Tag Labeling kit. Then, all analyses were performed by a Q-Exactive
   HF mass spectrometer (Thermo, United States) equipped with a Nanospray
   Flex source (Thermo, United States). For the subsequent database
   search, Proteome Discoverer (v.2.4) was utilized to thoroughly search
   all raw data against the sample protein database. The database search
   was performed with Trypsin digestion specificity, considering
   alkylation on cysteine as fixed modifications. To ensure reliable
   results, a global false discovery rate (FDR) of 0.01 was set, and
   protein groups were considered for quantification only if they
   contained at least 2 peptides.

2.7. Untargeted metabolomics from plasma by liquid chromatography (LC)-MS

   Plasma samples were prepared and analyzed in a random sequence.
   Progenesis QI software (v2.3, Non-linear Dynamics) was used to
   independently analyze data from each mode. Main parameters of 5 ppm
   precursor tolerance, 10 ppm product tolerance, and 5% product ion
   threshold were applied for qualitative analysis. Compound
   identification was conducted using databases such as The Human
   Metabolome Database (HMDB), Lipidmaps (V2.3), Metlin, EMDB, and PMDB,
   as well as self-built databases. A data matrix was created from
   positive and negative ion data and imported into R for Principle
   Component Analysis (PCA) to observe overall sample distribution and
   analysis stability. Orthogonal Partial Least-Squares-Discriminant
   Analysis (OPLS-DA) and Partial Least-Squares-Discriminant Analysis
   (PLS-DA) were used to distinguish metabolites that differed between
   groups, with 7-fold cross-validation and 200 Response Permutation
   Testing (RPT) used to prevent overfitting. Variable Importance of
   Projection (VIP) values from the OPLS-DA model ranked the contribution
   of each variable to group discrimination, and a two-tailed Student’s
   T-test verified significant differences in metabolites between groups
   with VIP values greater than 1.0 and p-values less than 0.05.

2.8. Quantification and statistical analysis

2.8.1. Differential expression analysis

   Differential expression analysis was performed using the DESeq2
   package, with a q-value <0.05 and a fold change >2 or fold change <0.5
   set as the threshold for significant differential expression analytes.
   R (v 3.2.0) was employed to conduct hierarchical cluster analysis of
   the differential expression analytes, illustrating the expression
   pattern of analytes across different groups and samples.

2.8.2. Pathway enrichment analysis

   To identify enriched pathways, Ingenuity pathway analysis (IPA, QIAGEN)
   was used to assess differentially expressed plasma analytes. The
   significance of each pathway was determined using hypergeometric
   probability (one-sided), while Fisher exact test was used to determine
   enrichment for proteins and metabolites. The p-values were corrected
   for multiple comparisons using the Benjamini-Hochberg method, with FDRs
   of <0.05 considered significant for proteins or metabolites,
   respectively. Fold change was estimated for significant molecules using
   the median of fold change relative to baseline, median of beta
   coefficients, and median of max (if up) or min (if down) fold change
   relative to baseline.

2.8.3. Pathway dynamic analysis

   Differential expressed analytes (FDR <0.05) were identified, and STRING
   analysis was used to explore the protein–protein interaction network
   before and after acute heat exposure. In PPI, proteins are represented
   as nodes, and edges between nodes represent interactions.

2.8.4. WGCNA analysis

   WGCNA was applied to analyze the co-expression modules and key analytes
   related to hub molecules. An adjacency matrix was transformed into a
   topological overlap matrix (TOM). Modules were identified with
   hierarchical clustering (minModuleSize =30). Module eigengenes were
   calculated. Module eigengenes (ME) and module memberships (MM) were
   used to determine key modules associated with hub molecules. The ME is
   defined as the first principal component of a given module and provides
   a representative expression profile. MM measured the relationship of a
   gene with the module eigengenes and reflects eigengene connectivity.
   The functions of metabolites in the key modules were investigated via
   Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment
   analysis.

2.8.5. Fuzzy c-mean clustering

   The data underwent log2-transformation and Z-score scaling before
   conducting fuzzy c-mean clustering with the ‘Mfuzz’ R package
   (v2.20.0). The ‘elbow’ method was utilized to determine the minimum
   centroid distance or a range of cluster numbers and to select the
   optimal number. Additionally, t-distributed stochastic neighbor
   embedding (tSNE) scatterplots were generated using the ‘Rtsne’ R
   package, with the parameters set as perplexity = 5 and theta = 0.05.

