1. Introduction
Inflammatory bowel disease (IBD), including Crohn’s Disease (CD) and Ulcerative Colitis (UC) are chronic multifactorial disorders which affect the gastrointestinal tract with a variable extent, typically leading to the development of symptoms, such as rectal bleeding, abdominal pain, diarrhea and weight loss [
1]. Although the pathogenesis of IBD remains still unknow, their development is considered as the result of genetic, environmental, gut microbiota and immunological factors [
2]. The diagnostic procedures of IBD are time consuming, requiring endoscopy, blood and stool exams, ultrasonography, or magnetic resonance imaging. Sometimes these procedures need to be repeated multiple times before a definitive diagnosis is reached, prolonging the discomfort of the patient [
3]. Therapeutic possibilities are limited, with rates of primary or secondary non-response to therapies reaching 50-60% of treated patients [
4,
5]. Moreover, despite ongoing research efforts, to date there are no biomarkers that can predict clinical response to available drugs, hence the need for continuous disease monitoring of IBD patients with invasive examinations, such as endoscopy.
In this respect, cell-free Nucleic Acids (cfNA) represent a promising field of research. In complement to the direct analysis of genomic material obtained from cells, the possibility of identifying and analysing cfNAs present in the circulation or other body fluids would allow the identification of asymptomatic individuals at high risk for IBD [
6], or could represent a promising new non-invasive biomarker to distinguish patients with active disease from those in remission or from healthy individuals, allowing the tracking of disease onset, progression and remission following therapy, and hopefully also to predict drug-response.
First findings of cfNA date back to 1948, when Mandel and Métais described it in human plasma [
7]. However, studies number grew only twenty years later, probably due to biotechnological and computational biology advances. Since then, their presence was detected in many easily obtainable bodily fluids, such as blood, saliva, stool and urine, assessing their ever-increasing role in new biochemical analysis. Among these, liquid biopsy reached wide interest, becoming part of the standard of care in many fields [
8], such as prenatal care with the so-called “non-invasive prenatal testing” (NIPT) [
9,
10], and oncology [
11,
12,
13,
14,
15].
All cfNA are, according to definition, free from origin cell and able to circulate into blood. Among them, human derived cfNAs can also travel with or without carriers, such as microparticles (MPs) [
16,
17,
18]. Within MPs, exososmes, appear to have regulatory and influential activity in many dysfunctional conditions, like cancer, chronic inflammation (e.g.IBD, arthritis), thrombosis, pregnancy diseases and sickle cell anemia [
19,
20,
21,
22], findings that point towards an ever-growing role in immune system regulation. Thanks to increasing basic science-based studies on this topic and to the development of sophisticated biotechnologies, today the clinical management of the above mentioned (and many more) pathologies is drastically changing. The aim of this review is to contribute with our results to provide basis for a new target of care for chronic inflammatory conditions like IBDs, that exploits the potential of circulating nucleic acid from understanding the pathogenesis of the disease, to the development of novel molecular therapeutics.
2. Cell-free DNAs (cfDNAs)
Nucleic acids are known to be packed into nucleus or floating into cytosol, but recently another location has been conferred to them: blood circulation and extracellular space. CfNAs can be released into circulation by various biological processes, so to explicate a regulatory role not only into cells they are normally contained by, but also at distance, in either a coding or a non-coding manner [
23,
24,
25]. Endogenous-derived nucleic acids must be distinguished from exogenous, which mostly derive from microbial (specifically bacterial) DNA. Endogenous nucleic acids result from many processes, among which apoptosis/necrosis, NETosis and active release (by microvesicles (MVs)/exosomes) are the most endorsed ones [
26,
27,
28]. These processes generate RNAs, that longly distracted our attention from cfDNA, nuclear (cf-ncDNA) and mitochondrial DNA (cf-mtDNA) in a linear shape. Nucleic acids can appear totally naked, bound to vesicles or in macromolecular structures (virtosomes, nucleosomes, Neutrophil extracellular traps (NETs)) [
27]. Another kind of cell free DNA, the extrachromosomal-circulating-circular DNA (eccDNA, see below), derives from other biogenesis mechanisms, mostly related to genomic instability, DNA damages and hostile cellular environments [
11], but also apoptosis [
29] and probably necrosis and pyroptosis. Similar to circulating DNA, which is findable either in a particulate (vesicles enclosed) and non-particulate form (naked DNA, nucleosomes), cfDNAs have been described as circulating free or MVs enclosed molecules, able to sustain propagation or even shut down inflammatory stimulus [
30,
31,
32].
