Cardiovascular complications of diabetes: role of non-coding RNAs in the crosstalk between immune and cardiovascular systems

HSPCs contribute to the homeostasis of the cardiovascular system [70], especially the subpopulation of CD34 + cells [71]. Indeed, CD34 + HSPCs contribute to vessel repair and their function is known to be altered in DM [72]. Both mobilization of CD34 + HSPCs from BM and their angiogenic function are impaired in patients with DM, a condition often referred to as ‘mobilopathy’ at least in part associated with the deregulation of ncRNAs [10, 70, 73‐76]. BM CD34 + HSPCs show an altered expression of miRNAs in DM, including miR-155 and miR-21 [14, 75] associated with lower cell survival and induction of apoptosis. A recent study from our group showed that, compared to healthy controls, BM-derived CD34 + HSPCs bear lower levels of miR-21 and higher levels of one of miR-21 targets, the tumor suppressor programmed cell death 4. Of note, these changes can be transferred to endothelial cells (ECs) via taurine upregulated gene 1, a lncRNA responsible for the sequestration of miR-21, delivering a death signal to ECs [77]. In addition, serum and PB circulating HSPCs from diabetic patients with CLI contain increased levels of miR-15a and miR-16 with consequential impairment in HSPCs migration and adhesion capabilities [78]. On the other hand, miR-210 overexpression improves HSPC hypoxia-induced, SDF-1-driven migration [79]. Interestingly, restoring the expression levels of miR-92a in CD34 + cells from patients with diabetic retinopathy has been associated with an anti-inflammatory effect [80]. Lastly, angiogenic miR-126 is modulated in DM in CD34 + cells and its transfer can impact the endothelium [81].

Neutrophils are continuously recruited to chronic inflammation sites in patients with DM, promoting inflammation through an excessive NET formation and impaired apoptosis which contribute to accelerated atherosclerosis and delayed wound healing respectively [82‐86]. miRNAs have emerged as important regulators of inflammation-modulating neutrophils functions [87, 88]. In particular, miR146a and miR-129-2-3p are two inflammatory miRNAs in neutrophils involved in DM complications. Ana BA et al. showed that miR-146a deficiency promotes NET formations in an atherosclerosis mouse model (Fig. 2a) [89]. Previous studies demonstrated that miR-146a also has an anti-inflammatory role in monocytes, by downregulating the production of pro-inflammatory cytokines [90]. Umehara et al. observed that diabetic mice display an increase of wound neutrophils in which the downregulation of miR-129-2-3p contributes to prolonged inflammation and impaired apoptosis through upregulation of Caspase (CASP)-6 and C-C chemokine receptor type 2, thus delaying diabetic wound healing (Fig. 2b). Accordingly, overexpression of miR-129-2-3p in the skin wound site of diabetic mice enhanced the wound healing process [86]. These observations point to miRNAs as crucial elements regulating the role of neutrophils in DM complications.

