Intertwined roles of microRNA-155 and metformin in osteoarthritis: Novel potential diagnostic, prognostic, and therapeutic modulators

Abstract

Osteoarthritis (OA) is a chronic joint disease characterized by cartilage degradation, synovial inflammation, and subchondral bone remodelling. Despite its increasing prevalence, effective diagnostic, disease-limiting, and therapeutic strategies remain unattainable. Recent studies have recognized the involvement of microRNA-155 (miR-155) in the pathogenesis of OA and most of its risk factors while also identifying the antidiabetic drug metformin as a potential modulator of disease progression. MiR-155, a key endogenous regulator of the immune system, mechano-transduction, and multiple genetic pathways, interacts with OA targets of cellular energetic and circadian homeostasis, promoting systemic and local articular inflammation, cartilage matrix degradation, and chondrocyte apoptosis. Metformin, widely used for type 2 diabetes, has demonstrated anti-inflammatory, anti-oxidative, and chondroprotective properties in OA, mainly through its activation of adenosine monophosphate-activated protein kinase and inhibition of nuclear factor kappa-B signalling. Enthrallingly, metformin targets the same cellular pathways as miR-155 with emerging evidence also suggesting miR-155 expression modulation, indicating synergistic, potentially disease-modifying effects in OA. This review highlights the central role of miR-155 in OA pathophysiology and its potential as a biomarker for disease diagnosis and progression. MiR-155 targeting – through microRNA therapeutics (mimics/antagomiRs) and/or metformin – could pave the way for innovative treatments, including novel articular delivery systems and cell-based therapies.

Key Words: Osteoarthritis; MicroRNA-155; Metformin; Glucagon-like peptide 1; Renin-angiotensin aldosterone system; Cyclooxygenase 2; Adenosine monophosphate-activated protein kinase

Core Tip: In osteoarthritis (OA) and its associated risk factors, elevated microRNA-155 (miR-155) regulates systemic and local joint inflammatory mediators, extracellular matrix degradation, and chondrocyte apoptosis, with context-dependent roles in cartilage repair. MiR-155 also shows promise as a diagnostic and prognostic biomarker. Metformin exerts cartilage-protective effects via adenosine monophosphate-activated protein kinase activation, nuclear factor kappa-B inhibition, and possible modulation of miR-155 expression. Combined miR-155-targeted strategies and metformin in OA could enhance anti-inflammation, cartilage-preservation, and joint regeneration through the development of innovative treatments, including novel articular delivery systems and cell-based therapies. Further research is needed to fully understand the molecular underpinnings of this relationship.

INTRODUCTION

Osteoarthritis: General considerations

Osteoarthritis (OA) is a degenerative joint disease characterized by the progressive degradation of articular cartilage, synovial inflammation, and changes to the subchondral bone leading to irreversible joint damage and deformity. OA with its rising prevalence primarily affects older adults, leading to pain, stiffness, and decreased mobility, all of which significantly contribute to additional years living with disability[1,2]. Etiologically, it is widely accepted that OA occurs as combination of many risk factors, such as increasing age, female sex, obesity, trauma, abnormal mechanical load as in heavy work and competitive sports, lack of nutrient supply, and genetic predisposition affecting cartilage, bone and synovial tissues[3]. Atherosclerotic (cardiovascular or cerebrovascular) disease and metabolic comorbidities, often in coexistence with OA, display certain associations sharing common pathogenetic pathways via obesity[4].

OA affects all joint components; the articular cartilage, the synovium, its capsule, and the subchondral bone with the surrounding muscular tissue and its ligaments are involved in the disease process[5]. The pathogenesis of OA results from age-and adiposity-compounded failure of synovial joint tissue to maintain metabolic homeostasis; when the destructive effects of accelerated chondrocyte and extracellular matrix (ECM) senescence, inflammation, metabolic insults, and mechanical stress outpace the joint’s ability to repair and regenerate important tissues, OA ensues[6,7].

Articular hyaline cartilage is the central component in all articular surfaces and in the adult is considered a permanent tissue with negligible turnover throughout adult life. It is composed by a single cell type, the chondrocytes, and the ECM they provide. The ECM makes up over 90% of the cartilage and is comprised primarily of collagen (type II collagen-Col2a1), proteoglycans (Aggrecan-ACAN), and water molecules[8]. Inflammation driven degradation of the ECM components, Col2a1 by metalloproteinase (MMP-3/MMP-13) and ACAN by aggrecanases [disintegrin and MMP with thrombospondin motif: (1) Adamts-4; and (2) Adamts-5], is regarded as a crucial pathogenic process in OA[8,9]. The successful replacement of ECM components by the resident chondrocytes is thus vital to a joint’s reparative process and is overwhelmingly mediated through mechano-regulation and growth factor release sequestered in pericellular matrix of the ECM in order to accommodate shifting metabolic requirements[8].

Articular cartilage anabolism and/or anti-catabolism signalling pathways involve transforming growth factor β1 (TGF-β1), insulin-like growth factor 1 (IGF-1), hypoxia-inducible factor 1 alpha (HIF-1α), and bone morphogenetic proteins (BMPs). Catabolic signals involve interleukin-1 (IL-1), IL-6, HIF-2α, and fibronectin fragments[10]. The molecular mechanisms underpinning OA thus are complex, involving inflammation, oxidative stress, disturbed mitochondrial dynamics, metabolic alterations, and aberrant cellular signalling.

In the following sections of this review, we will elaborate on how microRNA-155 (miR-155) as an endogenous regulator of immunity, inflammation, mechano-transduction, and multiple genetic pathways, interacts with OA targets of cellular energetic and circadian homeostasis that are also targeted by metformin. Furthermore, understanding how metformin influences miR-155 homeostasis in different cell contexts could help devise new therapeutic strategies to alleviate severe imbalances between anabolism and catabolism in articular hyaline cartilage and limit OA progression. Finally, microRNAs (miRNAs or miRs) therapeutics comprising miR-155 mimics and antagomirs in appropriate delivery systems and suitable molecule combinations, potentially also including cellular products to enhance cartilage regeneration will be reviewed.

MIR-155: GENERAL CONSIDERATIONS

Regulation of miR-155 biogenesis and expression

MiRNAs are small non-coding RNAs, 21-25 nucleotides long, produced in human cells. Over 2600 human and more than 250 viral miRNAs have been identified that regulate gene expression by binding to target messenger RNAs (mRNAs) and repressing translation[11,12]. Mature miRNAs contain a 2-8 nucleotide long seed sequence in their 5’ untranslated region (UTR) and can bind to a target mRNA’s corresponding 3’ UTR, thereby repressing (silencing) gene expression. Despite miRNAs comprising less than 3% of the human genome they can translationally repress and downregulate expression in the majority of the human genes and/or other ncRNAs[13,14]. Since a single miRNA can target multiple mRNAs, and conversely, an individual mRNA can interact with various miRNAs, the effects of miRNAs are both context-dependent and pleiotropic. Through intricate regulatory gene networks, miRNAs can modulate diverse host cellular processes in every conceivable way. MiRNAs mediate their effects between cells by exosomes, small extracellular vesicles (30-150 nm in diameter) secreted from different cell types, with miRNAs as part of their cargo[15]. Altered miRNA levels have been described in joint tissues and serum of OA patients, and are thought to mediate the effects of synovial inflammation, chondrocyte metabolic disruption, articular mechanical overload, as well as those of OA associated risk factors, critically linking miRNAs to OA pathogenesis[16-20].

The miR-155 is encoded by its host gene, MIR155HG, originally identified as the B-cell integration cluster gene and located on chromosome 21q21. Its biogenesis follows the regular miRNA biogenesis process as depicted in Figure 1[21-23]. As one of the first and most well studied miRNA discovered over 25 years ago, 22-nucleotide long miR-155 is an evolutionary preserved, pleiotropic, and multifunctional miRNA. With a central role in the regulation of the immune system, miR-155 is primarily expressed in the thymus and spleen with very low physiological levels[21,24-26] but swiftly expressed upon pathogen invasion or injury[25]. Anti-inflammatory cytokines (e.g. IL-10) and anti-inflammatory molecules (natural and synthetic glucocorticoids) will suppress its expression. Moreover, other miRNAs can regulate miR-155 expression such as miR-1 and miR-146 that can inhibit its processing to mature form or counterbalance its inflammatory functions, respectively[25]. The miR-155-miR-146 regulatory interplay is particularly important as through their respective binding with nuclear factor kappa-B (NF-κB) these 2 miRs can appositely counterbalance inflammation[26].

Figure 1 The regular microRNA biogenesis process. Schematic representation of microRNA-155 (miR-155) biogenesis. MIR155HG is transcribed in the nucleus by an RNA polymerase II (RNA pol II) to a primary miR-155 (pri-miR-155). This pri-miR-155 will be processed further by the Drosha-DiGeorge syndrome critical region gene 8 (microprocessor) complex in the nucleus to produce a 65-nucleotide stem-loop precursor miRNA (pre-miR-155). Exportin-5 exports the pre-miR-155 from the nucleus to the cytoplasm where it will be processed by the RNase III enzyme, Dicer, resulting in RNA duplexes of 22 nucleotides, the miR-155-5p and miR-155-3p. After Dicer cleavage, an Argonaute protein binds short RNA duplexes to form the core of the RNA-induced silencing complex (RISC) that incorporates the mature miR-155. Once bound to RISC the guide strand is retained and functional while the other (passenger strand) is degraded. The RISC/guide strand-mature miR-155 complex can now seek and recognize its target messenger RNAs and bind to it/them through complementary base pairing interactions on their 3′-UTR and its seed region (nucleotides 2 to 8) on the 5’UTR leading to translational repression and/or degradation. MIR155HG also encodes a long noncoding RNA-155 involved in the regulation of antiviral innate immunity and can function as a competing endogenous miRNA sponging and downregulating other miRNAs. Ago: Argonaute; DGCR8: Drosha-DiGeorge syndrome critical region gene 8; lncRNA-155: Long noncoding RNA-155; miR-155: MicroRNA-155; mRNA: Messenger RNA; RISC: RNA-induced silencing complex; RNA pol II: RNA polymerase II; TRBP: The human immunodeficiency virus transactivating response RNA-binding protein.

Demonstrating its pleiotropy, the MIR155HG gene transcription is cell-context dependent, containing multiple transcription factor binding sites that allow nuanced regulation of its transcription depending on which cell type responds to a shifting homeostasis (Table 1)[14]. Binding sites for NF-κB, mothers against decapentaplegic homolog 4, interferon-sensitive response element, interferon regulatory factors, activator protein 1 (AP-1), erythroblast transformation-specific proto-oncogene 1, Forkhead box protein P3, HIF-1α (3 sites), have been described[14]. The presence of three HIF-α binding sites confirms an important role for miR-155 in the anoxic environment chondrocytes reside (Table 1)[14,27,28]. Several other binding sites exist, and their number can vary depending on cell context.

Table 1 Transcription factors binding sites in the MIR155HG gene.

Number of MIR155HG transcription binding sites
Three (3) sites	Two (2) sites	One (1) site	Oncology-associated
Hypoxia-inducible factor-1 alpha	Nuclear factor kappa-B, forkhead box protein P3 (associated with oncogene special AT-rich binding protein-1), ETS proto-oncogene 1	RNA Polymerase II Subunit A, mothers against decapentaplegic homolog 4, interferon-sensitive response element, interferon regulatory factors, Activator protein 1, CCAAT/enhancer-binding protein alpha), ETS family member ELK3	DNA promoter methylation and transcription factor (SP1), v-myb myeloblastosis viral oncogene homolog, Annexin A2, breast cancer 1, early onset

ETS: Erythroblast transformation-specific.

To better comprehend the pleiotropy of miR-155 effects one needs to realize that miR-155 responses depend on its inducibility of expression (miR abundance or paucity) in different cell types and its seed complementarity for target mRNA binding site in those cells. Complementarity can vary between canonical binding sites with a perfect 6-8-mer seed match and non-canonical, perhaps weaker, binding sites. In addition, not all miR-155 targets are expressed in all cell types. This allows for cell-context dependent repressions and nuanced cellular effects[29].

MiR-155 functions

MiR-155 exerts its physiological functions through distinct expression profiles and with cell-context dependent pleiotropy. With a predicted target repertoire of over 900 genes, miR-155 is the most important miRNA in cardiovascular pathology but also displays decisive roles in many other physiological and pathological biological processes, such as hematopoietic cell line differentiation, inflammation response, immunity (especially in relation to viral and parasitic infections), and notably, tumor formation[24,30,31]. MiR-155 is centrally involved in the regulation of both the innate and adaptive arms of the immune system especially regarding proinflammatory responses, chemokine upregulation, immune interferon (IFN) response, immune cell activation, macrophage activity/polarization, monocytic differentiation, T helper 1 cells, T regulatory (Treg) cells, natural killer (NK) cells, dendritic cells, and cytotoxic T lymphocyte functions, thereby critically impacting functionality and shaping anti-tumor immunity[32].