Cells undergo cycles of birth and death, on a frequency based on tissue homeostasis. Cell death can occur unexpectedly (necrosis) or can be programmed (apoptosis); both processes, in the end, contribute to vesicle bound or naked cf-DNA release [
33]; thus, high-turnover tissues might contribute consistently to cfDNA circulation [
34]. Several studies demonstrated that the circulating fraction of total cell-free DNA varies from 0.1 to 89% [
35] suggesting that DNA circulation may occur on specific conditions. An unnatural high-turnover status is surely that of cancer; indeed, many studies associated cancer to abnormal cfDNA levels in serum [
11,
36]. Evaluations of length, patterns of methylation, nucleosomes alterations and fragmentomics revealed a ‘tissue-of-origin signature’ on cfDNA, proposing that most of it comes from mitotic cells, like those of cancer [
37]. Moreover, stability and half-life of cfDNA depend on protein association: long naked molecules (>10kb), like free DNA fragments shedded during apoptosis, are rapidly cleaved by DNase and lost, instead, shorter molecules (<100 bp) are bounded by nucleosomes, protecting them from nucleases and increasing half-life [
38], meaning that also clearance process contributes to DNA plasma levels; indeed, since DNAse activity in Crohn Disease is lower, here’s another reason of increased cfDNA plasma levels in IBD [
39].
2.1. NETs
NETosis is the process by which neutrophils exert their ‘license to kill’: the cell ‘sacrifices’ itself, releasing complex structures of genetic material that entraps mainly bacteria contributing to the cfDNA pool. It consists in nuclear or mitochondrial oxidised DNA, histones and protases release, which have pro-inflammatory capacities [
40,
41]. NETs also contain cathepsin-G and calprotectin, a well-known biomarker in IBD [
42,
43]. Immunogenicity of NETs consists primarily in stimulating many cell activities: activation of TLR of dendritic cells (resulting in INF-alpha synthesis), Immunoglobulin (Ig) class switching in B cells and T cell response boosting [
44]. It should be reminded that neutrophils can randomly circulate and be recruited from circulation to specific sites. This means that NETs can be released directly into circulation or in another tissue whose venous drainage can give back to systemic circulation NETs components or degrade them directly in site, without passing to blood. Thus, technically, not all of NETs released into the organism contribute to the circulating DNA pool [
45]. An IBD gut is known to undergo a sustained and chronic inflammation process, where both unbalanced T-cell priming and microbial dysbiosis are features. Although it is not clear who as first started the inflammatory processes, if NETs are the “primum movens” or the answer to a preexistent noxa, it is increasingly clear that both microbiota alterations and microenvironment participate to NETosis [
26] and so to mucosal damage. Specifically, in IBD, NETs seem to have pro-thrombotic abilities, through direct platelet activation [
46] but probably also by contributing to endothelial damage at blood-cell interface. However, NETs can regulate many cytokines secretion during inflammation, and menage the clearance of damage-associated-molecular-patterns (DAMPs) in mice [
47]. Recently, this mode has been extended also to eosinophils, cells able to release eosinophil extracellular traps (EETs), which apparently correlate with disease severity [
48]. Ultimately, NETs, along with calprotectin, do correlate with inflammation severity [
28,
49], but due to their multiple functions, they are not an optimal target as a therapy or as markers in IBD.
