The role of metformin in extracellular matrix-based three-dimensional cell culture Models: A mini-review on therapeutic potentials.

Given the pivotal role of the ECM in 3D models, this section examines its structural and functional properties and explores how its dynamic nature influences disease progression and therapeutic responses. The extracellular matrix (ECM) is a complex, dynamic network of structural and functional proteins (e.g., collagens, laminins, fibronectin) and polysaccharides (e.g., glycosaminoglycans) that provides mechanical support, regulates cell signaling, and influences cell fate [ 3 ]. In 3D cell culture models, the ECM is not a passive scaffold but an active modulator of disease progression and therapeutic response. Aberrant ECM remodeling—characterized by excessive deposition, abnormal cross-linking, or uncontrolled degradation—is a hallmark of diseases including cancer, fibrosis, and metabolic disorders [ 4 ]. In cancer, a stiff, fibrotic ECM promotes tumor progression and confers drug resistance [ 40 ]. In fibrosis, dysregulated ECM accumulation leads to progressive organ dysfunction [ 41 ]. ECM alterations in diabetes are well-documented contributors to microvascular complications and impaired tissue repair [ 4 , 42 ]. ECM composition profoundly influences cellular behavior and drug responsiveness. Natural matrices such as Matrigel support cell adhesion and differentiation, whereas decellularized ECM (dECM) preserves tissue-specific biochemical and biomechanical cues from native organs (e.g., liver, heart, tumor), rendering it ideal for disease modeling [ 33 ]. Synthetic scaffolds, although not biological ECM, can be engineered to replicate. Understanding the interplay between metformin and the ECM within these systems is essential for the development of targeted, context-specific therapies.

Extracellular Matrix Three-Dimensional AMP-activated protein kinase Transforming Growth Factor-beta Matrix Metalloproteinase Decellularized Extracellular Matrix Mesenchymal Stem Cell Guided Bone Regeneration Hepatic Stellate Cell Myofibroblast Lysyl Oxidase-Like 1 Tissue Inhibitor of Metalloproteinases 1 Alpha-Smooth Muscle Actin Epithelial-Mesenchymal Transition Hepatic Stellate Cells Myofibroblasts Human Umbilical Cord Mesenchymal Stem Cells 6-Phosphofructo-2-Kinase/Fructose-2,6-Bisphosphatase 3 organic cation transporters Polycaprolactone Poly (lactic-co-glycolic acid) gelatin-based hydrogels

Mahshid Zamani: Data curation, Writing – original draft. Nepton Soltani: Data curation, Investigation, Methodology, Writing – review & editing.

Although current evidence strongly supports the multifaceted role of metformin in 3D cell culture models and ECM regulation, several promising avenues remain for future exploration. To fully harness the potential of these advanced systems, future research should focus on the following directions: First, integrating ECM-based 3D models into preclinical studies—particularly those using organotypic cultures and patient-derived cells—could significantly enhance translational relevance and improve the prediction of clinical outcomes. These personalized models allow for the investigation of inter-individual variability in metformin response, which is critical for optimizing therapy in diverse patient populations. Second, the development of smart drug delivery systems, particularly those based on bioengineered or decellularized ECM (dECM) scaffolds, holds promise for enhancing metformin's tissue-specific targeting and sustained release. Such systems can be designed not only to deliver the drug but also to actively modulate the ECM microenvironment, thereby amplifying its therapeutic effects. Third, emerging technologies such as 3D bioprinting and organ-on-a-chip platforms offer powerful tools to study metformin's systemic effects in complex, multi-organ diseases like diabetes with cardiovascular or renal comorbidities [ 36 ]. These models enable real-time monitoring of drug distribution, cell-ECM interactions, and tissue-level responses in a controlled setting. Furthermore, incorporating immune and vascular components into 3D models will be essential for recapitulating the full complexity of the tumor microenvironment and fibrotic tissues. This advancement would allow for a more accurate assessment of metformin's immunomodulatory and anti-angiogenic effects within a physiologically relevant context. Finally, standardizing 3D culture protocols and ECM compositions across laboratories is crucial for ensuring reproducibility and facilitating broader adoption in drug screening and disease modeling. Addressing these challenges will pave the way for using 3D ECM-based models as robust platforms for personalized medicine and metformin repurposing. The accumulating evidence from 3D cell culture studies provides a strong rationale for translating metformin's ECM-modulating properties into clinical applications. Given its well-established safety profile, low cost, and broad availability, repurposing metformin for non-diabetic conditions—particularly those involving dysregulated ECM remodeling—holds considerable promise. In oncology, metformin could serve as an adjunct to conventional therapies by targeting the tumor microenvironment and enhancing drug delivery through modulation of ECM stiffness and composition. In fibrotic diseases, where current treatment options are limited, metformin's potential to attenuate fibrotic processes offers a novel therapeutic strategy. Moreover, its incorporation into bioengineered scaffolds and drug delivery systems may improve outcomes in regenerative medicine, especially for chronic wounds and bone defects in diabetic patients. Future clinical trials should focus on stratifying patient populations based on ECM-related biomarkers to identify individuals most likely to benefit from metformin.

A comprehensive literature search was conducted using PubMed, Scopus, Google Scholar, and Web of Science to identify relevant studies on the therapeutic effects of metformin in 3D cell culture models incorporating the extracellular matrix (ECM). The search encompassed articles published between 2009 and 2024 and employed the following keywords: “metformin,” “3D cell culture,” “extracellular matrix,” and “tumor microenvironment.” Studies were selected based on their relevance to the topic, with an emphasis on recent advancements in metformin research within 3D culture systems. Articles lacking 3D model components or ECM integration were excluded. The selected studies were analyzed to evaluate metformin's mechanisms of action, its interactions with ECM components, and its implications for disease modeling and therapeutic development.

This research received no specific grant from any funding agency. Institutional support was provided by 10.13039/501100003970 Isfahan University of Medical Sciences .

