Pneumonia is a lung infection caused by bacteria, viruses, or other microorganisms, primarily affecting the alveoli. Among these, bacterial and viral pneumonia are the most common. Recent studies have linked both forms to trained immunity. S. pneumoniae is the leading cause of community-acquired pneumonia and remains a life-threatening pathogen (138). Infection with S. pneumoniae induces histone H3K4me2 modifications in respiratory epithelial cells, which persist for at least nine days after bacterial clearance by antibiotics, these epigenetic changes result in altered cellular metabolism and lysosomal transport, enhancing bacterial adhesion and promoting secondary infections (28). Similarly, LPS exposure induces long-term phenotypic changes and trained immunity in lung-resident macrophages, this response provides significant protection against S. pneumoniae but offers limited defense against SARS-CoV-2 (23). In the context of viral infections, influenza virus has been shown to induce trained immunity in alveolar macrophages, conferring long-lasting anti-tumor immunity. This effect persists for at least 30, 60 and even 120 days after infection, significantly inhibiting pulmonary tumor growth. The protective mechanism depends on IFN-γ, NK cells, and epigenetic and metabolic reprogramming (139). SARS-CoV-2 infection–induced pneumonia may also be associated with trained immunity. SARS-CoV-2, the causative agent of COVID-19, is transmitted primarily through respiratory droplets, similar to influenza. Clinical manifestations range from asymptomatic infection and mild upper respiratory symptoms to severe pneumonia with respiratory failure and death (140). By enhancing innate immune responses via epigenetic and metabolic reprogramming, trained immunity may provide early antiviral protection in COVID-19, reducing viral replication and inflammation. Interestingly, S.aureus skin infections can induce eosinophil-mediated innate immune memory via IL-33 and C5a, leading to bone marrow reprogramming. This alters mature myeloid cell function and exacerbates lung inflammation in systemic allergy models (141). Future research should explore the role of trained immunity in pneumonia of different etiologies. Understanding how epigenetic and metabolic reprogramming modulate immune responses may offer new therapeutic strategies for the prevention and treatment of pneumonia.

Studies on trained immunity in gastrointestinal inflammation primarily focus on intestinal inflammation and pancreatitis. During maternal infection, elevated IL-6 crosses the placenta and acts on fetal intestinal epithelial cells. This proinflammatory signal induces epigenetic changes, promoting Th17 differentiation and expansion, enhancing resistance to intestinal infection but increasing risk of intestinal inflammation. These modifications include lasting changes in chromatin accessibility and transcriptional programs in intestinal epithelial stem cells, which may persist into adulthood (9). The integrity of the intestinal barrier is essential for maintaining host health. When this barrier is disrupted, commensal bacteria can translocate systemically and trigger inflammatory responses. One study showed that barrier damage allows Enterococcus faecalis, a member of the gut microbiota, to migrate to the bone marrow, where it induces trained immunity in myeloid progenitors via the Mincle receptor. This response may be either protective or detrimental, depending on context. Mice treated with heat-killed E. faecalis exhibited less weight loss and higher survival rates following infection with Candida albicans or influenza virus. However, in inflammatory models, trained immunity may aggravate disease, for instance, in dextran sulfate sodium-induced colitis, Mincle-deficient mice displayed milder pathological symptoms, indicating that trained immunity contributes to the pathology of inflammatory bowel disease (142). These findings offer novel insights into the complex interplay among the gut microbiota, immune system, and inflammation, which may inform new immunotherapies and improve understanding of diseases related to increased gut permeability. Moreover, inflammation is a major risk factor for pancreatic ductal adenocarcinoma. During pancreatitis, oncogenic kirsten rat sarcoma viral oncogene homolog (KRAS) mutations accelerate tumor progression. Notably, even transient inflammatory episodes can sensitize pancreatic epithelial cells to subsequent KRAS-driven transformation. This adaptive response involves persistent transcriptional and epigenetic reprogramming, enabling rapid activation of ADM during recurrent inflammation, thereby limiting tissue damage (31).