In oncogenesis, miR-155 is the most well described of oncomiRs, aberrantly expressed in breast cancer, non-small cell lung cancer, prostate cancer, squamous cell carcinoma, and B-cell lymphoma[33]. Its presence in those tumors indicates a pathogenetic role as well as diagnostic and prognostic potential[33]. Demonstrating its distinctive context dependent pleiotropy, on the one hand, miR-155 upregulation can indicate tumor occurrence and poor prognosis, but, on the other hand, it also promotes robust anti-tumor immune responses through its effects on NK-cells, CD8+ T cells, and immune checkpoint inhibitors. MiR-155 is essential for CD8+ T cell responses to virus, vaccination, and cancer. Its absence impedes the control of virus replication and tumor growth[34]. Moreover, the efficacy of dendritic cell-based cancer vaccines is improved through enhanced miR-155 expression[35].

MiR-155-5p has been shown to be elevated in circulation with obesity and metabolic dysfunction in humans[36] while also being a key mediator of diet-induced obesity in female mice[37]. A recent network analysis has confirmed miR-155-5p as a circulating biomarker that characterizes obese subjects affected by type 2 diabetes mellitus (T2DM) and its involvement in the pathogenesis and complications of both autoimmune type 1 diabetes mellitus and T2DM[38]. Importantly, weight loss following bariatric surgery leads to a reduction in miR-155-5p levels[39].

As a central regulator of immune responses[16], miR-155-mediated dysfunction of immune cells contributes to the development of inflammatory autoimmune diseases among which rheumatoid arthritis (RA)[40-44]. Upregulation of miR-155 has been reported in serum, blood, synovial fluid and joint tissues in patients with RA compared to OA and normal controls[45]. RA-associated miR-155 upregulation is seven to eight times higher compared to OA[40,45]. Already in 2008, Stanczyk et al[40] showed that MMP-3 induced by the pro-inflammatory cytokine IL-1β can be downregulated by elevated miR-155 in RA synovial fibroblasts and proposed a protective function for this miR that by local down-regulation of certain MMPs, excessive tissue damage due to RA inflammation can be controlled. Similar MMP results are reported in other tissues[46].

DYSREGULATION OF MIR-155 IN OA

Recent studies have demonstrated the important role of miRNAs in the pathogenesis of OA[47-50]. Below we will present novel evidence on how miR-155 initiates articular inflammation through ECM transduction of mechanical signals and how disrupted miR-155 levels trigger chondrocyte metabolic, energetic and circadian disruptions. Deeper understanding how these intricately interconnected miR-155 networks modulate OA could lead to precise and personalized therapeutic interventions.

MiR-155, obesity, and mechanical overload-induced inflammation in OA

Aging, obesity, and female sex are the risk conditions that most significantly predispose for OA (especially knee OA) development (Figure 2)[5]. The chronic low-grade systemic inflammation and general cell senescence that characterizes aging is intensified by the adipokine-mediated inflammatory milieu of obesity[51]. Sarcopenia, frequently associated with both aging and obesity, disrupts the biomechanical stability of weight bearing joints causing sustained mechanical stress that eventually leads to joint inflammation where destructive enzymes ultimately cause irreversible injury, thereby initiating and maintaining articular degradation[52]. Concurrent lack of exercise exacerbates muscular weakness and sarcopenia, while a dysregulated metabolic profile, chronic pain, and aging may additionally initiate circadian rhythm disruptions[53,54].

Figure 2 Risk factors in osteoarthritis. Aging, obesity, sarcopenia, type 2 diabetes mellitus (T2DM), and female sex, all important risk factors in osteoarthritis (OA), promote microRNA-155 (miR-155) overexpression in OA through chronic inflammation and mechanical overload that through numerous gene repressions/modulations perpetuate an established global inflammatory state and promote a local articular proinflammatory milieu with elevated metalloproteinase (MMP), disintegrin and MMP with thrombospondin motif (Adamts), Runx2, NOD-like receptor family pyrin domain-containing 3, caspase-1 (Cas-1), Gasdermin D resulting in chondrocyte catabolism and apoptosis, increased chondrocyte senescence, reactive oxygen species generation, autophagy disruption, ultimately contributing to augmented extracellular matrix degradation. Increased miR-155/nuclear factor kappa-B signaling promotes interleukin (IL)-1β/IL-6 secretion, chondrocyte hypertrophy, synovial M1 macrophage recruitment initiating synovial inflammation and subchondral bone remodeling. Adipokines, visfatin and resistin, potentiate miR-155 expression and effects. Piezo-Type Mechanosensitive Ion Channel Component 1 mediates mechanoelectrical transduction of inflammatory responses in the chondrocyte and up-regulates miR-155. Gene functions and gene abbreviations can be found in Table 2. AMPK: Adenosine monophosphate-activated protein kinase; Cas-1: Caspase-1; CASP3: Caspase 3; ECM: Extracellular matrix; GDF6: Growth differentiation factor 6; GSDMD: Gasdermin D; IGF-1: Insulin-like growth factor 1; IKBKE: Inhibitor of nuclear factor kappa-B kinase subunit epsilon; IL: Interleukin; MiR-155: MicroRNA-155; MMP: Metalloproteinase; MΦ: M1 macrophage; NF-κB: Nuclear factor kappa-B; NLRP3: NOD-like receptor family pyrin domain-containing 3; OA: Osteoarthritis; PIEZO1: Piezo-Type Mechanosensitive Ion Channel Component 1; PPARγ: Peroxisome proliferator-activated receptor γ; ROS: Reactive oxygen species; Runx2: Runt-related transcription factor 2; SOCS1: Suppressor of cytokine signaling 1; Sox9: SRY-related HMG box 9; STAT3: Signal transducer and activator of transcription 3; TNF-α: Tumour necrosis factor-alpha; T2DM: Type 2 diabetes mellitus; ↓: Decline; ↑: Increase.

The overrepresentation of female sex in knee OA may be partly attributed to a lower initial healthy cartilage volume and the subsequent postmenopausal hormonal profile which may render articular cartilage less conducive to regeneration[55]. MiR-155 is upregulated in female OA[56] and a female preponderance to altered miR-155 homeostasis in inflammation and obesity has been observed as female-derived innate lymphoid and T helper cells responded to various stimuli with higher miR-155 expression levels and numbers of miR-155-expressing cells than males[57]. Moreover, miR-155 elevated in obesity at risk for complications, is a key mediator of diet-induced obesity in female mice as its deletion prevented obesity development (Figure 2)[36,37].

In addition, obesity and T2DM-associated hyperglycemia with insulin resistance further exacerbate oxidative stress and result in a pro-inflammatory state with an increase in NF-κB activation as it has been demonstrated in adipose and vascular tissues[58]. In support to the above observations, up-regulation of miR-155 was also confirmed in biopsies of obese subjects where the induction of miR-155 was correlated with tumour necrosis factor-alpha (TNF-α) expression and to body mass index (BMI)[59]. Karkeni et al[59] and Tryggestad et al[60] showed that overexpression of miR-155 in adipocytes and macrophages respectively, induces a global inflammatory network response through chemokines secreted by inflamed adipocytes, that recruit leukocytes and increase macrophage migration in adipose tissue, robustly ushering the establishment of a proinflammatory status, at least partly via the miR-155-mediated downregulation of peroxisome proliferator-activated receptor γ (PPARγ) (Figure 2).

In an effort to counteract ECM degradation, hypertrophic chondrocytes appear displaying increased synthetic activity whereby degradation products, proinflammatory mediators (TNF-α, IL-1β and IL-6) and macrophages accumulate become activated and send stimulatory signals to the adjacent synovia[61]. Synoviocytes turn proliferative, M1 polarized macrophages accumulate further, synovia hypertrophies and becomes increasingly vascular. Furthermore, subchondral bone turnover increases and along with vascular invasion of the cartilage initiates bone remodelling and repair. Later, persistent inflammation, mechanical overload with abnormal joint mechanics will lead to osteophyte development[3].

MiR-155 is a proinflammatory regulator in clinical and experimental arthritis, significantly upregulated in OA and positively correlated to clinical stage, while its deficiency protected against the development of collagen-induced arthritis[17,42,62]. This miR-155 background provides the molecular basis for the inflammatory connection between abnormal mechanical loading and adiposity in OA as both have been reported associated with miR-155 dysregulation[63,64]. Chondrocytes sense the mechanical overloading through Piezo-Type Mechanosensitive Ion Channel Component 1 that mediates mechanoelectrical transduction of inflammatory responses, and up-regulate miR-155-5p which in turn, represses the downstream target gene growth differentiation factor 6 accelerating chondrocyte senescence and cartilage degradation (Table 2, Figure 2)[2,14,27,42,47,59,61,65-94].

Table 2 Direct gene targets of microRNA-155 relevant to osteoarthritis.

