During the early phase of sepsis, WAT undergoes intense lipolysis, driven by pro-inflammatory cytokines (e.g., TNF-α, IL-6) and stress hormones (e.g., catecholamines, cortisol) (7). This process is primarily orchestrated by the activation of HSL (34), resulting in a substantial release of released free fatty acids (FFAs) into the circulation. These FFAs not only serve as vital energy substrates but also function as signaling molecules that further modulate the activity of immune cells (35).

Under the synergistic drive of pro-inflammatory factors and stress hormones, white adipose tissue undergoes intense lipolysis. On one hand, cytokines such as TNF-α directly promote the activation of HSL by activating signaling pathways like ERK (36).these pathways lead to a sharp increase in plasma FFA levels (7).On the other hand, catecholamines phosphorylate and activate HSL via the classical β-adrenergic receptor-cAMP-PKA signaling axis, making this the rate-limiting step for lipolysis (34) The substantial amount of FFAs released not only serve as crucial energy substrates but also function as important signaling molecules. For instance, saturated fatty acids can further amplify the pro-inflammatory response by activating the TLR4 receptor on macrophages (35).

FFA released into the circulation are preferentially taken up by BAT, which is highly enriched with mitochondria. FFA undergo β-oxidation within BAT mitochondria, generating reducing equivalents that drive the electron transport chain (37) However, sepsis significantly induces and activates the unique UCP1 in BAT. Acting as a proton channel, UCP1 ‘uncouples’ oxidative phosphorylation, thereby diverting the chemical energy released from substrate oxidation away from ATP synthesis and converting it directly into heat (10). This nonshivering thermogenesis has been demonstrated in septic animal models to be a critical component of fever and the associated high energy expenditure (38).

Under the hypermetabolic conditions of sepsis, free fatty acids delivered to the liver not only fuel hepatocytes but are also partially converted into ketone bodies within hepatic mitochondria (19) The resulting ketones (e.g., β-hydroxybutyrate) serve as a highly efficient, water-soluble fuel that can be rapidly taken up and oxidized by the mitochondria of vital organs such as the heart and brain to generate ATP (39). The heart, in particular, utilizes ketone bodies as a preferred energy source, which helps sustain cardiac function under stress (40). Thus, the enhancement of ketogenesis during sepsis represents an evolutionarily conserved metabolic adaptation. It provides a critical alternative energy source for the body by converting stored fat into readily transportable and utilizable ketone bodies (41).

During the energy crisis of sepsis, the efficient utilization of the aforementioned lipid-based energy sources carries critical pathophysiological significance: namely, reducing the excessive reliance on skeletal muscle protein catabolism. It is well-established that during severe energy deficits, the body extensively breaks down muscle protein via the ubiquitin-proteasome pathway for energy production (20). Therefore, the active lipolysis, BAT thermogenesis, and hepatic ketogenesis observed in sepsis collectively constitute a metabolic adaptation strategy aimed at ‘protein sparing’. The essence of this strategy is disease tolerance—by prioritizing the mobilization of fat reserves to meet energy demands, the host maintains vital functions while maximizing the preservation of crucial structural and functional proteins (16). Substantial evidence indicates that, in the context of bacterial infections, interventions triggering this metabolic state (e.g., fasting) can significantly enhance host tolerance and improve survival rates through protein-sparing benefits (18).Furthermore, lipid metabolites such as ketone bodies may further contribute to tissue protection via direct anti-inflammatory mechanisms, among others (24).

Mitochondria-Associated Endoplasmic Reticulum Membranes (MAMs) as Hubs for Inflammatory Sensing and Signal Integration, MAMs are dynamic, specialized membrane contact sites between the mitochondrial outer membrane and the endoplasmic reticulum, enriched with proteins mediating calcium signaling, lipid transport, and inflammatory responses (42, 43).During sepsis, this structure becomes a central platform for sensing both internal and external danger signals. On one hand, pathogen-associated molecular patterns (e.g., LPS) can transmit their downstream signals to the MAMs domain via TLR4 receptors on the surface of adipocytes and macrophages (44).On the other hand, host-derived damage-associated molecular patterns (e.g., mtDNA) generated by cellular stress can leak into the cytosol, be recognized by cGAS, and activate the STING pathway. The activated STING protein is precisely localized to MAMs to initiate a robust type I interferon response (45). Upon sensing inflammatory signals, MAMs remodel the immunometabolic microenvironment of adipose tissue. By mediating Ca²⁺ flux between the ER and mitochondria (46), MAMs directly regulate cellular metabolic states, promoting the production of a series of lipid mediators (e.g., PGE2) and metabolites (e.g., succinate) in adipocytes (47). This reprogrammed metabolic environment, particularly specific lipid species (e.g., palmitoleic acid) and signaling molecules, provides instructional cues for immune cell fate. Among these, prostaglandin E2 (PGE2) has been shown to directly drive macrophage polarization towards an M2 anti-inflammatory phenotype via the cAMP-PKA pathway (48). Concurrently, these metabolic changes help prevent the abnormal accumulation of metabolites like succinate, indirectly supporting the stabilization of the M2 phenotype (49). Ultimately, through the synergistic action of these metabolic signals and Th2 cytokines, macrophages differentiate into the classical M2 phenotype, characterized by high expression of arginase-1 and mannose receptors, and secretion of IL-10 and TGF-β (50). This establishes an anti-inflammatory and pro-repair microenvironment within adipose tissue.

