Cellular crosstalk of fibroblast–myofibroblast transition in intestinal homeostasis and disease

Abstract

Fibroblasts are central components of the intestinal microenvironment and can acquire a myofibroblast phenotype through fibroblast-to-myofibroblast transition (FMT), a process linked to increased tissue mechanosensitivity and active extracellular matrix (ECM) remodeling. In this way, FMT participates in normal intestinal homeostasis and repair, but it can also contribute to fibrosis and tumor-associated stromal remodeling when sustained. More importantly, FMT is not a passive process but an active reprogramming event in response to physiological demands or pathological conditions. Its regulation is highly complicated by the temporal and spatial coordination of various intestinal cell types, ultimately determining disease onset and progression. In this context, targeting cellular interaction networks and remodeling the pathological microenvironment may provide opportunities to limit or redirect disease-associated stromal remodeling. In this review, we summarize the key intercellular dialogs that regulate FMT during the onset and progression of intestinal disorders and further analyze their regulatory mechanisms. Moreover, we discuss emerging therapies targeting cellular interaction networks to provide a new framework and research focus for treating inflammatory bowel disease (IBD), intestinal fibrosis, and related tumors.

Introduction

As a highly complex and dynamic microbial ecosystem in the human body, the intestine maintains homeostasis through complicated interactions among the epithelial barrier, stromal support, microbial community, and immune network mediated by metabolites and cytokines [1]. However, despite increasing attention to stromal cells in tissue homeostasis and inflammation, how fibroblasts undergo transition to myofibroblasts and interact with surrounding cells remains unclear [2].

Fibroblasts are the main producers and remodelers of the intestinal ECM, supporting tissue architecture and coordinating cellular interactions in homeostasis and disease [3]. A key event in this process is the upregulation of α-smooth muscle actin (α-SMA). Although a small fraction of α-SMA⁺ fibroblasts may be present at baseline, chemical or mechanical stimulation markedly increases α-SMA expression, driving their transition toward a myofibroblast phenotype [4]. Myofibroblasts are fibroblasts with smooth muscle-like properties and are characterized by α-SMA–mediated stress fiber formation, which results in enhanced contractility and a characteristic elongated, spindle-shaped morphology [5]. This transition is accompanied by an increased ability to synthesize ECM components such as collagen, fibronectin, and laminin, although the upregulation of ECM synthesis does not always occur in strict synchrony with phenotypic conversion [4].

In intestinal diseases such as IBD, inflammation and repair are activated almost simultaneously, with the inflammatory peak typically preceding the reparative phase [6]. Inflammatory signals promote repair, whereas activated fibroblasts increase inflammation, which is a positive feedback loop [7]. As inflammation subsides, reparative activity becomes more prominent. During this process, myofibroblasts generate contractile forces that facilitate wound closure and ECM remodeling [8], and subsequently undergo apoptosis or return to quiescence as sustained signals decrease [9]. However, when inflammation persists orthe intestine is subjected to repeated injury, myofibroblasts accumulate and exhibit resistance to apoptosis, leading to fibrosis [10]. Moreover, the chronic inflammatory microenvironment can induce malignant epithelial transformation, whereas tumor cells can reprogram adjacent fibroblasts into cancer-associated fibroblasts (CAFs), thereby further promoting tumor progression and metastasis [11].

Intestinal FMT is increasingly recognized not as an isolated fibroblast-intrinsic event, but as a microenvironment-regulated remodeling process coordinated by epithelial, immune, vascular, microbial, and mechanical cues. Classical profibrotic pathways such as transforming growth factor-β (TGF-β) signaling interact with mechanosensing, inflammatory mediators, metabolic reprogramming, and spatial niche signals to determine whether fibroblast activation remains reparative or becomes pathological. In the intestine, this regulation is strongly shaped by local tissue architecture, including cellular interactions along the crypt–villus axis [12], macrophage–fibroblast communication in the lamina propria [13], and fibroblast–immune cell networks that form lymphoid scaffolds [14]. Clinically, persistent stromal activation drives fibrotic remodeling in Crohn’s disease (CD) and contributes to tumor-associated microenvironmental changes that may reduce therapeutic responses [15, 16]. These observations suggest that complex multicellular interactions within diseased tissues may diminish the effectiveness of single-target therapies. A more systematic understanding of these intercellular networks may therefore help identify strategies to modify the microenvironment and improve therapeutic outcomes.

Therefore, this review discusses the intercellular dialogues that regulate intestinal FMT and highlights emerging strategies targeting these interactions. Understanding these mechanisms may guide microenvironment-oriented therapies for inflammatory and fibrotic intestinal disorders and related diseases characterized by abnormal stromal remodeling.

Cellular identity and plasticity in intestinal FMT

Cellular sources of intestinal stromal cells

Intestinal fibroblasts are spindle-shaped stromal cells with marked heterogeneity in origin, location, and functional state. During development, mesoderm-derived mesenchymal progenitors give rise to early fibroblast populations that support epithelial barrier formation and villus–crypt morphogenesis, a process partly modeled in organoid–stromal cocultures [17, 18]. However, whether these organoids fully recapitulate the in vivo complexity remains unclear [19].

During homeostasis, intestinal fibroblast renewal is mainly supported by tissue-resident mesenchymal progenitors, including Gli1⁺ cells at the crypt base and CD34⁺ cells near submucosal vessels, although their relative contributions may vary across regions and physiological states [20, 21].Under pathological conditions, resident fibroblasts expand and shift toward activated or myofibroblast-like phenotype. Other cell populations may also contribute to the stromal pool through context-dependent plasticity, including mesenchymal stem/stromal cell differentiation, epithelial–mesenchymal transition (EMT), endothelial–mesenchymal transition (EndoMT), pericyte activation, and, in certain contexts, macrophage-to-myofibroblast transition (MMT) [22–26]. It should be noted that EMT and EndoMT may generate cells with mesenchymal-like or stromal-like features, some of which can acquire myofibroblast-like properties under profibrotic or tumor-associated cues. However, whether these processes directly and consistently contribute to mature myofibroblast formation remains uncertain [27–29]. These potential cellular routes and their associated signaling cues are summarized in Fig. 1, emphasizing that intestinal myofibroblasts are better understood as an inducible functional state shaped by local microenvironmental signals rather than a fixed lineage-committed population.

Fig. 1.

Diverse cellular origins and signaling mechanisms drive myofibroblast formation. Multiple cellular sources and signaling pathways may contribute to myofibroblast formation. Resident fibroblasts and other stromal cells are likely the main sources, while direct conversion from epithelial, endothelial, pericyte, or macrophage lineages remains uncertain. Dashed arrows indicate lineage transitions for which direct in vivo evidence remains limited or controversial

Molecular definition and heterogeneity

Although fibroblast heterogeneity is now well recognized, traditional studies have largely relied on broad mesenchymal markers such as vimentin, fibroblast-specific protein 1/S100 calcium-binding protein A4 (FSP1/S100A4), and CD90 to define fibroblast populations and guide functional analyses [30–32]. However, these markers lack specificity, as they are also expressed by pericytes, SMCs, and certain immune subsets, making it difficult to precisely identify fibroblasts or capture their functional diversity. For instance, studies that used FSP1/S100A4 to trace fibroblast fate were later re-evaluated using lineage tracing and single-cell transcriptomics, revealing that these markers also label substantial numbers of immune cells, leading to the overestimation and misinterpretation of fibroblast functions [33]. Similarly, although vimentin and CD90 are widely used, their broad expression limits cross-model and cross-laboratory comparability.

Given these challenges, the use of single-cell RNA sequencing (scRNA-seq) has highlighted the weakness of the traditional “panmarker” framework. High-resolution atlases of intestinal and systemic fibroblasts have demonstrated extensive transcriptional diversity, with functionally distinct subsets displaying unique transcriptional profiles and spatial niches [34, 35](Table 1). However, scRNA-seq, which clusters cells according to highly variable gene expression profiles, often leads to complex and overlapping marker systems, making subpopulation delineation prone to noise and unclear true functional boundaries [36]. Integrating functional assays and anatomical context is therefore essential when interpreting these datasets. Spatial transcriptomics adds spatial information to transcriptional profiles and can help relate molecular patterns to tissue organization and local interactions. However, whether transcriptionally defined “niches” or “modules” represent stable functional units or simply reflect anatomical proximity remains unclear and requires further experimental validation [37]. Therefore, molecular classification should be interpreted together with functional assays and anatomical context. Multiomics and cross-cohort analyses further show that fibroblast programs include both conserved and tissue-specific components. For example, Pi16⁺ and Col15a1⁺ fibroblasts recur across organs and are considered conserved stromal subsets involved in homeostasis and early injury responses [38].In our view, rather than discarding traditional marker-based frameworks altogether, it is more helpful to refine and complement them with emerging molecular and spatial data. Such integrative approaches may improve our understanding of intestinal fibroblast heterogeneity in homeostasis, inflammation, fibrosis, and tumor-associated remodeling.