2.9. Enzyme-linked immunosorbent assay

   Plasma samples collected at baseline and 5 min, 30 min, 1 h, and 24 h
   after heat exposure were immediately centrifuged at 3,000 g for 10 min.
   To validate the key regulators identified by multi-omics analysis,
   serum levels of VWF, PF4, THBS1, HSP90AB1, and MPO were measured using
   enzyme-linked immunosorbent assay (ELISA) kits (Mlbio, Shanghai,
   China), following the manufacturer’s instructions. One-Way Repeated
   Measures ANOVA was employed to investigate the variability of
   parameters across different time points.

3. Results

3.1. Cohort characteristics and research design

   This study included 24 subjects. All were male, with a mean age of
   32.9 years, mean height of 170.9 cm, and mean weight of 67.7 kg
   ([54]Supplementary Table S1). The schematic diagram of experimental
   process is shown in [55]Figure 1A. In-depth multi-omics profiling was
   performed, including plasma proteomics (untargeted) and metabolomics
   (untargeted). After data curation and annotation, the final dataset
   contained 2,473 analytes, including 478 proteins and 1995 metabolites.

Figure 1.

   [56]Figure 1
   [57]Open in a new tab

   Study design and physiological response to acute heat exposure. (A)
   Overview of the study design. Subjects were exposed to an extreme high
   temperature environment of 50°C and 37–40% relative humidity for
   30 min. Subjects underwent a physical examination before and after
   acute heat exposure. Intravenous blood samples were collected before
   acute heat exposure (baseline) and at 5 min, 30 min, 1 h, and 24 h
   after acute heat exposure (recovery) for the multi-omics analysis. (B)
   Heart rate changes were detected during 30 min high temperature
   environment simulation training. (C) Cardiopulmonary function changes
   in response to acute heat exposure. Core Temperature, the forced vital
   capacity (FVC), vital capacity (VC), left ventricular ejection fraction
   (EF), and fractional shortening (FS) were detected at pre- and post-
   acute heat exposure. * represents p < 0.05 and *** represents p <
   0.001.

   During acute heat exposure, there was an observed increase in heart
   rate ([58]Figure 1B). Immediately after acute heat exposure, oral
   temperature rose to 37.2°C compared to 36.6°C at baseline.
   Cardiopulmonary function parameters with significant changes were also
   shown. The forced vital capacity (FVC), vital capacity (VC), left
   ventricular ejection fraction (EF), and fractional shortening (FS)
   decreased after acute heat exposure ([59]Figure 1C and
   [60]Supplementary Table S2). For the urinalysis, the urine specific
   gravity of the participants had no change after acute heat exposure,
   while the urine PH value decreased ([61]Supplementary Figure S1). As
   [62]Supplementary Figure S2 shown, the fatigue increased after acute
   heat exposure.

3.2. Proteins and metabolites expression signature

   The boxplot reflected the expression patterns of proteins in each
   sample, indicating normal proteins expression patterns ([63]Figure 2A).
   The omics datasets were assessed with principal component analysis,
   which suggested limited batch effects ([64]Figures 2B,[65]C).
   Additionally, the heatmap highlighted the expression of differentially
   expressed metabolites ([66]Figure 2D). Exposure to an extreme high
   temperature environment of 50°C induced extensive changes in 2473
   analytes spanning multiple omics layers, indicating system-wide changes
   ([67]Figure 2E). Levels of circulating plasma proteins and metabolites
   were altered compared to baseline across all time points, and a large
   proportion of plasma proteins and metabolites remained significantly
   different from baseline after 24 h of recovery.

Figure 2.

   [68]Figure 2
   [69]Open in a new tab

   Molecular response to acute heat exposure. (A) Visualization display of
   credible protein expression distribution by boxplot. (B,C) Principal
   components analysis using proteomics (B) and metabolomics (C) data. (D)
   The heatmap showed the expression of differentially expressed
   metabolites. (E) Multi-omics analysis to explore changes in response to
   acute heat exposure across molecular layers at all the time points.

3.3. Functional analysis of differentially expressed proteins and metabolites
at each time point

   As depicted in [70]Figure 3, the biological processes that were altered
   after heat exposure primarily encompassed immune response, complement
   activation, extracellular region, ECM-receptor interaction, and
   glycolysis. Similarly, as illustrated in [71]Figure 4, the metabolic
   pathways that were affected by heat exposure predominantly involved
   biosynthesis of unsaturated fatty acids, protein digestion and
   absorption, amino acid metabolism, cholesterol metabolism, and
   sphingolipid metabolism. These findings indicated substantial molecular
   reactions following acute heat exposure, emphasizing the need for
   further investigation into the temporal dynamics of these molecular
   responses.

Figure 3.

   [72]Figure 3
   [73]Open in a new tab

   The enrichment analysis of protein changes in response to acute heat
   exposure. (A) Go enrichment analysis. The dot color represented pathway
   significance. The dot size represented number of proteins enriched in
   pathway. (B) KEGG enrichment analysis. Length represented the percent
   of proteins enriched in pathway.