Allergic diseases have emerged as a global public health challenge, with rising prevalence imposing a significant socioeconomic burden. Chronic conditions such as allergic asthma, allergic rhinitis, food allergy, atopic dermatitis, and anaphylaxis impair quality of life and contribute to substantial healthcare expenditures (143). The treatment of allergic diseases has shifted from symptomatic control to a strategy combining immune modulation and targeted therapy. Allergen immunotherapy and biologics/monoclonal antibodies are emerging as key options, offering not only symptom relief but also the potential to induce tolerance and modify disease progression (144). Traditionally, allergic responses were thought to be driven primarily by the adaptive immune system, particularly IgE-mediated reactions. However, recent studies suggest that aberrant activation of the innate immune system also plays a critical role in the onset and progression of allergic diseases. Notably, trained immunity has been proposed as a potential mechanism. Early-life innate immune hyperactivation has been observed in allergic children, while asthmatic children display immune dysregulation characterized by reduced IFN-γ production and ILC2 expansion in response to rhinovirus or LPS stimulation (145). These phenomena may result from allergen- or virus-induced trained immunity, leading to long-lasting immune dysfunction and increased susceptibility to allergic inflammation. In allergic mouse models or patients with house dust mite–induced asthma, macrophages exhibit exaggerated production of TNF-α, CCL17, leukotrienes, PGE2, and IL-6 upon stimulation. This process depends on the TNF signaling pathway and involves 2-hydroxyglutarate accumulation and KDM1A-mediated demethylation, identifying potential therapeutic targets for allergic asthma (146). In a study using single-cell RNA sequencing, Li et al. (2022) identified an inflammatory neutrophil subpopulation in allergic asthma with molecular features of innate immune memory. Asthma may reprogram neutrophil populations, leading to expansion of G-CSFR⁺FcγRIIb⁺ neutrophils, which may represent memory-like neutrophils and serve as novel targets in neutrophil-dominant asthma (147). Two experimental studies further indicate that infection-induced reprogramming of innate immune cells may prevent asthma development. In one model, murine gammaherpesvirus 4 lung infection suppressed HDM-induced experimental asthma (HDM-EAA) through two mechanisms (1): downregulating costimulatory molecules CD80/CD86 on migratory dendritic cells, impairing Th2 activation; and (2) inducing embryonically derived alveolar macrophage apoptosis, while recruiting bone marrow–derived monocytes that differentiated into anti-inflammatory macrophages with sustained IL-10 production. These monocytes persisted in lung parenchyma for up to 28 days, suggesting the formation of long-term trained immunity via epigenetic reprogramming (148). In another study, neonatal peritoneal inoculation of EV-A71 virus induced glycolysis-dependent proinflammatory training in macrophages, these trained cells displayed prolonged (≥21 days) secretion of IL-6, TNF-α, and CCL17 and aggravated HDM-EAA. Adoptive transfer of EV-A71–trained BMDMs into naïve mice enhanced Th2 inflammation and allergen-specific IgE levels. Further analysis revealed temporal polarization of these trained macrophages: an early Th2-skewed phase, followed by a classical proinflammatory phenotype. This process was abrogated by the glycolysis inhibitor 2-deoxy-D-glucose (2-DG), confirming the central role of metabolic reprogramming in immune training (148). In addition to immune cells, other cell types such as airway epithelial cells may also contribute to trained immunity in allergic disease. Airway epithelial cells play key roles in asthma pathogenesis and respond to microbial compounds, allergens, and pollutants. Single-cell transcriptomic profiling in chronic rhinosinusitis revealed basal epithelial cells with Th2 cytokine memory characteristics, suggesting that epithelial cells may act as allergen memory reservoirs. Notably, microbial imprinting in the respiratory epithelium may dynamically regulate future immune responses by modulating cytokine output such as IL-1b and IL-8 (149), resembling the trained immunity seen in macrophages. Beyond asthma, food allergy has gained attention due to rising prevalence in Western countries. Trained immunity may influence food allergy development by reshaping the DCs–T cell interaction network and promoting Th2-skewed responses. In peanut-allergic infants, PBMCs exhibit excessive TNF-α production upon nonspecific stimulation. This hyperresponsive immune phenotype persists in allergic adolescents, marked by increased circulating dendritic/monocyte populations and elevated secretion of IL-6, IL-1β, and TNF-α upon LPS challenge (150). In summary, trained immunity provides a novel conceptual framework for understanding allergic diseases. It offers promising avenues for early intervention and the development of targeted therapies based on epigenetic and metabolic reprogramming of immune and non-immune cells.