Through the MAMs-STING axis, adipose tissue sensing of endogenous DAMPs (e.g., mtDNA) drives the polarization of macrophages toward an M2 phenotype (51). This shift establishes a local anti-inflammatory and pro-repair microenvironment (52, 53), which, by preserving adipose tissue homeostasis, systemically contributes to the maintenance of disease tolerance during sepsis (16).

Previous research has integrated the hypermetabolic phenomena observed in sepsis—such as lipolysis, ketogenesis, and BAT thermogenesis—into the “Metabolic Defense Priority” hypothesis (18). This hypothesis posits that the body prioritizes the consumption of renewable energy reserves (fat), even at a high energetic cost, to protect irreplaceable functional structures (muscle). Clinical observations directly validate the core metabolic features described by this hypothesis, demonstrating that septic patients exhibit significantly elevated plasma levels of FFAs and ketone bodies, the dynamics of which are closely associated with disease progression (7, 54).Furthermore, Baba et al. confirmed the activation of BAT in both rodent models and patients using imaging techniques (55),while Langley et al. revealed, through metabolomic studies, that survival outcomes are linked to specific metabolic states (54), Collectively, this clinical evidence provides robust support for the “Metabolic Defense Priority” hypothesis as a core host defense strategy.

These studies indicate that the metabolic phenomena observed during sepsis represent an evolutionarily conserved, active adaptive program. This ‘Metabolic Defense Priority’ program prioritizes the consumption of renewable fat reserves to spare structural muscle protein, aiming to maximize survival chances (18). It is executed through coordinated mechanisms: under the drive of inflammatory signals and catecholamines, white adipose tissue lipolysis releases FFAs into circulation (7, 56); a portion is converted by the liver into ketone bodies to provide an alternative fuel for the brain, reducing reliance on muscle catabolism (19); and another fraction is utilized by BAT for UCP1-mediated thermogenesis. By efficiently redirecting energy substrates through these pathways, the body delays sepsis-associated myopathy and preserves the function of vital muscles (57).

At the physiological level, the septic organism demonstrates the central role of fat mobilization and the achievement of protein sparing. At the clinical level, muscle loss has been consistently linked to adverse outcomes (57). At the experimental level, interventions that drive this metabolic switch directly enhance disease tolerance (18). This compelling body of evidence necessitates a paradigm shift in our understanding: the host is not a passive victim of infection, but rather actively orchestrates its energy resources through precise metabolic ‘switching’ (14).

Sepsis is a dynamic, multi-stage pathological process. The “Stage-Specific Adaptation” hypothesis proposes that the disease tolerance strategy executed by the host via the adipo-immune-metabolic axis is not static, but undergoes precise, temporally ordered reprogramming as sepsis progresses (acute, sub-acute, and chronic phases).

During the metabolic response to sepsis, the canonical lipolysis pathway serves as a crucial initial signal driving early energy mobilization. This pathway is primarily mediated by catecholamines via β-adrenergic receptors. Upon activation, the highly conserved signaling axis of Gs protein-adenylyl cyclase-cAMP- PKA is triggered, ultimately leading to the phosphorylation and activation of HSL, which hydrolyzes triglycerides to release FFAs (34).

Beyond classical endocrine regulation, adipose mobilization is also directly driven by non-canonical lipolytic pathways. These pathways are primarily activated by inflammatory signals (e.g., TNF-α, IL-1β) and operate independently of the classical β-adrenergic receptor-cAMP axis. Specifically, TNF-α exerts its effects by activating the JNK and ERK signaling pathways (60). The non-canonical lipolysis mediated by M1 macrophages represents the core driving force for the rapid energy mobilization from adipose tissue during early sepsis. A significant fate of the substantial amount of FFAs released via these lipolytic processes is their active uptake by BAT, facilitated by highly expressed fatty acid transporter proteins. Within BAT mitochondria, FFAs undergo β-oxidation; however, the resulting proton gradient is ‘short-circuited’ by UCP1. Consequently, the chemical energy is diverted away from ATP synthesis and is instead dissipated as heat (10). Furthermore, active BAT significantly accelerates the clearance of triglycerides and free fatty acids from the bloodstream. BAT thermogenesis, as a regulated energy expenditure pathway, safely dissipates the energy from excessively mobilized lipids in the form of heat, thereby mitigating lipotoxicity (61).

The products of lipolysis are not merely energy substrates but also crucial signaling molecules. Glycerol is transported to the liver, where it is catalyzed by glycerol kinase to serve as a key precursor for gluconeogenesis, helping to maintain blood glucose levels. Certain unsaturated fatty acids and their derivatives (e.g., oxidized fatty acids) can function as endogenous agonists for the nuclear receptor PPARγ. Binding and activation of PPARγ subsequently regulates anti-inflammatory gene expression programs, suppressing inflammatory responses in macrophages, among other cells (62). In contrast, saturated fatty acids (e.g., palmitic acid) can activate the TLR4 receptor, initiating pro-inflammatory signaling pathways such as NF-κB, thereby directly exacerbating the inflammatory response (63).