Like neutrophils, monocytes\macrophages are excessively recruited into chronic inflammation sites in patients with DM [91]. Two important lncRNAs, anti-inflammatory Diabetes Regulated anti-inflammatory RNA (DRAIR) and pro-inflammatory dynamin 3 opposite strand (Dnm3os) are involved in DM. DRAIR was found to be downregulated in monocytes of patients with DM compared with non-DM controls (Fig. 2c) [92]. DRAIR modulation was confirmed in vitro by culturing THP1 monocytes in normal glucose versus high glucose conditions. It was also shown that DRAIR silencing inhibited the expression of anti-inflammatory genes in THP1 monocytes. Furthermore, DRAIR silencing enhances monocyte-endothelial cell (EC) adhesion and expression of pro-inflammatory genes, such as IL-1β. Conversely, DRAIR overexpression in THP-1 cells increases monocytes/macrophage differentiation and the expression of anti-inflammatory target genes, such as cytoplasmic polyadenylation element-binding protein 2 gene and interleukin 1 receptor antagonist; at the same time, it also inhibits pro-inflammatory genes, such as tumor necrosis factor-α (TNF-α) and fc gamma receptor IIIb. Therefore, DRAIR has an anti-inflammatory role, and it has been observed that it modulates the inflammatory phenotype of monocytes via epigenetic mechanisms in humans. Indeed, its downregulation is associated with an inflammatory phenotype of monocytes in DM thus promoting chronic inflammation [92]. The nuclear lncRNA Dnm3os is significantly upregulated in CD14 + monocytes from patients with DM compared with healthy controls (Fig. 2c). A similar regulation of Dnm3os was observed in macrophages from high-fat diet-induced insulin-resistant mice, streptozotocin (STZ)-induced type 1 diabetic mice, and STZ-induced diabetic apolipoprotein E deficient mice (a model of accelerated atherosclerosis) compared with controls (Fig. 3a). Dnm3os overexpression increases the pro-inflammatory gene expression and phagocytosis in macrophages, while its silencing results in the opposite response. Dnm3os interacts with nucleolin which has a protective role in macrophages. Nucleolin expression levels are decreased under diabetic conditions, and this leads to Dnm3os increase, promoting inflammatory gene expression. This suggest that Dnm3os play a role in DM and accelerated atherosclerosis [93]. Emerging evidence indicates that NETs trigger NOD-, LRR- and pyrin domain-containing protein 3 (NLRP3) assembly in macrophages, thereby sustaining inflammatory response with cytokine production, e.g. interleukin (IL)-1β, IL-18 [94]. An excessive NLRP3 inflammasome activation causes tissue damage increasing the risk of developing atherosclerosis [95, 96]. Interestingly, miRNAs control the activity of NLRP3 inflammasome [97]. In particular, miR-146a and miR-155 are two of the most studied miRNAs in the field of diabetes and atherosclerosis which control NLRP3 inflammasome activity in macrophages. In intrarenal macrophages from diabetic mice, miR-146a deficiency increased NLRP3, IL-1 β, and IL-18 expression levels leading to accelerated diabetic nephropathy, ultimately confirming the protective role of miR-146a within the inflammatory signaling pathway (Fig. 3b) [98]. Also, miR-155 is involved in NLPR3 activation by regulating the mitogen-activated protein kinase kinase (MEK)/extracellular signal-regulated kinase (ERK)/nuclear factor kappa B (NFκB) pathway in oxidized LDLs (ox-LDL)-induced-macrophages. Lentivirus-mediated overexpression of miR155 in apolipoprotein E deficient mice accelerated atherosclerosis (Fig. 3b). Therefore, miR-155 may promote accelerated atherosclerosis via NLPR3 activation [99]. These two miRNAs appear to be an attractive target to attenuate diabetic inflammation by suppressing the NLRP3 inflammasome.

Several ncRNAs are important regulators of NK cell functionality. In particular, two ncRNAs are involved in the NK interferon-gamma (IFN-γ) production and activation respectively. Specifically, IFN-γ antisense 1 is overexpressed in NK cells upon activation, leading to increased IFN-γ secretion [100]. Moreover, Ni et al. found that NK cells in human PB express high levels of miR-362-5p [101]. This miRNA can up-regulate the expression of CD107a (a cell surface marker exhibited by NK-activated cells), promoting perforin and granzyme B secretion. Importantly, the production of IFN-γ and the activation of NK cells have a crucial role in vascular homeostasis and the occurrence of vascular diseases [102].

Several miRNAs, like miR-182, -148, and − 152, are known to affect dendritic cell activation. Worth mentioning, miR-181a controls the inflammatory response of DCs. Indeed, it has been observed that upon hyperlipidemia, the increase of miR-181a attenuated ox-LDL-induced DC immune inflammatory response. In addition, miR-181a inhibits the expression of an important inflammatory transcription factor, c-Fos, further corroborating miR-181a anti-inflammatory role, especially during vascular diabetic complications, like atherosclerosis [103]. Other studies confirm the importance of ncRNAs in DCs function. miR-148 and miR-152 regulate the antigen presentation of toll-like receptors (TLR) to T cells triggered by DCs [104], an important process that defends against infection, autoimmunity, and CVDs [105, 106].