2.2. Vesicle-bound DNA
Extracellular vesicles (EVs) or microparticles are cell-released membrane covered globules that include microvescicles (microparticles in literature), exosomes and others [
50], and can differ in size, composition, role and genesis processes [
51]. EVs are released from almost all cell types, including bacteria (Bacterial Extracellular vesicles, BEVs) [
50,
52], under specific circumstances. Their presence has been reported in saliva [
53], sperm [
54], milk [
55], urine, blood (serum and/or plasma) [
56] among others. Often, isolated vesicles were attributed to exosomes, since their exosome-like protein cargo; however, circulating vesicles are probably either MVs and exosomes. Indeed, studies revealed that luminal EV in IBD patients are of less than 500 nm in diameter [
27,
31], confirming presence of both exosomes and micro vesicles. In the last decade increasing interest has arisen, since it have been demonstrated that vesicles are shed from plasma membranes of cells in complete physiological conditions, and carry a significant amount of interesting biological material [
57]. Many pathological conditions started to benefit from MPs role, not just as possible biomarkers: EVs are elevated in pneumological and rheumatic diseases, while urinary EV seem to reflect acute kidney injury [
58]. Notably, their peculiar structure inspired new drug delivery systems [
59].
Exosomes are considered as single layer lipid membrane vesicles of 30-150 nm diameter [
27,
31]. Cellular trafficking generates exosomes, born from the so called multi vesicular body (MVB), that can become a late endosome, from which exosomes derive, or fuse to lysosomes and be degraded [
60], while macrovesicles derive from direct outward blebbing from plasma membrane [
57]; both are known to mediate intercellular communication. Exosomes can be considered a frozen image of cell conditions, since they reflect its content [
61].
In EVs originating from dendritic cells (DCs), besides MHC-I and MHC-II, CD86 costimulatory factor is present [
15]; specific integrins on EVs surface are a cell-specific signature. IECs derived exosomes contain β-defensines, antimicrobial molecules immunoglobulins and heat shock proteins (HSPs) [
15]. HSPs, a common finding also in DCs exosomes, are known to bind TLR2 and 4, as both Gram positive and negative do, leading to proinflammatory signaling [
62]. In contrast, exosomes deriving from granulocytes myeloid-derived suppressor cells are able to inhibit Th1 and induce Treg proliferation, modulating inflammation [
63]. Released exosomes are able to bind surface proteins of other cells or be internalised, in order to exert intercellular communication [
22].
Although our interest is directed to DNA, proteins found in exosomes play a fundamental modulatory effect on immune cells, thus inseparable from inflammation pathogenesis. So far, in IBDs exosomes proteins are known to be involved in (I) immunity regulation: T-regs induction as previously explained [
32] (II) regulation of intestinal barrier: annexin1 (ANXA1) exosomal levels in IBD patients serum correlates with systemic inflammation, conversely to the luminal exosomes ANXA1 enriched, that showed wound healing properties [
64]. (III) Regulation of intestinal microbiota: in paediatric IBD patients, vesicles from mucosa-luminal interface show altered proteome (ROS, MPO loaded vesicles), that apparently correlates with microbiota modifications (increased L-cysteine degradation, fungi proliferation, H2S production) and increase in microbiota defensive systems (Uracil-DNA glycosylase, a DNA damage repair system), features of aberrant host-microbiota interactions that finally lead to worsen disease activity [
65].
Haisheng Liu and coworkers [
27] analysed EV-DNA trough nano-flow-cytometry at a single vesicle level from human colorectal cancer cell line (HCT-15) and normal human colon fibroblast cell line (CCD-18Co), adding precious information to the currently limited knowledge on the topic. By the results achieved, it become clearer that cfDNA can be contained either on surface and inside a vesicle, in a double-strand shape (ds-cfDNA) or single stranded (ss-cfDNA). Length oscillates from 200 bp to 5000 and no histones are present; thus, surface EV-cfDNA is labile to DNase activity. Conversely, inner EV-cfDNA is protected by EV envelope and so very stable. Moreover, EVs exist in two size peaks: large EVs, that might be called microvescicles, of 80-200 nm diameter, containing less and smaller DNA fragments (0.2-2 Kbp), and small EVs (<100 nm), called exosomes, that carry longer and more DNA fragments.
4. CfDNA and eccDNA in inflammation
As described above, cfDNA are found either associated to MPs or free. Both display influential activity on the two immune system branches, with particular reference to the innate immune cells, as it follows.