Table 1.

Spatial and functional heterogeneity of intestinal fibroblast subsets

Fibroblast subsets	Representative markers	Spatial localization	Physiological functions	Pathological states	Species	References
Crypt-associated fibroblasts	PDGFRαlo, GREM1, RSPO3, WNT2/2B	Crypt base and pericryptal niche	Maintain ISC proliferation and self-renewal, crypt niche support	Expanded after injury; acquire CAF-like signatures in tumorigenesis	Mouse, conserved in human	[39–42]
Subepithelial telocytes / PDGFRαhi fibroblasts	PDGFRαhi, FOXL1, COL6A1, CD201, BMP ligands, WNT4/5	Beneath epithelium, crypt–villus axis, villus core	Orchestrate WNT–BMP gradients, regulate epithelial differentiation, villus homeostasis	Dysregulated in fibrosis and colitis; CAF-like phenotype in cancer	Mouse, conserved inhuman	[39, 41, 43]
Conserved fibroblast populations	PI16, DPT, CD34; COL15A1; CD81, CD90	Interstitial and stromal compartments across tissues	Reservoir for tissue-specialized fibroblast states	Potential source of activated fibroblasts during tissue perturbation	Mouse; conserved in human	[20, 38, 44]
Myofibroblasts / fibrosis-associated fibroblasts	ACTA2, POSTN, FAP, COL1A1,TWIST1, CD73 (NT5E), FGFR2	Submucosa, fibrotic foci	ECM organization and structural support	Major profibrotic subsets driving CD strictures and intestinal fibrosis	Human, mouse homologues	[41, 45]
CAF-like fibroblasts (SULF1⁺ CAFs, inflammatory/fibrotic CAF-like phenotype)	FAP, ACTA2, PDPN, COL1A1, SULF1, TGFB-responsive genes	Tumor stroma, adjacent to oncogenic epithelium	Stromal support, angiogenesis, immune modulation	Promote tumor progression, desmoplasia, angiogenesis, and immune evasion	Human, mouse	[42, 46, 47]
Inflammatory fibroblasts	IL11, IL24, CCL2, CCL19, CXCL2, CXCL12, CHI3L1,	Lamina propria, inflamed mucosa (ulcerative colitis, CD)	Crosstalk with immune cells, drive inflammatory signaling	Expanded in ulcerative colitis/CD; linked to fibrosis, tumor promotion, and resistance to anti-TNF therapy	Human, mouse	[34, 40, 48, 49]

Fibroblast plasticity and physiological versus pathological FMT

Although transcriptomic and multi-omics analyses have revealed extensive fibroblast heterogeneity, interpreting their roles in intestinal homeostasis and disease requires a functional perspective that distinguishes quiescent resident fibroblasts from activated stromal states. These states are not discrete but instead form a continuum of plasticity and may dynamically interconvert [50]. Quiescent resident fibroblasts maintain tissue stability by supporting baseline ECM synthesis, immune regulation, and epithelial renewal, whereas activated stromal states contribute to repair, inflammation, fibrosis, and tumor-associated remodeling. These fibroblast states are not fixed. Instead, they form a plastic continuum in which myofibroblasts, inflammatory fibroblasts (IAFs), and CAFs may overlap and interconvert depending on local cues [51].

Among these activated subsets, myofibroblasts represent the most characteristic effector phenotype of fibroblast activation. They were first described in granulation tissue as cells with features of both smooth muscle cells (SMCs) and fibroblasts, characterized by cytoplasmic actin microfilament bundles [5]. α-SMA was later established as a core marker of myofibroblasts [52]. However, some early-stage myofibroblasts display relatively low α-SMA expression despite measurable contractility and ECM remodeling [4, 53], and α-SMA is also expressed by pericytes and SMCs [54]. Additional markers can provide complementary information on different aspects of fibroblast activation. Fibroblast activation protein (FAP) and periostin are commonly associated with activated stromal populations, while collagen type I alpha 1 chain (COL1A1), collagen type III alpha 1 chain, and fibronectinreflect enhanced ECM production and remodeling. Transgelin is more closely linked to the acquisition of contractile properties, whereas platelet-derived growth factor receptor β (PDGFRβ) is frequently enriched in activated mesenchymal populations but is not restricted to fibroblasts. In our view, reliable identification should not rely on molecular markers alone but also consider morphological and functional features, including prominent stress fibers and focal adhesions, robust ECM synthesis, and contractile activity in vivo and in vitro.

FMT should be viewed as a plastic process regulated by biochemical signals, mechanical cues, and the inflammatory–metabolic milieu, characterized by multiple markers, functional properties, and spatial context. For example, emerging intestinal studies further suggest that Meflin-enriched fibroblast states may favor repair and restrain fibrosis [55], whereas TWIST1⁺FAP⁺ fibroblasts are associated with persistent ECM production in CD–related fibrosis [45].Within this continuum, physiological FMT and pathological FMT can be distinguished by the duration, spatial extent, and reversibility of fibroblast activation. Physiological FMT refers to a transient, spatially restricted, and self-limiting response to tissue injury, in which fibroblasts acquire contractile features, deposit and remodel provisional ECM, and facilitate wound closure [56]. As injury resolves, these activated cells undergo apoptosis or dedifferentiate toward a quiescent state, allowing stromal homeostasis to be restored [57]. In contrast, pathological FMT refers to sustained fibroblast activation that fails to resolve and progressively extends beyond the initial repair niche. Apoptosis-resistant myofibroblasts continue to produce and remodel ECM, leading to excessive matrix deposition and tissue stiffening, which in turn reinforce fibroblast activation [10, 58]. Such persistent FMT underlies chronic fibrotic remodeling and, in tumors, may contribute to the formation of an immune-exclusionary, tumor-supportive stroma and reduced therapeutic responsiveness [59]. Thus, physiological FMT represents a transient and reversible transition during tissue repair, whereas pathological FMT is characterized by sustained fibroblast activation, impaired myofibroblast resolution, and persistent stromal remodeling.

Signaling control and functional outcomes of FMT

In physiological settings, subepithelial myofibroblasts are present in low numbers but maintain intestinal stem cell renewal, regulate barrier integrity, and sense microbial or injury-associated signals to coordinate mucosal regeneration and immune responses [60]. In pathological situations, mesenchymal cells, predominantly fibroblasts, undergo extensive proliferation and FMT to generate α-SMA⁺ contractile cells with increased ECM synthesis and tensile force [4]. As discussed above, most myofibroblasts are regarded as a transient, context-dependent functional state rather than a fixed lineage [61]. Lineage-tracing and in vivo studies further show that their persistence can be sustained by chronic inflammation or tumor-associated stress through mechanobiological feedback involving TGF-β, integrins, and YAP/TAZ signaling, blurring the boundary between adaptive repair and pathological fibrosis or tumor-associated scarring [62]. This functional overlap underscores the challenge of preserving transient reparative responses across disease stages and tissue contexts while limiting sustained fibrosis or tumor-promoting stromal changes.

At the signaling level, FMT is regulated by multiple biochemical, mechanical, metabolic, and inflammatory cues, with TGF-β–SMAD2/3 signaling serving as a central driver [63].TGF-β signaling promotes fibroblast activation through canonical SMAD2/3-dependent pathways as well as noncanonical SMAD-independent pathways [64]. This signaling is further controlled by endogenous negative regulators such as SMAD7, which antagonizes TGF-β activity and helps maintain signaling balance [65].Other pathways, including MAPK, PI3K–AKT, RhoA–ROCK, YAP/TAZ, and TLR4–NF-κBalso contribute to fibroblast activation and matrix remodeling [66–70]. Mechanosensing is another essential regulator of FMT. Integrins and the Piezo1 ion channel can transfer mechanical cues that converge and integrate with TGF-β signaling events in a multinode network that amplifies FMT [71, 72].These signals are further integrated at the epigenetic and metabolic levels [73, 74], stabilizing gene expression and sustaining the myofibroblastic state. Accordingly, FMT is best examined within the broader cellular microenvironment rather than as an isolated cell state, allowing a more systematic understanding of intercellular interactions and context-dependent effects [44].