Gene symbol	Full gene name	Action	Ref.
AGTR1	Angiotensin II type 1 receptor gene	AGTR1 gene repression downregulates its translation, thereby lowering AT1R membrane expression and downstream signalling, like endogenous AT1R blockade redirecting Ang II towards its alternative receptor Ang II type 2 receptor	Yang et al[88]
ARG2	Arginase2	ARG2 upregulates MMP-3 and MMP-13 via the NF-κB pathway, causing destruction of OA cartilage; repressing ARG2 impairs oxidative phosphorylation, increases and stabilises HIF-1α and could represent an initial protective mechanism in OA	Dunand-Sauthier et al[80]
ATG3, ATG5, ATG14, FOXO3, GABARAPL1, MAP1 LC3, ULK1, RICTOR	-	Repression of autophagy related genes results in potent suppressor of autophagy in human chondrocytes	D'Adamo et al[47]
BACH1	BTB and CNC homology 1, basic leucine zipper transcription factor 1	Translational repression of BACH1 leads to potent anti-inflammatory, cytoprotective, antioxidant programs through heme oxygenase	Takada et al[92]
Bmal1	Brain and muscle Arnt-like protein 1	MiR-155 induction can repress Bmal1 directly in macrophages, endothelial cells and bone marrow mesenchymal stem cells. Disturbed Bmal1/HIF-1α interactions, a crucial pathway in chondrocyte homeostasis in hypoxia, leads to enhanced MMP-13 levels and catabolic chondrocyte effects. MiR-155-induced disturbance of Bmal1 disrupts endochondral bone formation via melatonin receptor 1/AMPKβ1/Bmal1 signaling axis	Curtis et al[74], Liang et al[75], Lee et al[76], Yu et al[77]
Cab39	Calcium-binding protein 39	Cab39 is a component of the trimeric LKB1-STRAD-Cab39 complex and regulates the activity of LKB1 and thus activates the phosphorylation of AMPK, thus, if repressed by miR-155, would hamper AMPK activation and negatively impact chondrocyte survival	Shi et al[69]
CASP3	Caspase 3	Knee OA synovial fluid increased miR-155-5p-induced inhibition of macrophage apoptosis by targeting CASP3	Li et al[61]
C/EBPβ	CCAAT/enhancer binding protein β	Repression of C/EBPβ and its downstream target genes nuclear factor erythroid 2-related factor 2, SOD1, and hemeoxygenase-1, thereby inducing reactive oxygen species generation. C/EBPβ repression leads to miR-143 repression, thus resulting in upregulation of hexokinase 2 expression. MiR-155 C/EBPβ repression could possibly lead to peroxisome proliferator-activated receptor γ downregulation	Karkeni et al[59], Onodera et al[78], Jiang et al[82]
COX-2	Cyclooxygenase 2	As COX-2 is highly expressed in osteocytes and gives rise to abnormal subchondral bone formation, it is of importance that NF-κB induces miR-155/COX-2 expression in macrophages and that miR-155 can bind COX-2, induce COX-2 reporter activity, and maintain mRNA stability, thereby potentially affecting subchondral bone remodeling	Yuan et al[89], Qiu et al[94]
Ets-1	E26 transformation-specific sequence-1	Ets-1 plays a role in OA inflammation and angiogenesis through its regulation of MMPs	Mahesh and Biswas[14]
GDF6	Growth differentiation factor 6	Activation of Piezo-Type Mechanosensitive Ion Channel Component 1 leads to the upregulation of miR-155-5p which represses the downstream target gene GDF6 and accelerates chondrocyte senescence and cartilage degradation	Qin et al[65]
HIF1A	Hypoxia inducible factor 1A	HIF1A is a direct target of miR-155 and decreases the HIF-1α mRNA. HIF-1α and miR-155 are together in a feedback loop whereby HIF1a induces miR-155 in hypoxia. Functional HIF-1α is required for energy production in chondrocytes though its control of basal glycolytic enzymes, phosphoglycerate kinase, and glucose transporter. The expression of HIF-1α, a crucial regulator in chondrocyte homeostasis in hypoxia, is decreased through Bmail1 disruption leading to enhanced MMP-13 Levels and catabolic chondrocyte effects	Zhang et al[27], Lee et al[76]
IGF1	Insulin growth factor 1	Negative regulation of IGF-1 downregulates IGF-1, a crucial molecule involved in repair of OA cartilage damage	Shen et al[72]
IKBKE	Inhibitor of nuclear factor kappa-B kinase subunit epsilon	MiR-155 targets IKBKE that participates in synovial inflammation, ECM destruction	Montagne et al[66], Long et al[67]
LEPR	Leptin receptor	MiR-155 repression of LEPR via inhibition of AMPK, which ultimately increases osteoclast activation and bone resorption of osteoclasts in alendronate-treated osteoporotic mice	Mao et al[91]
Mafb	Musculoaponeurotic fibrosarcoma oncogene family, protein B	Elevated miR-155 in hyperlipidemia improved glucose metabolism and the adaptation of β-cells to obesity-induced insulin resistance through suppression of Mafb and subsequent IL-6 – induced glucagon-like peptide 1 production in α-cells	Zhu et al[90]
MAPK	Mitogen-activated protein kinase	MiR-155/MAPK pathway signalling regulates chondrocyte activities and ECM degradation	Cazzanelli et al[93]
Mfn1/2	Mitofusin 1	Elevated miR-155 in OA can induce cellular senescence by regulating mitochondrial dynamics by promoting fusion through an increase in the expression of Mfn2	Wen et al[73]
PIK3R1	p85α regulatory subunit of PI3K	MiR-155 promotes chondrocyte apoptosis and catabolic activity through targeting PIK3R1-mediated PI3K/Akt pathway activation. Repressions modulate glycolysis via the PIK3R1-PDK/Akt-FOXO3a-cMYC axis	Kim et al[81]
PRKA	AMPK gene	MiR-155 targets 3’ UTR of AMPK mRNA, downregulating it. AMPK negatively controls mammalian target of rapamycin, the main negative regulator of autophagy, thereby suppressing autophagy and negatively impacting chondrocyte survival	Zhang et al[68]
Runx2	Runt-related transcription factor 2	As a direct target of miR-155, Runx2 repression in articular chondrocytes would be expected to slow OA progression. MiR-155 can regulate the RANKL expression via Runx2 mediated transcriptional inactivation; RANKL/RANK are key players in the bone remodeling process and subchondral bone remodeling	Wang et al[2], Komori[87]
RANKL	Receptor activator of nuclear factor kappa-B ligand
SHIP1	Src homology 2-containing inositol phosphatase-1	SHIP-1 repression in clinical and experimental arthritis increases the production of proinflammatory cytokines	Kurowska-Stolarska et al[42]
SIRT1	Sirtuin 1	MiR-155 directly targets and represses SIRT1, thereby indirectly downregulating and preventing AMPK activation and negatively impacting chondrocyte survival. SIRT1-downregulation activates the NLRP3 inflammasome	Hong et al[70], Lu et al[86]
Smad2/5	Mothers against decapentaplegic homolog 2/5	Elevated miR-155 in OA effectively increases Smad2 and reduces Smad5 transcription tipping the balance in the TGF-β1 (Smad2)/bone morphogenetic protein 2 (Smad5) signalling towards TGF-β1 and anabolism in an effort to counteract OA progression. MiR-155 induced Smad2 repression induces pyroptosis in knee OA via activation of the NLRP3/caspase-1 pathway	Gu et al[83], Chen et al[84], Shao et al[85]
SOCS1	Suppressor of cytokine signaling 1	Repression and decreased expression of SOCS1 contributes to the increased production of TNF-α, IL-1β and enhanced M1 macrophage polarization and apoptosis inhibition in OA	Li et al[61], Zhang et al[71]
SOD1/2	Superoxide dismutase 1/2	Binds and represses SOD1 3'UTR as well as diminishes SOD2 expression through FOXO3a repression	Bi et al[79]
TNF	Tumour necrosis factor	The miR-155 promoter was activated by TNF-α and at least partly through NF-κB leading to a direct, positive correlation between miR-155 expression and mRNA levels coding for TNF-α. Elevated miR-155 induces inflammatory response, chemokine expression, and macrophage migration, strongly participating in the establishment of a proinflammatory status which at least in part happens through PPARG downregulation	Karkeni et al[59]
PPARG	PPARγ

Ang II: Angiotensin II; Akt: Protein kinase B; AMPK: Adenosine monophosphate-activated protein kinase; AT1R: Angiotensin II type 1; Cab39: Calcium-binding protein 39; ECM: Extracellular matrix; HIF-1α: Hypoxia-inducible factor 1 alpha; IL: Interleukin; LKB1: Liver kinase B1; MiR-155: MicroRNA-155; MMP: Metalloproteinase; mRNA: Messenger RNA; NF-κB: Nuclear factor kappa-B; NLRP3: NOD-like receptor family pyrin domain-containing 3; OA: Osteoarthritis; PI3K: Phosphoinositide 3-kinase; TGF-β1: Transforming growth factor β1; TNF-α: Tumour necrosis factor-alpha.

As previously mentioned, miR-155 upregulation in female OA[56] correlates with TNF-α expression and BMI while the ability of TNF-α to induce the expression of miR-155 in various cell types among which synovial fibroblasts, preadipocytes, and adipocytes has been described[59]. The ensuing enhanced M1 macrophage polarization and apoptosis inhibition is also mediated via miR-155 upregulation [via suppressor of cytokine signaling 1 (SOCS1)/caspase 3 (CASP3) signalling pathways] perpetuating inflammation and favouring OA progression[61]. Moreover, miR-155 dysregulation has been reported in sarcopenia and frailty and evidence of a proinflammatory loop mediated by NF-κB and miR-155 that participates in the amplification of adipocyte inflammation has been demonstrated[59].

Intriguingly, when OA chondrocytes were exposed to the adipokines visfatin and resistin, miR-155 was reported elevated which along with significantly increased NF-κB activation induces the transcription of various catabolic factors (MMPs, Adamts, cytokines, chemokines, pro-inflammatory mediators) all contributing to ECM degradation while Col2a1 was reduced[64]. Those effects were potentiated when hydrostatic pressure (HP) and visfatin were applied simultaneously[63]. MIR155HG is potently upregulated by high HP in OA resulting in the downregulation of several known miR-155 targets of which inhibitor of nuclear factor kappa-B kinase subunit epsilon, is involved in the regulation of MMP-3 (Table 2, Figure 2)[66,67].

Moreover, as inflammation and autophagy are interconnected processes, autophagy regulation by miR-155 is an interesting target in OA. Aging, mechanical stress, and articular inflammation increase reactive oxygen species (ROS) in chondrocytes. Impaired autophagic clearance further raises ROS levels, inhibits ECM synthesis, activates MMPs leading to ECM degradation, disrupts chondrocyte mitochondrial function, and results in cell death[95]. MiR-155 Levels apart from potently modulating physiological autophagy, are in inverse correlation with the autophagic pathway in pathological situations such as aging and OA[47]. MiR-155 presents several seed-complementary sequences that can repress autophagy-related genes (ATG3, GABARAPL1, ATG5, ATG2B, LAMP2, FOXO3)[47] and adenosine monophosphate-activated protein kinase (AMPK) that negatively controls mammalian target of rapamycin 1 (mTOR1), the main negative regulator of autophagy (Table 2, Figure 2)[47]. AMPK appears repressible both directly[68] and indirectly via the calcium-binding protein 39 (Cab39)/AMPK and sirtuin 1 (SIRT1)/AMPK pathways whereby the elevated miR-155 levels in OA can negatively impact chondrocyte survival (Table 2, Figure 2)[69,70].

Finally, miR-155 elevation in OA as reported by several investigators is well in agreement with the observation that the application of miR-155 inhibitor could alleviate OA via the SOCS1/signal transducer and activator of transcription (STAT) 3 pathway[61,71]. Additionally, miR-155 was identified as a direct negative regulator of IGF-1[72], a crucial molecule involved in the repair of OA damage[73], whereby its repression worsens OA outlook (Table 2, Figure 2).

In summary, aging, obesity, and female sex, sarcopenia, lack of exercise, and T2DM, all important risk factors in OA, are linked via miR-155 overexpression that perpetuates an established global inflammatory state and promotes a local articular proinflammatory milieu resulting in chondrocyte senescence, autophagy disruption, ultimately contributing to augmented ECM degradation.

MiR-155 and circadian rhythm disruption

Cellular molecular clocks are synchronized to earth’s rotation, thereby a 24-hour cycle is applied to all biological processes, anticipating and controlling all physiological, behavioural, and biochemical functions. Several conditions result from and/or cause circadian rhythm disruptions. OA also displays distinct circadian pain and stiffness symptom variation thus circadian rhythm disruptions may be an integral part of OA pathophysiology. The mammalian circadian clock consists of a hierarchical central clock in the suprachiasmatic nucleus in the hypothalamus and several peripheral clocks in almost every tissue and cellular type. The rhythmic dimerization of the transcription factors brain and muscle Arnt-like protein 1 (Bmal1) and circadian locomotor output kaput (CLOCK) in the cell nucleus activate transcription via E-box elements, including transcription of the mammalian period (PER) and cryptochrome (CRY) genes that form the negative arm of the feedback loop. PER and CRY accumulate in the cytoplasm during the day and move into the nucleus at dusk to inhibit CLOCK and Bmal1, thereby inhibiting themselves and restarting a new cycle[96].

Dudek et al[96] elegantly showed Bmal1 disruptions in OA cartilage where loss of Bmal1 decreased the expression of chondrocyte crucial genes SRY-related HMG box 9 (Sox9), ACAN and Col2α1, inferring the protective role of chondrocyte Bmal1 in cartilage homeostasis and implying that circadian rhythm disruptions are strongly implicated in the pathogenesis of OA (Figure 3).

Figure 3 Schematic representation of microRNA-155 circadian clock effects and mitochondrial dynamics disruption. The 24-hour circadian locomotor output cycles kaput, brain and muscle Arnt-like protein 1 (Bmal1), period, and cryptochrome (CRY) feedback loop forms a cellular clock in cartilage. Loss of Bmal1 in osteoarthritic (OA) chondrocytes diminishes SRY-related HMG box 9, Aggrecan-(ACAN), and type II collagen (Col2a1), driving cartilage catabolism and degeneration. Elevated microRNA-155 (miR-155) in OA can directly repress Bmal1 and hypoxia-inducible factor 1 alpha, activating nuclear factor kappa-B-mediated inflammation and extracellular matrix (ECM) breakdown. Elevated miR-155 in OA increases the expression of mitofusin 2 and induce cellular senescence by promoting mitochondrial fusion. MiR-155 induced calcium-binding protein 39/p-adenosine monophosphate-activated protein kinase downregulations further aggravate senescence. CCAAT/enhancer-binding protein beta repression and downregulation of its downstream target genes nuclear factor erythroid 2-related factor 2, superoxide dismutase 1, and hemeoxygenase-1 induce reactive oxygen species (ROS) generation and disrupt ROS scavenging, perpetuating mitochondrial dysfunction. MiR-155 also impairs mitochondrial oxidative phosphorylation by repressing Arginase 2 (ARG2) and inhibiting ARG2-mediated increases of mitochondrial complex II activity at the electron transport chain. Gene functions and gene abbreviations can be found in Table 2. AMPK: Adenosine monophosphate-activated protein kinase; ARG2: Arginase2; Bmal1: Brain and muscle Arnt-like protein 1; CLOCK: Circadian locomotor output cycles kaput; CRY: Cryptochrome; C/EBPβ: CCAAT/enhancer-binding protein beta; ECM: Extracellular matrix; HIF-1α: Hypoxia-inducible factor 1 alpha; HMOX1: Hemeoxygenase-1; Mfn2: Mitofusin 2; MiR-155: MicroRNA-155; MMP: Metalloproteinase; NFE2 L2: Nuclear factor, erythroid 2 like 2; NLRP3: NOD-like receptor family pyrin domain-containing 3; OA: Osteoarthritic; OXPHOS: Oxidative phosphorylation; PER: Period; ROS: Reactive oxygen species; Runx2: Runt-related transcription factor 2; SOD1: Superoxide dismutase 1; Sox9: SRY-related HMG box 9; ↓: Decline; ↑: Increase.

Intriguingly, miR-155 induction can repress Bmal1 directly in macrophages, endothelial cells and bone marrow mesenchymal stem cells (MSCs)[74,75]. It is therefore plausible that the high levels of miR-155 in OA may target and repress Bmal1 thereby leading to a proinflammatory state through activation of the NF-κB complex, at the same time explaining OA’s circadian symptomatology[74,96].