When the lipolysis-thermogenesis axis becomes dysfunctional, the host’s metabolic defense system collapses, leading to rapidly deteriorating outcomes.1.Lipolytic Failure: Chronic inflammation induces insulin resistance and receptor desensitization in adipocytes (64). Concurrently, inhibitory signals such as TGF-β persistently suppress cAMP levels by upregulating PDE3B, directly curbing lipolysis. Coupled with macrophage infiltration and intrinsic dysfunction within the adipose tissue itself (65), these factors collectively lead to insufficient FFA release. This interruption in energy substrate supply forces the body to revert to a reliance on the catastrophic breakdown of skeletal muscle protein, accelerating cachexia and organ failure (57).2. hermogenic Program Failure: Brown adipose tissue becomes functionally exhausted, manifesting as reduced mitochondrial density, downregulated UCP1 expression, and loss of the crucial protective role provided by Treg cells. Consequently, the body loses its capacity for thermogenesis, struggling to maintain core body temperature.3. Systemic Metabolic Dysregulation: Due to impaired utilization and suppressed lipoprotein lipase activity (7). the FFAs generated from lipolysis are re-esterified in the liver, leading to hypertriglyceridemia. This condition is not only a hallmark of defective lipolysis but can also exacerbate inflammation and organ injury by activating immune cells, thereby creating a vicious cycle (66).

During the subacute phase of sepsis, when the metabolic paradigm shifts toward lipid mobilization, a balanced and moderate level of de novo lipogenesis plays a crucial physiological role (67). Catalyzed by key enzymes such as acetyl-CoA carboxylase (ACC) and fatty acid synthase (FAS), adipose tissue can utilize substrates like glucose to synthesize fatty acids de novo and esterify them into triglycerides for storage. This process does not oppose lipolysis but rather complements it, helping to prevent premature depletion of adipose reserves and thereby maintaining a sustained and controlled release of FFAs—a stable energy substrate for vital organs such as the heart and skeletal muscle.

However, this delicate equilibrium is highly vulnerable to disruption. Progressive insulin resistance impairs the normal stimulatory effect of insulin on lipogenesis (5), while persistent inflammatory signaling—exemplified by cytokines such as TNF-α—directly suppresses the expression and activity of ACC and FAS, thereby potently inhibiting lipid synthesis (68). When lipogenesis is severely impaired in the face of ongoing lipolytic demand, adipose tissue fails to adequately replenish its triglyceride stores, leading to functional exhaustion. Consequently, FFA release declines, forcing the body to revert to catastrophic protein catabolism in skeletal muscle to meet energy needs (57). This shift precipitates a chronic energy crisis and a wasting state, ultimately accelerating the progression toward multiple organ failure.

In summary, adipose tissue provides energy substrates via canonical and non-canonical lipolysis, while BAT clears lipotoxicity through thermogenesis. The functional integrity of this system is central to maintaining energy homeostasis and preventing lipotoxicity during sepsis. Its dysfunction, however, accelerates the progression towards cachexia and multiple organ failure.

Traditionally, the function of B cells was confined to antibody production. However, cutting-edge research has expanded their role into the realm of metabolic regulation. For instance, Gencer Sancar et al. discovered that B1a cells directly suppress adipose tissue lipolysis via the TGF-β/PDE3B axis, acting as a ‘brake’ to prevent lipotoxicity (78). Concurrently, Ippei Shimizu et al. demonstrated that regulatory B cells (Bregs) can ameliorate adipose tissue inflammation and insulin resistance (79). These findings collectively establish specific B cell subsets as crucial negative regulators in maintaining metabolic homeostasis Mechanistically, recruited or activated B cells secrete TGF-β, which acts on adipocytes to significantly upregulate the expression and activity of PDE3B (31).PDE3B subsequently hydrolyzes the intracellular second messenger cyclic AMP (cAMP), leading to decreased PKA activity. This cascade ultimately results in the inadequate phosphorylation and activation of HSL—the key rate-limiting enzyme in lipolysis—thereby potently inhibiting fat breakdown (80).This mechanism constitutes a robust negative feedback loop designed to prevent excessive and sustained lipolysis. Uncontrolled lipolysis releases an overflow of free fatty acids into the circulation, which can trigger lipotoxicity by activating pathways such as TLR4, ultimately impairing the function of the liver, pancreas, and vascular endothelium (35).

As the most abundant immune cells in adipose tissue, macrophages serve as a ‘bifunctional switch’ for lipolysis due to their remarkable functional plasticity.1. Pro-lipolytic Switch (M1 Phenotype): During early sepsis, M1 macrophages are activated and infiltrate adipose tissue in large numbers, driving lipolysis to fuel the immune response. These cells produce high levels of IL-1β. Research demonstrates that IL-1β strongly promotes fat breakdown by activating the ERK signaling pathway in adipocytes, leading to direct phosphorylation and activation of HSL (15).2. Anti-lipolytic Switch (M2 Phenotype): During disease progression or the resolution phase, M2 macrophages become dominant. They create an inhibitory microenvironment through the secretion of anti-inflammatory factors such as IL-10 and TGF-β (52). Furthermore, their high expression of metabolic enzymes like arginase-1 indirectly alters the metabolic state of adipocytes by depleting local substrates (81). The function of M2 cells is not to directly inhibit lipolytic enzymes, but rather to create conditions for terminating lipolysis and initiating tissue repair by suppressing the overall inflammatory tone and remodeling the metabolic microenvironment (52, 82).