Figure 4.

   [74]Figure 4
   [75]Open in a new tab

   The enrichment analysis of metabolism changes in response to acute heat
   exposure. The dot color represented pathway significance in KEGG
   enrichment analysis. The dot size represented number of metabolites
   enriched in pathway.

3.4. Time series system-wide proteomics data

   Cluster analysis identified four clusters of circulating plasma
   proteins with different longitudinal trajectories. The levels of some
   plasma proteins decreased after acute heat exposure and did not return
   to baseline within 24 h (cluster 1). The levels of some plasma proteins
   increased after acute heat exposure and continued to increase during
   recovery (cluster 2). There was a delayed increase in the levels of
   some plasma proteins after acute heat exposure (cluster 3). The levels
   of some plasma proteins increased after acute heat exposure, and
   quickly returned to baseline (cluster 4) ([76]Figure 5A). Pathway
   enrichment analysis was performed on each cluster ([77]Supplementary
   Figures S3, S4). As depicted in [78]Figure 5B, cluster 1 encompassed
   proteins associated with the immune-related pathways (immune response,
   antigen binding etc.), cluster 2 included proteins associated with the
   coagulation-related pathways (blood coagulation, fibrin clot formation,
   plasminogen activation), cluster 3 consisted of proteins associated
   with the cytoskeleton and extracellular matrix (structural constituent
   of cytoskeleton, extracellular region, extracellular space), and
   cluster 4 comprised proteins associated with oxygen transport (organic
   acid binding, hemoglobin complex etc.).

Figure 5.

   [79]Figure 5
   [80]Open in a new tab

   Time series system-wide proteomics data. (A) Clustering of longitudinal
   trajectories using significant plasma proteins (FDR <0.05). (B) Pathway
   enrichment analysis. Pathway direction is the median log2 fold change
   relative to baseline of significant molecules in each pathway (blue,
   downregulated; red, upregulated). Dot size represents pathway
   significance. (C–F) Heatmaps representing longitudinal trajectories of
   plasma proteins in cluster 1 (C), cluster 2 (D), cluster 3 (E), and
   cluster 4 (F). Protein names are listed on the right (blue,
   downregulated; red, upregulated).

3.4.1. Cluster 1

   Cluster 1 was enriched in molecules associated with the innate and
   humoral immune response ([81]Figure 5C). There was a sharp decrease in
   the plasma levels of immunoglobulin related fragments (i.e., IGHV3-73,
   IGLV6-57, IGHV3-38, IGHV3-15, IGHV3-9, IGHV3-43D, IGHV3-43, IGLL1,
   IGLV3-25, IGHV1-58, IGKV1-27, IGLV7-46, IGKV1D-33, IGKV1-33, IGHV1-45,
   IGLV2-11, IGKV1-8, FCGR3B, and IGLV3-1) and pattern recognition
   receptor (i.e., C-type mannose receptor 2). These data indicated that
   immune function was suppressed immediately by acute heat exposure and
   did not recovery in short term.

3.4.2. Cluster 2

   Cluster 2 was enriched in molecules associated with the coagulation
   cascade (platelet glycoprotein V, platelet factor 4, coagulation factor
   VIII, thrombospondin 1, coagulation factor XIII A chain, coagulation
   factor IX, coagulation factor X, coagulation factor XII, coagulation
   factor XIII B chain, coagulation factor V, von Willebrand factor),
   stress related proteins (serine/threonine-protein
   kinase/endoribonuclease, heat stress protein 90-beta), and
   glyceraldehyde-3-phosphate dehydrogenase, which is a key glycolysis
   rate limiting enzyme ([82]Figure 5D). These data implied elevated blood
   viscosity, risk of disseminated intravascular coagulation (DIC),
   anaerobic metabolism and the stress response induced by acute heat
   exposure.

3.4.3. Cluster 3

   Cluster 3 was enriched in proteins of the cytoskeleton and components
   of intercellular junctions (keratin, type II cytoskeletal 2 epidermal,
   vimentin, coronin-1A, alpha-actinin-1, keratin, type II, cytoskeletal
   1, keratin, type I cytoskeletal 10, keratin, type II cytoskeletal 6B,
   desmoglein-1, keratin, type II cytoskeletal 6C, matrix Gla protein)
   ([83]Figure 5E). This could be a result from the degradation of
   cytoskeleton, a visible component diffusely distributed throughout the
   cell, caused by acute heat exposure. The damaged cytoskeleton elements
   were latterly released to the plasma and therefore detected in a
   delayed manner.