Additional evidence supporting miR-155/Bmal1 interactions are provided by chondrocyte Bmal1/HIF-1α interactions[28]. The expression of HIF-1α, a crucial regulator in chondrocyte homeostasis in hypoxia, is decreased through Bmal1 disruption leading to enhanced MMP-13 levels and catabolic chondrocyte effects (Table 2, Figure 3)[27,28,76]. The inhibitory effects of miR-155 on HIF-1α are well established, suggesting therapeutic potential for the miR-155/HIF-1α pathway[76].

AMPK/Bmal1 interactions have been identified. AMPKβ phosphorylation, via melatonin acting on its melatonin receptor 1 in chondrocytes, increases Bmal1 expression, making melatonin a promising molecule in OA[77]. Moreover, when AMPK is endogenously or pharmacologically (via metformin) activated, Kir4.1 and Bmal-1 protein expression increases, further supporting metformin’s role in OA management[97].

Whether Bmal1 disruption in OA and metformin action in OA involve AMPK/Bmal1 activation via miR-155 is currently unknown (see Box 1).

MiR-155 and metabolic/energetic alterations

The coexistence of OA with metabolic comorbidities such as obesity, T2DM, cardiovascular disease implies common pathways for its pathogenesis through metabolic and energetic alterations. The aberrant inflammatory articular environment in OA imposes metabolic and energetic limitations in the resident chondrocytes that may play a key role in cartilage degeneration and OA progression[6]. As hyaline cartilage is an alymphatic, avascular and therefore hypoxic tissue, its adenosine triphosphate (ATP) energy requirements are provided mainly through anaerobic glycolysis (> 75%) while the remaining energy derives through mitochondrial oxidative phosphorylation (OXPHOS)[7]. OA chondrocytes display disrupted mitochondrial function with impaired OXPHOS, enhanced glycolysis, disturbed lipid and amino acid metabolism[6]. This metabolic reprogramming occurs not only in chondrocytes but also in synovial fibroblasts and immune cells of the joint. Compared with the healthy state, OA chondrocytes exhibit a metabolic switch from OXPHOS to enhanced anaerobic glycolysis, known as the Warburg effect. The ailing OA chondrocyte tries to strike a balance between its need of energy for anabolic processes and the push towards catabolism by the impending inflammation. The increased glucose uptake and glycolysis will provide the required energy for the enhanced anabolic activities while the downregulated OXPHOS will prevent additional oxidative stress by reducing excessive production of ROS due to the inflammation.

Mitochondrial dysfunction: Mitochondria play a crucial role in the maintenance of chondrocyte homeostasis. Impaired mitochondria morphology, function, and aerobic respiration have been observed in OA[6]. Mitochondrial fusion, controlled by mitofusin (Mfn) 1/2, generates long mitochondrial tubules promoting cellular senescence. Elevated miR-155 in OA can induce cellular senescence by regulating mitochondrial dynamics by promoting fusion through an increase in the expression of Mfn2[70]. Furthermore, miR-155-5p mimic treatment greatly downregulated the expression of Cab39 and p-AMPK, further promoting mitochondrial fusion and cell senescence (Figure 3)[70].

Additional mitochondrial disruptions observed in OA are generally ascribed to excessive ROS production. As miR-155 is a crucial regulator of the innate immune system, a host’s first line of defence against pathogens, the association of miR-155 with bursts of ROS generation in infected macrophages appears as an appealing ancestral generic immune response[98]. Elevated miR-155 in OA could therefore be implicated in excessive ROS generation. In aging and aged bone marrow, miR-155 has been reported to repress CCAAT/enhancer-binding protein beta (C/EBPβ) and its downstream target genes nuclear factor erythroid 2-related factor 2, superoxide dismutase 1 (SOD1), and hemeoxygenase-1 (HO-1), thereby inducing ROS generation in MSCs[78]. Moreover, persistent ROS generation through miR-155 mediated SOD1 repressions disrupting ROS scavenging may perpetuate mitochondrial dysfunction in OA. MiR-155 has been reported to target SOD1 3'UTR as well as diminish SOD2 expression through Foxo3a repression[79,99] while SOD1, 2 and 3 levels were found to be significantly decreased in OA cartilage (Table 2, Figure 3)[6].

Abnormal mitochondrial dynamics and function play a key role in modulating adaptive responses to hypoxia and appear related to arginine metabolism in the mitochondria. Arginase 2 (ARG2), localized in mitochondria, is specifically upregulated in OA chondrocytes and upregulates MMP-3 and MMP-13 via the NF-κB pathway, causing destruction of OA cartilage[100]. As ARG2 is a direct target of miR-155[80], this could be the way elevated miR-155 levels in OA impair OXPHOS by inhibiting ARG2-mediated increases of mitochondrial complex II activity at the electron transport chain[101]. In early OA, miR-155 elevation could also represent an effort by the ailing chondrocyte to counteract MMP-induced catabolism through ARG2 repression that also increases and stabilises HIF-1α, another chondrocyte protective factor (Figure 3)[27]. In addition, significantly decreased OXPHOS gene pathways have been observed in OA and will be discussed in the following paragraph[102].

MiR-155 is evidently crucially involved in shaping mitochondrial dynamics and function in OA.

Glycolysis: A transition from producing ATP via the anti-inflammatory dependent OXPHOS to pro-inflammatory dependent glycolytic energy has been observed in OA, potentially orchestrated by elevated miR-155 levels (Figure 4)[102].

Figure 4 Elevated microRNA-155 in osteoarthritis orchestrates the transition in energy production from anti-inflammatory oxidative phosphorylation generation to pro-inflammatory glycolysis-dependence. Early in osteoarthritis (OA), microRNA-155 (miR-155) upregulates glucose transporter and key glycolytic enzymes through CCAAT/enhancer binding protein β and transforming growth factor β1 (TGF-β1)/bone morphogenetic proteins 2 signalling modulations favouring TGF-β1, to counteract OA progression. Later in OA, through brain and muscle Arnt-like protein 1 and hypoxia-inducible factor 1 alpha repressions, miR-155 reduces TGF-β1/adenosine monophosphate-activated protein kinase/sirtuin 1/peroxisome proliferator-activated receptor γ coactivator 1 alpha/FOXO3 signalling in favor of mammalian target of rapamycin resulting in cartilage catabolism and inflammation. Gene functions and gene abbreviations can be found in Table 2. Akt: Protein kinase B; AMPK: Adenosine monophosphate-activated protein kinase; Bmal1: Brain and muscle Arnt-like protein 1; BMP2: Bone morphogenetic protein 2; C/EBPβ: CCAAT/enhancer-binding protein beta; ECM: Extracellular matrix; GLUT1: Glucose transporter; HK: Hexokinase; HIF-1α: Hypoxia-inducible factor 1 alpha; LDHA: Lactate dehydrogenase; MiR-155: MicroRNA-155; OA: Osteoarthritis; OXPHOS: Oxidative phosphorylation; PGC-1α: Peroxisome proliferator-activated receptor γ coactivator 1 alpha; PI3K/Akt: Phosphoinositide 3-kinase/protein kinase B; PK: Pyruvate kinase; ROS: Reactive oxygen species; SIRT1/PGC-1α/FOXO3: Sirtuin 1/peroxisome proliferator-activated receptor γ coactivator 1 alpha/FOXO3; TGF-β1: Transforming growth factor β1; ↓: Decline; ↑: Increase.

MiR-155 can augment glucose metabolism by upregulating the expression of glucose transporter (GLUT1) and key glycolytic enzymes, including hexokinase (HK), pyruvate kinase and lactate dehydrogenase A through the activation of PIK3R1-FOXO3a-cMYC pathway[81,103]. Additionally, miR-155 indirectly promotes HK2 expression by targeting C/EBPβ, a transcriptional activator of miR-143. Since miR-143 acts as a negative regulator of HK2, its repression by miR-155 leads to post-transcriptional upregulation of HK2[82].

In addition, GLUT1, which facilitates the transport of glucose across the chondrocyte membrane and HK, a key first step rate-limiting glycolytic enzyme and an intracellular glucose sensor, are upregulated through TGF-β1 which at the same time restricts OXPHOS. On the opposite end, BMP2 signalling promotes mitochondrial OXPHOS without affecting glycolysis or its enzymes[6,104].

MiR-155 balances TGF-β1/BMP2 signalling. TGF-β1 is a known direct target of miR-155 regulating cardiac fibrosis via the TGF-β1-mothers against decapentaplegic homolog 2 (Smad2) signalling pathway[105] while Smad5, an important downstream transcription factor for BMPs, was also found to be the target of miR-155. The elevated miR-155 reported in OA[16-20] effectively increases Smad2 and reduces Smad5 transcription tipping the balance in the TGF-β1/BMP2 signalling towards TGF-β1 and anabolism in an effort to mitigate cartilage degradation and counteract OA progression (Figure 4)[83].

This crucial role of miR-155 in chondrocyte glucose metabolism, including glucose uptake, glycolysis, and TCA cycle suggests that microRNA-based therapeutics can be used to modulate glucose metabolism reprogramming OA.

Lipids: Risk factors in OA such as metabolic diseases and comorbidities, e.g. obesity, T2DM, and cardiovascular diseases are almost always associated with lipid disorders. It is therefore plausible that disruption in lipid metabolism might contribute to the pathogenesis of OA[106]. Moreover, restoration of cholesterol homeostasis protected against cartilage destruction[107]. Several clinical and in vitro studies have examined the effect of omega-3 polyunsaturated fatty acids (n-3 PUFAs) in OA[108]. When n-3 PUFAs were added in vitro in IL-1β-stimulated synoviocytes, reductions of MMPs and Adamts have been reported along with inhibition of inflammatory molecule expression and signalling, altogether leading to a reduction in inflammatory response and IL-1β-mediated pyroptosis in chondrocytes[84]. The protective anti-inflammatory n-3 PUFA effects may be mediated via modulation of the NF-κB, NOD-like receptor family pyrin domain-containing 3 (NLRP3)/caspase-1 (Cas-1) and CTBP1/KDM5A pathways, all of which involve miR-155 targeting, thereby offering further support for this miR’s central role[85].

Connecting the different cellular and molecular pathways: The OA chondrocyte tries to balance its energy needs for anabolic processes against inflammatory-driven catabolism. Increased glucose uptake and glycolysis supply energy for anabolic activities, while reduced OXPHOS limits oxidative stress by decreasing ROS production amid inflammation. Centrally in this OA-induced metabolic shift that decides how ATP is generated, is the effect of miR-155 on the involved cellular energy sensors, AMPK and mTOR in the hypoxic articular cartilage under the influence of HIFs (Table 2, Figure 5)[6].

Figure 5 Schematic depiction of known microRNA-155 targets in osteoarthritis and associations with the four main osteoarthritis pathologies: Inflammation, extracellular matrix degradation, chondrocyte senescence and subchondral bone remodeling. MicroRNA-155 (miR-155) is a fundamental mediator of obesity outcomes and abnormal mechanical overload articular effects through aggravated inflammation, accelerated chondrocyte senescence and autophagy interference, while at the same time it disrupts circadian and metabolic chondrocyte homeostasis promoting further catabolism and inflammation through brain and muscle Arnt-like protein 1 (Bmal1)/adenosine monophosphate-activated protein kinase (AMPK) and Bmal1/hypoxia-inducible factor 1 alpha (HIF-1α) interactions. Finally, disrupted mitochondrial function with impaired oxidative phosphorylation and enhanced glycolysis is engendered through miR-155-induced alterations in AMPK/calcium-binding protein 39, Arginase 2, and CCAAT/enhancer-binding protein beta/nuclear factor erythroid 2-related factor 2, hemeoxygenase-1/superoxide dismutase 1 downstream signalling affecting mitochondrial dynamics, reactive oxygen species clearance and promoting key glycolytic enzymes. Moreover, miR-155 influences AMPK/sirtuin 1/HIF signalling balancing chondrocyte energy generation in hypoxia. Metformin can modulate Prkaa1, cyclooxygenase 2 and miR-155. The colours of the 4 main OA pathologies are the same as on the genes that influence them. Arrows indicate interactive effects. Detailed gene functions and gene abbreviations can be found in Table 2. AGTR1: Angiotensin II type 1 receptor gene; ARG2: Arginase2; BACH1: BTB and CNC homology 1, basic leucine zipper transcription factor 1; Bmal1: Brain and muscle Arnt-like protein 1; CASP3: Caspase 3; COX-2: Cyclooxygenase 2; C/EBPβ: CCAAT/enhancer-binding protein beta; ECM: Extracellular matrix; Ets-1: E26 transformation-specific sequence-1; GDF6: Growth differentiation factor 6; HIF-1α: Hypoxia-inducible factor 1 alpha; IGF-1: Insulin-like growth factor 1; IKBKE: Inhibitor of nuclear factor kappa-B kinase subunit epsilon; LEPR: Leptin receptor; Mafb: Musculoaponeurotic fibrosarcoma oncogene family, protein B; MAPK: Mitogen-activated protein kinase; Mfn: Mitofusin; MiR-155: MicroRNA-155; NF-κB: Nuclear factor kappa-B; RANKL: Receptor activator of nuclear factor kappa-B ligand; Runx2: Runt-related transcription factor 2; SHIP-1: Src homology 2-containing inositol phosphatase-1; SIRT1: Sirtuin 1; Smad: Mothers against decapentaplegic homolog 4; SOCS1: Suppressor of cytokine signaling 1; SOD: Superoxide dismutase; TNF-α: Tumour necrosis factor-alpha.