T lymphocytes, through their various subsets, are deeply involved in the immunometabolic dialogue within adipose tissue. Treg cells exert a direct inhibitory effect on lipolysis by expressing CD73, which converts local AMP into adenosine. Adenosine then binds to the A2B receptor on adipocytes, reducing intracellular cAMP levels and directly suppressing fat breakdown. Simultaneously, the IL-10 secreted by Tregs can enhance TGF-β production by B cells, indirectly reinforcing the suppression of lipolysis and forming a coordinated ‘Treg-B cell’ regulatory circuit. CD4+ T helper cells shape the global immune milieu through their signature cytokines: Th1 cells secrete IFN-γ, a classic signal for M1 macrophage polarization, thereby indirectly supporting a pro-lipolytic environment; Th2 cells, by secreting IL-4 and IL-13, drive M2 macrophage differentiation, indirectly contributing to inflammation resolution and the termination of lipolysis (50).

In contrast, cytotoxic CD8+ T cells often play a detrimental role in pathological states. During chronic inflammation, they infiltrate adipose tissue and, through the secretion of factors like IFN-γ and TNF-α, drive inflammation, recruit and activate macrophages, thereby exacerbating tissue damage and metabolic dysfunction (83).The key cellular players can be seen in the Table 1.

During sepsis, a critical ‘Insulin-INSR-Thermogenesis Axis’ is activated. This axis exhibits synergy with the TGFβ-PDE3b-cAMP pathway: insulin, via the PI3K-Akt pathway, can inhibit Smad2/3 phosphorylation, indirectly attenuating the TGFβ-mediated upregulation of PDE3b and preventing excessive suppression of lipolysis. Conversely, overactivation of the STING-ER stress-mtROS axis can damage the insulin receptor (INSR), leading to impaired BAT thermogenic function and creating a vicious cycle of ‘pathway imbalance → functional deterioration’. Contrary to its traditional role as an anabolic hormone, insulin’s signaling output in BAT is ‘reprogrammed’ towards energy consumption (84).

The core mechanisms of this axis are as follows:1. Signal Initiation and Program Execution: Upon binding to the insulin receptor on the BAT cell membrane, insulin activates the canonical IRS1-PI3K-Akt pathway. Activated Akt can directly act on the Ucp1 gene promoter, significantly upregulating its transcription and protein expression (85). Concurrently, insulin signaling synergistically promotes mitochondrial biogenesis (61), establishing the platform for thermogenesis.2. Energy Dissipation: The highly expressed UCP1 protein uncouples the mitochondrial respiratory chain, directly converting the chemical energy from substrate oxidation into heat (10). The specific process can be seen in the Figure 1.

Research confirms that insulin is necessary for activating BAT energy expenditure under conditions requiring thermogenesis. Therefore, in the context of sepsis, this axis consumes energy to maintain core body temperature and dissipates excess lipid energy.

The Insulin-INSR-Thermogenesis Axis does not operate in isolation; it forms an exquisite ‘Supply-Consumption’ synergy with systemic fat mobilization (lipolysis), collectively consolidating disease tolerance.1. Functional Coupling: Directed Energy Flow: The substantial amount of free fatty acids released into the circulation via lipolysis from white adipose tissue are efficiently taken up and cleared by brown adipose tissue as the primary ‘client’ (61). The actual oxidation of FFAs and uncoupling via UCP1 in BAT is primarily and directly driven by sympathetic nervous system activation (e.g., via β3-adrenergic receptors) (10).Insulin signaling, while classically anabolic in promoting lipid storage and glucose uptake in BAT (84), plays a critical indirect and supportive role in thermogenesis. Specifically, through pathways such as Akt, insulin upregulates UCP1 expression (85) and promotes mitochondrial biogenesis (86). This ‘primes’ BAT by replenishing intracellular substrate stores and enhancing its cellular capacity, thereby enabling a robust and efficient thermogenic response when catecholamine stimulation occurs. This creates an efficient metabolic pipeline: ‘WAT → circulating FFAs → BAT (insulin-activated → UCP1 thermogenesis)’, ensuring energy is rapidly and directionally utilized for survival-critical tasks.2. Homeostatic Synergy: Avoiding Metabolic Conflict: Although insulin inhibits WAT lipolysis at the systemic level (34), its thermogenic role in BAT exhibits tissue-specific differences. This synergy and dissociation enable precise regulation and systemic protection. Precise regulation is manifested by permitting moderate systemic fat mobilization while ensuring energy is diverted to BAT for ‘safe combustion’. Systemic protection is achieved by consuming excess, lipotoxic FFAs through thermogenesis, thereby protecting organs like the liver and pancreas, while the heat production itself maintains the core body temperature essential for survival.

Recent research on the Insulin-INSR-Thermogenesis Axis indicates: UCP1, as the core molecule of BAT thermogenesis, Igor Golic’s research showed that insulin can modulate mitochondrial function in brown adipose tissue, altering UCP1 expression and cellular thermogenic capacity (86). However, directly targeting UCP1 is highly challenging. Consequently, current research often focuses on its upstream pathways; for instance, Gómez-García found that CL316,243 significantly increased UCP1 protein expression in both BAT and white adipose tissue in mouse studies (87). To date, no safe and highly specific direct UCP1 agonist for humans has been identified. Agonists/inhibitors targeting upstream pathways (e.g., mirabegron derivatives) have been shown to activate BAT and improve metabolism in human clinical studies, but their application, particularly in obesity treatment, is limited by cardiovascular side effects such as increased blood pressure and heart rate (88).