3.4.4. Cluster 4

   Cluster 4 was enriched in hemoglobin (hemoglobin subunit alpha,
   hemoglobin subunit beta, hemoglobin subunit delta), carbonic anhydrase
   (carbonic anhydrase 2, carbonic anhydrase 1), plasminogen, S100A8, the
   anti-fibrinolytic system (plasminogen activator inhibitor 1) and
   antioxidants (peroxiredoxin-2, peroxiredoxin-6, kallistatin, superoxide
   dismutase 1) ([84]Figure 5F). These data implied that hemoglobin and
   the maintenance of acid–base balance were mobilized under acute heat
   exposure and the fibrinolytic system and antifibrinolytic system were
   activated, possibly due to increased coagulation.

3.5. Time series system-wide metabolomics data

   Cluster analysis identified six clusters of circulating plasma
   metabolites with different longitudinal trajectories. The levels of
   some plasma metabolites increased after acute heat exposure and did not
   return to baseline within 24 h (cluster 1). The levels of some plasma
   metabolites underwent a transient increase after acute heat exposure
   but followed by a decrease during recovery (cluster 2). The levels of
   some plasma metabolites decreased after acute heat exposure and
   returned to baseline within 24 h (cluster 3, cluster 4). The levels of
   some plasma metabolites increased after acute heat exposure and
   returned to baseline within 24 h (cluster 5, cluster 6) ([85]Figure
   6A).

Figure 6.

   [86]Figure 6
   [87]Open in a new tab

   Time series system-wide metabolomics data. (A) Clustering of
   longitudinal trajectories using significant plasma metabolites (FDR
   <0.05). (B,C,F) Heatmaps representing longitudinal trajectories of
   plasma metabolites in cluster 1 (B), cluster 2 (C) and cluster 5 (F).
   Metabolite names are listed on the right (blue, downregulated; red,
   upregulated). (D,E,G) Time chart representing longitudinal trajectories
   of plasma metabolites in cluster 3 (D), cluster 4 (E), and cluster 6
   (G). Metabolite names are listed on the right.

   Pathway enrichment analysis was performed on each cluster
   ([88]Supplementary Figure S5). Cluster 1 was associated with alanine,
   aspartic acid, and glutamic acid metabolism. Cluster 2 was associated
   with steroid hormone biosynthesis. Cluster 3 was associated with amino
   acid metabolism. Cluster 4 was associated with cholesterol metabolism.
   Cluster 5 was associated with the biosynthesis of unsaturated fatty
   acids. Cluster 6 was associated with sphingolipid metabolism.

3.5.1. Cluster 1

   Cluster 1 was enriched in alanine, aspartic acid, and glutamic acid
   metabolites (i.e., threo-hydroxyaspartic acid, aspartate,
   N-acetylaspartate, fumaric acid, and glutaric acid) and carnitine and
   acyl carnitine metabolites [i.e., 3-hydroxybutyric acid,
   (10Z)-12-hydroxyhexadecenoylcarnitine,
   (4E,6E)-undeca-4,6,9-trienoylcarnitine, 3-hydroxybutyrylcarnitine,
   5-methylnonanoylcarnitine, 7-methyldodecanoylcarnitine,
   decanoylcarnitine, nonanoylcarnitine, octanoylcarnitine] ([89]Figure
   6B). Accumulation of carnitine and acyl carnitine metabolites likely
   reflected disordered fatty acid metabolism. Accumulation of glutathione
   peroxide reflected oxidative stress. Accumulation of
   17-phenyl-18,19,20-trinor-prostaglandin E2 and 15(R)-pinane thromboxane
   A2 suggested endothelial damage and increased coagulation.

3.5.2. Cluster 2

   Cluster 2 was enriched in metabolites associated with tyrosine
   metabolism (i.e., N-methylamine, 3-fumarylpyruvate), pantothenate and
   CoA biosynthesis (i.e., 5,6-dihydrouracil, 2-ketoisovalic acid), and
   lysine decomposition (i.e., succinic acid, pipecolic acid) ([90]Figure
   6C). These data implied acute heat exposure was associated with the use
   of amino acids as an energy source. Accumulation of mineralocorticoid
   (i.e., 18-hydroxycorticosterone carbohydrate) suggested alterations in
   the regulation of water and salt balance.

3.5.3. Cluster 3

   Cluster 3 was enriched in amino acids (i.e., l-tyrosine,
   l-arginino-succinate, threonylglutamic acid, itaconic acid,
   N-acetyltryptophan, gamma-glutamylalanine, 2,5-dihydroxybenzenesulfonic
   acid, dimethylglycine, l-arginine, d-proline, isoleucyl-leucine,
   glutamylmethionine, aspterric acid, l-methionine, N-formylglycine,
   valyl-gamma-glutamate, N-acetylleucine, l-valine, 4-hydroxybutyric
   acid, salicylic acid, l-asparagine, l-glutamine, N8-acetylspermidine,
   isoleucyl-aspartate, l-histidine, l-lysine, l-threonine, leucine,
   serine, l-alanine, phenylalanine) ([91]Figure 6D). These data implied
   acute heat exposure was associated with the use of amino acids as an
   energy source, suggesting protein supplementation may be helpful after
   exposure to extreme heat.