Metformin hydrochloride (1,1-dimethylbiguanide hydrochloride; C 4 H 11 N 5 ·HCl; MW: 165.63 g/mol) is a white, crystalline, hygroscopic powder with a bitter taste [ 9 ]. It is highly soluble in water and 95 % ethanol but practically insoluble in non-polar organic solvents such as ether and chloroform. The compound exhibits two pKa values (2.8 and 11.5), indicating that it predominantly exists as a positively charged, hydrophilic species at physiological pH [ 9 ]. Its low lipophilicity (log P ≈ −1.43) accounts for its poor passive membrane permeability and reliance on active transport via organic cation transporters (OCTs), particularly OCT1 (Organic Cation Transporter 1; SLC22A1) and OCT2 (Organic Cation Transporter 2), for cellular uptake [ 10 ]. Metformin demonstrates an absolute oral bioavailability of approximately 50–60 % under fasting conditions and undergoes no hepatic metabolism, being excreted unchanged in the urine. It is chemically stable under recommended storage conditions, although heating may induce decomposition and release toxic fumes (e.g., nitrogen oxides) [ 11 ]. These physicochemical properties are critical for understanding its tissue distribution, cellular uptake, and formulation strategies in drug delivery systems. Given its favorable safety profile and well-established clinical use, metformin is a first-line therapeutic for type 2 diabetes, primarily acting through suppression of hepatic gluconeogenesis and enhancement of insulin sensitivity. Its effects are mediated via both AMPK-dependent and AMPK-independent pathways [ 12 ]. Metformin modulates mitochondrial respiration, thereby enhancing energy metabolism in the liver and peripheral tissues. It also reduces lipid secretion from intestinal cells and promotes fatty acid oxidation in adipose tissue and skeletal muscle, collectively contributing to improved metabolic profiles [ 12 ]. Additionally, metformin exerts cardioprotective effects, reducing the incidence of heart failure and associated mortality by improving myocardial energy metabolism and attenuating adverse cardiac remodeling [ 13 ]. Metformin demonstrates potential anticancer effects by inhibiting mitochondrial complex I, thereby inducing bioenergetic stress and altering metabolic pathways in cancer cells [ 14 ]. In breast cancer models, it induces cell cycle arrest at the sub-G1 phase, suggesting antiproliferative activity [ 15 ]. Furthermore, metformin reduces pro-inflammatory biomarkers and modulates adipokine secretion, offering therapeutic benefits in metabolic syndrome. It has also been associated with elevated levels of neurotrophic factors, implicating potential neuroprotective roles [ 16 ]( Fig. 1 ). Fig. 1 Pleiotropic effects of metformin beyond glycemic control, including AMPK-dependent and -independent modulation of metabolism, inflammation, fibrosis, cancer progression, and ocular complications [ 8 , 12 , [14] , [15] , [16] , [17] ]. Fig. 1 Pleiotropic effects of metformin beyond glycemic control, including AMPK-dependent and -independent modulation of metabolism, inflammation, fibrosis, cancer progression, and ocular complications [ 8 , 12 , [14] , [15] , [16] , [17] ]. In experimental models of diabetic complications, metformin has been shown to suppress retinal inflammation, oxidative stress, and glutamate excitotoxicity, thereby preserving retinal architecture and function [ 8 , 16 ]. These protective effects are largely attributed to AMPK-dependent downregulation of pro-inflammatory mediators such as NF-κB, ICAM-1, and IL-8 [ 8 ]. Additionally, metformin reduces histopathological damage in ocular tissues, including the retina and optic nerve, as supported by recent preclinical studies [ 17 ]. In glaucoma models, metformin reduces intraocular pressure by protecting trabecular meshwork cells from oxidative stress and fibrotic transformation—a process linked to integrin/ROCK signaling. Epidemiological data further support a reduced incidence of glaucoma among diabetic patients treated with metformin [ 17 ]. Although metformin exhibits diverse therapeutic effects, its actions are highly context-dependent and modulated by the cellular microenvironment. Three-dimensional (3D) cell culture models—particularly those incorporating the extracellular matrix (ECM)—offer a more physiologically relevant platform for investigating these context-specific effects compared to traditional 2D monolayer cultures. The following section outlines the major types of 3D models and their utility in disease modeling.

The scientific literature reveals significant contradictions regarding metformin's effects on ECM-related processes, particularly in angiogenesis and tissue remodeling. In vitro and in vivo studies have demonstrated paradoxical effects, wherein metformin exhibits both anti-angiogenic activity and simultaneous enhancement of pro-angiogenic mediators [ 108 ]. This duality is particularly evident in diabetic conditions, where metformin has shown pro-angiogenic effects in wound healing, cardiovascular disease, and tumor models [ 109 ]. The molecular pathways through which metformin affects ECM remodeling also exhibit conflicting patterns. Although AMPK activation is widely recognized as metformin's primary mechanism for reducing oxidative stress and inflammation in cardiac tissue [ 110 ], research has revealed that metformin's effects on collagen synthesis may operate independently of AMPK activation [ 39 ]. This suggests the existence of multiple, potentially parallel mechanisms through which metformin influences ECM dynamics, complicating our understanding of its therapeutic effects [ 39 ]. These contradictory findings have important implications for clinical applications, particularly in conditions where ECM remodeling plays a crucial role, such as wound healing and cardiovascular disease. In diabetic contexts, metformin has demonstrated beneficial effects on endothelial function and angiogenesis; however, the precise mechanisms and conditions under which these effects manifest remain unclear [ 109 ]. Among the limitations of the present article is the lack of access to the full text of some studies, which necessitated reliance solely on their abstracts. Additionally, although some articles initially appeared relevant based on their titles, their full texts did not align with the scope of our review and were therefore excluded. Finally, the mini-review format limits our ability to perform a systematic evaluation or meta-analysis of individual studies. Although we prioritized high-impact and recent references, the absence of quantitative data synthesis may overlook subtle variations in study outcomes. Future comprehensive reviews or meta-analyses could further clarify these relationships and refine therapeutic strategies. Indeed, our article serves as an introduction to future meta-analyses on this subject.

Metformin, a cornerstone therapy for type 2 diabetes, demonstrates broad therapeutic potential beyond metabolic regulation by modulating extracellular matrix (ECM) dynamics in three-dimensional (3D) cell culture models. Through mechanisms such as AMPK activation, TGF-β inhibition, and miRNA regulation, it influences ECM remodeling, offering promise in cancer, fibrosis, and regenerative medicine. However, challenges persist, including variable efficacy across tissues and discrepancies between in vitro and in vivo responses, underscoring the complexity of metformin's actions. Heterogeneity in experimental conditions—such as dosage, culture duration, and ECM composition—hinders direct comparisons and clinical translation. Although 3D models better replicate human pathophysiology than 2D systems, they often lack multi-organ interactions and immune components critical for systemic diseases. Despite these limitations, integrating metformin into ECM-based scaffolds and drug delivery systems shows significant potential for localized therapies, particularly in diabetic wound healing and bone repair. Its repurposing for non-diabetic conditions is appealing due to its safety profile, affordability, and accessibility. Future research should prioritize refining dosing strategies, clarifying tissue-specific effects, and validating findings in advanced in vivo-like models. Emphasizing personalized medicine approaches—combining patient-derived cells with sophisticated 3D platforms—could unlock metformin's full therapeutic potential for ECM-related diseases. This mini-review highlights the need for standardized, predictive 3D systems to bridge gaps between preclinical studies and clinical applications, ultimately advancing precision therapies for complex pathologies.