The TGFβ-PDE3b-cAMP axis represents a classic negative feedback pathway initiated by immune cells to maintain metabolic homeostasis. Through the molecular sequence of upregulating PDE3b → degrading cAMP → inhibiting HSL/ATGL activity, it provides an essential “braking” mechanism for the excessive fat mobilization seen in sepsis. This axis is one of the key balancing acts the body relies on to achieve disease tolerance, walking a tightrope between energy mobilization and tissue protection.

Core Mechanisms:1. Initiation of Immune Signal: Specific immune cell subsets, primarily regulatory B cells, are recruited or activated and secrete Transforming Growth Factor-β as the initiating signal.2. Receptor Binding and Signal Transduction: TGF-β binds to the TGF-β receptor II/I complex on the adipocyte membrane, activating its intrinsic serine/threonine kinase activity. This triggers the canonical Smad signaling pathway, leading to the phosphorylation of Smad2/3 and initiating downstream transcriptional reprogramming (89).3. Upregulation of the Key Effector Molecule: A central transcriptional target of TGF-β signaling is the gene for Phosphodiesterase 3B. The activated Smad complex directly binds to the PDE3b promoter, significantly upregulating its expression and activity. 4.Degradation of the Second Messenger and Halt of the Lipolytic Engine: PDE3b specifically hydrolyzes cyclic AMP, converting it to inactive 5’-AMP (34). The sharp decline in cAMP levels directly reduces Protein Kinase A activity. Inactive PKA cannot adequately phosphorylate and activate the two key rate-limiting enzymes of lipolysis—Hormone-Sensitive Lipase and Adipose Triglyceride Lipase—ultimately leading to significant suppression of fat breakdown. The specific process can be seen in the Figure 2.

Recent studies on the TGFβ-PDE3b-cAMP axis confirm that upregulating PDE3b expression or activity effectively inhibits lipolysis (90).Conversely, using PDE3 inhibitors stimulates lipolysis and increases plasma free fatty acid levels (91)These findings suggest that the activation of this axis holds pathophysiological significance far beyond simple signal transduction; it constitutes a core defensive strategy for maintaining homeostasis during the metabolic storm of sepsis. Firstly, as a “precision brake,” this pathway effectively prevents runaway lipolysis driven by pro-inflammatory signals and the premature depletion of energy reserves. Secondly, avoiding lipotoxicity is its most critical protective role. Uncontrolled lipolysis releases excessive FFAs into the circulation, leading to their ectopic deposition in non-adipose organs like the liver and pancreas, potentially causing mitochondrial dysfunction, insulin resistance, and even cell apoptosis. The essence of this axis is “Bidirectional Immunometabolic Crosstalk.” It exemplifies the immune system’s capacity not only to drive inflammation and metabolism but also to proactively downregulate these processes, showcasing a sophisticated layer of metabolic regulation vital for host defense.

In sepsis, the ‘STING-ER Stress-mtROS Axis’ forms the core of a vicious cycle that drives persistent inflammation and cellular apoptosis. Initial inflammation and hypoxia damage mitochondria, leading to the leakage of mtDNA into the cytosol (45), Cytosolic mtDNA is recognized by cGAS, which generates cGAMP to activate the STING signal (92). Activated STING not only translocates to MAMs to drive type I interferon expression, but its depletion from the endoplasmic reticulum also induces ER stress (93), This activates pathways like the UPR’s PERK-eIF2α branch, which, while attempting to restore homeostasis, paradoxically exacerbates inflammatory signaling (94). ER stress causes the release of calcium ions through channels like IP3R, which are then taken up in large quantities by adjacent mitochondria via MAMs. Mitochondrial calcium overload disrupts the electron transport chain, causing electron leakage and a burst in superoxide anion production, known as mtROS burst (95). Furthermore, activated STING can suppress mTORC1 activity, and this inhibition can alter the cell’s antioxidant defense capacity, potentially indirectly worsening oxidative stress (96).mtROS further exacerbates mitochondrial damage by oxidizing mtDNA itself, making it more recognizable to cGAS or more potent at activating NLRP3. On one hand, this creates a positive feedback loop, releasing more mtDNA to re-activate the cGAS-STING pathway (97). On the other hand, mtROS acts as a potent activation signal, directly triggering the assembly of the NLRP3 inflammasome, promoting the maturation of IL-1β/IL-18 and inducing pyroptosis (44). The specific process can be seen in the Figure 3.

The role of the STING-ER stress-mtROS axis in sepsis exhibits a distinct stage-dependent duality Early Phase: Moderate activation of the axis plays an adaptive defense role. Appropriate type I interferon production is crucial for coordinating early anti-infective immunity (98), while the initiated UPR aims to restore ER homeostasis and promote cell survival (94, 98).At this stage, inflammation and oxidative stress are controlled, protective responses designed to eliminate the threat. Advanced Phase: When sepsis becomes uncontrolled and progresses to a later stage, the axis descends into a self-sustaining vicious cycle. Persistent type I interferon response paradoxically promotes immune paralysis and tissue damage (96), Severe and irreversible ER stress switches its signaling from “pro-survival” to “pro-apoptosis” by activating pathways involving CHOP (99), Concurrently, explosive mtROS strongly activates the NLRP3 inflammasome, driving pyroptosis and leading to widespread death of adipocytes, immune cells, and parenchymal cells (44).On a metabolic level, mitochondrial failure directly impairs ATP synthesis (100), and disrupts normal lipid metabolism, exacerbating lipolytic failure (70). Ultimately, these cellular catastrophes converge into multiple organ dysfunction (101).