Cellular players and interaction networks in intestinal FMT

FMT in the intestine is regulated by a complex multicellular network in which epithelial, immune, vascular, and microbial cues integrate inflammatory, metabolic, and mechanical signals within the tissue microenvironment. The major cellular interactions that coordinate these processes are discussed in the following sections and summarized in Table 2 as key regulatory axes. Cellular senescence is discussed separately as a cross-cutting cell-state program that can modify several of these interactions.

Table 2.

Major intercellular axes regulating FMT in the intestine

Cellular interaction axis	Major mediators	Functional outcome in FMT	Representative context	References
Epithelial–fibroblast axis	TGF-β, WNT ligands, PGE2, extracellular vesicles (mtDNA)	Drive epithelial plasticity and stromal activation, supporting CAF formation	chronic inflammation and Tumorigenesis	[75–78]
Th17 / inflammatory T cell–fibroblast axis	TGF-β, IL-17 A, IL-21, amphiregulin	Amplify inflammatory fibroblast activation and ECM production	IBD-associated fibrosis	[79–82]
Tregs/ anti-inflammatory macrophages–fibroblast axis	IL-10, TGF-β (context dependent)	Limit excessive fibroblast activation and facilitate resolution of tissue remodeling	Tissue repair phase	[83, 84]
ILCs–fibroblast axis	IL-33, IL-13, IL-22, oxysterols	Coordinate barrier repair and stromal remodeling; context-dependent promotion of FMT	Mucosal immunity, colitis	[85–87]
Macrophage–fibroblast axis	TGF-β, IL-1β, PDGF, oncostatin M, integrinαvβ3, EVs	Promote fibroblast activation and myofibroblast differentiation	Chronic inflammation, intestinal fibrosis	[88–91]
Other innate immune amplification axis (neutrophils / MDSCs)	NET-associated proteases, ROS, chemokines	Enhance inflammatory signaling and indirectly promote FMT and matrix remodeling	Severe inflammation and tumor microenvironment	[92–95]
Pericyte–fibroblast axis	PDGF-B, TGF-β	Pericyte activation and detachment may contribute to stromal remodeling	Fibrosis and tumor stroma	[96, 97]
Endothelial–fibroblast axis	TGF-β, EndoMT-related signals	Generate mesenchymal-like cells and contribute to stromal expansion	Fibrosis and tumor angiogenesis	[24, 98]
Platelet–fibroblast axis	PDGF, TGF-β, TXA₂	Promote fibroblast activation and ECM deposition during tissue injury	Wound repair and fibrosis	[99–101]
Microbiota–stromal axis	Bacterial components and secreted products, microbial metabolites, EVs	Directly or indirectly modulate fibroblast activation, FMT, and ECM production through stromal, epithelial, and immune signaling	Intestinal homeostasis and dysbiosis	[102–105]

Intestinal epithelial cells (EpCs)

Intestinal EpCs and fibroblasts are not only spatially proximate but also engage in directed signaling interactions that coordinate intestinal homeostasis [50]. At the crypt base, Lgr5⁺ ISCs sustain epithelial renewal, while regionally specialized lamina propria fibroblasts provide instructive cues [106, 107]. Functionally, these fibroblasts act as paracrine modulators and can be distinguished as crypt-bottom fibroblasts (CBFs) and crypt-top fibroblasts (CTFs), each of which shapes intestinal stem cell behavior through distinct molecular cues. CBFs reside near ISCs, express relatively low levels of PDGFRα, and secrete canonical Wnt ligands along with inhibitors of bone morphogenetic protein (BMP) signaling to sustain stem cell proliferation and maintain an undifferentiated state [76]. In contrast, CTFs, which are enriched in the villus domain and characterized by high PDGFRα expression, secrete noncanonicalWingless/Integrated (Wnt) ligands and BMP factors that promote differentiation and directional epithelial renewal [37, 39]. Although this regional niche model aligns epithelial renewal with fibroblast-derived gradients, conditional Porcn deletion studies indicate that Wnt signaling from epithelial and subepithelial cells is not strictly indispensable, suggesting functional redundancy within a broader epithelial–stromal network [108].

Notably, intestinal EpCs are not passive recipients of stromal cues but active regulators that shape the surrounding microenvironment. For instance, through the platelet-derived growth factor (PDGF)–PDGFRα axis, epithelial-derived PDGF can influence PDGFRα⁺ fibroblasts and is associated with shifts toward an α-SMA⁺ phenotype, potentially supporting a responsive stromal pool during injury [75, 109]. Epidermal growth factor receptor (EGFR) ligands promote the contractile and secretory phenotype of subepithelial myofibroblasts [110], indian hedgehog (Ihh) maintains the myofibroblastic phenotype [111], and crypt-derived Wnt3a supports stem cell renewal while activating TGF-β/SMAD signaling to promote fibroblast conversion [112]. Epithelial turnover continuously prevents the development of pathological fibroblast-activating paracrine signals [113].

Epithelial plasticity further contributes to tumor-associated stromal remodeling and is associated with the emergence of heterogeneous CAF populations [114, 115].Among these, inflammatory CAFs (iCAFs) confer an immunosuppressive microenvironment [116], whereas myofibroblastic CAFs, characterized by high expression of α-SMA, promote invasion, angiogenesis, and metastasis [46, 117]. These subsets, which often coexist in tumors [118], are shaped by TGF-β and interleukin-1 (IL-1) [119]. Sustained epithelial plasticity may further support maintenance of the CAF pool and support a profibrotic niche [120].Changes in epithelial signaling programs can influence fibroblast behavior and stromal organization, linking epithelial stress responses to the emergence of myofibroblast-like populations and cancer-associated fibroblasts. Through reciprocal communication with stromal cells, epithelial signals therefore function as upstream regulators of FMT, coordinating tissue repair under normal circumstances but potentially driving pathological remodeling when regulatory balance is disrupted.

Immune cells

Beyond their role as ECM producers, intestinal fibroblasts actively participate in immune regulation. They secrete a variety of cytokines and growth factors to shape inflammatory responses [121, 122] and interact closely with macrophages, dendritic cells (DCs), and lymphocytes through paracrine and contact-dependent mechanisms [123–125].Although fibroblasts can express MHC-II, their capacity as functional antigen-presenting cells remains debated, as insufficient costimulatory signaling may favor tolerogenic outcomes [126, 127]. Fibroblast–immune interactions are bidirectional and highly context-dependent. Immune-derived mediators such as TNF-like ligand 1 A (TL1A), TGF-β, IL-1, and PDGF can promote fibroblast activation, while fibroblast-derived chemokines and growth factors regulate immune cell recruitment, differentiation, and retention.

Within the innate compartment, lamina propria innate lymphoid cells (ILCs) and regionally specialized fibroblasts form immune–stromal units [128]. Fibroblast-derived oxysterols act through G protein–coupled receptor 183 (GPR183) on ILC3s, guiding their positioning to cryptopatches and isolated lymphoid follicles and supporting the maturation of colonic lymphoid structures [87]. In turn, ILC3s can promote fibroblast activation and matrix remodeling in certain contexts [129].In the adaptive compartment, FRCs provide structural and signaling support for lymphoid organization, whereas T and B cells influence fibroblast fate, including FMT regulation. T-cell–derived cytokines and lipid mediators can either restrain or promote fibroblast activation, depending on subset identity and microenvironmental context. Proinflammatory T-cell populations, particularly Th17 (T helper 17) subsets, may enhance fibroblast proliferation, α-SMA expression, and ECM production through cytokine-driven feedback loops [79], whereas regulatory T-cell–associated signals such as IL-10, IL-22, and prostaglandins can counteract profibrotic pathways and maintain stromal quiescence [130–132]. B cells are also involved in this reciprocal network. Stromal fibroblasts help sustain B-cell survival and promote IgA class switching, supporting mucosal immunity [133]. In turn, plasma cell–derived growth factors can affect fibroblast activation, indicating that humoral responses contribute to the regulation of stromal remodeling [134].Clinically, B-cell depletion with rituximab may lead to a decrease in myofibroblast differentiation and an improvement in FMT, highlighting the potential therapeutic relevance of B-cell–targeted interventions [135, 136].