Innate immunity is a branch of immune system able to detect pathogen derived ‘non self’ molecules, known as pathogen associated molecular patterns (PAMPs), trough pattern recognition receptors (PRRs), as Toll Like Receptors (TLRs), NOD-like receptors (NLRs)(Janeway, 1989) [
77], explaining pathogen mediated inflammation. PAMPs free-inflammation, on the other hand, was better understood when Polly Matzinger [
78] proposed the ‘danger’ theory, claiming that molecules that reflect cellular health, released during cellular distress or tissue damage, called damage associated molecular patterns (DAMPs), can induce activation of the same immune cells, leading to ‘sterile-inflammation’. Since then, many molecules, including cfNAs have been identified as DAMPs. Specifically, both DAMPs and PAMPs are ligands of TLR, that are known to be located either on the plasmatic and endosomal membrane [
79]. Among these latter, that are known to be nucleic acids specific [
79], TLR9 is able to sense foreign (pathogens, mainly bacteria) and host DNA motifs, such as CpG islets [
80] and cause downstream NF-kB signaling.
cfDNA-binding molecules, such as histones, are the ligands of TLR2 and 4, whose activation results in the production of TNF-α, IL-6, IL-10 and MPO [
80,
81,
82]. Moreover, histones are able to elicit NETosis, hesitating in more histones release [
83] and so propagating inflammation in a loop; notably, core histones act as auto-antigens in systemic lupus erythematosus (SLE) [
84]. Nucleosomes, on the other hand, exert stimulation of different inflammatory pathways that still lead to cytokine secretion [
85,
86]. High-mobility group box1(HMGB1)-DNA complexes bind RAGE (receptor for advanced glycation end products) and are carried by early endosomes for TLR9 recognition, causing activation of DCs and B cells [
87].
Almost all nuclear components are massively exposed to extracellular environment or to blood stream during cell death and NETosis, displaying their immunogenicity especially in rheumatic disease, where ANAs abundantly circulate and eventually bind naked cfDNA, EV-cfDNA, forming immune complexes [
20], thus propagating autoimmunity-related complications.
Both nc and mtDNA are able to initiate cyclic guanosine monophosphate (GMP)– adenosine monophosphate (AMP) synthase (cGAS)- stimulator of interferon protein (STING), STING-NF-kB or protein absent in melanoma 2 (AIM2) cascade, either from direct binding to cell surface via MVs or leaking into cytosol, being sensed by genes stimulators [
88,
89]. Mitochondrial DNA, found in MVs or naked, is not protected by histones and so is smaller than ncDNA (30-80 bp) [
90]. Still, it is able to initiate innate immunity cells activation via TLR9 [
37,
91,
92], because of its particular similarity to bacterial DNA, or it can mediate communication between adaptive immunity cells, like T-lymphocytes and dendritic cells [
101]. Coherently with their role as DAMPs, cfDNAs (both exo and endogenous) are able to induce either primary and secondary hemostasis, thus being potentially involved in threatening conditions such as DIC (disseminated intravascular coagulation) [
46,
93].
As previously explained, many types of circulating nucleic acids are described across literature [
94,
95], but specifically circulating circular DNA is a less explored domain. Nonetheless, the few existing works already helped to gain a better insight. Wang et al. with their latest work highlighted concepts of great importance. They demonstrated that eccDNA is circular genetic fragment deriving from randomly chosen genome sequences, partly generated by cells undergoing apoptosis, that acts as a full-fledged DAMP, able to induce activation of innate immunity (in vitro DCs), throughout a cGAS-STING and TLR9 fashion, that results in INF type I, cytokines and chemokines production [
29].
Obermeier et al. showed that bacterial cfDNA rich in CpG motif activate TLR9, worsening the course of DSS-induced colitis [
96]. CpG motifs are typical of bacterial free DNA, but just recently attributed either to human cfDNA [
80]: in particular, the above mentioned micro eccDNA is an example. This demonstrates how endogenous and exogenous cfDNA share similarities, and that can both activate immune response in a sequence-dependent manner; however, Li et al. in 2012 [
82] added novelty to this model, demonstrating what Wang’s team later confirmed: eccDNA potency is to be attributed to its circular shape, but not sequence, since synthetic circular DNA triggered almost the same response and, above all else, that its linear counterpart was not nearly as potent as eccDNA [
29]. Specifically, it was proposed that DNA curvature, induced by high-mobility group box 1 proteins (HMGB1) and histones H2A, H2B, significantly enhanced TLR9 binding, in a stereo-specific less than sequence-dependent mechanism. Finally, these studies suggest that self-derived circulating DNA (both non-circular cfDNA in nucleosomes or naked, and eccDNA) are able to induce and sustain the inflammation machinery, especially the ‘sterile inflammation’, which is typical of autoimmune disease, where they actually seem to be involved [
84,
97].
