C-Reactive Protein as a Potential Signpost to Trained Immunity

It has been more than 30 years since markers of systemic inflammation (measured by the acute phase proteins: C-reactive protein [CRP] and serum amyloid A) have been associated with adverse clinical outcomes—initially in patients with acute coronary syndromes¹ and subsequently in apparently healthy men.² Significantly, the precise “drivers” of this subclinical inflammation are still not clear.

It seems plausible that the levels of these nonspecific markers of inflammation, which are released by the liver, reflect a cumulative “inflammatory load,” components of which also drive pathogenic processes relevant to atherosclerosis in parallel. In support of this notion, adverse outcomes pertain across a range of rheumatological and autoimmune diseases, and the precise nature of chronic inflammation seems to matter little. Processes of inflammation within the walls of arteries contribute to the manifestations of atherosclerotic cardiovascular disease (ASCVD), but these (sometimes small, focal) lesions seem more likely to be responding to systemic proinflammatory cues than driving the systemic inflammatory response that is reflected by elevated high-sensitivity C-reactive protein (hsCRP). In other words, systemic drivers of processes of inflammation probably serve to activate cells in the atherosclerotic plaque, as they do elsewhere. This interpretation is supported by events, such as experimental myocardial infarction, that drive inflammation and accelerate atherosclerosis.³

Mendelian randomization studies examining the effects of genetically determined CRP show that it does not have a causal role in coronary artery disease.⁴ In other words, while CRP shows a strong population-level correlation with coronary disease events, it is a “reporter” of relevant processes but is not driving them. The cytokine interleukin (IL)-6 plays a key role in immune responses and inflammation as a ligand that binds to specific receptors. Data accrued from animal models and epidemiological and genetic studies suggest a (likely causative) role in disease-relevant processes of atherogenesis. IL-6 also drives the production of CRP by the liver.

Trained immunity refers to a form of immunological priming that manifests in the innate immune system, which was traditionally thought to lack “memory.” Unlike the adaptive immune system (T and B cells), which respond to specific pathogens, trained immunity involves a functional reprogramming of innate immune cells such as monocytes, macrophages, and natural killer cells.⁵ This reprogramming occurs after an initial challenge and results in a primed state that accelerates and amplifies the response of these cells to subsequent stimuli. Significantly, several known risk factors for ASCVD—including infection, Western diet,⁶ hypercholesterolemia, and hyperglycemia⁷—promote states of trained immunity.⁸ The manifestations of trained immunity are patterns of altered immune cell function that alter both cellular metabolism (favoring aerobic glycolysis) and promote processes of inflammation while impairing pathways linked to repair and remediation. Cardinal features of trained immunity include increased propensity to generate IL-1β and IL-6, which can promote hepatic CRP production and ASCVD and have been identified as treatment targets. Although peripheral immune cells that mediate effects of trained immunity are short-lived, epigenetic modifications in bone marrow precursors provide and enduring record of training and a source of replenishment.

In this issue of JACC: Basic to Translational Science, Zhang et al⁹ undertake a case-control study in which they explore the features of trained immunity in the peripheral blood monocytes of patients with (>3 mg/L) and without (<1 mg/L) elevated hsCRP in the context of unstable angina. A total of 132 age-matched, sex-matched, body mass index–matched, and otherwise well-matched subjects were included. Although the rates of diabetes were equivalent between the groups, patients with higher-level CRP showed higher levels of fasting blood glucose. Patients with unstable angina as opposed to myocardial infarction were purposely recruited to minimize the elevation in CRP that occurs after myocardial infarction/necrosis. The median hsCRP level was 5.1 mg/L in the patients with elevated CRP and 0.4 mg/L among patients with the lower CRP criterion and was, therefore, broadly reflective of levels found in populations of patients with ASCVD. There were small, but statistically significant, differences in the numbers of neutrophils and monocytes, but these marginal differences, while demonstrating myeloid skewing, were not at clinically discernible levels. Similarly, differences in surface markers were largely unremarkable. However, functional analyses were more revealing. Following stimulation with lipopolysaccharide or PC3, CD14-positive monocytes from higher-CRP subjects had enhanced cytokine secretion, including the cytokines, IL-6, tumor necrosis factor α, and IL-1β, which are recognized markers of trained immunity. The levels of these poststimulation secreted cytokines correlated with the measured plasma level of hsCRP. Furthermore, analysis of metabolic activity in isolated monocytes under both basal and lipopolysaccharide-stimulated conditions from patients with unstable angina showed enhanced glycolysis in patients with higher-level CRP. Interrogation of metabolic profiles suggested monocytes from patients with higher-level CRP showed a Warburg-type metabolic profile, consistent with enhanced aerobic glycolysis, as previously documented in conditions of inflammation and trained immunity.