Hypoxia stabilises and translocates HIF-1α to the nucleus where it heterodimerizes with HIF-1β, binds to the hypoxia responsive element and initiates transcription of chondrogenesis anabolic genes. Functional HIF-1α is required for energy production in chondrocytes though its control of basal glycolytic enzymes, phosphoglycerate kinase, and GLUT1[109]. HIF-1α knockout prevents ATP generation and hampers hypoxia-induced matrix gene expression[110]. The adaptive dynamics of HIF-1α are orchestrated by miR-155, suggesting that this hypoxi-miR is indispensable for cellular metabolic adaptation to hypoxia[111].

AMPK is an energy-sensing heterotrimeric enzyme, activated in response to increased cellular adenosine monophosphate (AMP)/ATP ratio-either due to increased ATP consumption or decreased production. Furthermore, AMPK as a sensor of energy stress and mitochondrial damage can sense incoming environmental cues to control the dynamic flow between glycolysis, OXPHOS and mitochondrial fusion/fission required to maintain mitochondrial fitness for optimal cell growth. AMPK expression and activity are suppressed in OA while mTOR which normally is inhibited by AMPK, is upregulated, signifying increased chondrocyte catabolism and apoptosis[6]. AMPK is a direct target of miR-155[68] whereby the elevated miR-155 levels observed in OA could account for the low AMPK levels reported in OA, but this is currently not known[6]. The upregulated mTOR in OA further elevates miR-155 via the activation of the phosphoinositide 3-kinase (PI3K)/protein kinase B (Akt)/mTOR and RICTOR/Akt/mTOR signalling pathway impairing chondrocyte autophagy and promoting ECM degradation[112]. Indeed, AMPK’s and mTOR’s central roles in OA are supported by the protective effects of metformin, an AMPK activator and inhibitor of the abnormal activation of the PI3K/Akt/mTOR signalling (to be discussed in the following paragraphs)[1,113-115].

Apart from direct AMPK-miR-155 effects, the elevated miR-155 reported in later OA stages[16-20] represses Bmal1, disrupting circadian regulation and tipping the TGF-β1/BMP2 signalling away from TGF-β1-anabolism, effectively decreasing expression of the major ECM-related genes Sox9, ACAN, and Col2a1[96]. Due to loss of TGF-β1, AMPK activation to promote PPARγ coactivator 1 alpha (PGC-1α) and FOXO3 expression, which inhibit the synthesis of inflammatory factors in the synovial fibroblasts of human knee OA, does not take place, thereby promoting cartilage senescence and contributing to OA development and progression[116]. Furthermore, due to miR-155-induced interrupted Bmal1/HIF-1α interaction, normal chondrogenesis is disrupted with decreased chondrocyte proliferation and promotion of chondrocyte hypertrophy[28].

Moreover, SIRT1, a metabolic nicotinamide adenine dinucleotide biosensor, induced by oxidative stress, and mutually regulated by AMPK, is a direct miR-155 target and also impaired in OA[117]. Its specific repression promotes apoptosis and inflammatory response in both in vitro and in vivo models of hypoxic-ischemic brain injury[86]. MiR-155-mediated SIRT1 repression impairs its downstream targets, PGC-1α and FOXO1/O3 transcription factors, key regulators of mitochondrial function that promote mitochondrial biosynthesis and quality control and optimize mitochondrial OXPHOS efficiency and ROS generation. The AMPK/SIRT1/PCG-1α/FOXO pathway, is thereby severely disrupted, affecting ROS generation and chondrocyte catabolism. This disruption also activates the NLRP3 inflammasome, increasing pro-inflammatory cytokine secretion[118]. Metformin promotes SIRT1 expression in chondrocytes further supporting its protective effect in OA[119]. In the future, miR-155-dependent therapeutics may offer novel interventions by targeting those pathways.

In summary, miR-155 appears as the central mediator of obesity outcomes, abnormal mechanical overload articular effects, disrupted circadian chondrocyte homeostasis, perturbed mitochondrial dynamics, and metabolic/energetic alterations, initiating and perpetuating articular inflammation, ECM degradation, chondrocyte senescence, and subchondral bone remodeling resulting in OA (Figure 5). The interplay between the above molecular pathways still requires further research (see Box 1), but elevated miR-155 is the common denominator in OA pathophysiology[17].

MiR-155 interactions with other miR in OA

MiRNA interplay could facilitate chondrocyte metabolic switch towards enhanced glycolysis through the miR-155-mediated repression of miR-143, a negative regulator of HK2, thus resulting in upregulation of HK2 expression at the post-transcriptional level[82].

It has been shown that miR-155 and miR-146 demonstrate co-expression in inflammatory joint conditions. While miR-155 participates in pro-inflammatory processes, miR-146 is crucial in preventing dysregulated humoral immune responses and production of autoantibodies[26]. The balance between the two finetunes NF-κB-dynamics and shapes macrophage inflammatory responses[120].

Despite the notion that miR-155 is elevated in OA[17], Wang et al[2] reported that miR-155-ladden exosomes from synovial MSCs promote chondrocyte proliferation, migration, increased ECM protein (Col2a1, Sox9) secretion and inhibition of apoptosis. They also reported decreased expression of Runt-related transcription factor 2 (Runx2), a direct miR-155 target, linked to chondrocyte hypertrophy and ECM degradation in OA[87]. Other studies suggest that miR-155-mediated repression of miR-221 appears to favour anti-inflammatory responses through Janus kinase/STAT and PI3K/Akt signalling[121], and M2-macrophage shift[122]. Importantly, miR-221 silencing is robustly pro-chondrogenic and strongly enhances in vivo cartilage repair with no sign of collagen type X deposition, a marker of undesired hypertrophic maturation[123]. The silencing of miR-221 increases expression of tricho-rhino-phalangeal syndrome 1 gene that represses Runx2-mediated trans-activation thereby diminishing chondrocyte hypertrophy and ECM degradation and promoting ECM synthesis[124,125]. This observation could offer a more encompassing explanation in how miR-155-loaded exosomes could be used as a pro-chondrogenic disease modifying treatment in OA.

MiR-155 and the renin angiotensin aldosterone system

The renin angiotensin aldosterone system (RAAS) has long been recognized as a ubiquitous systemic homeostatic mechanism affecting cardiovascular, renal, and vascular functions[126]. Angiotensin II (Ang II) is the main effector in RAAS mediating its effects via the Ang II type 1 and 2 (AT1R/AT2R) receptors. Ang II mediates systemic pro-inflammatory, proliferative, and constrictive effects via the AT1R while AT2R signalling counteracts them. Ang II plays a significant role in the pathophysiology of OA as injecting it in rat knee joints recapitulates the inflammatory effects of monoiodoacetate induced arthritis while angiotensin converting enzyme 1 (ACE1) inhibition with captopril induces chondroprotection and Col2a1 synthesis and reduces leukocyte recruitment from synovium and chondrocyte death[127]. It appears thus, possible that that local and/or systemic RAAS modulation can affect the course of OA. Pharmacological systemic RAAS inhibition at the ACE1 or at the AT1R level has been in use in the management of cardiovascular and renal diseases since many decades[126].

Identified within various tissues, RAAS also operates locally producing tissue-specific effects. Locally produced Ang II and its receptors appear to mediate important functions in bone and cartilage[128]. Kawakami et al[129] and Tsukamoto et al[130] were the first to outline chondrocyte local RAAS. According to Tsukamoto et al[130,131] healthy hyaline chondrocytes from articular cartilages and meniscus of knee joints do not express AT1/2R. Furthermore, Kawakami et al[129] and Tsukamoto et al[130] reported expression of AT1R/AT2R mRNA in OA patients, while Tsukamoto et al[130,131] reported the expression of angiotensinogen and ACE1, AT1R and AT2R, though only in the chondrocytes undergoing hypertrophic differentiation in vivo and in hyaline cartilage only when the latter is under inflammatory stress. As the chondrocyte is situated in an avascular environment, Ang II is most certainly autocrine-produced[131].

In a series of elegant experiments, Tsukamoto et al[130,131] showed that activation of AT1R suppressed and activation of AT2R enhanced the expression of markers of chondrocyte hypertrophic differentiation, including collagen type X, MMP-13 and Runx2, suggesting that AT1R and AT2R can modulate differentiation into hypertrophic chondrocytes and contribute to chondrocyte degradation in OA[131]. Moreover, mechanical and shear stress can upregulate and activate AT1/2R independently of Ang II[132-135], although the effect of Ang II and mechanical loading is probably synergistic, especially when synovial inflammation co-exists[136].

These observations are, however, in contrast to the clinical effects of Renin inhibition with aliskiren[137], ACE inhibition with captopril[127,138] and perindopril[139], AT1R blockers losartan[140] olmesartan[141], and ACE2 agonist Diminazene aceturate[142]. According to Tsukamoto et al[130,131] blocking the ACE1 and/or AT1R and enhancing AT2R signalling should increase the expression of markers of hypertrophic chondrocyte differentiation, including collagen type X, MMP-13 and Runx2, and contribute to chondrocyte degradation in OA! A plausible explanation for this discrepancy is that the beneficial effect of systemic pharmacological RAAS inhibition on synovial tissue inflammation and the inflammatory effects of Ang II, may override any negative effects towards hypertrophic chondrocyte differentiation through local AT1R and AT2R actions, resulting in an overall positive outcome[141]. It has been reported that losartan exerts inhibitory effects on the inflammatory TGF-β1 signalling cascade, captopril increases expression of the protective AT2R, perindopril downregulates NF-κB, and aliskiren attenuates the OA-induced expression of IL-1 and TNF-α offering support to the above assumption[127,137-140]. A recent study confirmed that inhibition of the AGT/AGTR1a/IL-1β axis via Semaphorin 6D, a cardiovascular neuroeffector with chondroprotective potential, can significantly enhance ECM homeostasis, marked by increased ACAN, Col2a1 and decreased COL10A1, MMP-13, and Runx2 expression, thereby regulating ECM metabolism and chondrocyte hypertrophy[143].

An additional level of complexity is evident when we consider the elevated miR-155 levels in OA[17]. It is well known that the AGTR1 gene that codes for AT1R is a direct target of miR-155 and will therefore be repressed with a subsequent downregulation of the AT1R protein[88]. Considering the experiments of Tsukamoto et al[130,131] miR-155-mediated AT1R downregulation (and the ensuing Ang II activation of the unopposed AT2R) would aggravate OA progression, consequently confirming miR-155 elevation as a pathogenetic event in OA.

Kawakami et al[129] observed that while AT1R-mRNA was increased after IL-1 induction, the expression of the AT1R protein in the chondrocytes was not significantly enhanced by IL-1. Elevated miR-155 after IL-1 induction[144] could have bound to and inactivated AT1R-mRNA thereby impacting AT1R protein levels. MiR-155 was, however, not measured in that study.

How systemic RAAS effects through their influence on OA risk factors such as obesity, T2DM, and cardiovascular and renal conditions impact local synovial and chondrocyte RAAS and what the net effect becomes is not known. Notably, the impact of RAAS inhibition on miR-155 levels in OA has yet to be thoroughly studied (see Box 1). However, previous research suggests RAAS inhibition may reduce miR-155 expression[145,146], potentially enhancing anti-inflammatory and immunomodulatory functions and offering therapeutic benefits in OA[145,146].

Wingless-like/β-catenin signalling and RAAS Interactions: The role of miR-155

An added level of complexity in the role of RAAS in OA is evident when considering the wingless-like (Wnt)/β-catenin signalling pathway that also holds an important position in OA development[10].

Wnt signalling plays a crucial role in skeletal development involving MSC chondrogenic differentiation, chondrocyte and osteoblast development and maturation, and chondrocyte hypertrophy. Modest Wnt canonical (involving β-catenin) activation appears vital for chondrocyte maintenance as β-catenin inhibition in transgenic mice or Wnt knock-out animal models demonstrate cartilage degradation, chondrocyte hypertrophy and apoptosis, and promote OA[10].

Dickkopf-1 (Dkk-1) is an endogenous Wnt antagonist that can reduce cartilage damage and OA progression in mice with medial meniscus destabilization. Dkk-1 blocks Wnt-driven gene transcription, including Runx-2, MMPs, and Adamts 4 and 5. Dkk-1 helps preserve an equilibrium in the Wnt pathway as apposite activation of the latter in needed to maintain a healthy cartilage homeostasis[23].