Metformin exerts significant effects on tumor cell organization and behavior in 3D models, primarily by modulating cell-ECM interactions. It reduces sphere-forming capacity and targets cancer stem/progenitor cell populations, although it does not completely eliminate them [ 55 ]. In 3D spheroid models, metformin treatment decreases cancer cell invasiveness, with treated spheroids exhibiting reduced growth after 37 h compared to untreated controls [ 56 ]. Metformin potently inhibits cancer cell migration and invasion across multiple tumor types, largely by altering ECM adhesion and remodeling. In cervical cancer cells, it significantly suppresses migration by impairing filopodia and lamellipodia formation—structures critical for ECM sensing—through inhibition of key regulatory proteins, including focal adhesion kinase (FAK), Akt, Rac1, and RhoA [ 57 ]. Similarly, in ovarian cancer, metformin reduces wound healing and mesothelin expression while downregulating IL-6/STAT3 signaling, a pathway associated with ECM stiffness sensing [ 58 ]. Importantly, metformin suppresses matrix metalloproteinases (MMPs), which are essential for ECM degradation and metastasis. In esophageal carcinoma, 20 mM metformin inhibits cell migration and invasion by 87 % and 81 %, respectively, via suppression of AKT phosphorylation and MMP-9 protein expression [ 59 ]. In glioblastoma, metformin markedly suppresses TGF-β1-induced upregulation of MMP-9 [ 60 ]. In 3D invasion assays, metformin reduces invasion by approximately 30 % in SF268 cells and 50 % in U87 glioblastoma cells, while enhancing adhesion to collagen—a hallmark of reduced motility and increased ECM anchoring [ 61 ]. Metformin potently inhibits epithelial-mesenchymal transition (EMT), a key driver of tumor invasion and metastasis that is regulated by ECM composition and stiffness. It upregulates epithelial markers (e.g., E-cadherin, keratin 19) and downregulates mesenchymal markers (e.g., vimentin, N-cadherin) and transcription factors (Zeb1, Zeb 2) in breast cancer cells [ 62 ]. In ovarian cancer, metformin promotes epithelial marker expression and suppresses mesenchymal markers [ 58 ], while in cholangiocarcinoma, it upregulates E-cadherin and downregulates Snail and MMP-2 [ 63 ]. In hepatocellular carcinoma, metformin increases E-cadherin, decreases vimentin, and suppresses EMT [ 64 ]. These findings highlight metformin's role in stabilizing epithelial phenotypes and reducing invasive potential within ECM-rich microenvironments. Beyond its direct effects on cancer cells, metformin reprograms the tumor microenvironment (TME). In pancreatic cancer, it reprograms pancreatic stellate cells (PSCs) to reduce production of ECM components such as type I collagen and hyaluronic acid, thereby alleviating desmoplasia and decreasing tumor stiffness [ 2 , 65 ]. This stromal remodeling is associated with reduced metastasis and enhanced drug delivery. Metformin also modulates vascular-ECM interactions by downregulating PDGF-B, reducing microvessel density and leakage while enhancing mural cell coverage and perfusion [ 64 , 66 ]. This vascular remodeling, which relies on pericyte-ECM adhesion, increases the sensitivity of cancer stem cells to therapy and alleviates tumor hypoxia. Metformin modulates the immune landscape of the TME by suppressing the PI3K/Akt/PTEN pathway, thereby attenuating the effects of growth factors such as EGF and TGF-β—signals often initiated by ECM-integrin engagement. It also upregulates MHC-I expression on cancer cells, enhancing their recognition by cytotoxic T cells [ 67 , 68 ]. Collectively, these effects foster a less immunosuppressive and more therapy-responsive tumor microenvironment characterized by altered ECM signaling. In multicellular spheroid models, metformin disrupts the formation of large, cohesive clusters, causing cells to become more loosely attached—indicating impaired cell-cell and cell-ECM cohesion [ 37 ]. However, when used as a monotherapy in certain contexts, such as pancreatic microtumors containing cancer-associated fibroblasts, metformin's efficacy may be limited, although it can potentiate the effects of other treatments, including oxaliplatin or photodynamic therapy [ 69 ]( Table 1 ) (see Table 2 ). Table 1 Summary of Metformin's effects in ECM-Based 3D cell culture models. Table 1 Disease Category Cell Type/Model 3D Model Type ECM Type/Scaffold Key Metformin Effect Molecular Mechanism Ref. Cancer MCF-7 breast cancer cells Spheroid Matrigel Reduces sphere-forming ability and disrupts large cluster formation Downregulates integrin β1, increases collagen adhesion [ 37 , 55 ] Cervical cancer cells 3D culture – Inhibits migration and invasion Suppresses FAK, Akt, Rac1, RhoA; depletes filopodia/lamellipodia [ 57 ] Ovarian cancer cells 3D culture – Reduces wound healing and mesothelin expression Downregulates IL-6/STAT3 signaling [ 58 ] Esophageal carcinoma 3D model – Inhibits migration (87 %) and invasion (81 %) Suppresses AKT phosphorylation and MMP-9 expression [ 59 ] Glioblastoma (U87, SF268) 3D invasion assay Collagen-rich scaffold Reduces invasion by 30–50 %; increases adhesion to collagen Suppresses TGF-β1-induced MMP-9 upregulation [ 60 , 61 ] Breast, ovarian, cholangiocarcinoma, HCC 3D models Various Inhibits epithelial-mesenchymal transition (EMT) Upregulates E-cadherin; downregulates vimentin, Snail, Zeb1/2 [ 62 , 63 ] Pancreatic cancer 3D microtumor Desmoplastic stroma Reprograms PSCs; reduces collagen I and HA; alleviates desmoplasia Modulates TGF-β signaling and ECM remodeling [ 2 , 65 ] Fibrosis Hepatic stellate cells (HSCs) 3D hydrogel Collagen I Suppresses COL1A1 mRNA and α-SMA protein expression Inhibits TGF-β1/Smad3 signaling [ 38 , 39 ] Cardiac fibroblasts 3D culture Fibrin gel Reduces collagen deposition and expression of fibrotic genes (e.g., COL1A1, ACTA2) Inhibits TGF-β1/Smad3 signaling pathway [ 39 ] Renal fibroblasts 3D culture dECM (UUO-derived) Blocks angiotensin II-induced overproduction of ECM proteins (fibronectin, collagen I) Reduces fibronectin and collagen I expression [ 70 ] Orbital fibroblasts (OFs) 3D culture Tissue-derived ECM Inhibits TGF-β1-induced protein expression of α-SMA, collagen types I/II/III, and fibronectin Suppresses fibrotic signaling pathways [ 71 ] Myofibroblasts 3D model – Promotes transdifferentiation to lipo-fibroblasts Activates AMPK; increases BMP-2 protein and PPAR-γ phosphorylation [ 72 ] Metabolic Diseases Adipose-derived stem cells (ADSCs) 3D culture Fibrin-gel or spheroid Impairs adipogenesis; reduces lipid droplet fusion Activates autophagy; inhibits mTOR; alters ECM stiffness [ 73 , 74 ] Endothelial cells 3D co-culture – Inhibits pro-angiogenic factors (EMMPRIN, MMP-9) Glucose-dependent suppression of angiogenic signaling [ 75 ] Pancreatic beta cells 3D organoid – Protects from fatty acid-induced apoptosis Activates AMPK-mediated autophagy [ 76 ] Liver cells 3D model – Reduces cellular lipid content and fatty acid consumption Improves metabolic flexibility in ECM-rich environment [ 77 ] Multi-organ systems 3D organoids – Rescues mitochondrial dysfunction; improves glucose transport Enhances metabolic coupling in liver-islet models [ 78 ] Table 2 Applications of metformin-loaded biomaterials in 3D cell culture models. Table 2 Application Area Biomaterial Type 3D Model System/Cells Key Effect Mechanism/Outcome Ref. Diabetes Management Dual-encapsulated alginate beads 3D culture – 71 % glucose reduction for 29 days 122 % bioavailability vs. conventional [ 79 ] Gelatin-phenylboronic acid microparticles Smart system – Normalized glucose in 3h, sustained 8h Glucose-responsive release [ 80 ] Wound Healing & Tissue Repair Chitosan/gelatin nanofibers 3D diabetic wound Dermal cells Improved healing Sustained release, anti-inflammatory [ 34 ] Bone & Dental Regeneration Calcium phosphate cements 3D bone models MSCs Osteogenic differentiation, vascularization Scaffold supports tissue growth [ 81 ] PLA/PCL composites 3D dental models Stem cells Enhanced odontogenic differentiation Mineralized matrix formation [ 82 ] Gelatin-based hydrogels (GHMS) 3D bone models – Promoted tissue repair and vascularization Bioactive structural support [ 83 ] Cancer Therapy Lysozyme-functionalized nanoparticles 3D tumor spheroids Cancer cells Improved penetration into stroma Remodels dense tumor matrix [ 84 ] Immunomodulation & Regeneration PCL/PVA membranes GBR models Adipose stem cells, macrophages Bone regeneration; M1→M2 shift Pro-osteogenic and anti-inflammatory [ 85 ] – 3D co-culture UC-MSCs + macrophages Osteogenesis & M2 polarization Modulate PI3K/AKT/mTOR pathway; create pro-regenerative microenvironment [ 86 ] Sol-gel coatings 3D implant models Adipose-derived stem cells Enhanced proliferation and activity Improved biocompatibility [ 87 , 88 ] Summary of Metformin's effects in ECM-Based 3D cell culture models. Applications of metformin-loaded biomaterials in 3D cell culture models. Understanding fibrosis and its relationship with the extracellular matrix (ECM) is fundamental to evaluating metformin's therapeutic potential. Fibrosis is characterized by excessive deposition of ECM proteins, with collagens and fibronectin being the predominant components [ 41 ]. At the cellular level, fibroblasts are the primary cells responsible for producing and maintaining various ECM components, including collagen, elastin, and proteoglycans [ 89 ]. During fibrotic tissue remodeling, fibroblasts undergo activation and differentiate into myofibroblasts. These activated cells are characterized by their expression of α-smooth muscle actin (α-SMA) and increased production of ECM proteins. Myofibroblasts significantly contribute to structural and functional tissue alterations by increasing the deposition of ECM components, particularly collagen types I and III [ 90 ]. The regulation of ECM production and degradation involves complex molecular mechanisms. Fibroblasts secrete various growth factors, including TGF-β and TNF-α, as well as matrix metalloproteinases (MMPs). These metal-dependent proteolytic enzymes play crucial roles in ECM degradation, cell migration, differentiation, and tissue remodeling [ 89 ]. Metformin reduces the production and accumulation of extracellular matrix components across multiple studies. At concentrations of 1–10 mmol/L, metformin significantly decreases the expression of major ECM proteins, including collagen type I (COL1A1) and collagen type III (COL3A1) [ 2 ]. This reduction extends to other ECM components, such as elastin (ELN) and hyaluronic acid (HA), indicating a broad impact on matrix composition [ 2 , 91 ]. A key mechanism of metformin's antifibrotic action is its ability to suppress fibroblast-to-myofibroblast transition, as evidenced by decreased expression of α-smooth muscle actin (α-SMA) [ 92 ]. This effect has been demonstrated across multiple tissue types, including cardiac tissue, where metformin treatment significantly reduces fibrotic gene expression and ECM deposition [ 38 , 39 ]. At the molecular level, metformin's effects are mediated through suppression of key signaling molecules involved in ECM production. The drug reduces expression of transforming growth factor-β (TGF-β), platelet-derived growth factor-β (PDGF-β), and downstream signaling via SMAD2 [ 2 ]. Additionally, metformin decreases production of fibronectin, another major ECM component, particularly in response to PDGF-BB stimulation [ 51 ]. A particularly intriguing mechanism is metformin's ability to promote transdifferentiation of myofibroblasts—a process that directly reduces excessive ECM deposition and facilitates tissue repair. Through AMPK activation, metformin induces a phenotypic switch from myofibroblasts to lipofibroblasts, involving increased expression of bone morphogenetic protein 2 and phosphorylation of peroxisome proliferator-activated receptor-gamma [ 72 ]. This process may contribute to fibrosis attenuation and has been proposed as a potential step toward fibrosis resolution, particularly in lung tissue [ 43 ]. In cardiac tissue models, metformin administration significantly reduces expression of collagen types I and III and attenuates cardiac fibrosis [ 39 ]. Studies on orbital fibroblasts (OFs) in thyroid-associated ophthalmopathy demonstrate that metformin inhibits TGF-β1-induced expression of multiple fibrosis-related molecules, including α-SMA, various collagen types (COL1A1, COL2A1, COL3A1), and fibronectin (FN1) [ 71 ]. In renal tissue models, metformin effectively blocks angiotensin II-induced ECM overproduction in cultured renal fibroblasts, highlighting its direct action on ECM-producing cells. This effect extends to in vivo models, where metformin treatment reduces expression of fibronectin and collagen I in unilateral ureteral obstruction (UUO) mouse models [ 70 ]. In bone tissue models, metformin influences ECM composition differently, promoting the formation of a mineralized extracellular matrix rich in calcium and phosphorus deposits [ 93 ]. The development and validation of more sophisticated 3D culture methods represent a critical next step in understanding metformin's therapeutic potential. These advanced models are needed to better evaluate drug efficacy, optimal dosage regimens, and administration routes ( Table 1 ). The application of 3D culture techniques in diabetes research holds significant promise for both regenerative medicine and drug discovery [ 94 ]. These models are particularly valuable because diabetes affects multiple organs, and 3D systems can more accurately replicate the pathophysiology of diabetic complications using various formats, such as spheroids, organoids, and bioprinted constructs [ 94 ]. Researchers have successfully demonstrated practical applications, such as generating insulin-producing cells from human amniotic epithelial cells in 3D spheroid cultures, which exhibit glucose-dependent insulin secretion [ 49 , 95 ]. The response to metformin in 3D models varies depending on glucose concentration and cell type. Ghandour et al., using co-culture systems, have shown that metformin's effects on angiogenic factors are glucose-dependent. For instance, in endothelial cells, metformin inhibits pro-angiogenic factors such as EMMPRIN and MMP-9 across various glucose concentrations, whereas its effect on VEGF secretion is observed only under high-glucose conditions [ 75 ]. Metformin's effects in 3D culture models have been extensively studied across various tissue types, revealing diverse therapeutic mechanisms. In intestinal organoids, metformin inhibits cell proliferation through two primary pathways: AMPK activation and p53-dependent activation of REDD1, leading to mTOR inhibition and cell cycle arrest [ 50 , 96 ]. In pancreatic models, metformin exhibits significant protective effects, particularly in maintaining cell viability under high-glucose conditions [ 97 ]. The drug protects beta cells from fatty acid-induced apoptosis through AMPK-mediated autophagy activation [ 76 ]. The drug's effects on adipose tissue in 3D models are particularly noteworthy. Metformin impairs adipogenesis by enhancing stemness in human adipose-derived stem cells through autophagy activation and mTOR signaling inhibition—processes that alter ECM stiffness and composition [ 73 ]. Interestingly, metformin's effects on adipogenesis appear to be dose-dependent, with lower concentrations potentially promoting and higher concentrations inhibiting adipogenesis [ 74 , 98 ]. In liver models, metformin demonstrates significant antisteatotic properties, reducing cellular lipid content and fatty acid consumption [ 77 , 99 ]. The drug's effectiveness in 3D culture environments is enhanced by improved cell-cell communication and more frequent drug-cell interactions, which better mimic the native ECM-rich microenvironment [ 77 ]. Recent studies using multi-organ models have shown that metformin can rescue mitochondrial dysfunction and improve glucose transport in both liver tissue and organoid islets under high-glucose conditions [ 78 ]. This demonstrates the drug's potential to simultaneously address multiple aspects of metabolic pathology when studied in complex 3D systems ( Table 1 ).