T lymphocytes are major components of the adaptive immune system [107]. Several studies have shown that the immune-pathological effects mediated by ncRNAs on T lymphocytes play a role in the development of diabetic complications. In particular, four ncRNAs (miR-155, ENST00000411554, miR-29b, miR-10a, or -10b) are involved in T lymphocyte deregulation. In the PB of T2DM patients with retinopathy or proliferative retinopathy, intracellular miR-155, and serum transforming growth factor-β (TGF-β) levels are significantly upregulated while the percentage of CD4 + CD25 + forkhead box (Fox) P3 + T-regs was lower compared to non-diabetic retinopathy and healthy control group (Fig. 4a). It was shown that the expression of miR-155 was negatively related to the T-regs and TGF-β levels emphasizing that miR-155 may play an important role in the pathogenesis of T2DM retinopathy by regulating the Tregs via TGF-β [108]. Another study reported the altered expression of the lncRNA ENST00000411554 in the wound surface of foot ulcers from patients with T2DM and found an association with T lymphocytes impaired functions and immune regulation imbalance (Fig. 4b). In particular, Xu et al. observed that the diabetic wound was infiltrated with higher levels of CD3, CD8 lymphocytes, and inflammatory factors, i.e. IL-2, IL-10, IFN-γ, and TNF-α, compared to non-diabetic control. They performed a microarray on the tissue of diabetic foot ulcers to investigate the molecular mechanisms involved in immune dysfunction. Analyses showed the downregulation of 13 lncRNAs, and, among them, of lncRNA ENST00000411554, which is located in a key upstream regulatory region of the target gene Mitogen-Activated Protein Kinase (MAPK) 1. Interestingly, an inverse correlation between the lncRNA and MAPK1 was observed. Therefore, it was hypothesized that ENST00000411554 could be involved in immune-regulation imbalance of T2DM, thus regulating the expression of inflammatory factors in T-cell-induced responses by downregulating MAPK1 [109]. Several studies indicate miRNAs as regulators of T lymphocytes activities in diabetic nephropathy. As an example, renal miR-29b downregulation contributed to the upregulation of T-Box Expressed in T cells and IFN-γ at both mRNA and protein levels, promoting CD4 + IFN-γ + Th1 cell infiltration in db/db mice (Fig. 4c). Overexpression of miR-29b inhibited Th1-associated immune injury [110]. Similarly, miR-10a or -10b downregulation promoted the infiltration of T cells (CD3+), Th cells (CD3+/CD4+), and killer T cells (CD3+/CD8+) in the kidney from STZ-treated diabetic mice (Fig. 4c) [111]. Although several studies have investigated the role of B lymphocytes, more investigations are needed to understand the role of B-lymphocyte ncRNAs in this context.

The cells in the adult human heart and vasculature are continuously engaged in cell-to-cell communications (CCC) to regulate both single-cell functions and coordinated responses needed for homeostasis [112]. CCC is ensured by a variety of biochemical mediators and electrical impulses [113, 114]. Diseases disrupt CCC, and cells might not interact properly or decode molecular messages properly. The study of altered cellular crosstalk might facilitate an understanding of disease pathogenesis and progression. Although protein-mediated interactions, such as ligand-receptor, receptor–receptor, and receptor-extracellular matrix interactions, have been better studied so far [114, 115], CCC can also take place via ncRNAs, and miRNAs in particular. miRNA-based CCC takes advantage of the interaction between miRNAs released by “message sender cells” and the miRNAs “target genes” present in the “message recipient cells”. miRNA exchanges are mostly realized via extracellular vesicles (EVs) (Fig. 5), although several alternative extracellular carriers of miRNAs, such as lipoproteins and argonaute-2, have been described [115]. Moreover, miRNAs can also be exchanged via connexin gap junctions, without being released extracellularly.

The term EVs denotes a highly diverse family of membrane vesicles of different biogenesis and sizes: apoptotic bodies (> 1000 nm), microvesicles (MVs) (150 nm − 1 μm), and small EVs (sEVs, circa 30–150 nm). sEVs, also called “exosomes”, are currently of particular interest to the biomedical community. sEVs are actively secreted and contain endosomal membrane markers, such as tetraspanins, plus a composite molecular cargo; that can include RNAs and proteins [116‐118], as well as mitochondria fragments and cytosol-derived small molecules (< 1500 Da), including metabolites as sugars, amino acids, lipids, nucleotides, and N-glycans [119, 120]. Once released by their parent cells, sEVs protect their molecular cargo from degradation until they pass it in a functionally active status to neighboring cells through binding, fusion, or endocytosis [117]. Alternatively, sEVs are transported in biological fluids to reach distant organs [121] or be excreted [118]. sEVs shuttle biologically active miRNAs from their parent cell to recipient cells, thus spreading the miRNA regulatory actions and profoundly influencing biological functions, including pro and anti-angiogenic responses [122‐124].