The above ILC- and T/B-mediated signals can induce transient remodeling of the epithelial barrier and stroma [87, 137]. Whether repair resolves or progresses to fibrosis or malignancy depends largely on myeloid signaling in the lamina propria, as sustained myeloid activation can prolong fibroblast activation and tissue remodeling. Current evidence indicates that macrophages and fibroblasts engage in close and reciprocal interactions in fibrotic and tumor-associated remodeling [138].Activated fibroblasts help recruit and retain monocytes by releasing chemokines such as C-C motif chemokine ligand 2 (CCL2), as well as growth factors that support their differentiation into macrophages [124, 139]. They also produce macrophage migration inhibitory factor, which limits macrophage egress and promotes their local accumulation [140]. Macrophage-derived mediators such as TGF-β, PDGF, and oncostatin M promote fibroblast proliferation and myofibroblast transition [90, 141, 142], contributing to stromal activation. Other myeloid cells, including DCs, neutrophils, mast cells, and myeloid-derived suppressor cells, also regulate fibroblast activation through cytokines, proteases, reactive oxygen species, and stress-related pathways.

Vascular-associated cells

Pericytes and endothelial cells (ECs) constitute a vascular-associated regulatory axis that intersects closely with fibroblast activation and FMT. Pericytes are specialized stromal cells closely associated with the microvascular basement membrane and share a mesodermal progenitor origin with ECs (e.g., Flk1⁺ progenitors) [143]. Single-cell and lineage-tracing studies have shown that during human intestinal development, pericytes emerge after endothelial network formation, following a maturation trajectory parallel to that of myofibroblasts. Early pericyte progenitors share partial transcriptional features with fibroblasts and primitive α-SMA⁺ cells and are linked to Wnt6 signaling. Lineage tracing has further revealed that pericytes arise not from fibroblasts but from mesodermal Wnt6⁺ precursors [34]. Despite fundamental differences in origin, localization, and molecular markers [144], both pericytes and fibroblasts synthesize ECM, an overlap that once led to their misclassification as “specialized fibroblasts” [145]. Notably, several studies have reported that pericytes can transdifferentiate into ECs, fibroblasts, or macrophages. However, labeling and marker limitations likely lead to false-positive labeling, and such results need to be validated by spatial omics and advanced lineage tracing [146]. Although direct lineage evidence for pericyte-to-myofibroblast transition (PMT) in the intestine is lacking, disease models in the eye, kidney, and lung have demonstrated that pericytes can differentiate into α-SMA⁺ myofibroblasts via the TGF-β–SMAD2/3–AKT/mTOR signaling pathway, providing mechanistic insights relevant to intestinal fibrosis [96, 147].

In addition to pericytes, ECs critically shape fibroblast activation and FMT. ECs, which codevelop with pericytes, dynamically regulate FMT through three major mechanisms: (1) generating myofibroblast-like cells via EndoMT [24]; (2) shaping fibroblast fate through vascular factor secretion [148]; and (3) modulating local TGF-β signaling via mechanobiological and metabolic reprogramming [98]. The most prominent of these processes is EndoMT, in which microvascular ECs exposed to IL-1β, TNF-α, and TGF-β1 undergo phenotypic switching—losing CD31 and VE-cadherin expression while acquiring α-SMA and collagen—thereby contributing to tissue repair and fibrosis [24, 149]. TGF-β1 is regarded as the master regulator in this process, whereas the role of IL-1β is less consistent and is often considered auxiliary, acting through NF-κB–SMAD crosstalk [24]. Endothelial metabolism and stress also influence FMT, as released mtDNA activates the cGAS–cGAMP axis to inhibit Yes-associated protein 1 (YAP1) and impair endothelial regeneration, whereas the cGAS–STING pathway directly promotes EndoMT [98, 150]. In addition, senescent ECs release galectin-3, which enhances SMAD3 signaling through transforming growth factor-β receptor type 1 (TGFBR1) to drive FMT [151]. Thus, even in the absence of EndoMT, ECs modulate FMT through the release of paracrine cues, suggesting that inhibiting EndoMT and restoring endothelial regeneration may represent promising antifibrotic strategies. Conversely, fibroblast activation also alters vascular stability, permeability, and immune cell trafficking, establishing reciprocal communication between the vascular compartment and the stromal matrix [152, 153].

Platelets also participate in stromal regulation beyond their classical hemostatic role, particularly during acute intestinal injury when vascular disruption and repair responses are rapidly initiated. In experimental colitis, platelet-derived thromboxane A2promotes intestinal myofibroblast proliferation and migration, whereas platelet-specific COX-1 deletion reduces myofibroblast accumulation and collagen deposition [154]. Platelets can also influence fibroblasts indirectly through immune cells. Platelet-derived C-X-C motif chemokine ligand 4 (CXCL4) promotes the emergence of profibrotic SPP1⁺ macrophages, which subsequently enhance fibroblast activation and ECM production [155]. Other α-granule factors, such as thrombospondin-1, have shown profibrotic activity in experimental models outside the intestine, although their direct contribution to intestinal FMT remains to be established [156].

Microbiota

The intestinal commensal microbiota contributes to host homeostasis by supporting nutrient metabolism, barrier integrity, and IgA production [157, 158]. Recently, microbial components and metabolites also interact with epithelial, immune, and stromal cells, thereby shaping the intestinal microenvironment [159]. Notably, the causal relationship between microbial dysbiosis and FMT remains unclear. Clinical investigations suggest that dysbiosis may precede fibrotic remodeling and is frequently associated with FMT [160]. Thus, the microbiota should be incorporated into fibroblast interaction networks with an emphasis on microbial cues that distinguish physiological FMT from pathological FMT.

Specific bacteria may contribute to stromal activation. Adherent-invasive Escherichia coli and chronic Salmonella Typhimurium infection promote intestinal collagen deposition and fibrotic remodeling, and may induce a myofibroblast-like phenotype in fibroblasts through both direct and indirect mechanisms [161–163]. Beyond intact bacteria, bacterial peptidoglycan–polysaccharide complexes can also stimulate collagen and profibrotic cytokine production in intestinal myofibroblasts, while microbiota-derived dipeptidyl peptidase-4 activates human intestinal myofibroblasts through PI3K–AKT signaling and aggravates experimental fibrosis [103, 164]. Proteolytic microbiota and bacterial proteases can activate protease-activated receptor 2 (PAR2) [165], while PAR2 signaling may promote collagen production and profibrotic activation in human colonic fibroblasts [166]. Within the intestinal context, Faecalibacterium prausnitzii-derived extracellular vesicles (EVs) attenuate intestinal fibrosis through macrophage metabolic reprogramming [104].

Microbial metabolites add another layer of regulation to intestinal FMT, with effects that are highly context dependent. Butyrate can suppress fibroblast activation through histone deacetylase inhibition, yet under certain conditions may instead promote FMT and matrix remodeling via G protein-coupled receptor 43 signaling [46, 105]. Indole-3-acetic acid inhibits fibroblast differentiation and ECM deposition through PI3K/AKT/mTOR signaling [167], whereas trimethylamine N-oxide promotes FMT through protein kinase R-like endoplasmic reticulum kinase (PERK) activation [168]. In patients with CD, microbiota-derived L-aspartate further enhances TGF-β1-induced FMT and ACTA2 expression [169]. Other microbial metabolites may also influence stromal responses, although the strength of evidence varies. The secondary bile acid deoxycholic acid (DCA) can directly activate Ca²⁺/PKC–COX-2/PGE₂ signaling in colonic fibroblasts, but its role in intestinal FMT remains uncertain [170]. In contrast, succinate–SUCNR1 signaling has been more directly linked to fibroblast activation and intestinal fibrosis [171]. Polyamines appear to act mainly through indirect mechanisms, as spermidine and putrescine can modulate intestinal immune and barrier responses, while their direct effects on FMT remain unclear [172]. Evidence from Helicobacter pylori-associated gastric cancer likewise suggests that microbial cues can induce CAF-like stromal activation, although its relevance to intestinal FMT is indirect [173, 174].