Trained immunity exerts long-lasting effects on cell function through epigenetic regulation and metabolic reprogramming ( Table 1 ) and shows bidirectional roles in immune regulation: disease-induced trained immunity can, in turn, exacerbate pathology, creating a self-reinforcing vicious cycle. Accordingly, new prevention and treatment strategies should aim to enhance its beneficial effects and intervene early in harmful reprogramming. These findings offer new perspectives for targeted therapies, though underlying mechanisms and clinical applications require further investigation.

Unlike classical adaptive immune memory, which relies on antigen receptor gene rearrangement, trained immunity operates through epigenetic modifications that increase chromatin accessibility at gene loci following primary stimulation. Upon secondary exposure, this enables rapid transcriptional activation of genes involved in inflammation, antimicrobial defense, and stress responses, without altering the underlying DNA sequence. In eukaryotic cells, genomic DNA is compacted into nucleosomes by structural histone proteins, rendering most gene regions inaccessible to transcription factors. Transcription initiation requires nucleosome displacement at promoter regions to expose regulatory elements, allowing transcription factor binding and recruitment of RNA polymerase (74, 151). Chromatin accessibility is regulated by epigenetic modifications, including histone methylation, histone acetylation, and DNA methylation, among which histone modifications are well-established drivers of inflammatory memory.