The adult human heart comprises billions of cells and approximately a third of them are cardiomyocytes, whilst the remainder comprises ECs and other vascular cell types, fibroblasts, resident macrophages, and neural cells [125, 126]. Recent evidence shows that different types of cardiovascular cells use sEVs as a means to communicate with each other as well as with other cells present in the cardiac microenvironment, such as immune cells and adipocytes [122, 124, 127, 128]. Moreover, evidence is accumulating for different types of sEVs being simultaneously released by a single cell type [129]. sEVs can differ substantially in their cargo and membrane composition [130, 131]. Consequently, the pool of sEVs that populate the heart’s extracellular space and contribute to cell-to-cell communications is diverse and variable. The coordination of sEV-mediated responses is important for maintaining physiology, but it can also mutate to support disease propagation [132, 133]. This can be understood considering that the amount and quality of sEVs secreted from a cell are influenced by both its activation status and the external environment to which the cell is exposed [134, 135], including hypoxia and D-glucose concentration [136]. As an example, Wang et al. showed that rat diabetic cardiomyocytes produce anti-angiogenic sEVs in vitro. This is in net contrast with the proangiogenic program activated in recipient ECs by their non-diabetic sEV counterparts [124]. Also, macrophages and ECs communicate via EVs, and changes in the macrophage phenotype contribute to reduced microvascular repair and impaired myocardial perfusion by promoting atherosclerosis [132], chronic inflammation and fibrosis [130, 137, 138].

The chronic inflammation induced by DM dysregulates the constant interaction of immune cells and cardiovascular cells [57]. CCC in DM has been reported to be driven mainly by sEVs RNA cargo [139‐142]. Indeed, higher levels of EVs have been found in human and animal models of type 1 and 2 DM [141]. miRNAs are the most studied ncRNA cargo of sEVs with a demonstrated functional role. sEVs-miRNAs have been shown to act on pancreatic β-cells, the insulin-producing cells. In detail, Zhang et al. showed that miR-223-3p and − 5p were two to three times increased in the medium of cultured pancreatic β-cells after high glucose stimulation, which modulated Glucose transporter 4, favoring insulin sensitivity [142]. Of note, several studies revealed bidirectional crosstalk between β-cells and immune cells via EVs, both in physiological and pathological conditions [143‐146]. As such, metabolic regulatory miRNAs, such as miR-29a-3p [147], miR-155-5p [148], miR-503 [149], and miR-210-3p and − 5p [150], can be carried by EVs and delivered into insulin-responsive cells and organs via paracrine or endocrine routes.

Macrophage-derived sEVs also play a role in diabetic complications. It has been reported that high glucose levels in pro-inflammatory macrophages induce the production of EVs carrying high levels of IL-1β, inducible nitric oxide synthase, human antigen R (HuR), miR-21-5p, miR-486-5p, and TGF-β mRNA [151‐155]. These biomolecules can be transferred to target cells and subsequently induce renal and cardiac injury and dysfunction. In particular, two miRNAs closely related to cardiac fibrosis and diastolic dysfunction, miR-122-5p [156‐158] and miR-21-5p are sorted specifically into EVs by the retinol-binding protein HuR [159]. Additionally, sEVs can effectively transmit pathogenic miRNAs to target cells, including vascular smooth muscle cells (VSMCs), ECs, neutrophils, and macrophages, leading to vascular stenosis, dysfunction, and inflammation, promoting atherosclerosis and thrombosis [157]. Likewise, recent mouse studies suggest that T cells might also have a role in DM by communicating with β-cells through the transfer of pro-apoptotic sEVs-miR-142-3p/5p and miR-155 [158]. As mentioned above, CD34 + HSPCs can transfer a negative pro-apoptotic signal including ncRNAs to the ECs via sEVs [14].

Exposure to numerous noxious stimuli, such as other cardiovascular risk factors, is also considered an important pathogenic event in DM [160]. ECs activation entails the expression of cell adhesion molecules like intercellular adhesion molecule 1 (ICAM-1) and vascular cell adhesion molecule 1 (VCAM-1) that serve as triggering signalling to the leukocyte adhesion on the endothelium. The uptake of LDLs into the intima and its oxidation, with the enhanced infiltration of monocytes/macrophages, promotes the formation of foam cells that induce cytokine and chemokine production and the additional recruitment of circulating immune cells [61]. As an example, miR-205-3p/-5p can regulate the adhesive properties of ECs, most likely by targeting the tissue metallopeptidase inhibitor 3, thereby triggering the release of soluble TNF-α, increasing monocyte adhesion [161]. Similarly, numerous miRNAs (miR-181b-5p, miR-103, miR-126-3p, miR-10a-5p, miR-92a-3p, let-7 g-3p/-5p, miR-195-5p) regulate expression in ECs of cytokines with pro or anti-inflammatory potential by indirectly targeting NF-kB and consequently affecting the immunophenotyping [162‐164].