A recent study has revealed a regulatory role for MIR155HG, the gene coding for miR-155, as a suppressor of bone marrow MSC differentiation into osteoblasts via the positive regulation of two pathways, the miR-155-5p/β-catenin signalling and the DKK1/β-catenin signalling[23]. Similarly, elevated miR-155 inhibits RA synovial fibroblast viability and induces apoptosis and cell cycle arrest through elevation of β-catenin expression and Wnt pathway activation[147]. Elevated levels of β-catenin in chondrocytes promoted chondrocyte senescence, through downregulation of SIRT-1[148]. Collectively these result support a plausible regulation of the Wnt/β-catenin signalling via elevated miR-155 levels in OA.

Enthrallingly, overexpression of either β-catenin or various Wnt ligands induce the expression of all RAAS genes while blockade of Wnt/β-catenin signalling can simultaneously repress multiple RAAS genes underscoring the importance of this interaction between Wnt signaling and RAAS activity in OA[149]. Moreover, fibroblast growth factor 23 (FGF23)-elevated in OA-has been shown to increase intracellular expression and secretion of Ang II and activate cardiac RAAS to promote cardiac hypertrophy and fibrosis[150]. Increased FGF23 in OA may modulate Wnt/β-catenin signalling in chondrocytes, thereby providing an additional, important RAAS activation pathway[151].

Importantly, losartan abrogates FGF23-induced cardiac hypertrophy and alleviates OA cartilage defects further implicating dysfunctional FGF23/Wnt/β-catenin/RAAS signalling in OA[151,152]. Similar dual Wnt/RAAS activation is observed in diabetic kidney disease, a major microvascular complication of diabetes where combination therapies (e.g., Wnt inhibitors + RAAS blockers) have been proposed[153].

Whether elevated miR-155 in OA, alone or in combination with FGF23, activates the Wnt/β-catenin signalling that in turn activates local articular RAAS aggravating OA is not known. This essential knowledge gap needs to be addressed (see Box 1).

MiR-155 and glucagon-like peptide 1 receptor agonists

The discovery of glucagon-like peptide 1 (GLP-1) receptors in synovial tissue and cartilage and their involvement in the downregulation of proinflammatory cytokines (TNF-α, IL-6, IL-8), inhibition of MMPs, blockade of cellular inflammation signalling pathways, c-Jun N-terminal kinase, AP-1, and NF-κB pathways[154], and chondrocyte apoptotic pathways[155,156], has unravelled GLP-1-receptor agonists (RA) effects unrelated to weight loss in OA[157-159].

As miR-155 is a central pro-inflammatory mediator, its inhibition may unmask its robust anti-inflammatory capabilities. MiR-155[29,60,61] has been reported repeatedly as the messenger of macrophage-mediated inflammation in many tissues and conditions[89,98,101,120,122]. In agreement with these observations, treatment with liraglutide diminished miR-155[160] while semaglutide alleviated inflammation-induced endothelial progenitor cell inflammation and function by inhibiting miR-155 expression in macrophage exosomes[161]. However, miR-155 levels in OA after GLP-1RA treatment have not yet been evaluated or correlated to clinical symptoms (see Box 1).

Additional miR-155 effects related to GLP-1 homeostasis might also be relevant. In hyperlipidemia, elevated miR-155 in β-cells increases GLP-1 production via musculoaponeurotic fibrosarcoma oncogene family, protein B repression and IL-6 increase. Elevated GLP-1 improves glucose metabolism and helps β-cells adapt to obesity-induced insulin resistance (Figure 5)[90]. Moreover, miR-155 knockout mice display decreased GLP-1 secretion from islet L-cells and lower basal GLP-1 plasma levels which may contribute to adipose tissue inflammation, obesity progression, and dyslipidaemia.

Apparently, as evidence appears to support the existence of a miR-155/IL-6/GLP-1 loop, more research is needed to shed light in the miR-155/GLP1-RA interaction (see Box 1).

METFORMIN: A NOVEL THERAPEUTIC APPROACH FOR OA

Metformin is a synthetic biguanide compound derived from Galega officinalis in 1920. Since the 1950s it has been the first-choice oral antihyperglycemic agent in the treatment of T2DM demonstrating well-documented gluco-metabolic benefits and a longstanding evidence-based, reassuring record of clinical safety[162,163]. Beyond its T2DM indication, metformin has been used in various unrelated conditions and diseases spanning from cancer to aging, obesity, and neurodegenerative diseases, polycystic ovary syndrome to anti-bacterial/viral/parasitic/malarial and coronavirus disease-19 repurposing[162,164-167].

In addition, metformin’s role as a disease-modifying drug[1,6] in OA has been documented repeatedly and associates with activation of AMPK, the enzyme pivotally involved in cellular energy and glucose availability in all cells including the chondrocyte[113,114,168-170]. AMPK, an established miR-155 target[68] crucially involved in the pathogenesis of OA and most of the risk factors for OA development, connects miR-155 with metformin effects. This review aims to present how the roles of miR-155 and metformin intertwine in OA[171,172]. For a comprehensive review of metformin’s mode of action[162,163,173].

In short, our understanding of metformin’s mechanism of action still remains incomplete but it has repeatedly been shown to activate AMPK, the master regulator of energy balance and metabolism[163]. In the gastrointestinal tract, metformin slows intestinal glucose absorption, attenuates glycaemic response, modulates intestinal gluconeogenesis, and enhances glucose management through GLP-1 secretion[173]. Moreover, it shifts gut microbiota towards short chain fatty acid (SCFA)-producing bacteria in T2DM patients further reducing serum glucose levels, improving insulin sensitivity, and mitigating inflammation. In addition, SCFA can enhance GLP-1 secretion[174]. In the liver, it blocks hepatic glucose production[163] and mediates its effects on appetite and body weight through the biosynthesis of Lac-Phe[175]. It targets brown adipose tissue function through HIF-1α and attenuates the proinflammatory responses in M1 macrophages[109]. Moreover, metformin exerts important beneficial effects on chronic inflammation though its action on several proinflammatory pathways and immune cell types (Figure 6)[176].

Figure 6 Schematic known and potential metformin actions on chondroprotection, immunomodulation, circadian rhythmicity, and pain reduction resulting in improved osteoarthritis management. Metformin reduces the catabolism and apoptosis of chondrocytes, regularizes the circadian disruptions observed in osteoarthritis (OA), and decreases the infiltration and M1 polarization of synovial macrophages. Metformin is able to modulate microRNA-155 expression but whether this ability mediates its effects systemically and in OA is currently unknown. Light orange and purple boxes indicate molecular metformin effects while light green boxes indicate metformin tissue effects. Detailed gene functions and gene abbreviations can be found in Table 2. AMPK: Adenosine monophosphate-activated protein kinase; Bmal1: Brain and muscle Arnt-like protein 1; CLOCK: Circadian locomotor output cycles kaput; Cas-1: Caspase-1; COX-2: Cyclooxygenase 2; ECM: Extracellular matrix; ERK: Extracellular regulated protein kinases; GI: Gastrointestinal; GSDMD: Gasdermin D; HIF-1α: Hypoxia-inducible factor 1 alpha; IL: Interleukin; MiR-155: MicroRNA-155; MMP: Metalloproteinase; mTORC1: Mammalian target of rapamycin C1; MΦ: M1 macrophage; NF-κB: Nuclear factor kappa-B; NLRP3: NOD-like receptor family pyrin domain-containing 3; Nrf2: Nuclear factor erythroid-2-related factor 2; PER: Period; PTEN: Phosphatase and tensin homolog; ROS: Reactive oxygen species; Runx2: Runt-related transcription factor 2; SIRT1: Sirtuin 1; Sox9: SRY-related HMG box 9; TNF-α: Tumour necrosis factor-alpha; ↓: Decline; ↑: Increase.

Recent discoveries point to the lysosome as a new site of metformin action, where it binds the membrane protein presenilin enhancer 2, forming a complex with ATP6AP1 to inhibit v-ATPase and activate AMPK independently of AMP[177]. This mechanism underlies metformin’s effects on postprandial glucose, hepatic steatosis, mTORCH1 inhibition, and lifespan extension, with fewer side effects at low doses (Figure 6)[177].

The idea of using metformin as a pharmaceutical intervention for OA was entertained for the first time in 2014 by Mohammed et al[178] in combination with meloxicam, a COX inhibitor, reporting significant improvements in knee OA symptomatology and insightfully discussing metformin’s AMPK activation and COX expression inhibition. Metformin, indeed, downregulates cyclooxygenase 2 (COX-2) via the AMPK/phosphatase and tensin homolog pathway[179,180]. Later studies showed mixed results: Barnett et al[181] found no significant association between prescription of metformin and OA diagnosis while a matched cohort study from Taiwan reported that the combination of metformin and COX-2 inhibition was associated with lower joint replacement surgery rates[182]. Wang et al[183] reported beneficial effect on long-term knee joint outcomes, independent of weight loss, in obese knee OA patients after metformin intervention.

These encouraging epidemiological results prompted basic molecular studies in experimental OA animal models. AMPK/SIRT1 activation and inhibition of mTORC1, NF-κB with pro-inflammatory factor (MMPs, Adamts, TNF-α, NLRP3, Cas-1, Gasdermin D, calgranulins, and IL-1β) reduction and ACAN/Col2a1 increases have been reported as main chondroprotective metformin effects along with its anti-apoptotic effect on articular cartilage[113,184]. Metformin also suppresses inflammatory cell death factors, macrophage inflammatory response, and induction of the autophagosome-lysosome fusion phenotype through AMPK activation[177]. Moreover, another important symptom in OA, osteoclast-mediated abnormal subchondral bone remodelling, could be attenuated by metformin through the AMPK/NF-κB/extracellular regulated protein kinases signalling pathway[185], COX-2 inhibition[180], and through decrease in Runx2 expression and atypical angiogenesis[186].

Several clinical studies have repeatedly confirmed reductions in total joint replacements[169,183,187-189], reduction in OA development[1] and attenuation of OA symptoms[114,190-193] in patients with diabetes, impaired glucose tolerance, and/or overweight or obesity. Overall, in OA, metformin reduces the catabolism and apoptosis of chondrocytes and decreases the infiltration and M1 polarization of synovial macrophages (Figure 6). The therapeutic effect of metformin in obese and/or diabetic OA appears stronger than that in normal weight subjects which could be mediated by reduced leptin secretion and improved leptin sensitivity (vide infra)[91,194].

Metformin also regulates circadian clock and metabolic rhythms through AMPK/SIRT1 activation, both integral parts of OA pathophysiology[113,195,196]. Impaired AMPK activity along with reduced SIRT1 expression and activity in white adipose tissue co-occurs with impaired mRNA and/or protein expression of core clock components (CLOCK, Bmal1 and PER2), adversely impacting circadian function and promoting dysregulation of several metabolic genes[196]. Bmal1 disruptions in OA cartilage decrease chondrocyte gene expression (Sox9, ACAN and Col2α1), inferring the protective role of chondrocyte Bmal1 in cartilage homeostasis[96]. Silencing of Prkaa1 (gene for AMPKα1) also leads to decreased Bmal1 protein expression while AMPK activation via metformin restores it, reversing the above dysregulations and further supporting metformin’s role in OA management[97,196]. Moreover, metformin-induced AMPK activation mitigates ROS and alleviates age-related pathologies via reactivation of the Bmal1/nuclear factor erythroid-2-related factor 2 (Nrf2)/antioxidant pathway (Figure 6)[197].

THE SYNERGISTIC POTENTIAL OF MIR-155 AND METFORMIN IN OA

Whether the mechanisms underlying the above metformin effects in OA and other conditions where metformin use has been proposed-involve miR-155 has not been investigated and therefore remains unknown (Figure 6, see Box 1).

A few studies have explored how metformin affects miR-155 expression which appears to be cell-context dependent and pleiotropic and, at times, even contradictory. On the one hand, in NK-cells, metformin has been shown to increase miR-155 levels, enhancing IFN-γ synthesis, thereby significantly potentiating NK cell activation and anti-tumour cytotoxicity[198]. Interestingly, metformin-induced miR-155 increase targets Anexelekto, a molecule highly implicated in cancer metastasis, an observation that directly supports metformin’s anti-cancer properties[199]. Moreover, the use of metformin in high fat-induced inflammation in endothelium upregulates miR-155 and improves the inflammatory response through miR-155/miR-146 regulatory interplay, appositely regulating NF-κB[200].

On the other hand, metformin improves cardiac hypertrophy by reducing miR-155 levels and Akt signalling[171]. Moreover, metformin-induced AMPK activation and decreased miR-155 levels upregulate the leptin receptor and improve leptin sensitivity and inflammatory cytokines (TNF-α/IL-1β) in bone tissue[91]. Metformin-induced miR-155 down-regulation could help explain the enhanced metformin treatment effects seen in T2DM patients with OA[172].