For instance, IFN-γ stimulation of HeLa cells increases H3K4me2 at the DRA promoter, which persists after withdrawal and across multiple cell cycles (152). Similarly, Primary S. pneumoniae infection induces H3K4me2 enrichment in respiratory epithelial cells (28). In murine EpSCs, inflammatory signals activate STAT3, which translocates to the nucleus and binds H3K4me1- and H3K27ac-marked chromatin, establishing an epigenetically poised state (6). Histone modifications also mediate trained immunity in vitro: in β-glucan–trained monocytes, promoters and enhancers of inflammation- and immunity-related genes gain H3K4me1, H3K4me3, and H3K27ac marks, maintaining transcriptionally active chromatin and enabling enhanced cytokine production upon restimulation (153). In contrast, LPS-induced tolerance is marked by attenuated transcriptional responses upon secondary LPS challenge. Tolerized macrophages fail to deposit active histone marks (e.g., H3K4me1, H3K27ac) at the promoters of tolerant genes, resembling the immune paralysis seen in sepsis, interestingly, β-glucan injection can reverse LPS-induced tolerance by restoring H3K27ac deposition and reactivating macrophage function. The BET family small-molecule inhibitor I-BET151 (GSK1210151A) is primarily used to suppress inflammation and for anti-tumor purposes. Evidence supporting its role as an mTOR inhibitor remains largely preclinical (154). Its limitation is the inability to reverse effects once tolerance is established, making it potentially ineffective in monocytes that have already experienced inflammatory activation. Nonetheless, β-Glucan and I-BET151 offer a theoretical basis and potential targets for developing immunomodulatory strategies against tolerance-related diseases such as sepsis (155).Similarly, BCG-induced trained immunity involves the remodeling of H3K9me3, H3K4me3, and H3K27ac histone marks (156, 157). Genome-wide H3K27ac ChIP-seq analyses revealed enrichment of this active histone modification at genes encoding the oxLDL receptor, a marker of atherosclerosis, and other inflammation-related genes in monocytes following BCG vaccination, suggesting transcriptional activation of these pro-atherogenic genes (157). The trained phenotype and gene-specific H3K4me3 enrichment persist for at least three months and up to one year post-vaccination (156). In atherosclerosis, oxLDL-induced trained immunity in monocyte-derived macrophages also promotes H3K4me3-mediated activation of genes involved in inflammation, chemotaxis, and foam cell formation (102). The histone methyltransferase inhibitor 5′-deoxy-5′-methylthioadenosine (MTA) reverses histone methylation and prevents chromatin remodeling, effectively abrogating the trained immunity phenotype induced by oxLDL. This finding offers a potential strategy for modulating inflammation in atherosclerosis (102, 158, 159). Histone methylation is regulated by a complex network of enzymes, including histone methyltransferases and demethylases, which modulate gene expression by adding or removing methyl groups at specific histone residues (160). The broad-spectrum Jumonji histone demethylase inhibitor JIB-04 reduces trained immunity by modulating the repressive H3K9 mark (161). To date, evidence for its therapeutic effects as an mTOR inhibitor is limited to pharmacological and animal studies, with no established clinical trials (162). Similarly, histone acetylation, catalyzed by histone acetyltransferases (HATs) and HDACs, plays a regulatory role in trained immunity (163). Class I/II HDAC inhibitors used in cancer therapy, including vorinostat (suberoylanilide hydroxamic acid) and others such as ITF-2357 and SAHA, have shown anti-inflammatory effects in inflammatory diseases (164). Vorinostat is not a classical mTOR antagonist but has been used clinically in combination with mTOR inhibitors or to modulate mTOR-related pathways (165). ITF-2357 has advanced in multiple clinical trials, with recent progress in some indications; for example, a phase II study (NCT00928707) showed that SAHA combined with hydroxyurea was well tolerated (166). In trained monocytes, elevated NAD⁺/NADH ratios and lactate accumulation modulate HDAC activity and influence downstream gene expression (167, 168). In THP1 cells from rheumatoid arthritis patients, LPS stimulation increases H3 and H4 acetylation at the CCL2 promoter, which can be reversed by TNF-α inhibitors such as etanercept and adalimumab (77). Besides. In SLE, hydroxychloroquine suppresses trained immunity by inhibiting H3K27ac and H3K4me3 modifications at inflammation-related genes (169).

Although less studied, DNA methylation also contributes to trained immunity. In human embryonic kidney cells, persistent TNF-α stimulation activates TET enzymes, which demethylate the CALCB enhancer and IL32 promoter, establishing a trained-like transcriptional state (170). In trained macrophages, long-term changes in gene expression are associated with altered DNA methylation patterns, particularly 5-methylcytosine (171). Whether modulating DNA demethylation can similarly correct inflammatory dysregulation remains an open question. Certain metabolites, including fumarate, α-ketoglutarate (α-KG), and succinate, regulate TET family demethylase activity, although direct DNA methylation changes have not been consistently observed in trained immunity models (50, 172). In conclusion, both histone modifications and DNA methylation serve as central regulatory mechanisms in trained immunity. Targeting these epigenetic pathways may offer a precise strategy to control excessive trained immune responses and alleviate inflammation-related diseases.

Trained cells not only enhance gene expression via epigenetic remodeling but also reprogram metabolic pathways to meet elevated energy demands. In quiescent states, immune cells primarily rely on oxidative phosphorylation (OXPHOS) and fatty acid oxidation (FAO) for energy. Upon activation, monocytes and macrophages shift toward glycolysis, lipid metabolism, and amino acid metabolism to support proinflammatory functions.