In addition to miRNAs, lncRNAs, and circRNAs can also be exchanged between cells via EVs, and play a regulatory role in DM complications. For example, lncRNA Nexilin F-actin binding protein antisense RNA 1 (NEXN-AS1) circulates in sEV and suppresses the TLR4/NF-kB pathway, reducing endothelial activation and monocyte recruitment in human vascular ECs [165]. Further, its overexpression inhibits pro-inflammatory pyroptosis-related biomarkers known to drive atherosclerosis, such as NLRP3, CASP-1, IL-1β, IL-18, Gasdermin D [166]. Moreover, lncRNA growth arrest specific 5 (GAS5), long intergenic non-protein coding RNA 1005 (LINC01005), and metastasis associated lung adenocarcioma transcript-1 (MALAT-1) have also been detected in sEVs and may play a role in DM pathogenesis [167‐169]. In detail, silencing lncRNA GAS5 alleviates apoptosis and fibrosis in diabetic cardiomyopathy by targeting miR-26a/b-5p [170]. Besides, sEVs-LINC01005 from ECs has been reported to promote VSMC phenotype switch, proliferation, and migration by regulating the miR-128-3p/Kruppel-like factor (KLF) 4 axis [168]. On the contrary, sEVs-MALAT-1 derived from ECs promoted anti-inflammatory macrophage polarization [169, 171].

Of relevance, the circRNA homeodomain-interacting protein kinase 3 (circHIPK3) released by ECs-sEVs has been reported to promote the proliferation of VSMCs induced by high glucose via the miR-106a-5p/FoxO1/VCAM1 pathway [172]. Moreover, sEVs- circRNA euchromatic histone-lysine-N-methyltransferases 1 from pericytes could suppress the generation of NLRP3 inflammasomes in high glucose-treated ECs by upregulating the levels of the transcription factor nuclear factor I A, which in turn inhibited the apoptosis of ECs and reduced its angiogenic activity, ultimately attenuating high glucose-induced retinal vascular dysfunction [173].

Interestingly, aging and DM share common pro-inflammatory pathways, such as the NF-κB and NLRP3 signaling pathways, which are regulated by several miRNAs. Among the most remarkable miRNAs regulating age-mediated inflammation, miR-146a stands out, as it can reduce NF-κB DNA binding activity [195] and has been proposed as a biomarker of healthy aging [196]. Inflammation is a common ground in several CVDs and sex differences in the transcriptomic regulation of inflammatory genes and pathways in CVDs have been previously reported [181, 197, 198]. Circulating levels of miR-146a decline in aged people, and this phenomenon is particularly evident in men [196]. In turn, miR-146a levels decrease after estrogen treatment in mouse-isolated splenic lymphocytes [199], and it could act as a sex-biased inflammation-associated biomarker, modulating T2DM and other aging-associated diseases [200]. Also, miR-34a, a key promoter of vascular aging related to DM-associated vascular complications, shows sex differences in its expression [201]. Of note, it induces a systemic inflammatory state [202]. Indeed, miR-34a expression is elevated in several cardiac pathologies, and its therapeutic inhibition in a mouse model of dilated cardiomyopathy provides more protection in females than in males [203]. Particularly, in association with DM, sex was found to be a determinant of plasma miR-34a levels in patients with left ventricular diastolic dysfunction [204].

Other miRNAs, such as let-7 g and miR-221 have been related to glucose metabolism [205‐207]. Thus, different studies support a sex-dependent regulation of physiological functions by ncRNAs in proinflammatory states that underlie cardiovascular or metabolic diseases, among others, also show sex differences, as women with metabolic syndrome display higher circulating levels of let-7 g and miR-221 than men [207]. In addition, mouse studies have revealed a male-specific induction of miR-23a-3p, miR-27b-3p, miR-130a-3p, miR-133a-3p, miR-143-3p, and let-7e-5p, together with a corresponding downregulation of downstream protein targets involved in mitochondrial metabolism, thereby probably contributing to sex-biased maladaptive cardiac remodeling [208].

Some of the sex differences in transcriptomics are evident from fetal age. In particular, analysis of miRNA expression patterns in healthy placentas performed in the Environmental Influences on Early Ageing (ENVIRONAGE) study shows a different miRNA expression pattern in the placenta of mothers of girls and boys, and enrichment analyses showed sex-dependent differences in the regulation of biological processes, including metabolic and immunological pathways [209]. In newborn girls, the higher expression of placental miRNA-210, miR-146a, miR-34a, and miR-222, among others, is critical for specific cellular processes in aging and inflammation [210]. It is important to note that some of the alterations that occur during pregnancy can affect the future development of pathologies in both the mother and the newborn [211]. In particular, women with gestational DM have a higher risk of developing future T2DM [212]. In this regard, during gestational DM, the miRNA expression profile changes in human placental ECs [213]. Thus, the hyperglycemic intrauterine environment could affect placental function through elevated levels of proinflammatory cytokines, as well as through transcriptional modifications, including those mediated by ncRNAs.