Indirect evidence points to the involvement of miR-155 in mediating metformin effects in OA. MiR-155 is an indispensable and significantly upregulated proinflammatory regulator in clinical and experimental arthritis and its deficiency protects against the development of collagen-induced arthritis[17,42,62]. On the one hand, metformin reduces the catabolism and apoptosis of chondrocytes as well as decreases the infiltration and M1 polarization of synovial macrophages via AMPK activation and mTORC1 inhibition[194]. On the other hand, macrophage M1 polarization and reduced macrophage apoptosis in the proinflammatory environment of knee OA synovial fluid is mediated via miR-155-5p/SOCS1 and miR-155/CASP3 pathways respectively[61]. Furthermore, frugoside, a cardiac glycoside compound delayed OA progression in mice via inhibiting miR-155-modulated synovial macrophage M1 polarization, and co-transfection of miR-221-3p/miR-155-5p mimics in M2-macrophages increased M1-specific IL-12 secretion[122,201]. These studies indicate that miR-155 associates with the same pathways affected by metformin and that pharmaceutical miR-155 modulation replicates these effects[122,201].

Hepatic Bmal1 expression is necessary to mediate metformin effects on AMPK activation and blood glucose reduction highlighting the strong association between the two[202]. In severely obese, insulin resistant db/db mice, reduced AMPK/SIRT1 expression impairs the expression core clock component mRNA (CLOCK, Bmal1 and PER), linking circadian dysfunction to both metabolic disorders and OA[96,196]. Metformin restores AMPK and SIRT1 levels and normalizes core clock components[196]. Similarly, in another study, AMPK activation via metformin results in increased Kir4.1 and intermediate core clock component Bmal1 protein expression in diabetes[97]. Aging disrupts the Bmal1/Nrf2/antioxidant pathway, increasing ROS levels; metformin can restore antioxidant responses and mitigate damage in cells, tissues, and organs by activating AMPK to reactivate this pathway[197]. Interestingly, Bmal1 possesses two miR-155-binding sites in its 3'-UTR and can mutually repress each other’s expression, clearly assigning circadian rhythmicity to macrophage function and other tissues[74]. Bmal1/miR-155 pathway can regulate the aging and osteogenic differentiation ability of bone marrow MSCs as well as control endothelial cell apoptosis and inflammatory response in atherosclerosis[75,203]. It is yet unclean whether metformin-mediated AMPK activation and Bmal1 dynamics involve miR-155 homeostasis (Figure 6, see Box 1). This is clearly a field requiring further research.

To conclude, accumulating recent and consistent evidence across pre-clinical and human studies strongly supports a favourable effect of metformin on chondroprotection, immunomodulation, circadian rhythmicity, and pain reduction in OA with the potential to become a disease modifying drug. Its therapeutic effect in obese and/or diabetic OA appears stronger than that in normal weight subjects which could be mediated by differential miR-155 effects and/or miR-155-mediated leptin sensitivity improvement (Figure 6, see Box 1)[91,194]. Further studies are needed to determine whether metformin's effects on key OA molecular targets, which overlap with miR-155 targets, are mediated through miR-155 modulation (see Box 1).

DIAGNOSTIC AND PROGNOSTIC PRESPECTIVES OF MIR-155

MiR-155 holds promise as a reliable diagnostic biomarker for OA, offering potential for early, non-invasive detection through blood or synovial fluid analysis. Its levels can reflect the degree of inflammation and cartilage degradation in OA patients, serving as an indicator for disease presence and severity. Several miRNAs, among which miR-155 are reported robustly elevated in OA peripheral blood mononuclear cells (PBMCs) compared to their levels in the PBMCs of healthy controls[17,19].

Recently, miR-155 could differentiate erosive hand OA from psoriatic arthritis and healthy subjects with high accuracy which improved when combined with C-reactive protein[204]. However, findings are not entirely consistent – one study reported lower miR-155 expression in OA and osteoporotic fractures compared to controls, suggesting variability across patient populations and disease phenotypes[50].

Elevated miR-155 expression may also serve as a prognostic marker, helping clinicians predict disease progression. MiR-155 levels increased with increasing OA stage and level of severity in two studies[17,18] but decreased in another[20] underlying the need for more research. It is conceivable that patients with high levels of miR-155 may experience faster cartilage degeneration and more severe joint dysfunction, allowing for more personalized treatment plans. Moreover, synovial fluid miRNA signatures could be more appropriate as those would reflect the local articular environment and would be less prone to interference than plasma miRNAs[16].

While miR-155 can indeed be of diagnostic and prognostic help in OA, its expression is significantly lower in OA than in arthritis of autoimmune etiology (e.g. RA)[16,40,42] or arthritis with greater expression of inflammatory cytokines [e.g. Lyme arthritis (LA)][205]. MiR-155 levels could distinguish antibiotic-refractory LA or RA from antibiotic-responsive LA or OA with miR-155 downregulation in the latter and persistent miR-155 elevation in the former, thus being able to predict risk for a refractory disease course[206].

Using miR-155 levels alone or in combination with other inflammatory markers, in differential diagnosis and to evaluate clinical intervention/treatment could improve risk stratification and therapeutic response in OA. Decrease of miR-155 expression has been reported after celecoxib[207], hyaluronic acid (HA)[208] and balneotherapy interventions[209]. Metformin has also been shown to reduce miR-155 levels[171] but whether this decrease reflects a metformin treatment effect in OA has not yet been evaluated.

Further research is necessary in order to establish whether miR-155 alone or in combination with other miRNAs can be reliably used to finetune diagnostic and prognostic accuracy in OA.

THERAPEUTIC PERSPECTIVES

MiR-155 modulation

MiRNA expression can be altered temporarily with exogenous miRNA mimics or inhibitors for overexpression or downregulation, respectively. MiRNA masks, miRNA sponges, miRNA inhibitors (antagomiRs) provide a transitory effect by interfering with the miRNA binding spot with an exogenous sequence. Delivery of naked miRNAs faces challenges such as low retention and bioavailability at target site due to RNA nuclease degradation susceptibility, lysosomal degradation, low cell membrane permeability, high dose requirement, and rapid renal clearance.

To overcome these issues, viral vectors, lipid-based carriers, polymeric nanoparticles, MSC-based and/or exosome-based delivery, nanoparticles, and hydrogel-based delivery systems are being utilized[210,211]. Long-term and more stable modulation is possible using miRNA mimics/inhibitors in lentiviral vectors, eliminating the need for repeated transfections[212]. Lentivirus genetic interference, possible off target effects, immune effects and unknown miRNA network effects remain important concerns.

Alternative delivery systems such as hydrogel-based systems might work better for cartilage regeneration. Hydrogels with their high-water absorbing capacity, plasticity and ability to fill cartilage defects, appear particularly desirable for cartilage regeneration in OA[213]. Hydrogels are being extensively researched and used for the fabrication of biocompatible vehicles for tissue engineering and cartilage repair allowing precise and triggered elution with negligible invasiveness and minimal or no toxicity[211]. These hydrogel systems can be loaded with cells, exosomes, cytokines, miR-modulating constructs, some are able to home and recruit synovium-derived stem cells[214] and/or combine with cartilage-targeted cationic nanocarriers that electrostatically interact with anionic glycosaminoglycans in the ECM, ensuring passive penetration and prolonged retention in cartilage[215,216]. As exosomal transfer of osteoclast-derived miRNAs to chondrocytes contributes to OA progression, MSC-derived exosomes, that can be modified to carry specific single or multiple miRNAs targeting various OA pathogenetic pathways, can be ideal for delivering therapeutic cargo[15].

Given miR-155’s role in OA pathogenesis, its targeted inhibition – or in some contexts, overexpression – may provide therapeutic benefits[16-20]. As a well-established oncomiR, miR-155 appears involved in the pathophysiology of many tumours and plays instrumental role in autoimmune diseases, thus, miR-155 inhibition rather than overexpression could be of benefit. In general, the suppression of miR-155 has shown antitumor effects on transformed cells by promoting cell death mechanisms, inhibiting proliferation and cell growth, and preventing invasion and metastasis[217]. Moreover, miR-155 inhibition reversed the drug resistance in certain cancer types, thereby improving treatment response[218]. However, increased miR-155 levels in many immune cell types [CD4+ and CD8+ T cells, M1 tumour-associated macrophages (TAM), dendritic cells, and NK cells] strike a balance with other cell types (Treg) cells, M2 TAMs, cancer immune evasion cells: Myeloid-derived suppressor cells to ensure adequate immunosurveillance, therefore, global miR-155 inhibition, could interfere with these vital functions and worsen patient outlook and prognosis[210,219,220]. It is thus, desirable to study how miR-silencing can be cell-specific, precise, and preferably also time-controlled as global interference carries the risk of unintentional adverse effects.

Robust evidence supports miR-155 pathogenicity in arthritis of various etiologies[42] corroborated by the fact that miR-155-deficient mice appear resistant to a collagen-induced arthritis model[221,222]. As miR-155 is driving autoantibody production in RA[223], its pathogenetic effect in RA is potentially larger than in OA. A recent study has reported the successful use of polyethylene glycol liposomes encapsulating antagomiR-155-5p in decreasing joint inflammation and ameliorating the arthritis in a preclinical moder of RA through decreased expression of miR-155-5p that restored monocyte polarization into anti-inflammatory macrophages[224]. The authors also reported that prophylactic antagomiR administration was more effective than curative intervention. Whether this approach could be applied with miR-155 in OA is not known.

Studies using miRNA modulation in OA have, however, started to emerge using injectable hydrogel systems for miRNA-based intra-articular delivery[214,216,225]. There has been no study yet with either agomiR-155 or antagomiR-155 in a hydrogel system in OA but local administration of a temperature-sensitive hydrogel (Pluronic F-127) loading miRNA-155-5p antagomiR promoted angiogenesis and accelerated wound healing in diabetic mice[226]. This study described a miR-155-dependent pathway in M1-polarized macrophages that in diabetic mice inhibit angiogenesis and blocked it with the above antagomiR[226]. A similar pathway in synovial macrophages has been reported in OA and could be restored in a similar way using miRNA-155-5p antagomiR-loaded hydrogel system[61].

Despite the observation that miR-155 levels are elevated in OA and thus, reducing miR-155 levels could alleviate articular inflammation, exogenously increasing miR-155 local availability could exploit other miR-155-dependent pathways and elicit therapeutic effects. Wang et al[2] reported that miR-155 containing exosomes could achieve therapeutic effects in OA via decreased expression of Runx2 involved in chondrocyte hypertrophy and ECM degradation. This likely occurs via the known silencing effect of miR-155 on miR-221 and subsequent inhibition of the latter’s anti-chondrogenic effects[123]. Similarly, Uysal et al[227] confirmed elevated expression of miR-155 in human synovial fluid-derived MSC exosomes and proposed this miR as a therapeutic target to restore damaged cartilage tissue. In another example, as miR-155 targets and represses BTB and CNC homology 1, a transcriptional repressor of cytoprotective and antioxidant HO-1, increased levels of the latter have been reported to reduce the severity of OA-like changes in OA chondrocytes[92]. Therefore, developing a hydrogel delivery system loaded with miR-155 exosomes could present a therapeutic modality in OA. The choice of exosomes, however, must be based on known OA pathogenetic pathways to avoid unwanted effects. As an example, adipocyte-derived exosomal miR-130b-3p under T2DM conditions suppresses mitochondrial function in chondrocytes through targeting the AMPKα1/SIRT1/PGC1-α pathway and exacerbates OA injury[228]. This study exemplifies that not every exosome or MSC is beneficial when planning treatment. It is imperative to understand the molecular background of the product and how it will impact the main diagnosis and associated comorbidities.

MiRNA modulation in OA could also be achieved indirectly without exogenous miRNA therapeutic intervention as through balneotherapy[209], HA injections[208], and celecoxib treatment, all of which resulted in decreased miR-155 levels with a tendency to greater reduction in clinical responders[207].

Understanding ECM and chondrocyte affinities in vitro and in vivo to choose the most appropriate hydrogel delivery system with good biocompatibility, superior mechanical properties and minimal toxicity, knowing which miRNAs is/are most suitable for encapsulation and delivery to achieve successful cartilage rejuvenation and repair, are issues to be considered and solved prior to developing future organelle-based, cell-free OA treatments. As we mentioned earlier in this review, one miRNA can target multiple mRNAs – and vice versa – effective therapeutic design requires a deep understanding of the complex regulatory web that miRNAs form.