Cellular senescence

Cellular senescence adds another layer to the regulation of FMT, and its effects appear to depend largely on how long the senescent state persists and in which cellular context it occurs. Transient senescence can contribute to normal repair. In wound-healing models, senescent fibroblasts and endothelial cells promote myofibroblast differentiation through platelet-derived growth factor-AA secretion [175]. By contrast, persistent senescence may sustain a profibrotic microenvironment through prolonged senescence-associated secretory phenotype (SASP) signaling. In fibrostricturing IBD, senescent intestinal EpCs promote fibroblast activation and increase α-SMA, collagen I, and fibronectin expression [113], while fibrotic ileal tissues from patients with complicated CD exhibit SASP-associated fibroblast programs together with enhanced resistance to apoptosis [176]. Senescent immune cells may also contribute to this profibrotic milieu through persistent inflammatory and SASP-related signals; however, direct evidence that they drive intestinal FMT is still lacking.

Senescence also extends into the tumor microenvironment. Senescent intestinal fibroblasts can promote epithelial proliferation and invasion through growth differentiation factor 15 [177], while macrophage-derived IL-1β induces CAF senescence in colorectal cancer and generates an IL-6/CXCL12-rich SASP associated with therapy resistance [178]. Persistent senescence may therefore help maintain a microenvironment in which fibroblast activation and matrix remodeling fail to resolve. As fibroblasts continue to remodel the ECM, the local mechanical environment also changes.

Taken together, the epithelial-, immune-, vascular-, microbial-, and senescence-associated interactions discussed above highlight the highly interconnected nature of FMT regulation. Among these cellular interaction pairs, mechanical cues deserve particular attention because they are closely intertwined with the soluble biochemical signals described above. Mechanotransduction is not an isolated pathway; rather, it can emerge from multicellular interactions and feed back to strengthen them [91]. Changes in ECM stiffness, integrin-mediated adhesion, Piezo1 activity, and YAP/TAZ signaling can influence the establishment and persistence of these interactions, while activated fibroblasts continue to remodel the matrix and reshape the local mechanical environment. Under sustained injury, this reciprocal coupling may help maintain FMT and drive fibroblast activation from a transient repair response toward a more persistent pathological state [179].

Spatiotemporal dynamics of FMT-related cellular interactions in homeostasis and disease

To better understand the multicellular regulation of FMT, the following section outlines how these interactions are organized across space and time in homeostasis, repair, fibrosis, and tumorigenesis (Fig. 2). A clearer view of these dynamics may help identify stages at which disease progression can be limited or tissue remodeling redirected. Current evidence suggests that FMT is not simply a consequence of disease progression, but an active part of the processes that shape tissue homeostasis and remodeling. Future studies should clarify the timing and localization of these intercellular interactions, particularly during transitions between homeostasis, inflammation, fibrosis, and tumorigenesis.

Fig. 2.

Spatiotemporal dynamics of FMT during intestinal homeostasis, injury, fibrosis, and tumorigenesis. (A) During homeostasis, epithelial, microbial, immune, and stromal signals maintain fibroblast quiescence and balanced ECM remodeling. (B) Acute inflammation and tissue injury activate fibroblasts through cytokines, microbial products, immune-cell signals, and vascular-associated cues, resulting in transient FMT. With successful repair, activated myofibroblasts undergo apoptosis or return to quiescence, thereby restoring stromal homeostasis. (C) When injury persists, impaired myofibroblast resolution leads to sustained activation and excessive ECM deposition. Increased matrix stiffness further reinforces FMT through mechanotransduction, establishing a self-sustaining profibrotic loop. EMT, EndoMT, and MMT may also contribute to myofibroblast accumulation, although their direct lineage contribution remains uncertain. (D) During tumorigenesis, persistent epithelial, immune, vascular, and mechanical signals promote CAF accumulation and stromal remodeling, creating a stiff, proangiogenic, and tumor-supportive microenvironment

Intestinalhomeostasis

In the steady state, myofibroblasts are sparse and largely restricted to specific subepithelial niches [180]. By producing ECM and secreting signals such as Wnt, they form a regulatory network that sustains epithelial proliferation, differentiation, and barrier function, thereby preserving mucosal integrity [181]. Interstitial cells of Cajal, a specialized myofibroblast-like population between the submucosa and muscularis, serve as pacemaker cells that control gastrointestinal motility, electrical conduction, and neurotransmitter responses [75, 182].Overall, myofibroblast activation remains limited and spatially confined. With the epithelial barrier intact, most microorganisms remain within the intestinal lumen, and their effects on host tissues are thought to be mediated mainly by microbial metabolites, particularly short-chain fatty acids (SCFAs) [183, 184]. At the same time, EpCsproduce prostaglandin E₂ (PGE₂) via cyclooxygenase-1 [77]. Through PGE receptors on adjacent fibroblasts, epithelial-derived PGE₂ limits fibroblast migration and counteracts profibrotic activation [185, 186]. These regulatory signals help restrain α-SMA expression and contribute to the maintenance of stromal quiescence [187].

In addition, the lamina propria contains a range of immune cells, with macrophages, Tregs, and ILC3s contributing to mucosal immune tolerance and barrier integrity. Steady-state macrophages and Tregs produce immunoregulatory mediators such as IL-10 and TGF-β, which help dampen inflammatory signaling and limit excessive fibroblast activation [83, 84]. In parallel, ILC3s respond to microbiota-derived tryptophan metabolites (e.g., indole-3-aldehyde) and produce IL-22, which cooperates with stromal cells to support epithelial barrier function [188]. Beyond immune regulation, vascular-associated stromal cells also contribute to tissue stability. Pericytes help maintain microvascular integrity by regulating endothelial junctions, basement membrane composition, and local shear stress responses [189–191].

Acute inflammation and tissue injury

During acute intestinal injury, disruption of epithelial integrity initiates a coordinated repair program involving epithelial, stromal, vascular, microbial, and immune components. Rather than acting independently, these signals converge within a spatially restricted injury niche and collectively regulate FMT, enabling rapid but tightly controlled matrix remodeling.

Firstly, loss of epithelial barrier integrity allows microbial products to translocate into the lamina propria, triggering innate immune responses. These microbial signals stimulate immune cells to release profibrotic cytokines such as TGF-β1 and IL-1β, reinforcing a local inflammatory and profibrotic microenvironment [49, 70, 192]. At the same time, bacterial cell wall components can directly engage toll-like receptors (TLRs) on fibroblasts to induce FMT through innate immune signaling [164, 193]. Among these pathways, the LPS–TLR4–NF-κB axis plays a prominent role by enhancing fibroblast contractility, relieving SMAD7-mediated inhibition of TGF-β signaling, and promoting connective tissue growth factor (CTGF)expression, thereby reinforcing profibrotic signaling [102]. Other TLR family members also participate in shaping the fibrotic phenotype. Consistent with this, fibroblast-specific MyD88 deletion in a dextran sulfate sodium (DSS) colitis model markedly reduces collagen and fibronectin deposition and attenuates fibrosis [194]. However, because TLR signaling networks are highly redundant and shared across multiple cell types, blocking a single pathway is often insufficient to durably suppress FMT.

Microbial signals also reshape the immune microenvironment, which in turn reinforces stromal remodeling. For example, excessive oxysterol accumulation during colitis promotes aggregation of ILC3s at inflammatory sites, exacerbating disease progression [87]. In parallel, CD90^hiCD74⁺ fibroblasts can secrete IL-33 to activate ILC2s, leading to IL-13 release that drives macrophages toward a profibrotic phenotype and further stimulates FMT [85, 86]. Through these reciprocal interactions, ILCs coordinate short-term and reversible remodeling of both the epithelial barrier and the stromal compartment.PD-1⁺ Th17 cells promote FMT by releasing TGF-β that activates SMAD3 signaling in fibroblasts, whereas fibroblast-derived IL-6 enhances PD-1⁺ Th17 differentiation, forming a positive feedback circuit that exacerbates fibrosis [79]. In addition, Th17-derived IL-17 stimulates fibroblasts to upregulate CXCL1 and CXCL6, recruiting neutrophils and monocytes that further amplify inflammation and indirectly promote FMT [80].Concurrently, macrophages, neutrophils, and mast cells in the lamina propria release TGF-β, PDGF, and TNF-α, which strongly drive local fibroblasts toward myofibroblasts [95]. Beyond soluble factors, direct cell–cell contact and mechanical signaling also contribute to fibroblast activation. Under IL-4/IL-13–driven profibrotic conditions, macrophage integrin αvβ3 physically interacts with fibroblasts and activates mechanosensitive Piezo1 channels via Ca²⁺ influx, thereby increasing α-SMA expression and contractility [91]. Cadherin-11–mediated adhesion similarly forms specialized macrophage–fibroblast junctions that locally activate TGF-β1 signaling and reinforce FMT [195].In addition, under conditions of endoplasmic reticulum stress, PERK⁺ plasma cell–like DCs can directly engage fibroblasts and induce the expression of α-SMA and fibronectin [196].