The epidemic of diabetic cardiovascular complications underlines the need for the translation of current preclinical knowledge and novel technologies into actual clinical practice. Recently, ncRNAs were shown to be powerful therapeutic tools for different disease settings by affecting the onset and progression of several human disorders [236]. Therefore, researchers focused on the development of innovative next-generation technologies to exploit these ncRNAs to treat such diseases, especially those not susceptible to pharmaceutical interventions [237].

The use of ncRNA-based therapies has several advantages compared to other compounds. NcRNAs, and in particular miRNAs, can naturally modulate different specific targets within one pathway thereby exerting a broader biological effect. Thus, ncRNA therapies might have the potential to boost therapeutic effects compared to those drugs acting on one single target gene. Additionally, chemical modifications can easily be adapted to prolong the stability and half-life of ncRNA drugs in circulation, facilitating the effective delivery to the site of interest without degradation (Fig. 6).

Strategies to target the level of ncRNAs can be classified into two approaches: suppressing upregulated ncRNAs or enhancing levels of downregulated ncRNAs. Technologies to modulate miRNA expression include miRNA mimics to elevate their expression; concerning miRNA inhibition, miRNA sponges exert their function by sequestration of endogenous miRNAs, while inhibitory chemical compounds can compromise miRNA action through the interaction with components of miRNA biogenesis pathways. Moreover, the most pursued strategy relies on base pairing of complementary antisense oligonucleotides (ASOs) with ncRNA sequences leading to duplex formation or subsequent degradation (e.g. anti-miRNAs). LncRNAs can be targeted through modified ASOs such as locked nucleic acids (LNAs) with modified phosphorothioate backbones, gapmers composed of two LNA flanking regions and a central DNA stretch that activates ribonuclease H cleavage, or small interfering RNAs inducing their degradation [237, 238]. Other modulation tools include genome-editing strategies such as Clustered Regularly Interspaced Short Palindromic Repeats (CRISP)/CRISPR associated protein 9 (Cas9) or viral-based approaches [239, 240]. Although there is an increasing number of preclinical reports regarding ncRNA therapies, only a small amount have proceeded into clinical trials [241, 242]. Key challenges have to be faced to successfully translate ncRNA therapeutics into the clinic. Off-target effects result in suppressing genes other than the desired target due to sequence similarities or overdosing along with undesired immune responses. Additionally, special delivery systems are required to target the organ and cell type of interest and naked oligonucleotides have to be chemically modified to resolve issues such as degradation by nucleases or their inefficient uptake due to negative charge or size [243‐245]. In this respect, recently, it has been shown that lipid- or polymer-based nanoparticles, as well as exosomes, are promising delivery tools (Fig. 6) [246‐250].

Nonetheless, a plethora of RNA-based drugs are already under clinical development, most of them aiming to modify mRNA levels in the liver, central nervous system, or muscle and some of them have even been approved by the Food and Drug Administration or European Medicines Agency (e.g. Viltolarsen, Nusinersen, Patisiran) [251‐254]. In addition, tremendous progress has also been made in the field of miRNA-based drugs which entered clinical testing [238, 255] and lncRNA-based formulations are also gaining more interest as promising treatment options. One of the ncRNA compounds, which entered the clinical phase, is an inhibitor against miR-92a (MRG-110) [256, 257] that was shown to exhibit beneficial effects in wound healing and CVDs such as LI, MI (NCT03603431, NCT03494712). Another example is anti-miR-21, which was evaluated for its capability to reduce kidney fibrosis in patients with Alport syndrome (NCT03373786) and will be further examined in a clinical phase II study (NCT02855268) [258]. Moreover, CDR132L, a synthetic oligonucleotide inhibiting miR-132 indicated as a key regulator in cardiac remodeling processes [259, 260], entered recent clinical study phase II (NCT04045405, NCT05350969) for safety and efficacy assessment in ischemic HF patients. Concerning DM, AstraZeneca performed a phase I/IIa clinical trial (NCT02612662, NCT02826525) with an N-Acetylgalactosamine (GalNAc)-conjugated anti-miRNA against miR-103/107 (RG-125) [261]. miR-103/107 is an insulin sensitizer already tested in preclinical models that show promise as a compound to treat patients with T2DM and nonalcoholic fatty liver diseases.