Metformin preconditioning for MSC-based and exosome-based therapies in OA

A recent meta-analysis evaluating the efficacy and safety of MSCs for the treatment of knee OA has shown that MSCs are advantageous in targeting analgesia and leading to functional improvement in OA[229]. However, when considering cellular delivery systems, the choice of an appropriate cellular source and delivery system is challenging. The success of stem cell therapy in repairing damaged articular cartilage depends on maintaining the viability and functionality of the injected cells at the disease site, over a longer period. Presently, less than 5% of the naked injected MSCs remain viable on site[230]. Recently, a cell-compatible, biodegradable, injectable, and self-healing polysaccharides-based hydrogel (Suc-CS/Ald-HA) has been reported that could sustain encapsulated human adipose stem cells viable and functional even after 14-days post incubation[230].

While allogeneic MSCs are easy and swift to procure, and can be mass produced, their long-term benefits are less appealing[70]. Compared with autologous and syngeneic MSCs, allogeneic origin MSC are marred with alloimmunization with alloantibody development, low cell survival caused by immune rejection and are eliminated from the heart by five weeks after implantation, with their functional benefits being lost within five months[231]. Autologous MSCs, especially in a conducive hydrogel delivery system, are expected to be chondrogenically more effective through better local integration and paracrine signalling thanks to complete immune compatibility and absence of immune rejection, which subsequently improve survival. However, the impact of age can be significant, especially regarding chondrogenesis[232].

Moreover, in patients with chronic kidney disease (CKD) who may need kidney transplantation in the future, it is imperative to consider that autologous cell transplantation is preferred over allogenic sources, as the latter may increase the risk of allosensitization. It is, therefore, crucial to ensure adequate functionality of autologous MSCs when treating age-related conditions. Granulocyte colony stimulating factor activated autologous PBSCs contain MSCs, albeit in low numbers, and have been shown to be successfully used in knee OA[233,234].

Metformin preconditioning and co-culture of autologous MSCs has been shown to rescue aging MSCs, restore senescence and enhance their regenerative properties[235-237]. Metformin treatment significantly improved the reduced proliferation, accelerated senescence, and increased DNA damage in CKD MSCs[238]. In a murine OA model, metformin and human adipose MSC co-culture increased the expression of immunoregulatory mediators and inhibited the expression of proinflammatory molecules delivering superior chondroprotective and antinociceptive effects[168]. Moreover, exosomes derived from metformin-pretreated bone marrow MSCs showed potential as a therapeutic modality for diabetic wounds via miRNA-dependent pathway[239]. Metformin could, therefore, be used to rejuvenate MSCs used as cellular vehicle for exosome delivery or as sources of exosome production to be loaded in hydrogel delivery systems[239].

Alternatively, intra-articular metformin could be delivered loaded in ECM-based biomimetic scaffolds or combined into porous microspheres. In these, platelet derived growth factors and kartogenin can be encapsulated, allowing adhesion, proliferation, differentiation, and potentially also rejuvenation of endogenous MSCs resulting in upregulated osteogenic and chondrogenic gene expression[240,241]. Several hydrogel constructs are being developed to deliver sustained high-dose metformin with increased efficiency and achieve successful in situ MSC rejuvenation to improve therapeutic results in OA[242,243]. Interestingly, metformin phonophoresis, as ultrasound absorbable gel, has shown statistically and clinically significant improvement in pain intensity, knee ROM, and function in knee OA[244].

The is still, a knowledge gap regarding the appropriate metformin delivery method as well as metformin dosage in MSC cultures depending on the MSC cell origin and the desired differentiation direction as chondrogenic or osteogenic differentiation may require different dosing. Research is also necessary to determine whether metformin MSC co-culture is more efficient and/or safer than intra-articular metformin delivery with in situ MSC recruitment and rejuvenation (see Box 1).

Moreover, elevated miR-155-5p has been observed in elderly individuals[245] and their aged bone marrow-derived dendritic cells[246] and MSCs, promoting senescence through inhibited mitochondrial fission and increased mitochondrial fusion via the AMPK signalling pathway[70,78]. AMPK activation reversed these effects while transplantation of anti-miR-155-5p-aged MSCs had a better capacity to attenuate cardiac remodelling and restore heart function in mice following infarction[70]. With its unique potential to downregulate miR-155 expression through AMPK activation, it is necessary to research whether metformin effects on MSC rejuvenation involve decreased miR-155 levels and whether metformin use more is effective and/or safer than direct miR-155 inhibition (see Box 1).

Finally, combining metformin with miR-155 targeting using antagomiRs or miRNA mimics in various applications (intra-articularly, per os, co-culture), may provide synergistic effects that further enhance therapeutic outcomes in OA[247,248]. Metformin’s versatile ability to modulate inflammatory pathways and enhance tissue repair opens new avenues for disease-modifying treatments in OA.

FINAL REMARKS

Although miRNAs constitute less than 3% of the human genome, more than 2588 miRNAs regulate over 60% of human genes. Their control is central and crucial in various cellular processes, including cell growth and apoptosis[249]. With its 239 experimentally validated targets in various cell types and tissues, well-studied miR-155-5p plays a critical role in OA pathogenesis, diagnosis, and prognosis[250]. Furthermore, as miRNA modulation will most certainly have therapeutic implications, elements, compounds or molecules with potential to influence miRNA homeostasis should be evaluated[30,31,35].

The review has identified miR-155 as an important element in OA, noting that aging, sarcopenia, obesity, diabetes mellitus and female sex, all of which are significant OA risk factors, have common underlying pathologies that also involve this miRNA. The elevated miR-155 levels observed in OA set off chain reactions that recapitulate the pathophysiology on OA (Figures 2, 3, 4, and 5). AMPK (Prkaa1) with a central role in OA, is both a direct and indirect target of miR-155[68-70] whereby the elevated miR-155 Levels in OA could account for low AMPK levels[6], in turn upregulating mTOR which further elevates miR-155 via the activation of the PI3K/Akt/mTOR and RICTOR/Akt/mTOR signalling pathways, suppressing chondrocyte autophagy, impacting chondrocyte survival, and promoting ECM degradation (Figures 3, 4, and 5)[112].

Elevated miR-155 levels in OA mediate pro-inflammatory effects through elevated TNF-α/IL-1β, IL-6, MMPs, Adamts, NLRP3, Cas-1, Gasdermin D with associated pyroptosis of chondrocytes[84], synovial recruitment and M1 polarization of macrophages, and osteoclast-mediated abnormal subchondral bone remodelling through targeting of various genes (Figures 2, 3, and 5)[93]. Importantly, elevated miR-155 levels in OA potently suppress autophagy in human primary chondrocytes by repressing several components/genes involved in different stages of autophagy (Figures 3 and 5)[47].

Moreover, an important observation for OA pathogenesis is that miR-155 elevates COX-2 expression both directly[89,94] and indirectly via inhibition of a COX-2 expression repressor (Figure 5)[251,252]. Elevated COX-2 expression in osteocytes stimulates aberrant subchondral bone formation in the development of OA and is intimately associated with pain and inflammation[253].

MiR-155 also influences OA’s circadian pathology through Bmal1 repression that significantly affects chondrocyte senescence and survival[74,75,96,203]. Bmal1’s direct associations with AMPK and their mutual restoration by metformin strongly implies miR-155 involvement in their homeostasis (Figures 3 and 6)[197,254].

Metabolically, mir-155 orchestrates the transition between anti-inflammatory OXPHOS towards pro-inflammatory aerobic glycolysis (Figure 4). MiR-155 elegantly choreographs the direct targeting of TGF-β1/BMP2 signalling[81,82], adaptive HIF-1α dynamics, and the regulation of key genes involved in glycolysis to upregulate GLUT1 and central glycolytic enzymes[101-104]. Enhanced glycolytic activity produces ROS, leading to mitochondrial dysfunction while altered lipid metabolism additionally disturbs mitochondrial dynamics and increases oxidative stress in chondrocytes. MiR-155-directed AMPK/SIRT1 repression disturbs important ROS scavenging PCG-1/FOXO and Nrf2/SOD1/HO-1 signalling pathways, perpetuating mitochondrial dysfunction in OA[78,79,117]. While these miR-155-mediated metabolic changes collectively disrupt chondrocyte functions, promote inflammation, and accelerate cartilage degeneration in OA, they also present an opportunity to strategically modulate OA pathogenesis and progression via miR-155 manipulation.

The role of RAAS in OA is increasingly being recognized while the beneficial disease-modifying effects of pharmacological RAAS inhibition have in recent years been explored intensively. Elevated miR-155 in OA mediates RAAS’ inflammatory articular and chondrocyte effects via increased FGF23 and Wnt/β-catenin signalling, both of which can be inhibited by metformin. Metformin increases Klotho and downregulates FGF23[255] as well as affects the Wnt pathway at various levels[256].

Everything considered, as we have now identified miR-155 as a central molecule behind crucial aspects of the OA pathophysiology, its local targeted modulation at the articular or cellular level would be desirable. Given its centrality, pleiotropy, and importance in regulating the immune system, cell-or tissue-specific drug targeting will be more beneficial than non-specific inhibition of global miR-155 expression which would certainly have severe off-target effects[220].

Metformin (Figure 6)[171], celecoxib[207,252], and HA[208], all impact on miR-155 levels. The contradictory effects of metformin on miR-155 may reflect this miR’s is cell-context dependency and pleiotropy. The conflicting results for celecoxib could be attributed to the timing and the site of miR-155 measurements but the Dong et al[207] study better reflects the clinical reality in OA. In either case, the clinical efficacy of metformin, celecoxib and HA, has long been established and the involvement of miR-155 in mediating the drugs’ effects supports its involvement in the pathogenesis of OA. Further research is needed to dissect the pleiotropic nature of this miR and devise the best delivery vehicles and drug combinations to achieve cure in OA (see Box 1).

Box 1: Research questions needing answers

MiR-155 levels post RAAS inhibition in OA; do treatment effects correlate to miR levels? MiR-155 levels post metformin treatment in OA; do treatment effects correlate to miR levels? MiR-155 levels post GLP-1RA treatment in OA; do treatment effects correlate to miR levels? Are the elevated miR-155 levels observed in OA the reason for the low AMPK expression and activation levels reported in OA? Does Bmal1 disruption in OA involve miR-155? Does AMPK/Bmal1 activation after metformin treatment involve miR-155 modulation? Do n3-PUFA effects in OA relate to miR-155 levels? Does miR-155 activates Wnt/β-catenin pathway that in turn activates local articular RAAS? MiR-155 levels after COX-2 inhibition; do treatment effects correlate to miR levels? Is metformin’s impaired effect in lean/non-T2DM patients related to impaired miR-155 suppression compared to obese/T2DM patients? MiR-155 levels post total joint replacement; do treatment effects correlate to miR levels? Is prophylactic intra-articular miR-155 antagomir administration in hydrogels effective in OA? Is metformin MSC co-culture is more efficient and/or safer than intra-articular metformin delivery with in situ MSC recruitment and rejuvenation? Do metformin effects on MSC rejuvenation involve miR-155? Is metformin use more is effective and/or safer than direct miR-155 inhibition?

CONCLUSION

In conclusion, realizing that miR-155 architects the molecular dynamism in OA and understanding how these complex and dynamic networks influence OA pathophysiology will enable the development of novel therapeutic approaches. We have in this review emphasized and outlined the central role of miR-155 in OA and highlighted the fact that novel potential disease modifying treatments, such as metformin, COX-2 inhibitors, pharmacological RAAS inhibition, and potentially also GLP-1RAs, appear to mediate their effects through miR-155 pathway networks.

The combination of exogenous (miR mimics/inhibitors, masks, sponges) and/or single molecule pharmaceutical miRNA modulation (metformin, COX-2 inhibition, pharmacological RAAS inhibition, GLP-1RAs), in synergism with cellular strategies, holds a great disease-modifying and regenerative promise in OA. Tissue engineering systems comprising hydrogel delivery modules and scaffolds with cartilage affinity and potentially also stem cell recruiting, along with integrated stem cells, exosomes, and miRNA modulating pharmaceutical compounds, administered appropriately systemically or intra-articularly, could provide substrates for hyaline cartilage regeneration under the protective shield of the pharmacological molecules. Alternatively, aged autologous MSCs could be rejuvenated and primed when co-cultured with metformin, COX-2 inhibitors or GLP-1RAs instead of administering the drugs systemically, thereby avoiding patient side effects. Modulating relevant miRNAs in vivo and/or in vitro in MSC cultures before delivering regenerative treatment as cells or exosomes, systemically or in intra-articularly, will revolutionize OA care and provide disease-modifying approaches that are currently non-existent.

This is by no means an easy task and ample research is necessary to evaluate the spectrum of potential downstream consequences associated with administrated doses, potential cytotoxicity, and long-term effects. Treatment precision demands an accurate understanding of the complexity of miRNA network regulation and context-dependent miRNA modulation and delivery in cellular systems (immune or other) and/or tissues (healthy or inflamed) in order to avoid adverse off-target effects.