Tissue injury is frequently accompanied by vascular damage, further expanding the signaling network that regulates stromal activation. Endothelial disruption and platelet activation initiate a hemostasis–repair cascade in which fibrin scaffolds form at the wound site [197].Coagulation proteases such as thrombin and factor Xa activate protease-activated receptor 1 (PAR1) signaling in fibroblasts, stimulating autocrine PDGF and CTGF production and promoting fibroblast proliferation, migration, and collagen synthesis [101, 198]. Platelet-derived mediators, including PDGF, TGF-β, and Adenosine diphosphate, further enhance FMT [99, 100, 199]. Concurrently, endothelial barrier disruption and inflammatory cues recruit monocytes and neutrophils, among which Ly6C⁺ macrophages secrete profibrotic mediators, while neutrophils release neutrophil extracellular traps (NETs) that directly induce FMT, together amplifying matrix remodeling [35, 92, 200].Activated myofibroblasts subsequently remodel the provisional fibrin clot through contractile stress fibers and ECM deposition, gradually replacing it with a stable collagen matrix.

Despite these multiple activating signals, epithelial–stromal communication simultaneously imposes regulatory constraints that limit excessive fibrosis. During epithelial regeneration, PGE₂–EP4 signaling promotes Lgr5⁺ intestinal stem cell expansion and crypt repair, whereas reduced stromal PGE₂ temporarily releases its inhibitory restraint on fibroblasts, allowing transient myofibroblast accumulation to facilitate wound closure [201, 202]. Under normal circumstances, once epithelial repair is completed and inflammatory signals subside, myofibroblasts undergo apoptosis or revert to a quiescent fibroblast state, preventing excessive matrix deposition [203].

Taken together, FMT during acute intestinal injury represents a rapid, spatially confined, and self-limiting process coordinated by a dynamic network of epithelial, microbial, immune, vascular, and stromal signals. These interconnected pathways ensure that fibroblast activation is precisely timed and locally restricted, enabling efficient tissue repair while maintaining mucosal homeostasis.

Chronic inflammation and fibrotic remodeling

When tissue injury persists, the regulatory balance that normally restrains stromal activation is progressively disrupted. Sustained activation of TGF-β, IL-1α, and other inflammatory pathways maintains fibroblast activation and shifts tissue repair toward fibrotic remodeling [204, 205].Epithelial signaling changes represent an early component of this transition. In chronic disease, aberrant activation of the Wnt/β-catenin pathway may induce EMT-like changes in EpCs that contribute to ECM remodeling and luminal narrowing [206]. Studies in renal and pulmonary fibrosis have shown that EpCs undergoing partial EMT can express profibrotic mediators such as TGF-β, thereby influencing stromal activation and disease progression [207, 208]. Although EMT-derived cells are not generally considered a dominant source of stable myofibroblasts, these findings suggest that epithelial plasticity can contribute to a microenvironment that favors persistent FMT. In the intestine, sustained epithelial reprogramming has been linked to chronic inflammation and may provide a bridge between fibrotic remodeling and tumorigenesis.

Immune cells further amplify this profibrotic network. Th17-associated cytokines, including IL-17, IL-21, and amphiregulin, promote fibroblast proliferation, α-SMA expression, and myofibroblast differentiation, whereas blockade of Th17 differentiation or IL-17 signaling alleviates experimental intestinal fibrosis [81, 82, 209]. However, regulatory Th17 subsets can produce IL-10 and partially counteract myofibroblast activation, underscoring the context-dependent effects of T-cell regulation [210]. Myeloid cells also contribute to sustained stromal remodeling. Macrophage–fibroblast crosstalk forms a feedback circuit that supports fibroblast activation. Beyond this paracrine interaction, several studies have reported macrophages with α-SMA⁺, ECM-producing features under fibrotic conditions, raising the possibility of a MMT-like process [89, 211–213]. In intestinal fibrosis, CD68⁺/α-SMA⁺ or F4/80⁺/α-SMA⁺ macrophage-like cells have been detected in fibrostenotic CD and experimental colitis. Glial cell line-derived neurotrophic factor has also been reported to facilitate this phenotypic shift through Src activation [26]. This process is further amplified by glycolytic reprogramming, which leads to hypoxia-inducible Factor 1α stabilization and peroxisome proliferator-activated receptor gamma coactivator 1α downregulation, resulting in mitochondrial DNA leakage [214, 215]. These findings suggest a possible role for MMT-like process in chronic intestinal stromal remodeling, but their lineage origin and in vivo relevance remain unclear. The resulting phenotype appears to be plastic and context dependent rather than a stable terminal lineage. In addition to direct phenotypic transitions, macrophages can regulate fibroblast activity through extracellular vesicles enriched with regulatory microRNAs [216, 217]. Other innate immune populations further reinforce this process. Neutrophils, mast cells, and polymorphonuclear myeloid-derived suppressor cells (PMN-MDSCs) can promote fibroblast activation and matrix remodeling through neutrophil extracellular traps (NETs), proteases, reactive oxygen species (ROS), and chemokine signaling [92–94, 218, 219].

As chronic inflammation progresses, stromal regulation becomes increasingly uncoupled from the normal injury-resolution program. A subset of myofibroblasts escapes apoptosis and remains persistently activated, providing a cellular foundation for sustained fibrosis and inflammation-associated tumorigenesis. In this setting, Th2- and Th17-associated cytokines such as IL-13 and IL-17 promote fibroblast survival and expansion across intestinal layers [35, 220]. Meanwhile, inflammatory mediators including TGF-β and TNF-α can induce epithelial and ECs to undergo EMT and EndoMT, may providing additional sources of myofibroblast-like cells [221]. Ultimately, these processes lead to excessive contractility, collagen cross-linking, and ECM deposition, ultimately resulting in tissue stiffening, extension of fibrotic lesions, and remodeling of intestinal architecture [222].

Thus, chronic intestinal inflammation establishes a multicellular regulatory network in which epithelial plasticity, adaptive and innate immune responses, and stromal feedback loops collectively sustain FMT. Unlike the transient activation observed during acute repair, FMT in this setting becomes temporally persistent and spatially expanded, gradually spreading beyond the initial injury site and driving progressive fibrotic remodeling.

Tumorigenesis in the intestine

In tumorigenesis, FMT becomes embedded within a complex multicellular regulatory network involving epithelial, immune, and vascular compartments. Unlike the transient and spatially confined activation observed during tissue repair, stromal remodeling in tumors is persistent and progressively reorganized within the tumor microenvironment. Persistent epithelial reprogramming may help sustain a microenvironment that favors CAF activation and stromal remodeling. In this setting, tumor cells can further influence neighboring EpCs through extracellular vesicles containing mitochondrial DNA (mtDNA), which increase oxidative stress and activate TGF-β signaling, thereby reinforcing epithelial plasticity and stromal activation [78]. At the same time, chronic inflammatory stress can induce epithelial mutagenesis, increasing the likelihood of malignant transformation.

Within the tumor microenvironment, fibroblasts display marked functional heterogeneity. Some CAF subsets primarily regulate immune responses. IL-1β and TGF-β signaling can drive mesothelial cells toward MHC-II⁺ antigen-presenting CAFs that promote regulatory CD4⁺ T-cell responses and immune evasion, whereas CCL19⁺ fibroblastic reticular cells can organize tertiary lymphoid structures and support antitumor immunity [223, 224].These immune-modulating populations coexist with ECM-producing, contractile, myofibroblast-like CAFs. Single-cell and spatial transcriptomic studies in colorectal cancer further support the coexistence of inflammatory, immune-regulatory, and matrix-remodeling CAF programs [225]. Together, these findings indicate that fibroblast functions in tumors are shaped by cellular identity, spatial organization, and local microenvironmental cues rather than by a single defining marker.