Clinical outcomes of COVID-19 are worse in the presence of CVDs and risk factors, such as DM (Fig. 7). Severe COVID-19, DM, and CVDs share a common denominator, an increased inflammatory status, which is also thought to underlie male-biased cardiovascular complications in COVID-19 [299, 300]. Assessing cardiovascular inflammatory status through evaluation of ncRNAs levels appears as a promising strategy to aid in personalized healthcare, prevent fatal cardiovascular damage, and thus improve clinical outcomes of COVID-19 patients. SARS-CoV-2 can affect multiple organ systems directly or indirectly through cytokine storm and the adaptive immune system activation [301, 302], dysregulation of the renin-angiotensin-aldosterone system (RAAS) as a result of virus-mediated suppression of angiotensin-converting enzyme (ACE) 2 expression [303], damage to ECs [304], vascular inflammation and sepsis [305]. Cardiovascular damage is frequently observed in severe COVID-19 patients and it is dictated directly by the amount of viral load or indirectly by the extent of the patient’s immune response. Cardiovascular damage is also dependent on the presence of co-morbidities associated with RAAS deregulation (Fig. 7) [306]. Although severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) has shown major tropism to the pulmonary cells, recent reports confirmed its presence in myocardial biopsies of COVID-19 patients with myocarditis [307‐309]. Cardiac tissues of COVID-19 patients with high viral invasion had increased expression of pro-inflammatory cytokines [307]. However, this finding was not related to the prominent invasion of the cells, thereby indicating that the presence of SARS-CoV-2 in the heart may not fundamentally cause an inflammatory response, which is in line with clinical myocarditis. The long-term consequences of SARS-CoV-2 infection and associated cardiovascular outcomes require advanced examination.

ACE2 is a part of RAAS and plays a critical role in the development of DM, hypertension, and HF. Disturbance of RAAS caused by loss of ACE2 in diabetic patients results in declined vascular function and systolic dysfunction that are dependent on the angiotensin II/AT1 receptor [310]. Since SARS-CoV-2 uses human ACE2 as cell entry receptors and human proteases as entry activators, RAAS is an important component of COVID-19 pathogenesis. Detailed mechanisms of SARS-CoV-2 entry into host cells have been recently reviewed in [311]. Imbalanced ACE/ACE2 and RAAS activation are responsible, at least in part, for COVID-19 progression which can lead to cardiac damage by inducing a hyperinflammatory response characterized by high levels of IL-6, IL-1, and TNF-α [312]. A recent meta-analysis that pooled 76,993 patients from 10 studies has demonstrated that hypertension, renal, liver and cerebrovascular diseases, obesity, smoking history, and current smoking are associated with the severity and poor prognosis of COVID-19 [313]. The mechanisms which increment vulnerability to COVID-19 infection in patients with DM may occur through ACE inhibitors or angiotensin receptor blockers which modulate ACE2 expression or through compromised immunity and elevated furin levels [314]. However, the Council on Hypertension of the European Society of Cardiology emphasized that there is no evidence supporting harmful effect of ACE inhibitors (ACE-I) or angiotensin receptor blockers (ARBs) during the COVID-19 outbreak. The Council strongly advise physicians and patients to continue their anti-hypertensive therapy treatment since there is no scientific or clinical basis for stopping ACEi or ARBs due to Covid-19 ¹.

A study by Abe et al. has shown that COVID-19 patients with DM had an increased incidence of acute myocarditis, acute HF, acute MI, and atrial fibrillation [315, 316]. Increased serum levels of IL-6, IL-8, and IP-10 were associated with a high mortality rate in COVID-19 patients [315]. The interplay between COVID-19 inflammatory response and DM may help to explain the poor cardiovascular prognosis of patients after SARS-CoV-2 infection, suggesting that interventions targeting innate immune pathways could potentially benefit COVID-19 patients with DM and other cardiovascular risk factors. Overall, applying therapeutic strategies to improve cardiovascular health may improve outcomes in COVID-19 patients, including patients with diabetes [317‐319].