3.5.4. Cluster 4

   Cluster 4 was enriched in bile acid metabolites, including
   chenodeoxycholic acid, 3β-hydroxy-5-cholenoic acid, hyodeoxycholic
   acid, glycoursodeoxycholic acid, cholic acid, deoxycholic acid,
   deoxycholic acid, glycocholic acid, taurochenodesoxycholic acid,
   taurocholic acid and taurodeoxycholic acid ([92]Figure 6E). These data
   implied acute heat exposure was associated with alterations in bile
   acid metabolism.

3.5.5. Cluster 5

   Cluster 5 was enriched in free fatty acids, including dihomo-linolenic
   acid, dihomo-alpha-linolenic acid, 4,7,10,13,16-docosapentaenoic acid,
   trans-11,cis-15-octadecadienoic acid, 4,5-dehydro docosahexaenoic acid,
   docosahexaenoic acid, alpha-eleostearic acid, docosapentaenoic acid
   (22n-3), 9-hydroxylinoleic acid, adrenic acid, 2-hydroxyhexadecanoic
   acid, (E)-10-hydroxy-8-decenoic acid, eicosadienoic acid, erucic acid,
   linoleic acid, arachidonic acid, heneicosanoic acid, eicosapentaenoic
   acid, α-linolenic acid, and stearic acid ([93]Figure 6F). Accumulation
   of free fatty acids might be due to the inhibition of heat production
   and reduce of fatty acid digestion.

3.5.6. Cluster 6

   Cluster 6 was enriched with products of sphingomyelin metabolism [i.e.,
   sphinganine, sphinganine 1-phosphate, sphingosine 1-phosphate,
   sphingosine 1-phosphate (d16:1-P) ([94]Figure 6G)]. Metabolites of
   sphingomyelin are important signaling molecules, which may be involved
   in regulating cell proliferation, immune function, inflammatory
   reactions, and oxidative stress.

3.6. Identification of hub molecules involved in the response to acute heat
exposure

   The levels of some plasma proteins continued to increase after acute
   heat exposure, implying they are key regulators of heat exposure. To
   assess the interaction of these plasma proteins and identify potential
   plasma proteins modulating the response to acute heat exposure,
   protein–protein interaction analysis was performed. Several important
   nodes, corresponding to VWF, PF4, THBS1, HSP90AB1, and MPO, were
   revealed ([95]Figure 7A). These proteins were also validated using
   ELISA, which showed their levels increased after acute heat exposure
   ([96]Figures 7B–[97]F and [98]Supplementary Table S3).

Figure 7.

   [99]Figure 7
   [100]Open in a new tab

   Hub molecules associated with acute heat exposure. (A) Protein–protein
   interaction after acute heat exposure produced by STRING. Line
   thickness indicates the strength of data support for each interaction.
   Red colored proteins were associated with acute heat exposure. (B–E)
   Plasma HSP90β, VWF, PF4, THBS1, and MPO levels before and after acute
   heat exposure were detected by enzyme-linked immunosorbent assay. (G)
   Clustering dendrogram of differentially expressed metabolites.
   Clustering was based on dissimilarity of metabolites. (H) Heatmap of
   the correlation between metabolites modules and the plasma proteins
   HSP90, VWF, PF4, THBS1, and MPO.

   Additionally, weighted correlation network analysis measured the
   co-expression relationship between VWF, PF4, THBS1, HSP90AB1, MPO and
   plasma metabolites. Hierarchical clustering with dynamic tree cut
   methods were applied to identify metabolite modules ([101]Figure 7G).
   Pearson’s correlation coefficient showed HSP90AB1, VWF, PF4 and THBS1
   were associated with metabolite modules, identifying HSP90AB1, VWF, PF4
   and THBS1 as hub proteins. HSP90AB1 was correlated with the green
   metabolite module and the cyan metabolite module. VWF was correlated
   with the lightcyan metabolite module and the blue metabolite module.
   PF4 was correlated with the yellow metabolite module, the purple
   metabolite module, and the turquoise metabolite module. THBS1 was
   correlated with the yellow metabolite module, the purple metabolite
   module, the turquoise metabolite module, and the green-yellow
   metabolite module ([102]Figure 7H). KEGG pathway enrichment analysis
   was performed on the metabolites in the key modules correlated with the
   hub proteins ([103]Supplementary Figure S6). These modules were
   enriched in fructose and mannose metabolism, choline metabolism,
   cholesterol metabolism, histidine metabolism, carbohydrate digestion
   and absorption, and sphingolipid metabolism. The data implied a
   regulatory role of HSP90AB1, VWF, PF4 and THBS1 to metabolites in acute
   heat exposure.