During the preparation of this work the authors used [Grammarly] in order to improve language clarity and grammar. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication.

The interaction between metformin and the extracellular matrix (ECM) is bidirectional: while metformin actively remodels the ECM, pre-existing biomechanical and biochemical properties of the ECM—such as stiffness, composition, and tissue-specific origin—profoundly modulate cellular responses to the drug. This context-dependent interplay is critical for understanding metformin's efficacy in physiologically relevant 3D models. For instance, in hepatocellular carcinoma (HCC), increased ECM stiffness attenuates metformin's antiproliferative and anti-invasive effects by activating integrin β1 and upregulating miR-17–5p, which suppresses PTEN and activates the PI3K/Akt/MMP pathway—a signaling cascade that overrides metformin's inhibitory actions [ 40 ]. Similarly, in pancreatic cancer, the dense desmoplastic stroma—rich in collagen I and hyaluronic acid—acts as a physical barrier that limits metformin penetration and reduces its bioavailability, thereby diminishing its therapeutic efficacy [ 2 ]. However, metformin can reprogram cancer-associated fibroblasts (CAFs) within this stroma, inducing ECM remodeling and enhancing drug delivery, thereby establishing a positive feedback loop [ 2 ]. In fibrotic liver models, baseline ECM composition influences metformin's efficacy. In matrices exhibiting high cross-linking and lysyl oxidase (LOX) activity, metformin's capacity to suppress TGF-β1/Smad3 signaling and reduce collagen deposition is diminished, suggesting that advanced fibrotic states may be less responsive to therapy [ 43 ]. This underscores how the extent of pre-existing ECM alterations can determine therapeutic outcomes. In adipose tissue, metformin reduces aberrant ECM synthesis and LOX-mediated collagen cross-linking, thereby improving tissue compliance [ 44 ]. Notably, its effects are depot-specific, with greater reductions in collagen deposition observed in epididymal white adipose tissue (eWAT) compared to inguinal white adipose tissue (iWAT). A positive correlation exists between reduced hydroxyproline levels and improved glycemic control [ 44 ]. In cardiovascular contexts, metformin counteracts hyperglycemia-induced ECM dysfunction by restoring the endothelial glycocalyx (GCX), downregulating adhesion molecules (e.g., ICAM-1, E-selectin), and reducing endothelial stiffness [ 42 ]. In reproductive tissues, it ameliorates ECM imbalances in polycystic ovary syndrome (PCOS) by modulating MMP/TIMP ratios, thereby promoting folliculogenesis [ 45 ]. I. Modulation of miRNAs: Metformin upregulates miR-33a, which targets and downregulates c-MYC, thereby reducing ECM deposition and fibrosis [ 46 ]. It also enhances DICER expression—a key enzyme in miRNA biogenesis—altering the expression of energy metabolism-related miRNAs that suppress ECM-associated genes [ 46 , 47 ]. Modulation of miRNAs: Metformin upregulates miR-33a, which targets and downregulates c-MYC, thereby reducing ECM deposition and fibrosis [ 46 ]. It also enhances DICER expression—a key enzyme in miRNA biogenesis—altering the expression of energy metabolism-related miRNAs that suppress ECM-associated genes [ 46 , 47 ]. In cancer, metformin upregulates miR-34a and miR-200, inhibiting epithelial-mesenchymal transition (EMT) and collagen synthesis. It also modulates miR-143–3p to suppress ECM deposition [ 47 , 48 ]( Fig. 2 ). II. Inhibition of Aerobic Glycolysis: Metformin suppresses PFKFB3 (6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase 3), a key regulator of glycolytic flux, thereby reducing lactate production and glucose consumption. This inhibition is mediated via AMPK/mTOR signaling, resulting in decreased COL1A1 mRNA and α-SMA protein expression [ 47 ]. This mechanism helps prevent pathological ECM remodeling in fibrotic conditions ( Fig. 3 ). Fig. 2 Modulation of miRNAs by metformin in fibrotic contexts. Metformin upregulates miR-143–3p, which directly targets and inhibits ERK5 (MAPK7), leading to reduced extracellular matrix (ECM) deposition, suppression of fibrotic signaling, and decreased proliferation in hepatic and renal fibroblasts [ 47 , 48 ]. Fig. 2 Inhibition of Aerobic Glycolysis: Metformin suppresses PFKFB3 (6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase 3), a key regulator of glycolytic flux, thereby reducing lactate production and glucose consumption. This inhibition is mediated via AMPK/mTOR signaling, resulting in decreased COL1A1 mRNA and α-SMA protein expression [ 47 ]. This mechanism helps prevent pathological ECM remodeling in fibrotic conditions ( Fig. 3 ). Modulation of miRNAs by metformin in fibrotic contexts. Metformin upregulates miR-143–3p, which directly targets and inhibits ERK5 (MAPK7), leading to reduced extracellular matrix (ECM) deposition, suppression of fibrotic signaling, and decreased proliferation in hepatic and renal fibroblasts [ 47 , 48 ]. Metformin suppresses PFKFB3 (6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase 3), a key regulator of glycolytic flux, thereby reducing lactate production and glucose consumption. This inhibition is mediated via AMPK/mTOR signaling, resulting in decreased [ 49 , 50 ]. This mechanism helps prevent pathological ECM remodeling in fibrotic conditions ( Fig. 3 ).” III. AMPK Activation : Metformin inhibits mitochondrial complex I, reducing ATP production and increasing the AMP/ATP ratio, thereby activating AMPK. Activated AMPK subsequently inhibits mTOR signaling and suppresses ECM protein synthesis [ 39 ]. Fig. 3 Inhibition of aerobic glycolysis as an antifibrotic mechanism of metformin. Metformin suppresses PFKFB3 (6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase 3)—a key regulator of glycolytic flux—via AMPK/mTOR signaling. This results in reduced lactate production, decreased expression of collagen I/III and α-SMA, and attenuation of pathological ECM remodeling in fibrotic tissues [ 47 , 51 ]. Fig. 3 AMPK Activation : Metformin inhibits mitochondrial complex I, reducing ATP production and increasing the AMP/ATP ratio, thereby activating AMPK. Activated AMPK subsequently inhibits mTOR signaling and suppresses ECM protein synthesis [ 39 ]. Inhibition of aerobic glycolysis as an antifibrotic mechanism of metformin. Metformin suppresses PFKFB3 (6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase 3)—a key regulator of glycolytic flux—via AMPK/mTOR signaling. This results in reduced lactate production, decreased expression of collagen I/III and α-SMA, and attenuation of pathological ECM remodeling in fibrotic tissues [ 47 , 51 ]. Additionally, at clinically relevant concentrations, metformin binds PEN2, inhibiting vacuolar ATPase (v-ATPase) and activating AMPK independently of cellular AMP levels—further suppressing collagen and fibronectin production [ 52 , 53 ] ( Fig. 4 ). IV. TGF-β Inhibition: Metformin inhibits TGF-β1-induced Smad2/3 phosphorylation, thereby preventing myofibroblast differentiation and accumulation of ECM proteins. Through AMPK activation, it disrupts TGF-β/Smad3 signaling, reducing expression of α-SMA and fibronectin—a critical antifibrotic mechanism [ 44 , 54 ] ( Fig. 5 ). Fig. 4 AMPK-dependent and -independent pathways in metformin-mediated ECM regulation . Metformin activates AMPK through mitochondrial complex I inhibition (increasing AMP/ATP ratio) or via PEN2/v-ATPase binding. Activated AMPK suppresses mTOR and TGF-β1/Smad3 signaling, leading to reduced synthesis of collagen, fibronectin, and other ECM proteins [ 39 ]. Fig. 4 Fig. 5 Metformin inhibits TGF-β signaling—a central driver of fibrosis—by suppressing Smad2/3 phosphorylation . This results in downregulation of key fibrogenic mediators, including α-SMA (myofibroblast marker), LOXL1 (collagen cross-linking enzyme), TIMP1 (inhibitor of matrix degradation), and ECM-related processes such as epithelial-mesenchymal transition (EMT) and activation of hepatic stellate cells (HSCs) and myofibroblasts (MFBs) [ 53 , 54 ]. Fig. 5 TGF-β Inhibition: Metformin inhibits TGF-β1-induced Smad2/3 phosphorylation, thereby preventing myofibroblast differentiation and accumulation of ECM proteins. Through AMPK activation, it disrupts TGF-β/Smad3 signaling, reducing expression of α-SMA and fibronectin—a critical antifibrotic mechanism [ 44 , 54 ] ( Fig. 5 ). AMPK-dependent and -independent pathways in metformin-mediated ECM regulation . Metformin activates AMPK through mitochondrial complex I inhibition (increasing AMP/ATP ratio) or via PEN2/v-ATPase binding. Activated AMPK suppresses mTOR and TGF-β1/Smad3 signaling, leading to reduced synthesis of collagen, fibronectin, and other ECM proteins [ 39 ]. Metformin inhibits TGF-β signaling—a central driver of fibrosis—by suppressing Smad2/3 phosphorylation . This results in downregulation of key fibrogenic mediators, including α-SMA (myofibroblast marker), LOXL1 (collagen cross-linking enzyme), TIMP1 (inhibitor of matrix degradation), and ECM-related processes such as epithelial-mesenchymal transition (EMT) and activation of hepatic stellate cells (HSCs) and myofibroblasts (MFBs) [ 53 , 54 ]. Collectively, these mechanisms underscore metformin's broad influence on ECM dynamics and its therapeutic potential in 3D models of cancer, fibrosis, and metabolic disorders .

