Exploring unconventional targets in myofibroblast transdifferentiation outside classical TGF- β signaling in renal fibrosis

We propose that the key initiators of renal fibrosis are myofibroblasts which originate from four predominant sources—fibroblasts, pericytes, endothelial cells and macrophages. Increased accumulation of renal interstitial myofibroblasts correlates with an increase in collagen, fibrillar proteins, and fibrosis severity. The canonical TGF- β pathway, signaling via Smad proteins, is the central molecular hub that initiates these cellular transformations. However, directly targeting these classical pathway molecules has proven challenging due their integral roles in metabolic process, and/or non-sustainable effects involving compensatory cross-talk with TGF-β. This review explores recently discovered alternative molecular targets that drive transdifferentiation into myofibroblasts. Discovering targets outside of the classical TGF-β/Smad pathway is crucial for advancing antifibrotic therapies, and strategically targeting the development of myofibroblasts offers a promising approach to control excessive extracellular matrix deposition and impede fibrosis progression.


Introduction
Classically, the process of fibrosis in the kidney is broken down into five distinct stages.

Associated pathway Models
Fibroblasts -Demonstrated to contribute ~94% of myofibroblast formation (Asada et al., 2011) Kif26b Kinesin Family Member 26B review highlights myofibroblast transformation, with a focus on the cell of origin and the less-explored molecular pathways driving this change.
As the main producer of fibrotic ECM, myofibroblasts are distinguished from quiescent fibroblasts by elevated levels of alphasmooth muscle actin (α-SMA) (Vierhout et al., 2021).Throughout the fibrotic process in the kidney, myofibroblasts generate collagens (type I, III, IV, V, and VI) and produce fibrillar matrix components such as fibronectins, fibrillins, and deposit connective matrix elastins and tenascins, as well as smaller molecules like proteoglycans and matricellular proteins such as MMPs and TIMPs (Klingberg et al., 2013).Protein matrices aid in tissue degradation, blood flow inhibition, and structural stiffening.Myofibroblast release of cytokines and chemokines like IL-1α, IL-1 β, IL-6, TNF-α, and CCL2 also exacerbate interstitial damage (Baum and Duffy, 2011;Vierhout et al., 2021).It's noteworthy that the process of renal fibrosis and the action of the myofibroblast vary across different types of renal insults and diseases, encompassing allograft rejection, glomerular nephritis, proteinuria, and tubulointerstitial disease (Mack and Yanagita, 2015).
While subject to controversy, fibroblasts, pericytes, endothelial cells, and macrophages collectively contribute significantly to myofibroblast formation in the renal interstitium during fibrosis.The ongoing debate over myofibroblast origin is fueled by differing methodologies used for cell fate tracing (Humphreys et al., 2010), and is further complicated by the impact of renal pathology on myofibroblast origin.LeBleu et al. reported that the contribution of epithelial-to-mesenchymal transition (EMT) to myofibroblasts is less than 5%, indicating a considerably lower role for this transition compared to demonstrated transformations of fibroblasts, pericytes, endothelial cells, and macrophages (Oldfield et al., 2001;Iwano et al., 2002;LeBleu et al., 2013;Falke et al., 2015).In several studies, renal perivascular fibroblasts have been shown to contribute to the myofibroblast pool during fibrosis, as much as 94% as determined by P0-Cre lineage-labelling (Asada et al., 2011).Pericytes, constituting approximately 2.5%-5% of all kidney cells also destabilize capillaries, leading to peritubular capillary rarefaction (Schrimpf and Duffield, 2011).In rat angiotensin IIinduced kidney injury, pericyte transformation contributed to approximately 87% of myofibroblast formation (Faulkner et al., 2005).Conversely, in mice with unilateral ureteral obstruction (UUO), pericyte transformation accounted for either 6% or 0% of myofibroblast formation (LeBleu et al., 2013).Endothelial-tomesenchymal transition (EndoMT) determined by double labelling with pan-endothelial marker, platelet endothelial cell adhesion molecule 1 (PECAM-1), αSMA, and Tie2-YFP lineage-traced cells in rodent models of kidney disease discovered that the EndoMT process contributed to ~40% of all myofibroblasts (Zeisberg et al., 2008).Circulating monocyte-derived macrophages play a key role in renal injury, secreting pro-inflammatory cytokines and undergoing myofibroblast transformation.Studies on obstructive nephropathy reveal their contribution to the myofibroblast pool, ranging from 35% to as much as 90%, as observed through red fluorescent transgene tracing controlled by Acta2 and identification using F4/80 and α-SMA markers (LeBleu et al., 2013;Vierhout et al., 2021).
Transforming growth factor beta (TGF-β), a well-established profibrotic molecule, is the main driver that guides cellular transdifferentiation to form the myofibroblast (Yang et al., 1999;Iwano et al., 2002;Li and Bertram, 2010;Mariasegaram et al., 2010;Qin et al., 2011;Wu et al., 2013;Loeffler and Wolf, 2015;Wang et al., 2016;Humphreys, 2018).However, TGF-β has effects beyond fibrosis, and under certain circumstances, its inhibition can stimulate tumorigenesis (Garber, 2009).In fact, the use of TGFβ-neutralizing monoclonal anitbody, Fresolimumab, led to an increased rate of cutaneous squamous cell carcinoma development in patients, which promptly regressed upon antibody treatment cessation (Lacouture et al., 2015).Complicating its suitability as a drug target, TGF-β's influence on active fibrogenic myofibroblast transdifferentition and proliferation is also cell-dependent and dose-dependent (Hathaway et al., 2015;Meng et al., 2016).Furthermore, receptor targeting by conditional blockade through the deletion of TGF-β type II receptor (TGF-IIR) on matrix-producing interstitial cells has failed to effectively reduce fibrosis severity in a model of mouse unilateral ureteral obstruction and in an aristolochic acid renal injury model suggesting the involvement of alternative signaling pathways (Neelisetty et al., 2015).
TGF-β/Smad pathway targets remain ineffective or have unintended impacts in renal fibrosis (Klinkhammer et al., 2017).Emerging molecular targets beyond the canonical TGF-β axis show promise for fibrotic repair (Meng et al., 2016).This review, though non-exhaustive, highlights these targets and associated pathways, providing a foundation for understanding the dynamic nature of myofibroblast formation in kidney fibrosis (Duffield, 2014).It may reveal potential regenerative interventions to modulate nonessential pathways.Table 1 includes additional emerging targets beyond the reviewed molecules to capture a broader framework of these novel pathways.