On analyzing monocyte gene expression, 506 showed increased expression and 473 decreased expression. Because trained immunity is a priming phenomenon, the differential expression even in the ostensibly basal state raises a question as to the nature of the secondary stimulus in these patients, particularly because they were selected to be without material myocardial injury. Notwithstanding, the pattern of changes in relation to inflammation and energy utilization were consistent with a trained immunity state. Notably, genes related to the NLRP3 inflammasome zone were enriched, including IL-6 and IL-18. A similar, if more pronounced, proinflammatory profile was demonstrated after exogenous stimulation.

The findings provide an intriguing potential link between trained innate immunity and the clinically useful, but hitherto nonspecific, feature of elevated hsCRP. Furthermore, trained immunity potentially explains how varied inflammatory and metabolic stimuli could converge in epigenetic changes that are recorded in bone marrow stem cells. Analyses from our laboratory (K. Boden, PhD, et al, June 2025, unpublished data) suggest that multiple and diverse inflammatory stimuli to trained immunity converge on patterns of shared gene expression profiles, indicative of common states of trained immunity. If this is the case, exciting new possibilities emerge. First, the utility of hsCRP is further enhanced by the identification of plausible underlying processes that drive its chronic elevation (as distinct from episodic, large fluctuations that accompany acute illness). Moreover, if the drivers of trained immunity converge and are aggregated as epigenetic modifications in identifiable cells, new possibilities emerge for treatments. These could be directed toward correcting such modifications in addition to current approaches that seek to inhibit the downstream effector molecules, such as IL-6 and IL-1β.

The authors can be congratulated on gathering a series of patients with relatively tightly constrained levels of hsCRP in the context of unstable angina without myocardial injury. The relatively large number of patients allowed for impressively precise matching between the groups. Nonetheless, the design of the study (observational case-control) cannot show that trained immunity causes higher levels of CRP. However, the alignment of secretory profile, gene expression, chromatin accessibility, and altered metabolism is all consistent with a trained immunity state in peripheral monocytes, which is at least is associated with elevated hsCRP. The possibility that hsCRP is reporting on trained immunity status is intriguing. The study was not designed to track cardiovascular outcomes/event rates with measures of trained immunity, but in sufficiently large numbers of patients, given the relationship with hsCRP, such an association is likely to emerge. On the other hand, the physiological relevance of the trained immunity demonstrated here (through the stimulation of cells with lipopolysaccharide) has uncertain in vivo mechanistic significance, though trained immunity identified in this way drives atherosclerosis in animal models.⁷

The study does not address the precise epigenetic modifications, or their loci, and the actual epigenetic signatures are worthy of further investigation. As the authors point out, there may be additional genetic vulnerabilities in relation to degree of inflammation risk. Because these cells are mediators of pathology, susceptibility to greater or lesser degrees of inflammatory programming could be material, particularly because many of the cardiovascular disease single nucleotide polymorphisms identified in genome-wide association studies reside in gene regulatory elements.

Recent metanalysis has shown that, in patients with—or at high risk of—atherosclerotic disease and intensively lowered low-density lipoprotein cholesterol, hsCRP emerges as the more potent marker of future risk.¹⁰ Given that CRP is not driving disease, it becomes even more pressing to understand and map upstream inflammatory pathways, and to define which cell type(s) and organ(s) contribute processes of inflammation that are manifest in elevated hsCRP. The work of Zhang et al provides an important new insight that shows an association between features of trained immunity in monocytes and elevated hsCRP. Future work will doubtless elaborate these possibilities, identify specific epigenetic modifications, broaden the scope of clinical applicability, and examine potential causation.

Funding Support and Author Disclosures

Dr Choudhury is the UK Chief Investigator for the ZEUS clinical trial of Ziltivekimab (sponsor NovoNordisk); he has undertaken paid consultancy on matters related to cardiovascular inflammation for NovoNordisk; on the Scientific Advisory Board for Tourmaline Bio and Nodthera; and his laboratory is supported by the British Heart Foundation, the Medical Research Council (UK), the NovoNordisk Foundation, and the Kusuma Trust.