Trained immunity is the epigenetically-mediated functional reprogramming of innate immune cells (monocytes, macrophages, NK cells) and their bone marrow progenitors following initial pathogen or inflammatory exposure, enabling enhanced responsiveness to subsequent unrelated challenges through persistent metabolic reconfiguration (Warburg Effect), chromatin remodeling (H3K4me3, H3K27ac marks), and altered transcriptional landscapes that outlive individual cellular lifespans. This represents adaptive memory in a system previously considered incapable of learning.
Think of trained immunity as a fire station that rewires itself after a major blaze. The first big fire (pathogen exposure) doesn't just teach the firefighters — it physically renovates the station. They install more fuel pumps (GLUT1 upregulation), rewire the alarm system to be more sensitive (epigenetic modifications at inflammatory gene loci), and stock the trucks with pre-loaded equipment (metabolic reprogramming). Even when the original crew retires, the new recruits inherit these renovations because the station blueprints themselves have been rewritten in the bone marrow "headquarters."
The station now responds faster and stronger to ANY alarm — not just fires similar to the original. This is brilliant when facing real emergencies (infections), but problematic if the alarm becomes too sensitive. A small kitchen fire (minor inflammation) now triggers a full battalion response, causing collateral damage to the neighborhood (metaflammation, chronic low-grade inflammation). The station has learned, but it hasn't learned nuance — it's just learned to react bigger and faster, which can be a double-edged sword.
Trained immunity operates through interconnected epigenetic, metabolic, and transcriptional reprogramming:
graph TD
A["Pathogen exposure: β-glucan, BCG, LPS"] --> B["PRR activation: Dectin-1, TLR4"]
B --> C["AKT/mTOR/HIF-1α pathway activation"]
C --> D[Metabolic shift to aerobic glycolysis]
D --> E[Increased GLUT1 expression]
E --> F[Enhanced glucose uptake & lactate production]
F --> G[Acetyl-CoA accumulation]
G --> H["Histone acetylation: H3K27ac"]
B --> I["Histone methylation: H3K4me3"]
H --> J["Open chromatin at IL-6, TNF, IL-1β loci"]
I --> J
J --> K[Enhanced transcriptional accessibility]
K --> L[Bone marrow progenitor reprogramming]
L --> M["Long-lived trained phenotype: weeks to months"]
M --> N[Faster, stronger response to secondary challenge]
C --> O[IRF5 upregulation]
O --> P[Sustained pro-inflammatory gene expression]
Initial Exposure Phase:
- Pathogen-associated molecular patterns (β-glucan via Dectin-1, BCG via TLR4, oxidized LDL) activate pattern recognition receptors
- AKT pathway → mTOR → HIF-1α stabilization (even under normoxic conditions)
- Shift from oxidative phosphorylation to Aerobic Glycolysis (Warburg Effect)
- GLUT1 transporter upregulation increases glucose uptake 3-5 fold
- Glycolytic flux generates lactate and acetyl-CoA intermediates
Epigenetic Imprinting:
- Acetyl-CoA from glycolysis fuels histone acetyltransferases
- H3K27ac marks deposited at promoters and enhancers of IL-6, TNF-α, IL-1β
- H3K4me3 (trimethylation) marks established at inflammatory gene loci
- DNA methyltransferases activity altered, creating heritable methylation patterns
- Chromatin accessibility increases at immune response elements
Transcriptional Rewiring:
- IRF5 (interferon regulatory factor-5) expression sustained for weeks
- Long non-coding RNAs (lncRNAs) maintain open chromatin architecture
- NF-κB binding sites remain accessible despite stimulus removal
- Metabolic genes (GLUT1, hexokinase-2, lactate dehydrogenase) constitutively elevated
Bone Marrow Progenitor Reprogramming:
- Hematopoietic stem cells (HSCs) and myeloid progenitors acquire trained phenotype
- Methylation patterns transmitted through cell division
- Short-lived monocytes (2-3 day lifespan) continuously replaced with pre-trained cells
- Explains persistence beyond individual cell turnover (3-12 months documented)
Secondary Challenge Response:
- Trained cells produce 2-10 fold more IL-6, TNF-α, IL-1β upon re-stimulation
- Response kinetics accelerated: cytokine production within 2-4 hours (vs. 6-12 hours)
- Enhanced phagocytosis and pathogen clearance
- Cross-protection against heterologous pathogens
Revolutionary Immunological Paradigm:
Trained immunity fundamentally challenges the traditional innate/adaptive dichotomy. Leo Pruimboom emphasized this as critical: the innate immune system is NOT "stupid" — it learns, adapts, and remembers through non-genetic mechanisms. This updates clinical understanding of vaccination (BCG provides non-specific protection), early-life immune education, and chronic inflammatory states.