Glycolysis is intrinsically upregulated in the trained immunity process, independent of external stimuli (173). β-glucan-trained monocytes exhibit increased glucose uptake and lactate production, reflecting enhanced glycolytic flux (110). Similarly, BCG-trained monocytes show marked increases in glycolysis and glutaminolysis. Pharmacological or genetic inhibition of key glycolytic enzymes impairs trained immunity (109). The Akt/mTOR/HIF1α axis orchestrates this metabolic reprogramming by modulating HIF1α activity and glycolytic enzyme expression (174, 175). Similarly, inhibition of Akt/mTOR/HIF1α signaling suppresses β-glucan and BCG-induced trained immunity (109, 110). 2-DG, and metformin are established mTOR inhibitors. Studies using rapamycin-loaded high-density lipoprotein nanoparticles (mTORi-HDL) showed that they prevent post-transplant monocytes from developing trained macrophages upon exposure to alloantigens and HMGB1. mTORi-HDL accumulates in the graft and is taken up by myeloid cells; three intravenous injections reduce inflammatory cell infiltration, prolong graft survival, and suppress macrophage proinflammatory responses (176). A phase I trial determined that 2-DG combined with docetaxel at 63 mg/kg/day is clinically tolerable, showing manageable toxicity and modest antitumor activity (177). Experimental studies indicate that metformin activates Adenosine 5’-monophosphate (AMP)-activated protein kinase and/or inhibits mTORC1 via pathways including regulated in development and DNA damage responses 1 (178). Besides, mTOR activation occurs on lysosomal surfaces (179), where trained immunity is marked by lysosomal gene activation (155). Lysosomotropic agents like chloroquine and hydroxychloroquine impair lysosomal function and robustly inhibit trained immunity (169). The mTOR metabolic signaling pathway regulates HIF-1α activity. In HIF-1α knockout mice, the absence of HIF-1α signaling results in defective epigenetic programming and impaired production of proinflammatory cytokines (180). Succinate, a glycolytic intermediate, stabilizes HIF-1α by inhibiting its degradation, thereby sustaining chromatin accessibility. As an epigenetic modulator, succinate also suppresses histone and DNA methylation, supporting prolonged expression of glycolysis-related genes (109, 181). Collectively, the Akt/mTOR/HIF-1α axis coordinates metabolic and epigenetic reprogramming to drive and maintain trained immunity, representing a potential target for therapeutic intervention. In β-glucan–trained cells, the PPP is also upregulated. However, PPP inhibition has minimal impact on trained immunity, suggesting a limited role in this process (50). Several metabolism-targeting drugs are used not only in immune-related conditions such as cancer but also in inflammation research. 2-DG and dichloroacetic acid inhibit glycolysis and effectively block trained immunity induction in vitro (182). Glyceraldehyde 3-phosphate dehydrogenase (GAPDH), a key glycolytic enzyme, also exerts immunomodulatory effects. Through metabolic reprogramming, GAPDH regulates trained macrophages and reduces allergen-induced airway inflammation in murine models of allergic asthma. Mechanistically, GAPDH promotes the transition of M2 macrophages to a proinflammatory M1 phenotype, thereby suppressing Th2 cell activation and exerting anti-inflammatory effects (183).

In addition to glycolysis, enhanced glutamine metabolism is a critical metabolic feature of trained immunity. Inhibition of glutaminase significantly suppresses β-glucan and BCG-induced trained immunity (50, 109). Both stimuli promote glutaminolysis through HIF-1α activation, increasing intracellular α-KG levels. α-KG enters the TCA cycle, supplying energy to support cell function (184). Supplementation with glutamine, arginine, and tryptophan further enhances immune responses in trained cells. These amino acids contribute to DNA synthesis, redox homeostasis, protein translation, and nitric oxide production, thereby improving innate immune cell function (185, 186).