Metformin, a first-line therapeutic for type 2 diabetes, has garnered substantial attention for its potential applications beyond glycemic control. Recent reviews [ 1 , 2 ] highlight its therapeutic benefits in diverse diseases, mediated through mechanisms including metabolic regulation, anti-inflammatory effects, and modulation of extracellular matrix (ECM) properties. The extracellular matrix (ECM), a complex network of proteins and polysaccharides, provides essential structural and biochemical support for cells, playing a pivotal role in maintaining tissue homeostasis and regulating cellular signaling in both physiological and pathological processes [ 3 ]. Dysregulation of ECM structure and function is a hallmark of numerous diseases, including cancer, fibrosis, and metabolic disorders [ 4 ]. In three-dimensional (3D) cell culture systems, which more accurately recapitulate the physiological microenvironment than traditional 2D cultures, ECM dynamics are critical for cell proliferation, differentiation, and migration [ 5 ]. Emerging evidence indicates that metformin can alter ECM composition and biomechanics, thereby influencing cell-ECM interactions, remodeling processes, and mechanotransduction pathways. These findings suggest novel therapeutic avenues for metformin in conditions characterized by dysregulated ECM remodeling, such as tumor microenvironments and fibrotic diseases. Notably, these effects align with broader metabolic intervention strategies, including magnesium supplementation, which aim to improve ECM-related outcomes in metabolic disorders, underscoring the interconnectedness of metabolism and ECM homeostasis in disease pathogenesis [ 6 , 7 ]. Despite promising findings, significant research gaps persist in understanding how the extracellular matrix (ECM) modulates metformin's therapeutic effects within 3D cell culture models. First, the precise mechanisms by which ECM composition and stiffness influence metformin's actions—particularly on AMPK/mTOR signaling, metabolic reprogramming, and epithelial-mesenchymal transition (EMT)—remain incompletely elucidated. Second, the absence of standardized ECM-based 3D models limits reproducibility and hinders cross-study comparisons and clinical translation. Third, most current models fail to incorporate key components of the tissue microenvironment, such as immune cells and stromal fibroblasts, despite metformin's well-documented immunomodulatory and anti-fibrotic properties. This mini-review addresses these gaps by synthesizing current evidence on metformin's role in ECM-based 3D systems across cancer, fibrosis, and metabolic diseases. We highlight how these advanced models reveal context-dependent drug responses and provide a more physiologically relevant platform for evaluating metformin's therapeutic potential beyond traditional 2D cultures [ 8 ].