Conversely, other studies propose that preconditioning with methylated and unmethylated genomic DNA isolated from a probiotic mixture (VSL#3), followed by DSS colitis induction in mice, have anti-inflammatory effect, in a TLR9-dependent way, compared to control (TLR9 ko mice), that didn’t show any effect [
98]. Likewise, Műzes et al. revealed that pretreatment with a single intravenous injection of colitic cfDNA in DSS-induced colitis mice showed increased TLR9-macroautophagy response and up-regulation of Baclin1 expression in the colon, with a decreased disease activity as compared to normal cfDNA injection [
99]. These findings confirm the suspect that cfDNA exerts different behaviors depending on its origin (distressed cells or normal) and on the features of the local immunobiological milieu (inflamed or normal).
As previously mentioned, autophagy can occur following the activation of TLR9 [
100]. Since DAMPs can activate TLR9, especially via CpG oligonucleotides [
100], it can be stated that cfDNA sensing and autophagy are related: injection of cfDNA triggered increase of TLR9 mediated autophagy, that apparently supported cellular fitness within an inflamed environment, reasonably explaining its protective effect [
101]. However, origin of such cfDNA and mechanisms underlying the anti-inflammatory response are still being elucidated.
In 2021, Zhao et al. demonstrated that levels of EVs containing ncDNA, mtDNA and proteins were increased in both plasma and colon lavage from DSS induced active colitis in mice, and in IBD patients, and that these EV were secreted by IECs [
102]. Macrophages cultures enriched in EV from IBD patients showed inflammatory phenotype and activation of STING pathway, leading to cytokines synthesis. Further analysis declared that dsDNA is necessary to STING activation [
31,
102]. EV-cfDNA levels adequately correlated whit DAI and CDAI (disease activity index, Crohn disease activity index), suggesting that inflammation and cell damage trigger EV release and vice versa, and allowing to definitely confirm that EV-cfDNA could be an activity marker. These results propose that either EVs-cfDNA and cf-circulating DNA [
28] are probably involved in immune stimulation to develop colitis in both mice and human, throughout different mechanisms but reaching the same result.
Much evidence suggests that EVs carrying nuclear or cytosolic auto-antigens, generated during apoptotic processes [
103], undergo biochemical modification that lead to formation of modified self-molecules that break self-tolerance [
104] and trigger autoimmune response. Posttranslational biochemical modification, such as citrullination, can occur in several cellular processes including apoptosis, autophagy and NETosis, contributing to auto-antigen formation in autoimmune disease [
103,
105]. This model has been approved for Sjögren syndrome, SLE, rheumatoid arthritis (RA), type-1 diabetes mellitus (T1DM) [
105,
106]. Specifically, citrullinated proteins are found in synovial fluid EVs from RA patients, neutrophils deriving from RA patients show citrullinated vimentin, that is a known autoantibody target in RA, and citrullinated histones [
107]. These evidences suggest that EVs derived or DNA bound citrullinated proteins, the so-called “citrullinome” correlate to inflammation severity in the above-mentioned chronic inflammatory disease. Coming to IBD: higher expression of citrullination peptides have been found in colonic biopsies of IBD patients, confirming correlation between citrullination and inflammation [
108]; however, these results are not significant and further analysis are needed. Moreover, M.-L. Liu et al. preliminary analysis revealed that tobacco smoke extract-treated neutrophils release MVs with an inhibitory effect of macrophage phagocytosis [
109,
110]. Hence, high apoptotic rate and NETosis, characteristic of autoimmune disease (AID) and chronic inflammatory diseases, correlate with increased MVs release and clearance slowdown, resulting in accumulation of apoptotic cells derived auto-antigens and immune complexes, worsening SLE and cutaneous LE [
109]. In conclusion, it is clear how MVs and an unbalanced, non self-resolving immune system can truly mediate modification of local and distant microenvironment, concurring to pathogenesis and progression of AID.