3.7. Multi-omics prediction of body temperature and cardiopulmonary
parameters

   Multi-omics data at the time point immediately after acute heat
   exposure was used to assess the relevance of clinical data for
   immediately detecting acute heat exposure. Correlation analysis
   identified a set of biomarkers that were highly predictive of changes
   in body temperature, FVC VC, EF, and FS ([104]Figure 8). Body
   temperature was correlated with metabolites of ascorbic acid and
   N-acetyl-d-tryptophan. FVC was correlated with the metabolites of
   acetylphenol sulfate and chlorobenzoic acid. VC was correlated with the
   plasma protein adhesion G protein-coupled receptor L4. EF and FS were
   correlated with the plasma proteins integrin alpha-5 and creatine
   kinase M-type, metabolites of amdoxovir and docosahexaenoic acid. These
   biomarkers have been previously associated with body temperature and
   the cardiopulmonary system, highlighting their clinical value
   ([105]27–30).

Figure 8.

   [106]Figure 8
   [107]Open in a new tab

   Multi-omics prediction of body temperature and cardiopulmonary
   function. (A) The proportion of plasma proteins and metabolites
   associated with body temperature and cardiopulmonary parameters. (B)
   Heatmap showing the correlations between plasma analytes and body
   temperature and cardiopulmonary parameters. * represents p < 0.05, **
   represents p < 0.01 and *** represents p < 0.001.

4. Discussion

   The increasing global temperatures and incidence of occupational
   exposure to extreme high temperatures require enhanced understanding of
   how heat exposure impacts human health ([108]31). This study
   investigated the short-term tolerance of humans to extreme high
   temperatures and characterized the detailed series of events that occur
   in response to acute heat exposure (50°C). Longitudinal multi-omics
   profiling of plasma components revealed acute heat exposure affected
   physiological processes in a time-dependent manner. These physiological
   processes included the immune response, coagulation, acid–base balance,
   oxidative stress, cytoskeleton maintenance, and energy metabolism. Time
   series data showed variations in the dynamics of these physiological
   processes and the molecular changes underlying the response to acute
   heat exposure. HSP90AB1, VWF, PF4, and THBS1 were identified as key
   regulators of the physiological processes involved in the response to
   acute heat exposure. A small number of blood-based analytes were
   predictors of changes in cardiopulmonary function, with potential for
   use as a clinical tool to evaluate the health risks of heat exposed
   populations.

4.1. Physiological performance in acute heat exposure

   In this study, the physiological performance of the subjects was
   detected. We found that during heat exposure, subjects exhibited
   typical physiological reactions, such as increased heart rate, to
   ensure proper circulation of tissues. However, after heat exposure,
   subjects suffer some impairment of cardiac function, which may be
   related with the fluid loss or the impaired vascular function ([109]18,
   [110]32). Furthermore, our study revealed a decrease in lung function
   following heat exposure. This finding aligns with a previous study that
   found a consistent association between each 1°C increase in lifetime
   summer mean temperature and a 1.07% decrease in forced vital capacity
   (FVC) ([111]33). These findings suggested that the subjects included in
   our study exhibited some adaptive responses to acute heat exposure but
   experienced a slight decline in their cardiopulmonary function. To
   mitigate these effects, it is advisable to implement preventive
   measures such as adequate hydration before, during, and after heat
   exposure, and monitoring individuals for signs related to heat
   exposure. Additionally, ensuring proper acclimatization to heat and
   providing education on heat exposure management can help enhance the
   resilience of individuals facing acute heat exposure.

4.2. Immune response in acute heat exposure

   Although there were no significant changes in the vital clinical signs
   of the subjects after extreme high temperature exposure, the changes in
   molecular reactions still indicated the potential increased risk after
   heat exposure. With respect to the immune response, the present study
   showed that synthesis of immunoglobulin was reduced after heat
   exposure, indicating the decreased immune function even short time
   exposure to extreme high temperature. This finding is consistent with
   previous reports. An investigation of individuals who worked in the
   foundry showed chronic heat exposure was associated with a decrease in
   the number of white blood cells ([112]34). Additionally, Chen et al.
   ([113]35) found the strong associations between heat waves and
   emergency department visits for intestinal infection. Studies of
   patients hospitalized for heat stroke during sustained heatwaves have
   reported a high incidence of urinary, blood or lung infections during
   recovery ([114]36–38). This phenomenon may be related to abnormal
   development and decreased number of lymphocytes ([115]39). Findings
   from a mouse model with heat exposure showed that whole blood collected
   30 days after recovery from heat exposure exhibited an
   immunosuppressive phenotype ([116]40).