Recent single-cell multi-omics studies have revealed that exosomes secreted by adipocytes during the chronic phase of sepsis can carry STING and trigger a potent type I interferon response via the cGAS-STING pathway, persistently amplifying inflammation (102, 103)Conversely, using specific small-molecule STING inhibitors (e.g., H-151) can effectively block this signaling axis, significantly attenuating tissue damage and improving outcomes in inflammatory and sepsis models (104). This collectively confirms that targeting STING-mediated communication between adipose and immune cells represents a potential therapeutic strategy.

The three key pathways described above—the Insulin-INSR-Thermogenesis Axis, the TGFβ-PDE3b-cAMP Axis, and the STING-ER Stress-mtROS Axis—orchestrate the function of the adipo-immune-metabolic axis through intricate crosstalk. A balance among these pathways enhances disease tolerance, whereas their imbalance leads to organ failure. Their stage-dependent nature provides molecular targets for precise intervention.

In summary, the Insulin-INSR axis supports energy expenditure and substrate storage, the TGFβ-PDE3b axis provides critical feedback inhibition to prevent lipotoxicity, and the STING-mtROS axis integrates organelle stress with immune activation. Their balanced interaction facilitates disease tolerance (e.g., coordinated fuel supply and inflammation buffering in the subacute phase), whereas their dysregulation (e.g., sustained STING activation coupled with insulin resistance) propels the transition to metabolic exhaustion and organ failure. The stage-specific activity of these pathways offers a molecular basis for the ‘Stage-Specific Adaptation’ hypothesis.

This stage is dominated by the pro-lipolytic arm of ‘Bidirectional Immunometabolic Crosstalk’, embodying the early rapid response mode of the ‘Stage-Specific Adaptation’ hypothesis. Its core feature is the host’s initiation of the fastest defense mechanisms to counter the sudden life-threatening insult. Although fat mobilization is initiated, the drastic breakdown of skeletal muscle protein via the ubiquitin-proteasome pathway becomes dominant (20), providing the most rapid energy substrates for the brain and immune cells. Immunologically, a pro-inflammatory response holds absolute sway, with neutrophils and M1 macrophages infiltrating adipose tissue and releasing high levels of TNF-α, IL-1β, and IL-6 (105), They directly ‘accelerate’ energy mobilization via the IL-1β→ERK-HSL pathway (15).The metabolic-immune profile of this phase represents an adaptive program designed for a rapid response to a life threat (18), Consequently, clinical management focuses primarily on fluid resuscitation, source control, and antibiotic administration (106), rather than sophisticated metabolic support.

This stage is the core manifestation of the ‘Metabolic Defense Priority’ hypothesis, where fat mobilization and protein sparing peak, and bidirectional immunometabolic crosstalk achieves a precise balance. Once the acute threat is initially controlled, the body switches to an ‘endurance race’ mode, and the function of the adipo-immune-metabolic axis peaks, exhibiting refined regulation. Metabolically, lipolysis reaches its zenith and becomes the central energy source, coupled with significantly enhanced hepatic ketogenesis (17), Together, they execute the ‘Metabolic Defense Priority’ strategy, aiming to maximize protein sparing and protect muscle function (18). Concurrently, brown adipose tissue is activated, consuming FFAs via UCP1-mediated thermogenesis to maintain body temperature. Immunologically, the landscape undergoes profound remodeling, shifting towards M2 macrophages, Tregs, and Bregs, with rising levels of anti-inflammatory cytokines like IL-10 and TGF-β (17), This exemplifies precise ‘Bidirectional Immunometabolic Crosstalk’: while M1/IL-1β signaling persists, the Breg/TGF-β → PDE3b-cAMP axis is activated as a crucial ‘brake’ to prevent runaway lipolysis. Therefore, this phase represents the ‘golden window’ for clinical nutritional support and metabolic intervention, requiring the provision of nutrients suitable for fat oxidation and aggressive protein supplementation to prevent excessive muscle breakdown (107, 108).

The transition mechanisms outlined in the ‘Stage-Specific Adaptation’ hypothesis fail, and the adipo-immune-metabolic axis collapses into a vicious cycle of ‘metabolic exhaustion - immune paralysis’. If sepsis persists unresolved, the body enters a state of exhaustion, and axis function breaks down. Metabolically, adipose tissue exhibits severe lipolytic failure due to insulin resistance and direct inflammatory suppression (109, 110).Accompanied by hypertriglyceridemia (111).Widespread mitochondrial dysfunction leads to disordered energy metabolism, forcing the body back to a reliance on the catastrophic breakdown of skeletal muscle protein, ultimately resulting in severe cachexia and multiple organ failure (20). Immunologically, this stage is characterized by immune paralysis, featuring lymphocyte apoptosis, exhaustion, and anergy (4)Treatment at this stage is extremely challenging and necessitates aggressive exogenous intervention, including meticulous metabolic (e.g., insulin therapy), targeted nutritional support (e.g., specific lipid supplementation and adequate protein), and attempted immunomodulation, aiming to break this lethal vicious cycle (107, 112). The specific process can be seen in the Figure 4.