Tumor progression is also accompanied by extensive vascular remodeling. Pericytes contribute to shaping the vascular microenvironment by regulating vessel maturation and permeability and by modulating immune cell infiltration through paracrine and contact-dependent signaling [226]. Sustained tumor growth requires continuous neovascularization to support metabolic demand, making angiogenesis a key component of tumor expansion, metastasis, and premetastatic niche formation [227]. Clinically, antiangiogenic strategies remain debated, particularly in colorectal cancer, where increasing attention has shifted toward vascular normalization to improve perfusion and immune cell infiltration rather than simply blocking vessel formation [228]. Emerging evidence further suggests that preventing pathological EndoMT may help limit matrix stiffening and hypoxia while improving drug delivery.

Within this environment, tumor cells and stromal fibroblasts establish reciprocal signaling loops that reinforce stromal activation. TGF-β and TNF-α signaling can reprogram surrounding fibroblasts toward a CAF phenotype [229], while CAFs reciprocally influence tumor behavior through pathways such as Wnt, hepatocyte growth factor (HGF), and CXCL12 signaling [230, 231]. By remodeling the ECM and secreting proinflammatory and proangiogenic factors, CAFs establishan immunosuppressive niche that promotes tumor growth and metastasis [232]. Moreover, increased ECM deposition enhances tissue stiffness and activates mechanotransduction pathways such as YAP/TAZ signaling in both fibroblasts and tumor cells, thereby reinforcing contractility, matrix production, and tumor cell invasiveness [233, 234].

Collectively, FMT in tumors reflects sustained, spatially structured remodeling driven by coordinated epithelial, immune, vascular, and stromal signals, which collectively shift the tumor microenvironment toward an immunosuppressive and pro-tumorigenic state.

Dynamic transitions of FMT

Notably, although the specific roles of FMT during the physiological and pathological stages of the intestine have been increasingly revealed, intestinal diseases involve complex and dynamic processes rather than representing a strict continuum from homeostasis to inflammation, fibrosis, and tumor development. Emerging reports have revealed that disease processes are nonlinear and discontinuous and may include abrupt transitions between pathological states. For example, in some animal models, fibrotic phenotypes can emerge before acute colitis is completely resolved [235]. Under chronic inflammation, EpCs can also develop a precancerous phenotype together with a fibrotic phenotype [236]. Therefore, intestinal pathological states frequently overlap, and FMT is a spatiotemporally dynamic process that can shift from a transient, reversible response to persistent pathological activation through coordinated epithelial, immune, endothelial, and fibroblast interactions during different disease stages (Table 3). Targeting key signaling pathways or disrupting pro-transitional intercellular communication may prevent pathological microenvironments from coopting FMT to provide potential precise therapies for IBD, intestinal fibrosis, and cancers of the intestine.

Table 3.

Spatiotemporal features of FMT across physiological and pathological conditions

Stage	Spatiotemporal and Histological Features	Major Cellular Interaction Pairs	DominantPhenotype/Outcome	References
Intestinal homeostasis	Minimal/basal, reversible FMT; crypt–villus gradient; microbiota-mediated equilibrium	Epithelium–fibroblast, ILC3–fibroblast, T/B-cell–fibroblast, microbiota–fibroblast, pericyte–endothelial cell	Barrier maintenance and mild ECM remodeling	[39, 111, 185, 237, 238]
Acute inflammation and tissue injury	Transient, wound-edge activation	Epithelium–fibroblast, macrophage/neutrophil–fibroblast, platelet–fibroblast, ILCs–fibroblast	Wound closure, provisional ECM formation	[80, 87, 90, 99, 203, 239]
Chronic inflammation and fibrotic remodeling	Persistent activation; stiff ECM; EMT/EndoMT may involvement	Epithelium–fibroblast, Th17–fibroblast, B-cell–fibroblast, macrophage–fibroblast, endothelial–fibroblast	Collagen accumulation and tissue scarring	[24, 41, 79, 135, 176, 206]
Tumorigenesis in the intestine	CAF heterogeneity; high-tension, immunosuppressive stroma	Epithelium–CAF, iCAF–myoCAF, endothelial–CAF, tumor cell–fibroblast, immune cell–CAF	Immune-exclusionary stroma and metastasis	[46, 114, 225, 229, 240]

Emerging advances and cell interaction–based strategies for targeting FMT

Given the central role of TGF-β signaling in fibrotic remodeling, this pathway has been considered a potential target for modulating FMT. However, the pleiotropic functions of TGF-β in immune regulation and tissue repair limit the feasibility of systemic blockade [241], and inhibition of a single signaling node often shows limited efficacy in complex inflammatory–fibrotic environments. For example, the SMAD7 antisense oligonucleotide strategy failed to demonstrate the expected therapeutic benefit in later clinical studies. Mechanical signaling represents another important driver of FMT [242]. The RhoA–ROCK pathway functions as a key mechanotransduction hub linking cytoskeletal contraction, inflammation, and ECM remodeling. Yet, because Rho-associated coiled-coil containing protein kinase (ROCK) signaling is broadly involved in fundamental cellular processes across multiple tissues, systemic inhibition lacks specificity and may cause off-target effects, which has limited its clinical application [243]. Together, these examples highlight the limitations of single-target therapies in fibrotic diseases and tumor-associated stromal remodeling. Network redundancy, compensatory signaling, and the spatiotemporal dynamics of tissue repair can all reduce therapeutic efficacy or interfere with physiological regeneration [112]. These challenges have prompted growing interest in therapeutic strategies that consider the multicellular networks regulating FMT, including cell transplantation, engineered exosomes, smart drug-delivery systems, and standardized microbiota transplantation, which aim to modulate the fibroblast-centered microenvironment at different stages of FMT, with the goal of promoting tissue repair, limiting fibrosis and chronic inflammation, and ultimately preventing inflammation-associated tumorigenesis (Fig. 3).

Fig. 3.

Emerging therapeutic strategies targeting pathological FMT. Cell transplantation, EV-based therapy, smart biomimetic delivery platforms, and fecal microbiota therapy converge to suppress pathological FMT, restoring fibroblast quiescence and ECM homeostasis

Cell transplantation therapy

Cell transplantation therapy provides a novel strategy for targeting FMT through two main mechanisms. Engineered immune cells (e.g., FAP-CAR-T [244] and FAP-CAR-macrophages [245]) selectively eliminate pathogenic cells, interrupting the progression from inflammation to fibrosis and carcinogenesis. Although primarily applied in cardiac fibrosis models, these strategies may also be applicable to intestinal fibrosis, as FAP⁺ fibroblasts constitute the major ECM-producing population in the gut [45]. In parallel, stem cells, particularly MSCs, modulate immune responses and promote tissue repair, restoring the balance of immune–stromal interactions. Their applications have expanded from hematologic malignancies [246] to solid tumors such as colorectal cancer [247]. Interestingly, MSCs exhibit considerable therapeutic potential for IBD [248]. In summary, these reports support cell transplantation targeting intercellular interactions as a promising approach to modulate pathological FMT.

Despite this promise, several challenges limit clinical translation. Engineered immune cell therapies may cause off-target tissue injury or excessive immune activation in the complex intestinal microenvironment [249, 250]. MSCs have great anti-inflammatory and regenerative potential; however, their long-term homing properties and phenotypic stability are still unclear [251]. Even under certain pathological conditions, they may be reprogrammed by tumors to expand CAFs and deposit ECM [252]. In addition, the therapeutic effect of MSCs is transient, and stable and controllable delivery systems are needed [253]. Moreover, most experimental evidence derives from animal models, which cannot fully recapitulate the complexity of human immune–stromal interactions and intestinal ecosystems and therefore limit their clinical application [254]. To solve these problems, advanced platforms that can better mimic human intestinal architecture and cellular networks, such as organoids and gut-on-a-chip models, provide useful tools for safety assessment and translational development of cell-based therapies [255, 256].

EV-based therapy

Emerging as indispensable mediators of intercellular communication beyond chemical and mechanical cues, EVs are generally involved in immune regulation, tissue repair, and tumor progression [257]. In intestinal fibrosis, EVs are increasingly considered key regulators of FMT [258]. miR-103a-3p is enriched in EVs derived from the AD-MSCs of patients with CD and targets TGFBR3 to increase the phosphorylation of SMAD2/3 and further promote fibroblast activation [259]. In contrast, conditioned medium from UC/PL-MSCs dramatically suppresses fibroblast activation and ameliorates intestinal fibrosis, and these effects may be largely mediated by MSC-derived EVs [260]. Consistently, bone marrow–derived MSC (BMSC) exosomes inhibit the CCN2–TGF-β signaling pathway and thereby block FMT to prevent fibrotic remodeling [261]. Moreover, engineered EVs carrying miR-200b target the EMT transcription factors ZEB1 and ZEB2, thereby suppressing FMT and fibrotic progression [206]. Overall, these findings identify EVs as key regulators of FMT and identify engineered EVs as promising tools for precision therapy.