4.3. Coagulation in acute heat exposure

   Coagulation disorders accompany heat-related illness, and severe DIC
   caused by heatstroke can be fatal ([117]41–43). Findings from the
   present study imply that coagulation, fibrinolytic and
   anti-fibrinolytic processes are mobilized in the initial phase of heat
   exposure. In particular, coagulation-related factors, including VWF,
   PF4, and THBS1, remained elevated 24 h after acute heat exposure. VWF
   binds the glycoprotein Iba on platelets causing platelet adhesion,
   activation and aggregation, which is crucial for hemostasis and
   thrombosis ([118]44, [119]45). PF4, also known as chemokine (C-X-C
   motif) ligand 4, has the ability to bind with negatively charged
   polyanions such as VWF, resulting in the formation of immune complexes
   that can contribute to the development of thrombosis ([120]46). THBS1,
   which is produced by platelets when stimulated by thrombin, plays a
   role in platelet aggregation ([121]47). The increase in coagulation
   related factors reflects a hypercoagulant state suggesting that
   modulating coagulation early after exposure to high temperature may be
   beneficial for preventing heat-related illness. In previous studies,
   various anticoagulants, such as antithrombin III, recombinant
   thrombomodulin and tissue factor pathway inhibitors, had been used to
   prevent heatstroke ([122]48, [123]49).

4.4. Acid–base balance in acute heat exposure

   Acid–base balance is regulated by hemoglobin and carbonic anhydrase
   levels, which increased in response to acute heat exposure in the
   present study. Hemoglobin in red blood cells binds oxygen and releases
   hydrogen ions ([124]50). Carbonic anhydrase catalyzes the reversible
   hydration of carbon dioxide (CO[2]) to form carbonic acid (H[2]CO[3])
   in the lungs, kidneys, and red blood cells ([125]51). Previous studies
   showed that weekly heat training increases elite cyclists or
   cross-country skiers’ hemoglobin mass, enhancing their ability to use
   oxygen ([126]52–54). In the present study, the increases in hemoglobin
   and carbonic anhydrase were temporary and returned to baseline levels
   the day after acute heat exposure, indicating that repeated heat
   exposures are needed to maintain increased levels of hemoglobin and
   carbonic anhydrase.

4.5. Oxidative stress in acute heat exposure

   Oxidative damage may be induced by chronic and long-term heat exposure,
   which enhances GST and GPX enzymatic activity and nitric oxide
   concentration ([127]55, [128]56). Consistent with this, the present
   study showed prooxidants such as MPO and S100A8 were upregulated after
   acute heat exposure. Interestingly, some antioxidants (peroxiredoxin-2,
   peroxiredoxin-6, kallistatin, superoxide dismutase 1) were induced to
   resist harmful effects. Peroxiredoxin-2, peroxiredoxin-6 and superoxide
   dismutase are antioxidant enzymes that protect organisms from oxidative
   damage caused by reactive oxygen species (ROS) ([129]57, [130]58).
   Kallistatin antagonizes TNF-α-induced oxidative stress, and its active
   site is crucial for stimulating antioxidant enzyme expression
   ([131]59). Previous reports showed the negative effects of heat
   exposure on oxidative status might be alleviated by dietary antioxidant
   supplementation ([132]60, [133]61). The present study adds to this
   growing body of evidence.

4.6. Energy metabolism in acute heat exposure

   Energy metabolism was altered by acute heat exposure. Levels of
   circulating plasma metabolites, such as amino acids, free fatty acids,
   bile acid, carnitine and sphinganine, were changed. Amino acid levels
   decreased over time as they were catabolized for energy generation.
   Acylcarnitines and fatty acids accumulated after acute heat exposure.
   This may be due to the transformation of beige adipocytes into white
   adipocytes, which can inhibit adipocyte heat production and reduce
   fatty acid digestion ([134]62). Sphinganine levels increased briefly,
   confirming the role of sphingosine as a potent regulator of signaling
   pathways. Levels of the bile acids chenodeoxycholic acid, cholic acid
   and deoxycholic acid decreased, which might be associated with the
   decrease in that was observed in high-temperature environments
   ([135]63).