Clinical Applications:
Protective Trained Immunity:
- BCG vaccination induces trained immunity persisting 3-12 months, reducing all-cause mortality in children by 30-50% through heterologous protection
- β-glucan supplementation (from mushrooms, yeast) can prime innate immunity before high-risk periods (surgery, chemotherapy)
- Early-life microbial exposures (hygiene hypothesis) establish beneficial trained immunity setpoints
Maladaptive Trained Immunity:
- Western diet (high glucose, oxidized LDL, AGEs) chronically activates trained immunity pathways
- Contributes to metaflammation — the cells are "trained" to be hyperinflammatory
- Explains why metabolic syndrome patients show exaggerated inflammatory responses to minor triggers
- immunosenescence paradox: aged individuals show both immune exhaustion AND trained hyperreactivity (context-dependent)
Five Metamodel Integration:
- Metamodel 1 (chronic low-grade inflammation): Maladaptive trained immunity perpetuates inflammatory tone
- Metamodel 3 (immunometabolism): Glycolytic shift is both marker and mechanism
- Metamodel 5 (Evolutionary mismatch): Modern antigens (oxLDL, processed foods) create inappropriate training
Diagnostic Markers:
- Monocyte GLUT1 expression (flow cytometry)
- Histone modification ChIP-seq at inflammatory loci
- Ex vivo cytokine production capacity after LPS stimulation
- lactate:pyruvate ratio in circulating monocytes
- No single clinical test available; research-level assessment
Intervention Strategies:
- Reset maladaptive training: Intermittent fasting (depletes glycolytic substrates), ketogenic states (beta-hydroxybutyrate inhibits NLRP3)
- Support protective training: Controlled cold/heat exposure, β-glucan supplementation, fermented foods
- Metabolic modulation: Target mTOR (rapamycin analogs), HIF-1α stabilization (controversial), restore oxidative phosphorylation
- Timing matters: Early intervention (first 3-6 months of training) more effective than attempting reversal
Patient Populations:
- Autoimmune conditions: Investigate whether disease flares reflect trained immunity in tissue-resident macrophages
- Cardiovascular disease: Atherosclerotic plaques contain trained macrophages perpetuating inflammation
- Type 2 Diabetes: Hyperglycemia sustains trained immunity glycolytic substrate
- Recurrent infections with hyper-inflammatory responses: Consider maladaptive training rather than adaptive deficiency
- Trained immunity persists 3-12 months despite monocyte lifespan of only 2-3 days — effect mediated through bone marrow progenitor reprogramming
- Key epigenetic marks: H3K4me3 (trimethylation) and H3K27ac (acetylation) at inflammatory gene promoters/enhancers
- GLUT1 expression increases 3-5 fold in trained monocytes, enabling sustained Aerobic Glycolysis even under normoxic conditions
- IRF5 (interferon regulatory factor-5) is central transcriptional mediator — remains elevated for weeks post-exposure
- BCG vaccination induces heterologous protection: reduces non-TB infections and all-cause mortality by 30-50% in children
- β-glucan (yeast, mushroom extract) induces trained immunity within 7 days; peak effect at 7-14 days
- Trained cells produce 2-10 fold more IL-6, TNF-α, IL-1β upon secondary challenge (measured by ex vivo LPS stimulation)
- Western diet components (oxidized LDL, high glucose, AGEs) can induce maladaptive trained immunity contributing to chronic low-grade inflammation
- Mechanistic dependence on Warburg Effect: glycolytic intermediates (acetyl-CoA) directly fuel histone acetylation
- First described 2011 (Netea lab); represents paradigm shift away from "innate = non-specific, no memory" dogma
- Can be induced by diverse stimuli: β-glucan, BCG, oxidized LDL, uric acid crystals, Western diet