Despite our growing understanding of the adipo-immune-metabolic axis in sepsis, significant blind spots remain in the current knowledge of its key cellular subtypes. In the B-cell domain, the dynamic changes of subsets like B1, B2, and B10 across different septic stages and their specific roles in regulating lipolysis are not clearly defined; The functional differences between tissue-resident macrophages (e.g., Lipid-Associated Macrophages) and monocyte-derived infiltrating macrophages in sepsis have not been distinguished. The pathological significance of their spatial distribution (e.g., whether they form “crown-like structures”) still requires validation (113).The impact of adipocyte depot specificity (subcutaneous vs. visceral) and preadipocyte differentiation fate on long-term sepsis outcomes currently lacks direct evidence. Furthermore, the traditional M1/M2 dichotomy is inadequate to cover the full functional heterogeneity of macrophages. The roles of novel macrophage subtypes revealed by single-cell sequencing (e.g., lipid-associated macrophages or other metabolically regulatory states) in sepsis have not been sufficiently elucidated (113).The spatiotemporal dynamics—specifically, the spatial distribution (e.g., forming “crown-like structures” around dead adipocytes vs. diffuse infiltration) and temporal evolution of different macrophage subtypes/states across sepsis stages (114) require more research. Most critically, how sepsis influences the differentiation fate of adipose precursor cells, thereby determining the long-term plasticity and function of adipose tissue (115), remains largely unknown.

Current research confirms that the functions of key metabolites are highly context-dependent, yet a complete framework to guide precise intervention is still lacking. For instance, the groundbreaking study by Youm et al. demonstrated that the ketone body β-hydroxybutyrate can directly inhibit the NLRP3 inflammasome by acting as an HDAC inhibitor, suggesting ketone bodies play a crucial role as inflammatory ‘brake’ signals beyond their classical function as an alternative fuel (24). However, this protection is not universal. Research by Wang et al. further revealed that a ketogenic state, which is protective in bacterial infections, might be detrimental in viral infections (18). These findings highlight that ketone bodies have transcended their classic role to become important signaling molecules. While they can exert potent organ protection via HDAC inhibition and NLRP3 blockade ( (24), extremely high blood ketone levels are also a marker of poor prognosis, indicating a narrow ‘therapeutic window’ (100). The type of fatty acid dictates its biological effect: Saturated fatty acids can drive inflammation and lipotoxicity via pathways such as TLR4 activation (35). Among unsaturated fats, Omega-6 fatty acids (e.g., arachidonic acid) are primary precursors for potent pro-inflammatory eicosanoids (e.g., prostaglandins, leukotrienes). In contrast, Omega-3 fatty acids (e.g., EPA, DHA) actively promote inflammation resolution by serving as substrates for pro-resolving mediators (e.g., Resolvins, Protectins) and by activating anti-inflammatory pathways such as those mediated by PPARs (116).The role of adenosine is most contentious: locally and early, it exerts powerful anti-inflammatory and organ-protective effects via the A2A receptor (117), However, at the systemic level and in later stages, sustained adenosine signaling becomes a key mediator driving immune paralysis and secondary infections (4), making antagonism of its signaling (e.g., using A2A receptor antagonists) a potential strategy to restore immune function.

The lack of clarity regarding precise mechanistic regulation is further amplified by the disparities between animal models and human sepsis, hindering the translation of basic research findings to the clinical setting.

Differences in Metabolic Pathways:1.BAT Disparity: Adult mice possess substantial amounts of highly active brown adipose tissue, which is a core organ for maintaining body temperature in response to cold and environmental challenges. In contrast, the amount and activity of BAT in healthy adult humans are significantly lower than in mice and decline with age and obesity. Human thermoregulation relies more on skeletal muscle, vasoconstriction, and potentially beige adipose tissue. Consequently, the universality and importance of BAT-driven protective thermogenesis pathways, discovered in murine models, may be overestimated for human sepsis. 2. Basal Metabolic Rate and Response Dynamics: Mice have an extremely high metabolic rate, leading to more intense and rapid metabolic responses to starvation and stress. Humans, however, have a relatively lower metabolic rate and a more protracted disease course. Therefore, the rapid and drastic metabolic switches observed in mice may occur more slowly and gradually in humans. Disease stage definitions based on murine timelines may not accurately correspond to human disease progression. Furthermore, the reliance on and regulation of alternative fuels like ketone bodies likely differ between species.3. Lipoprotein Profiles: The lipoprotein profile of mice differs from that of humans. Consequently, mechanisms underlying hypertriglyceridemia and adipose tissue lipolytic failure identified in mouse models may not fully reflect the reality of lipid disorders in human sepsis.

Differences in Infection Models:1. Limitations of Standard Models: Although the mouse CLP model is considered the ‘gold standard’, it still represents a standardized, focal peritoneal infection. The LPS model provides only a single, potent inflammatory stimulus and does not recapitulate the natural process of bacterial proliferation and clearance. Therefore, mechanisms of adipo-immune-metabolic regulation discovered in a single model might only be applicable to a specific subset of human sepsis (e.g., abdominal origin for CLP) and cannot be generalized to sepsis stemming from other infection sources.2. Lack of Clinical Comorbidities: Experimental mouse models typically use young, genetically identical animals without comorbidities. Human patients, however, often present with multiple underlying conditions. A strategy proven to improve survival by enhancing lipolysis in healthy young mice might trigger catastrophic lipotoxicity and potentially accelerate mortality in septic patients with pre-existing conditions like diabetes and dyslipidemia.