4.7. Potential regulators and biomarkers associated with acute heat exposure

   Levels of some circulating plasma proteins (PF4, F13A1, F13B, F9, F12,
   F10, F5, F8, GP5, VWF, THBS1, MPO, GAPDH, HSP90AB1, and ERN1) continued
   to increase during the recovery phase after acute heat exposure,
   implying these proteins are potential regulators of important
   physiological processes involved in the response to acute heat
   exposure. Protein–protein interaction and WGCNA analyses identified
   HSP90AB1, VWF, PF4, and THBS1 as hub proteins. In previous literature,
   HSP90AB1, VWF, PF4, and THBS1 have been documented to exert substantial
   influence on various aspects of cell function. HSP90AB1, also known as
   HSP90β, serves as molecular chaperones to maintain protein stability
   and has protective roles in response to various kinds of cellular
   stress, such as heat exposure ([136]64, [137]65). VWF, PF4, and THBS1
   play a critical role in coagulation and regulate immune cells and
   inflammatory responses ([138]66–68).

   After heat exposure, we found that the changes in the levels of certain
   plasma molecules were significantly correlated with the changes in
   clinical parameters. Notably, some of these molecules have previously
   been associated with clinical parameters, suggesting their potential
   clinical value in predicting functional changes following heat
   exposure. One of such molecules is N-Acetyl-d-Tryptophan, which showed
   a positive correlation with body temperature after acute heat exposure.
   This molecule is a derivative of tryptophan and can be metabolized into
   5-hydroxytryptamine (5-HT), and activation of 5-HT receptors can lead
   to an increase in body temperature ([139]69, [140]70). Furthermore, we
   observed that the increase of creatine kinase M-type leptin and the
   decline in docosahexaenoic acid were found to be correlated with the
   decline in cardiac function after heat exposure. Creatine kinase, an
   enzyme involved in ATP production and utilization, has been extensively
   studied as potential risk marker for cardiovascular events ([141]71).
   Docosahexaenoic acid is an omega-3 fatty acid, which is found to have
   positive effects on cardiac function and significantly decreases the
   overall mortality in patients diagnosed with coronary heart disease
   ([142]72). Additionally, we found adhesion G protein-coupled receptor
   L4 (ADGRL4) was positively associated with pulmonary function after
   acute heat exposure. ADGRL4 is an orphan adhesion GPCR expressed in
   endothelial cells that can induce vascular normalization and immune
   suppression ([143]73), which may contribute to protecting lung
   function.

4.8. Limitations

   This study had some limitations. The sample size was small and did not
   include females or the older adult. Owing to the safety consideration
   for subjects, the max tolerance time was not detected, and the relative
   humidity were not set at a high level, which had obvious effect on
   human’s tolerance time of extreme high temperature. Therefore,
   molecular changes associated with life-threatening heatstroke could not
   be explored under the current experimental setting. Furthermore, while
   this study provided a comprehensive analysis of molecular changes in
   response to acute heat exposure, the potential regulators and
   biomarkers identified require additional validation in larger and more
   diverse populations. Additionally, although the study identified
   potential biomarkers predictive of changes in body temperature and
   cardiopulmonary parameters, the clinical significance and practical
   applications of these findings are yet to be determined. Further
   studies are needed to evaluate the usefulness of these biomarkers in
   real-world settings.

5. Conclusion

   Overall, this study offered insights into key research blanks in health
   impact assessment of heat exposure. Longitudinal multi-omics profiling
   identified thousands of molecules that were affected by acute heat
   exposure, and time-series clustering and network analysis revealed
   crosstalk between inflammation, coagulation, immunity, acid–base
   balance, metabolism, and oxidative stress. Potential regulators and
   biomarkers for acute heat exposure were revealed. These findings have
   significant implications for both occupational population exposed to
   heat and the general population during extreme heat environment. It
   will help to improve the prediction of short-term health outcomes in
   acute heat exposure and develop prevention strategies to heat-related
   injuries and mitigate the adverse effects of heat exposure.

Data availability statement

   The datasets presented in this study can be found in online
   repositories. The names of the repository/repositories and accession
   number(s) can be found in the article/[144]Supplementary material.

Ethics statement

   The studies involving humans were approved by The Ethics Review
   Committee of West China Hospital of Sichuan University. The studies
   were conducted in accordance with the local legislation and
   institutional requirements. The participants provided their written
   informed consent to participate in this study.

Author contributions

   JWe: Conceptualization, Data curation, Formal analysis, Investigation,
   Validation, Writing – original draft, Writing – review & editing. JC:
   Data curation, Formal analysis, Investigation, Writing – original
   draft, Writing – review & editing. LW: Data curation, Formal analysis,
   Investigation, Methodology, Writing – original draft. CL: Data
   curation, Methodology, Writing – original draft. YZ: Data curation,
   Investigation, Writing – original draft. JWu: Conceptualization, Formal
   analysis, Funding acquisition, Visualization, Writing – review &
   editing. JL: Conceptualization, Formal analysis, Funding acquisition,
   Project administration, Supervision, Writing – original draft, Writing
   – review & editing.

Acknowledgments