- immunosenescence involves paradoxical state: some trained immunity pathways hyperactive (inflammaging) while adaptive immunity declines
- immunometabolism — the metabolic basis (glycolytic shift, Warburg Effect) is both mechanism and marker of trained immunity
- Warburg effect — aerobic glycolysis is THE metabolic signature of trained innate cells; lactate and acetyl-CoA drive epigenetic changes
- GLUT1 — glucose transporter upregulated 3-5 fold in trained monocytes; enables sustained glycolytic flux
- IRF5 — transcription factor mediating trained immunity; remains elevated for weeks and maintains inflammatory gene accessibility
- metaflammation — maladaptive trained immunity is a key mechanism perpetuating chronic low-grade inflammation in metabolic disease
- immunosenescence — trained immunity offers partial explanation for "inflammaging" — innate hyperreactivity despite adaptive decline
- epigenetics — histone modifications (H3K4me3, H3K27ac) and DNA methylation are the molecular memory substrates
- HIF-1α — stabilization (via mTOR pathway) drives glycolytic reprogramming central to trained immunity induction
- NF-κB — transcription factor whose binding sites remain accessible in trained cells; enables rapid inflammatory gene activation
- IL-6 — cytokine production increased 2-10 fold in trained cells upon re-stimulation; key readout marker
- TNF-α — similarly amplified in trained immunity; contributes to both protective and pathological responses
- macrophages — long-lived tissue-resident macrophages can maintain trained phenotype for months; relevant in atherosclerotic plaques
- monocytes — primary circulating cell type exhibiting trained immunity; continuously replenished from trained bone marrow progenitors
- NK cells — natural killer cells also exhibit trained immunity (particularly after cytomegalovirus exposure)
- beta-glucan — prototypical training agent from fungal cell walls; induces protective trained immunity via Dectin-1 receptor
- mTOR — mechanistic target of rapamycin; key signaling node linking pathogen sensing to metabolic reprogramming in trained immunity
- lactate — end-product of aerobic glycolysis; accumulation itself can modify immune function and histone lactylation (emerging mechanism)
- chronic low-grade inflammation — maladaptive trained immunity is a mechanistic driver in conditions like metabolic syndrome, atherosclerosis
- hygiene hypothesis — early-life microbial exposures may establish beneficial trained immunity setpoints; absence creates inappropriate immune education
- Western diet — high glucose, oxidized LDL, and AGEs induce maladaptive trained immunity contributing to non-communicable disease burden
- BCG — Bacillus Calmette-Guérin vaccination is most-studied inducer of protective trained immunity; provides heterologous protection beyond TB
- oxidative phosphorylation — metabolic state inversely correlated with trained immunity; restoration (fasting, ketosis) may "untrain" cells
- Acetyl-CoA — glycolytic intermediate that fuels histone acetyltransferases; direct mechanistic link between metabolism and epigenetics
- bone marrow — site of hematopoietic progenitor reprogramming; explains persistence of trained phenotype beyond individual cell lifespan
- Dectin-1 — pattern recognition receptor for β-glucan; canonical pathway for inducing trained immunity
- TLR4 — Toll-like receptor activated by LPS and BCG; alternative training pathway
- atherosclerosis — trained macrophages in plaques perpetuate inflammation; may explain why metabolic factors accelerate CVD
- Module 1 — Trained immunity introduced as foundational update to innate immune understanding; connection to metaflammation and immunometabolism
- Module 4 — Clinical immunology master class; Leo Pruimboom emphasis on paradigm shift: innate system CAN learn