The funding organization(s) played no role in the study design; in the collection, analysis, and interpretation of data; in the writing of the report; or in the decision to submit the report for publication. There is no confits of interest.

Although metformin-loaded biomaterials have been extensively explored for regenerative medicine and systemic glucose control, their integration with three-dimensional (3D) cell culture models offers a unique opportunity to study the drug's therapeutic effects in a physiologically relevant context. These systems are not merely passive carriers but active components that can modulate the extracellular matrix (ECM) microenvironment and serve as platforms for disease modeling. Beyond its local therapeutic effects in wound healing, metformin-loaded polymeric biomaterials have also been engineered to achieve sustained systemic release for glycemic control, offering a promising strategy for diabetes management. Preclinical studies demonstrate that these advanced delivery systems consistently outperform conventional metformin formulations in both glucose-lowering efficacy and duration of action. For instance: • Temperature-responsive microneedle systems have reduced blood glucose levels to 7.6 mM (within normal range) in diabetic mice, with metformin release directly correlated with glucose reduction [ 100 , 101 ]. • Niosome-based delivery systems achieved a maximum blood glucose reduction of 45.89 % (vs. 25.21 % with free metformin), with effects sustained for 6–8 h [ 102 , 103 ]. • Advanced dual-encapsulation platforms using lipid-core systems in alginate beads demonstrated 122 % relative bioavailability and maintained 71 % glucose reduction after 29 days [ 79 ]. • PLGA-based nanoparticles reduced glucose levels from 529.9 mg/dL to 286.5 mg/dL in diabetic animal models, demonstrating systemic benefits with lower drug doses [ 104 , 105 ]. • Glucose-responsive smart systems, such as gelatin-phenylboronic acid microparticles, normalized glucose levels within 3 h and maintained control for 8 h [ 80 ]. • Advanced microneedle technologies achieved 90 % metformin release within 8 h and demonstrated anti-obesogenic effects, including improved glucose homeostasis and reduced inflammation [ 106 ]. Temperature-responsive microneedle systems have reduced blood glucose levels to 7.6 mM (within normal range) in diabetic mice, with metformin release directly correlated with glucose reduction [ 100 , 101 ]. Niosome-based delivery systems achieved a maximum blood glucose reduction of 45.89 % (vs. 25.21 % with free metformin), with effects sustained for 6–8 h [ 102 , 103 ]. Advanced dual-encapsulation platforms using lipid-core systems in alginate beads demonstrated 122 % relative bioavailability and maintained 71 % glucose reduction after 29 days [ 79 ]. PLGA-based nanoparticles reduced glucose levels from 529.9 mg/dL to 286.5 mg/dL in diabetic animal models, demonstrating systemic benefits with lower drug doses [ 104 , 105 ]. Glucose-responsive smart systems, such as gelatin-phenylboronic acid microparticles, normalized glucose levels within 3 h and maintained control for 8 h [ 80 ]. Advanced microneedle technologies achieved 90 % metformin release within 8 h and demonstrated anti-obesogenic effects, including improved glucose homeostasis and reduced inflammation [ 106 ]. Importantly, these metformin-releasing scaffolds can be incorporated into 3D cell culture models to evaluate their efficacy in a disease-relevant setting. For example: • Diabetic wound healing scaffolds containing metformin can be co-cultured with patient-derived fibroblasts and endothelial cells to assess angiogenesis and ECM remodeling in 3D spheroids or organoids [ 34 ]. • Metformin-loaded biomaterial scaffolds—including calcium phosphate cements, PLA/PCL composites, calcium phosphate–chitosan matrices, and gelatin-based hydrogels (GHMs)—have demonstrated significant potential in 3D disease models for bone and dental tissue engineering by supporting cell viability, enhancing osteogenic and odontogenic differentiation, and promoting vascularization and tissue regeneration [ 81 , 83 , 107 ]. • In cancer research, lysozyme-functionalized metformin-loaded nanoparticles have shown enhanced penetration into tumor spheroids by remodeling the dense tumor stroma [ 84 ]. Diabetic wound healing scaffolds containing metformin can be co-cultured with patient-derived fibroblasts and endothelial cells to assess angiogenesis and ECM remodeling in 3D spheroids or organoids [ 34 ]. Metformin-loaded biomaterial scaffolds—including calcium phosphate cements, PLA/PCL composites, calcium phosphate–chitosan matrices, and gelatin-based hydrogels (GHMs)—have demonstrated significant potential in 3D disease models for bone and dental tissue engineering by supporting cell viability, enhancing osteogenic and odontogenic differentiation, and promoting vascularization and tissue regeneration [ 81 , 83 , 107 ]. In cancer research, lysozyme-functionalized metformin-loaded nanoparticles have shown enhanced penetration into tumor spheroids by remodeling the dense tumor stroma [ 84 ]. This synergy between metformin-loaded biomaterials and 3D culture systems paves the way for personalized therapeutic strategies, wherein drug delivery platforms serve not only for treatment but also as tools for predictive disease modeling and drug screening. Moreover, metformin's ability to influence mesenchymal stem cells (MSCs) enhances the regenerative potential of biomaterial scaffolds. In human umbilical cord MSCs (UC-MSCs), metformin promotes osteogenic differentiation and drives M2 macrophage polarization via the PI3K/AKT/mTOR pathway, creating an immunomodulatory microenvironment conducive to tissue repair [ 86 ]. In periodontal applications, metformin-loaded polycaprolactone/polyvinyl alcohol membranes pre-seeded with endometrial stem cells have demonstrated enhanced bone regeneration in guided bone regeneration (GBR) models [ 85 ], while sol-gel coatings on metallic implants have enhanced adipose-derived stem cell proliferation and metabolic activity [ 87 , 88 ]. These advances demonstrate that metformin-loaded biomaterials are not merely delivery vehicles but integral components of advanced 3D models, reinforcing the central theme of this review: the dynamic interplay between metformin, the ECM, and the 3D cellular microenvironment.