During metastasis, EVs secreted by CAFs or tumor cells activate fibroblasts in distant tissues, inducing FMT and a premetastatic niche to promote cancer dissemination [262]. To inhibit this process, miR-138-5p and the antifibrotic drug pirfenidone were encapsulated in BMSC-derived EVs, which target TGFBR1 to suppress TGF-β signaling and thereby block FMT and its protumorigenic actions [263]. However, the stability and targeting capacity of EVs within the complex human intestinal microenvironment remain unknown [264]. The lack of standardized isolation and purification protocols combined with the heterogeneity of EVs leads to variable efficacy and side effects [265]. Engineering approaches can finely control the cargo of EVs to increase their stability and specificity [266]; however, immunogenicity and high production costs remain major obstacles to clinical application.

Smart biomimetic delivery platform

Based on these mechanistic insights, several therapeutic strategies have been explored, including anti–TGF-β antibodies [267], ROCK inhibitors [268], and PAR2 inhibitors [269]. Although these approaches suppress FMT, reduce fibrosis, and delay tumor progression in preclinical models, their clinical application has been limited by the complex intestinal microenvironment, short drug half-life, and lack of adequate local delivery. Therefore, addressing these issues has been the focus of recent studies. Recent advances in hydrogels, nanoparticles, and smart biomaterials have improved drug stability and intestinal targeting by fine-tuning the spatiotemporal release of drugs to better match the dynamic nature of FMT [270, 271].

Inflammation and repair are closely related under pathological conditions and constitute a self-perpetuating feedback loop that promotes disease progression. Coordinated regulation of these two processes has become important for improving therapy and has driven the development of intelligent drug delivery systems. For example, TNF-α–neutralizing antibodies and HGF can be coloaded into a thermosensitive hydrogel for local delivery of both an anti-inflammatory signal and a prorepair signal, thereby ameliorating the microenvironment and promoting rapid tissue regeneration [272]. More sophisticated designs focus on temporal release, in which a carboxymethyl chitosan–PEG nanocomposite hydrogel sequentially releases Nox4i and liposomal doxorubicin, either by suppressing FMT and enhancing immune infiltration first or increasing the efficacy of chemotherapy later, thereby achieving synergistic effects among matrix modulation, immunotherapy, and the potentiation of chemotherapy [273]. Another design focuses on dynamic signal regulation, in which an FXIIIa-crosslinked LA peptide hydrogel can control the release of anti-scar peptides first by camouflaging cell membranes that can adsorb TGF-β, thereby preventing pathological scar formation through stage-specific intervention during the process of wound healing [274].

Camouflaging cell membranes into biomimetic delivery systems is a breakthrough in nanomedicine that has been realized by cloaking nanocarriers with natural membranes derived from cancer cells, platelets, red blood cells, stem cells, or exosomes, conferring immune evasion and cell-specific targeting properties [275, 276]. Such biomimetic camouflages can improve the stability and bioavailability and allow for controlled accumulation through receptor–ligand interactions. Importantly, a gold nanoparticle system coated with myofibroblast membranes codelivered IL-4 and membrane-derived factors, which inhibited infection and inflammation through epigenetic modulation, angiogenesis, and epithelial regeneration [277]. In summary, this novel strategy may provide an effective approach to bypass the constraints of conventional nanomedicines on permeability and the biological environment for targeted accumulation.

Fecal microbial therapy

Fecal microbiota therapy has attracted increasing attention as an innovative therapeutic approach that involves transferring various microbial communities from healthy donors into recipient intestines to restore microbiome homeostasis. Unlike probiotic supplementation, microbiota transplantation reshapes the broader microbial community and can also alter the abundance and activity of microbiota-derived metabolites [278].

An increasing number of studies have suggested the involvement of gut microbiota dysbiosis in IBD. Compared with healthy individuals, patients with IBD display decreased richness and evenness of microbial communities, i.e., the loss of butyrate-producing bacteria such as Faecalibacterium [279]. To correct this imbalance in microecology, researchers have attempted to reshape the gut microbiota by transferring target microbes. In DSS-induced colitis models, the transfer of spore-forming Clostridium could restore microbial homeostasis by decreasing the abundance of pathogenic Escherichia coli and Shigella and increasing the abundance of Faecalibacterium to increase butyrate production, promote epithelial repair, attenuate inflammation, and suppress activated fibroblasts to accelerate mucosal healing [280]. Similarly, the transplantation of microbial consortia such as Faecalibacterium prausnitzii, Bacteroides faecis, and Roseburia intestinalis mitigates fibroblast overactivation by restoring the Treg/Th17 balance [281], whereas Lactobacillus acidophilus directly reduces α-SMA expression in fibroblasts [282]. These findings suggest that microbiota transplantation might have promising therapeutic effects on IBD and could modulate fibroblast activities in the gut microenvironment.

Although animal studies have demonstrated the therapeutic effects of microbiota transplantation, its clinical application remains limited because of low colonization efficiency, heterogeneous efficacy, and possible immune reactions. To solve these problems, recent studies have focused on more refined and controlled strategies, such as the use of microbiota-derived exosomes to deliver functional molecules ornanoparticle platforms to load target metabolites for stepwise and sustained release [283, 284]. These novel technologies not only overcome the limitations of conventional microbiota transplantation technology but also provide more flexible and precise ways to modulate the interaction network of microbiota–metabolite–fibroblasts.

Conclusion and prospects

Recent advances in intestinal biology have shifted our view of FMT from a fibroblast-intrinsic response toward a dynamic process shaped by continuous communication among epithelial, immune, vascular, stromal, and microbial compartments. This network-level organization may help explain why therapies directed at a single cell type or signaling pathway often show limited efficacy in the highly integrated intestinal microenvironment. FMT emerges from interconnected inputs including cell–cell contact, paracrine signaling, metabolic circuits, ECM mechanics, and EV-mediated communication. An important unresolved question is how this response shifts from a transient and adaptive process during tissue repair to persistent activation that sustains fibrosis or tumor-associated remodeling.

The influence of individual cellular populations on FMT is highly context dependent. The same cell type or signaling molecule may support repair in one setting but promote pathological remodeling in another, depending on disease stage, tissue location, and local microenvironmental cues. This functional plasticity is evident across immune, epithelial, vascular, and stromal compartments and argues against rigid classifications of “profibrotic” and “antifibrotic” cells or mediators. Therapeutic strategies will therefore need to consider when and where these interactions occur, rather than simply suppressing fibroblast activation or a single signaling pathway.

Important limitations remain. Single-cell and multi-omics approaches have greatly expanded our understanding of fibroblast heterogeneity, but they provide only a partial view of how cellular interactions evolve over time in intact tissue. In addition, many proposed cellular interactions are inferred from transcriptomic associations, experimental models, or evidence from extra-intestinal tissues, and their causal and lineage-resolved relevance in the human intestine remains to be established. Future studies combining organoid–stroma cocultures, spatial and temporal multi-omics, intravital imaging, and lineage tracing should help define when fibroblast activation remains adaptive and when it becomes difficult to resolve. Computational and AI-based approaches may further assist in reconstructing these multicellular interaction networks and identifying transition points that are difficult to capture experimentally.

Amore precise understanding of these spatiotemporal cellular dialogues may ultimately shift therapeutic thinking from broadly suppressing FMT toward modulating fibroblast plasticity according to disease stage and tissue context. Such an approach could preserve the reparative functions of transient FMT while limiting the persistent stromal remodeling that contributes to intestinal fibrosis and tumor progression.

Author contributions

HY wrote the manuscript; LZ, PX, YS and KY discussed and revised the manuscript; SW conceptualized the study and revised the manuscript. All authors read and approved the final manuscript.

Funding

This work was supported by the National Natural Science Foundation of China (Grant No. 82271850), Natural Science Foundation of Jiangsu (Grant No. BK20240850), Research Project of the Jiangsu Commission of Health (Grant No. K2023062), and Science and Technology Support Program (Social Development Project) of Zhenjiang (Grant No. SH2025002).

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.