Lipid metabolism also plays a vital role in trained immunity. β-glucan-induced training depends on mevalonate, a key intermediate in the cholesterol biosynthesis pathway. Monocytes from patients with HIDS, characterized by mevalonate accumulation, exhibit a trained immunity, like phenotype. Statins, which inhibit mevalonate production, effectively block β-glucan-induced training (50). Additionally, statins and liver X receptor antagonists suppress cholesterol biosynthesis and reduce mevalonate availability, thereby preventing mTOR-mediated trained immunity (96, 187). In addition, fluvastatin, an inhibitor of the rate-limiting enzyme HMG-CoA reductase, also attenuates β-glucan–induced training in vitro (184). In contrast, aldosterone promotes trained immunity by enhancing fatty acid synthesis, an effect that can be blocked by aldosterone receptor antagonists (188). Lipid signaling also regulates PPARs, SREBPs, mTOR, and various lipid mediators, modulating the metabolic state, inflammatory response, and functional output of trained immune cells (189–192). In summary, metabolic reprogramming in trained immunity involves the coordinated regulation of glucose, amino acid, and lipid metabolism. These pathways not only meet the energetic demands of immune activation and inflammation but also shape long-term gene expression through epigenetic mechanisms. Regulation of key signaling pathways such as HIF-1α, mTOR, and Akt, along with control over amino acid and lipid metabolism, provides a promising framework for developing therapeutic strategies against trained immunity–associated hyperinflammatory diseases.

Conventional vaccines primarily rely on the stimulation of antigen-specific T and B cells to elicit adaptive immune responses against specific pathogens. With growing insights into trained immunity, several classical vaccines have been shown to induce long-lasting functional reprogramming of innate immune cells, resulting in enhanced nonspecific protection upon subsequent infections. MV130, a sublingually administered inactivated polybacterial vaccine composed of 90% Gram-positive and 10% Gram-negative bacteria (193), rapidly induces DCs to secrete proinflammatory cytokines such as IL-6, TNF-α, and IL-1β. It also promotes Th1 and Th17 responses against both intracellular and extracellular pathogens, followed by IL-10 production to limit excessive inflammation (194). In a virus-induced asthma model, MV130 demonstrated nonspecific protection by enhancing neutrophil-mediated immunomodulation, thereby preventing asthma exacerbation during rhinovirus infection (195). A phase III randomized placebo-controlled trial involving 120 wheeze-prone children showed that daily sublingual administration of MV130 for six months significantly reduced wheezing episodes, symptoms, and medication use, with protective effects lasting another six months (196). In addition, allergen–mannan conjugate vaccines have been shown to reprogram monocytes into tolerogenic DCs via epigenetic and metabolic remodeling, supporting the induction of trained tolerance in allergen immunotherapy (197).

Vaccines typically consist of antigens and adjuvants. Due to the weak immunogenicity of antigens alone, adjuvants are essential to amplify and sustain immune responses (198). Many live-attenuated or inactivated vaccines act as self-adjuvants by stimulating PRRs. BCG is a classic example. BCG induces trained immunity via chromatin remodeling and metabolic rewiring, enhancing the functionality of monocytes and neutrophils. Upon secondary exposure to unrelated pathogens, trained cells exhibit increased cytokine production and antimicrobial capacity (199–201). A phase III clinical trial (NCT03296423) demonstrated that BCG vaccination enhances inflammatory cytokine secretion via epigenetic modulation, improving resistance to respiratory viral infections in the elderly. BCG also boosts influenza-specific antibody responses and modulates cytokine production in adults (202). Vaccination with AS03-adjuvanted influenza vaccine increases chromatin accessibility in innate immune cells and enhances interferon signaling, conferring broad antiviral resistance in vitro against dengue and Zika viruses. These findings suggest that AS03 functions not only as an adaptive immunity enhancer but also as a trained immunity inducer, offering broad-spectrum antiviral potential (203).