Several studies highlighted that altered levels of circulating cardiovascular, immunological, and inflammatory ncRNAs could be potential biomarkers of COVID-19 severity and myocardial damage. As an example, Garg et al. observed that inflammation-associated miR-155-5p, profibrogenic miRNA miR-21-5p, and cardiomyocyte-associated miRNAs miR-499-5p and miR-208a-3p were upregulated in the serum of severe COVID-19 patients requiring invasive ventilation in comparison to healthy controls and influenza-acute respiratory distress syndrome patients [320]. Another study showed that circulating myomiRNA miR-133a was upregulated in the plasma of severe compared to mild COVID-19 patients [321]. In particular, miR-133a overexpression was associated with intensive care unit (ICU) mortality within 28 days. Interestingly, myocardial injury correlated with an increase in circulating miR-133a levels [321]. In this setting, the main source of circulating miR-133a is most likely to be the cardiomyocytes, since they express high levels of this miRNA. However, miR-133a is also expressed in other cell types, such as neutrophils [319], that may be contributing sources of plasmatic miR-133a levels, considering that neutrophil degranulation and extravasation play a role in myocyte damage. Indeed, it was observed that circulating levels of miR-133a inversely correlate to neutrophil counts and directly correlate to myeloperoxidase, a neutrophil activation marker associated with cardiovascular outcomes in COVID-19 patients [322]. Other inflammation and cardiovascular-related miRNAs were modulated, e.g. miR-21-5p, miR-142‐3p, and miR‐146a‐5p were downregulated, and miR-15b-5p was upregulated, in red blood cell‐depleted whole blood of severe COVID-19 patients compared to moderate COVID-19 patients.

De Gonzalo-Calvo et al. [323] showed that miRNAs implicated in immune/inflammatory response and coagulation were altered in the plasma of COVID-19 patients. In particular, miR-27a-3p, miR-27b-3p, miR-148a-3p, miR-199a-5p and miR-491-5p were upregulated and miR-16-5p, miR-92a-3p, miR-150-5p, miR-451a and miR-486-5p were downregulated in ICU patients compared to ward patients. Moreover, three miRNAs (miR-148a-3p, miR-451a, and miR-486-5p) discriminated between ICU and ward patients. miR-16-5p, miR-92a-3p, miR-98-5p, miR-132-3p, miR-192-5p, and miR-323a-3p were downregulated between non-survivors and survivors to ICU stay. Two miRNAs (miR-192-5p and miR-323a-3p) distinguished between ICU non-survivors and survivors. Of note, a recent investigation has shown the potential of miRNAs, including miR-145a-5p and miR-181-5p, to define endotypes in critically ill COVID-19 patients [324], which suggests their usefulness in personalized medicine.

Recently, our investigations indicated that decreased plasma levels of miR-144-3p and − 5p were associated with both mortality and disease severity grade in three independent cohorts of hospitalized and non-hospitalized COVID-19 patients [325]. Thus, circulating miR-144 emerges as a noninvasive tool for early risk-based stratification and mortality prediction in COVID-19. This result is noteworthy in light of the reduced plasma levels of miR-144 observed in patients with DM with cardiac dysfunction [223].

Additional investigations have proposed that the disease not only impacts the circulating miRNA profile during the acute phase but also in “long-COVID-19”. Plasma levels miRNAs implicated in diverse immune mechanisms, e.g. miR-9-5p or miR-146a-5p, or miR-221-5p, have been associated with pulmonary dysfunction and radiologic abnormalities after hospital discharge in survivors of severe COVID-19 [326].

Collectively, the results show the association between COVID-19 severity and the alterations in the circulating miRNA profile, which may also contribute to the inflammatory effects of the disease. Indeed, circulating miRNAs could serve as biomarkers for clinical decision-making, but also as targets for developing specific therapies. For instance, Meidert et al. has proposed different EV-derived miRNAs, including let-7e-5p, miR-146a-5p, and miR-3168, as key regulators of signaling pathways implicated in inflammation and immune function in COVID-19 patients [327]. Results from independent publications support the participation of immune- and inflammatory-related miRNAs in the modulation of these mechanisms [328, 329]. In the context of metabolic disease, Wang et al. identified three antiviral miRNAs, miR-7-5p, miR-24-3p, and miR-223-3p, showing anti-inflammatory and cardio-protective roles, and remarkably decreased levels in serum exosomes of patients with DM compared to non-diabetic controls. The altered levels of these exosomal miRNAs in circulating exosomes weakened the inhibition of SARS-CoV-2 replication in patients with DM [49, 330, 331]. Therefore, miRNAs could serve as potential therapeutic candidates to prevent the progression from mild to severe COVID-19 disease in patients with DM by reducing viral replication.