Three-dimensional (3D) cell culture is an in vitro model in which cells proliferate within a scaffold or self-assembled matrix, enabling interactions in all three spatial dimensions. Unlike traditional two-dimensional (2D) monolayer cultures, 3D systems more accurately recapitulate the structural organization, cell-cell and cell-ECM interactions, and functional heterogeneity of native tissues [ 18 ]. This enhanced physiological relevance renders 3D models powerful tools for disease modeling, drug screening, and regenerative medicine. The fundamental distinction between 2D and 3D cultures lies in their spatial architecture. In 2D systems, cells are confined to flat, rigid substrates, resulting in unnatural morphology, altered cytoskeletal organization, and restricted cell-cell and cell-ECM interactions [ 19 ]. The artificial mechanical stress in 2D environments—up to a million-fold greater than that in soft tissues—can profoundly influence gene expression, signaling pathways, and drug responses [ 20 ]. In contrast, 3D cultures enable cells to maintain their native 3D morphology and assemble into complex multicellular structures, thereby providing a more accurate representation of in vivo tissue behavior [ 18 ]. Notably, approximately 50 % of genes exhibit altered expression upon transition from 2D to 3D culture, underscoring the substantial biological divergence between these models [ 20 ]. The integration of ECM components into 3D models further enhances their physiological fidelity. ECM-based systems are particularly valuable in disease research, enabling the study of dynamic processes—including tumor-stroma crosstalk, hypoxia development, and fibrotic remodeling—that are frequently absent in 2D cultures [ 21 ]. In cancer research, 3D models provide critical insights into tumor biology, including mechanisms of invasion, drug resistance, and the role of the tumor microenvironment (TME) [ 21 , 22 ]. These models facilitate the formation of hypoxic cores and nutrient gradients, thereby mimicking in vivo tumor architecture and enabling more accurate assessment of therapeutic efficacy. Similarly, in neurodegenerative disorders such as Alzheimer's disease, 3D models incorporating brain-mimetic ECM components offer enhanced platforms for studying neuronal network formation and disease progression [ 23 , 24 ]. Beyond oncology and neuroscience, 3D cultures are increasingly employed to model metabolic disorders, liver diseases, endometriosis, and infectious diseases, offering deeper insights into cellular and molecular mechanisms under pathophysiological conditions [ 25 , 26 ]. Their capacity to support co-cultures and patient-derived cells further enhances their translational relevance. 3D cell culture systems can be broadly categorized into scaffold-based and scaffold-free models [ 27 ]. The primary distinction lies in whether cells are supported by exogenous materials or permitted to self-assemble naturally. Three-dimensional (3D) cell culture systems can be broadly categorized based on the presence or absence of an exogenous scaffold that supports cell growth and organization. This classification is not mutually exclusive with the use of advanced platforms (e.g., bioreactors or bioprinting), as both scaffold-based and scaffold-free models can be integrated into such systems. Scaffold-free systems rely on the intrinsic ability of cells to self-assemble and produce their own endogenous extracellular matrix (ECM) in the absence of external biomaterials [ 28 ]. The most common examples include: • Spheroids: free-floating multicellular aggregates formed via hanging drop, low-adhesion plates, or spinner flasks [ 29 , 30 ]. • Organoids : self-organizing 3D structures derived from stem cells that recapitulate organ-specific architecture and function [ 29 , 31 ]. Spheroids: free-floating multicellular aggregates formed via hanging drop, low-adhesion plates, or spinner flasks [ 29 , 30 ]. Organoids : self-organizing 3D structures derived from stem cells that recapitulate organ-specific architecture and function [ 29 , 31 ]. Importantly, spheroids and organoids can also be cultured within exogenous scaffolds (e.g., Matrigel, collagen hydrogels). In such cases, they are classified as scaffold-based models, as the external matrix actively influences cell behavior [ 29 , 31 ]. Scaffold-based systems utilize natural or synthetic matrices that mimic the biochemical and biophysical properties of the native ECM [ 32 ]. These include: • Natural hydrogels (e.g., Matrigel, collagen, fibrin), which provide a supportive 3D architecture while permitting nutrient diffusion and facilitating cell migration [ 32 ]. • Decellularized ECM (dECM) derived from tissues, which preserves tissue-specific biochemical and mechanical cues [ 33 ]. • Synthetic polymers (e.g., PCL, PLGA), which, although not inherently biological, can be functionalized with biological ECM components (e.g., collagen, fibronectin) to create ECM-mimetic microenvironments for cell culture and controlled drug delivery [ 34 ]. Natural hydrogels (e.g., Matrigel, collagen, fibrin), which provide a supportive 3D architecture while permitting nutrient diffusion and facilitating cell migration [ 32 ]. Decellularized ECM (dECM) derived from tissues, which preserves tissue-specific biochemical and mechanical cues [ 33 ]. Synthetic polymers (e.g., PCL, PLGA), which, although not inherently biological, can be functionalized with biological ECM components (e.g., collagen, fibronectin) to create ECM-mimetic microenvironments for cell culture and controlled drug delivery [ 34 ]. Beyond scaffold classification, advanced technological platforms enhance the physiological relevance of 3D models by enabling dynamic control over the cellular microenvironment. These platforms are not a third scaffold category, but rather engineering approaches applicable to both scaffold-free and scaffold-based systems. • Bioreactor systems (e.g., spinner flasks, perfusion bioreactors, hollow fiber systems) provide controlled fluid flow, gas exchange, and mechanical stimulation. Hollow fiber bioreactors, for instance, have demonstrated up to 19.4-fold higher cell yields compared to 2D cultures [ 35 ]. • Microfluidic “organ-on-a-chip” devices integrate permeable membranes and microscale fluid dynamics to simulate interstitial flow, vascular perfusion, and multi-tissue crosstalk [ 32 ]. • 3D bioprinting enables spatially precise deposition of cells, biomaterials, and metformin-loaded carriers to construct tissue-like architectures with controlled ECM composition [ 36 ]. Bioreactor systems (e.g., spinner flasks, perfusion bioreactors, hollow fiber systems) provide controlled fluid flow, gas exchange, and mechanical stimulation. Hollow fiber bioreactors, for instance, have demonstrated up to 19.4-fold higher cell yields compared to 2D cultures [ 35 ]. Microfluidic “organ-on-a-chip” devices integrate permeable membranes and microscale fluid dynamics to simulate interstitial flow, vascular perfusion, and multi-tissue crosstalk [ 32 ]. 3D bioprinting enables spatially precise deposition of cells, biomaterials, and metformin-loaded carriers to construct tissue-like architectures with controlled ECM composition [ 36 ]. Critically, the choice of platform modulates metformin's effects in a context-dependent manner. For example, in Matrigel-based spheroids, metformin disrupts structural integrity [ 37 ], whereas in tumor- or liver-derived dECM models, it suppresses fibrotic gene expression [ 38 , 39 ]. This underscores the need to consider both ECM composition and culture platform when evaluating metformin's therapeutic potential.