Trained immunity links cardiovascular disease and COVID-19

This editorial refers to ‘Atherosclerosis licenses for an exceeding immune response in COVID-19 disease by interferon priming in circulating myeloid cells’ by J. Leberzammer et al., https://doi.org/10.1093/cvr/cvaf268. The link between infections and atherosclerotic cardiovascular diseases (ASCVD) is known for almost a century, starting with the observation of excess cardiovascular mortality during influenza epidemics.1 Interestingly, influenza, pneumonia, and several other infections not only increase the risk to develop a myocardial infarction in the immediate post-infection period, but this increased risk can persist for months and even years.1 Recently, it has been suggested that one of the mechanisms causing this association is trained immunity,2 that describes the hyper-responsive innate immune cell phenotype that can develop after an initial exposure to microbial stimuli.3 In other words, innate immune cells such as monocytes or macrophages remember their first microbial encounter and exhibit a heightened immune response upon a subsequent stimulation. This functional hyper-responsiveness is underlined by epigenetic reprogramming induced during the initial stimulation, with the enrichment of specific histone marks that promote an open transcriptionally permissive chromatin structure. In humans, a trained immunity state can persist for months to years due to modifications at the level of hematopoietic stem and progenitor cells in the bone marrow, giving rise to long-term production of trained monocytes and neutrophils.3 Patients suffering from coronavirus disease 2019 (COVID-19) also have a strongly increased risk for ASCVD, that persists for several years.4 Cheong et al. demonstrated recently the presence of hyper-responsive trained monocytes up to 1 year after COVID-19 in humans, associated with epigenetic reprogramming of circulating hematopoietic stem cells, and they suggested that this contributes to the increased cardiovascular risk of these patients.5 Interestingly however, the link between COVID-19 and ASCVD is bi-directional, because patients with established ASCVD also have a higher risk to develop severe COVID-19 with a poor outcome, even after vaccination.6 In the current issue of Cardiovascular Research, Leberzammer et al. now provide data that suggest that an inappropriate induction of a trained immunity programme in patients with ASCVD is a likely mechanism for the severe form of COVID-19 they can develop upon infection.7 Indeed, trained immunity induction is not limited to infections. It has been recently shown that risk factors for ASCVD, such as hypercholesterolemia and hyperglycaemia, can induce maladaptive trained immunity changes in the innate immune cells.8 Hence, monocytes isolated from patients with established ASCVD have a trained phenotype, with an increased production of inflammatory cytokines when stimulated ex vivo, for example with lipopolysaccharides, and enrichment of specific histone modifications.8 Leberzammer et al. report that isolated monocytes from patients with ASCVD also show a hyperinflammatory response to ex vivo exposure to SARS-CoV-2, with increased expression of cytokines known to mediate a trained immunity phenotype, such as interleukin (IL)-1β and IL-6. Additionally, they performed single-cell RNA-sequencing (scRNA-seq) of peripheral blood mononuclear cells from 11 hospitalized patients with moderate COVID-19, with (n = 5) and without (n = 6) ASCVD. Monocytes showed enhanced expression of genes involved in inflammation in the ASCVD patients, with transcription factor (TF) activity analysis implicating STAT2 and IRF2 as upstream regulators, pointing towards a central role for interferons in these effects.7 In their in vitro experiments, Assays for Transposase-Accessible Chromatin sequencing of isolated monocytes of ASCVD patients before exposure to SARS-CoV-2 revealed an enrichment of type I interferon associated genes among the differentially accessible regions. In addition, the hyperinflammatory response to SARS-CoV-2 was reduced by co-administration of the JAK-STAT inhibitor Baricitinib, although the cytokine response still appeared to be higher than in monocytes from individuals free from ASCVD. These findings strongly suggest a trained immune monocyte phenotype pre-disposing to a heightened response to SARS-CoV-2 in patients with ASCVD. Leberzammer et al. also paid particular attention to neutrophils in their scRNA-seq analysis of patients with moderate COVID-19: the low-density neutrophils showed up-regulation of genes associated with inflammasome activation and alarmins, and concomitant immunosuppression due to up-regulation of anti-inflammatory genes such as IL-1 receptor type 2 and arginase 1, earlier described in patients with severe COVID-19.7 TF analysis again implicated the importance of the type I interferon pathway. These neutrophil-related findings are particularly interesting in light of previous studies that reported trained immunity in neutrophils: 3 months after vaccination with Bacille Calmette–Guérin, isolated neutrophils showed a trained hyperinflammatory phenotype, which was associated with an enrichment of the activating histone modification histone 3 lysine 4 trimethylation (H3K4me3) at promoter sites of genes involved in JAK-STAT signalling.9 Together with the previous data that COVID-19 induces trained immunity,5 the results presented by Leberzammer et al. now propose the framework that trained immunity contributes to the bi-directional connection between ASCVD and COVID-19, which is illustrated in Figure 1: the presence of trained innate immune cells in patients with ASCVD licenses for an excessive immune response to SARS-CoV-2 that could result in more severe COVID-19, whereas COVID-19 itself also induces trained immunity, potentially contributing to the increased CVD risk associated with COVID-19. The scRNA-seq study of Leberzammer et al. has a few limitations, in additional to the small sample size. First, they compared COVID-19 patients with and without ASCVD who were selected for similar (moderate) COVID-19 severity. Although this is understandable, it precludes the conclusion that ASCVD resulted in more severe COVID-19 in these patients. Therefore, the hypothesis that the hyperinflammatory monocyte and neutrophil phenotype in ASCVD patients really pre-disposes to more severe COVID-19 remains to be proven. An additional limitation is that one of the patients was placed in the ASCVD group because of severe asymptomatic carotid artery stenosis, which was not ruled out in the control patients. However, the authors have corroborated their most important scRNA-seq results by analysing bulk-RNAseq data from an independent larger cohort of patients with COVID-19 with and without ASCVD. Also, the in vitro experiments showing differences in chromatin accessibility and a hyperinflammatory response to SARS-CoV-2 in monocytes from ASCVD patients strongly support a role for trained immunity. An overview of the ASCVD and COVID-19 are linked by trained immunity. Circulating monocytes and neutrophils from patients with established ASCVD have a hyperinflammatory trained immunity phenotype. This probably results from the fact that cardiovascular risk factors, such as hypercholesterolemia, can induce central trained immunity in haematopoietic stem cells (HSCs) in the bone marrow. These trained innate immune cells license for an enhanced inflammatory response to SARS-CoV-2, which can worsen COVID-19. At the same time, SARS-CoV-2 itself can induce trained immunity, which can contribute to the development of atherosclerosis and to the increased risk of ASCVD in patients with COVID-19. The framework that ASCVD pre-disposes to severe COVID-19 via trained immunity can potentially be used in the future to develop novel strategies to protect ASCVD patients from severe infections. This requires the answer on several outstanding questions, including (i) identifying the trigger for trained immunity in ASCVD patients, (ii) providing a more detailed description of the molecular (epigenetic) pathways of trained immunity, also on the level of bone marrow progenitors, and (iii) assessing whether this framework can be extrapolated to other infections. N.P.R. is financially supported by a Vici grant of the Netherlands Organisation for Health Research and Development (ZonMw; 09150182410037), and by a consortium grant of the Dutch Heart Foundation (AtheroNeth; 01-001-2024-0601). M.G.N. is supported by a Spinoza Grant of the Netherlands Organization for Scientific Research (NWO).