In studies investigating the adipo-immune-metabolic axis in sepsis, the vast majority of foundational experiments conventionally use young male mice by default, aiming to circumvent variability associated with the female estrous cycle. However, this practice creates a significant disconnect from clinical reality and risks developing therapeutic strategies that are ineffective or even harmful for specific demographic groups, such as women and the elderly.

1. Sex-Specific Differences: Sex hormones play a crucial role. Estrogen may confer protection mediated by Estrogen Receptor α (ERα). Specifically, 17β-estradiol possesses immunomodulatory and vasoprotective properties and may enhance both innate and adaptive immunity via ERα signaling (118, 119), On the other hand, the influence of the sex chromosome complement is often entirely overlooked. Females, due to the X-chromosome immune gene dosage effect, may possess a more robust and diverse immune response capability (120).2. The Age Dimension: The standard use of young models fails to recapitulate key features of the aged host. This includes immunosenescence—characterized by the coexistence of inflammaging and T cell exhaustion (112) and metabolic aging, featuring dysfunctional adipose tissue and attenuated mitochondrial function (121).

These model limitations constrain the depth of mechanistic investigation, and the resultant insufficient mechanistic understanding directly hinders clinical translation, thereby creating a vicious ‘research-translation gap’.

Sepsis is not a single disease but a highly heterogeneous syndrome driven by different infection sources, pathogens, host genetic backgrounds, and comorbid states (2, 122).This heterogeneity is the primary obstacle to clinical translation.1. The Dynamic Disease Course vs. Timing of Intervention: The progression of sepsis is dynamic, and the function of the adipo-immune-metabolic axis evolves from compensation to decompensation. According to the ‘Stage-Specific Adaptation’ hypothesis, patients have distinct characteristics in the acute, subacute, and chronic phases. However, it is clinically challenging to precisely determine a patient’s current stage, making correctly timed interventions difficult.2. The Double-Edged Sword Effect of Metabolic Interventions and Safety Concerns: Targeting metabolic pathways is attractive but fraught with risk, as metabolites themselves are pleiotropic and concentration-dependent. Supplementing ketone bodies may be protective early on by inhibiting HDAC and the NLRP3 inflammasome, but in later stages, it could exacerbate immune paralysis and acidosis. Driving lipolysis can supply energy, but excessive free fatty acids can induce lipotoxicity, impairing liver and pancreatic function. Adenosine has anti-inflammatory effects early but contributes to immune paralysis in later stages.

Currently, no single or panel of metabolites/adipokines has been validated for bedside use in stratifying, staging, or predicting the prognosis of septic patients. Many potential biomarkers, such as FABP5 identified by Rong Wu et al. as having a pro-inflammatory role in sepsis, remain in the animal model stage, far from clinical verification and application (123), Many molecular targets effective in animal models (e.g., STING, specific B-cell subsets) lack safe and effective agonists or antagonists for human use. Directly modulating these pathways may lead to unforeseen off-target effects. Intervention tools are also severely lacking. Currently, there are no safe, specific UCP1 agonists available for human use (124). Non-specific drugs (e.g., β3-adrenergic receptor agonists) can cause serious cardiovascular side effects, such as tachycardia and hypertension (125, 126).

Pre-existing adipose tissue inflammation and insulin resistance may predispose patients to a more severe septic course. The diminished anti-inflammatory buffering capacity can exacerbate the early cytokine storm, while impaired metabolic flexibility may disrupt the protective “Metabolic Defense Priority” response, increasing the risk of lipotoxicity and accelerated catabolism. Clinically, this heterogeneity makes it difficult to predict which obese patients will develop these maladaptive responses (122).

The foremost clinical controversy is the repeated observation of an “obesity paradox,” where overweight and moderately obese septic patients often exhibit lower mortality rates than their normal-weight counterparts. This starkly contradicts the expected pathophysiological harm, creating a significant challenge in prognostication and therapeutic strategy. The paradox is likely explained not by obesity being protective per se, but by confounding factors such as greater energetic reserves, differences in adipokine profiles, and the fact that BMI is a crude measure that does not differentiate between fat distribution, muscle mass, or the severity of underlying metabolic dysfunction. This paradox underscores a critical gap in our ability to risk-stratify septic patients based on simple anthropometrics and highlights the need for biomarkers that reflect functional adipose and immune health rather than mere mass (5).

In summary, the main clinical challenge lies in integrating the dual narrative of obesity in sepsis: it can be a source of pathological priming yet also associate with surprising resilience. This necessitates a shift from viewing obesity as a uniform risk factor toward personalized assessments that account for metabolic health, body composition, and functional immune-metabolic responses to tailor prognostication and therapy.

In conclusion, for the key pathways discussed in this review—such as the STING-ER Stress-mtROS Axis and the TGFβ-PDE3b-cAMP Axis—the core challenge for clinical translation lies in achieving ‘Tissue-Specific Regulation’. The unresolved key problem is how to inhibit pathological effects while preserving physiological function.