Overall, the above-mentioned studies reveal that in IBD microenvironment, damage is mainly directed to colonocytes and enterocytes, which have been demonstrated to maximally contribute to cfDNA levels in situ and in plasma of induced colitis. Disruption of intestinal epithelial cells (IECs), lining the surface of intestinal mucosa, results in a leaky gut, a way that may increase translocation of exogenous and microbial cfDNAs to circulation. In vivo studies of induced colitis in mice suggest that probably all those mechanisms of cell stress could act in concert to increase cfDNA plasma levels over time with disease progression [
28]. What seems clear is that IECs, immune system and microbiota fit together into this loop, in which cfDNAs play a dominant role across all the disease progression over time, as Maronek et al. reviewed [
28], which would imply that eccDNA in circulation has a similar role in disease progression. Consequently, the amount of cfDNA could correlate with DAI and be used as a diagnostic and monitoring tool instead of invasive examinations to assess disease severity. Finally, studies highly stressed how damaged or even normal genetic material can truly affect immune activity; however, although propagation mechanisms are ever clearer, initiation ones aren’t so. That’s the reason why IDB are still a tricky and obscure condition to heal.
7. Concluding remarks and future perspectives
In this review we performed a comprehensive overview of the role of cell-free DNA and circular DNA in inflammatory bowel disease, ranging from potential diagnostic to therapeutic applications. As we reviewed above, different types of cfDNAs can be detected in different body fluids, such as blood or feces, potentially representing a novel noninvasive and easily accessible biomarker. Rising from “passive” processes of cell death (necrosis, mechanical damage, etc), “active” processes of cell death (apoptosis, NETosis), defects in DNA repair mechanisms, oxidative stress and inflammatory environment, especially for eccDNA, and processes of cell-to-cell communications (microvescicles, exosomes, nanovesicles), the concentration of cfDNA seems to be associated with the severity of disease in IBD patients and murine models of colitis.
In each process, cfDNAs represent detectable signals of the processes which are ongoing in the organism, and of its own reaction, thereby, giving researchers the possibility of both passively monitoring or actively controlling these processes. Growing evidences show that cfDNAs play important roles in several inflammatory diseases, including IBD, most of whom are still unknown, requiring further research. The creation of in vitro models would allow us a better understanding of the exact meaning of these processes. Notably, Moller et al. created a dual-fluorescence biosensor cassette, which upon the delivery of pairs of CRISPR/Cas9 guide RNAs is able to generate in human cells eccDNA from intergenic and genic loci, of different sizes, thus, allowing researchers to study the cellular impact, persistence and function of eccDNAs in different tissues, such as human intestinal epithelium [
72].
Furthermore, one goal for the future will be represented by focus on the rapid progression of single cell-based studies on immune system disorders, autoinflammatory or autoimmune diseases. In the past decade, single-cell RNA sequencing (scRNA-seq) has driven a true revolution in the field of immunological research, since it allows a better understanding of the heterogeneity associated with individual immune cells and immunological responses at the molecular level both under physiological and pathological conditions [
222]. The development of similar technologies for the purification and sequencing of cfDNA, and especially eccDNAs will open new roads to translational research in the field of immunology and cancer. Besides, it is urgently needed to innovate more bioinformatics tools for data analysis and integration from different technologies and data types, in order to standardize the immune cells annotation and to utilize the whole spectrum of available data.
Therefore, cfDNA may represent the much-needed noninvasive biomarker, from liquid biopsy, for early diagnosis, diseases monitoring, prediction of drug response, and early detection of progression to dysplasia in IBD.
Herein, we reviewed the current state of art regarding the development of molecular therapies in IBD, paying attention to the role played by cGAS-STING signaling, TLR9 signaling, and discussing several other potential oligonucleotide-based therapies in IBD. We also discussed the rationale of employing NPs for drug delivery and provided novel interesting insights into the use of different molecules, such as eccDNA for the treatment of IBD.
However, further research is needed to better characterize the origins, functions, and biological features of cfDNA and eccDNA, which could contribute to the elucidation of IBD pathogenesis and to the development of novel molecular therapeutic strategies.