At the beginning of the COVID-19 pandemic in 2019, the absence of specific immunity and vaccines drew attention to BCG vaccination. It was considered a potential source of heterologous protection through cross-reactivity with viral antigens, nonspecific activation of lymphocytes, and enhancement of innate immunity via trained immunity. Animal studies have suggested that BCG reduces viral load and mitigated immunopathology in SARS-CoV-2 infection models (204). Unreviewed observational studies also reported lower COVID-19 mortality in countries with universal BCG vaccination, such as South Korea and Japan, compared with countries without such practice, including Italy and the United States (205). However, these associations are indirect and may be confounded by genetic, environmental, vaccine strain, and policy factors. Large-scale studies have generally shown that BCG does not prevent COVID-19 infection or severe disease, and some even indicated an increased risk of symptomatic COVID-19 among vaccine recipients (206–208). The role of BCG in COVID-19 protection remains uncertain, with several unresolved issues. For example, studies in children showed stronger effects when vaccination occurred before 9 months of age, but the influence of age at vaccination on COVID-19 protection is unknown (209). Trained monocytes gradually lose their enhanced cytokine responses, and the duration of heterologous immunity remains unclear (210). It is also unknown whether patients receiving intravesical BCG for bladder cancer gain protection against COVID-19 (211). Overall, the use of BCG in COVID-19 highlights the potential of trained immunity in broad antiviral defense, but faces challenges of inconsistent clinical evidence, uncertain durability, and limited vaccine resources. Other vaccines have also exhibited potential cross-protective effects. Influenza vaccination, by inducing trained immunity, was associated with reduced SARS-CoV-2 infection risk, particularly among healthcare workers and elderly populations (212). Additionally, varicella-zoster virus vaccines may also offer partial protection against COVID-19 (213).

Although trained immunity-based vaccines have not received the same level of global attention as other pandemic countermeasures, multiple clinical trials are ongoing to assess their potential benefits in the context of COVID-19. If proven effective, these vaccines could serve as a proof-of-concept for their use during future pandemics, especially as bridging strategies when pathogen-specific vaccines are unavailable or under development. Given their ability to induce trained immunity, vaccines and vaccine adjuvants should not be viewed solely as passive immunogenic agents. Instead, they represent potential tools for targeted activation of innate immune memory. Through rational antigen design and adjuvant optimization, it may be possible to strategically induce trained immune states to support infection control, immune modulation, and even chronic disease intervention.

Beyond epigenetic and metabolic regulation, advances in nanomedicine have opened new avenues for modulating trained immunity. Nanomaterials not only enable efficient delivery of immunoregulatory agents but can also be engineered to target the bone marrow and myeloid precursors, thereby inducing or suppressing trained immune responses. For example, mTORi-HDL suppress trained immunity in myeloid cells and promote the expansion of anti-inflammatory Ly6C^- macrophages, alleviating graft rejection and prolonging graft survival (214). Combined delivery of CD40 co-stimulation inhibitors using TRAF6i-HDL nanoparticles further disrupts both trained immunity and co-stimulatory pathways in myeloid cells, significantly improving transplant acceptance and inducing long-term immune tolerance (176). In addition, bone marrow–targeted nanomaterials engineered for spleen or inflammation site accumulation (e.g., tumors or atherosclerotic plaques) have demonstrated therapeutic potential. These particles either induce monocyte training in the spleen or reprogram myeloid cells at disease sites toward anti-inflammatory phenotypes (215). Although hepatic accumulation of certain nanomaterials may cause adverse effects, targeted delivery to the spleen and inflammatory foci offers promising immunomodulatory and anti-inflammatory benefits (216). In summary, chemical functionalization of nanomaterials allows the modulation of epigenetic and metabolic programs in myeloid progenitors, enabling the induction or inhibition of trained immunity. This strategy holds therapeutic potential for enhancing antimicrobial defense or suppressing maladaptive immune responses in diseases such as atherosclerosis and rheumatoid arthritis. Apart from nanotechnology, molecular targeting therapies are also being explored. IL-1β and GM-CSF are key mediators of trained immunity (180). Monoclonal antibodies against IL-1β have shown efficacy in inflammation-related cardiovascular diseases, reducing thrombotic events and attenuating trained immune responses following myocardial infarction (117). Anti–GM-CSF antibodies are currently under clinical investigation for the treatment of rheumatoid arthritis and related conditions (217). Moreover, RNA interference offers a novel approach to block trained immunity, associated signaling pathways. However, its clinical translation remains in early stages (218).