Fibroblast to myofibroblast
The yes-associated protein (YAP) signaling pathway has been shown in renal models to regulate fibrosis (An et al., 2022).YAP is part of the Hippo signaling pathway known to control cell growth, proliferation and organ size (Piccolo et al., 2014).It also plays a role in promoting fibroblast activation and myofibroblast differentiation induced by Smad/TGF-β signaling.Interestingly, the extent of TGFβ stimulation appears to dictate YAP's function.Short exposures to TGF-β lead to YAP translocating into the nucleus, where it upregulates profibrotic and myofibroblast transdifferentiationrelated genes.However, with prolonged and high-level TGF-β exposure, mammalian STE20-like protein kinase 1/2 (MST1/2) activation inhibits YAP translocation, effectively blocking a profibrotic response.This intricate regulatory mechanism involves MST1/2, acting as a negative regulator of YAP.Notably, knockout of Mst1/2 in platelet-derived growth factor receptor-α (PDGFRα)+ specific cells leads to increased YAP expression, concomitant myofibroblast accumulation, and heightened renal fibrosis.Given that approximately 80% of PDGFRα+ or α-SMA + fibroblasts also happen to be YAP + cells, this signaling pathway emerges as a potentially dominant mechanism to drive the transition of fibroblasts to myofibroblasts (An et al., 2022).
Associated with YAP molecular signaling, another molecule emerges-transcriptional coactivator with PDZ-binding motif (TAZ), a transcriptional coactivator also within the Hippo pathway.Recent studies conducted in both mouse and human tissues have unveiled a notable pattern: following kidney transplant injury and in mouse models of UUO and ischemiareperfusion injury (IRI) injury, the levels of YAP and TAZ in myofibroblasts surge dramatically.However, when myofibroblast YAP/TAZ is in short supply, fibrosis in unilateral ureteral obstruction-induced kidneys is diminished.Conversely, an excess of fibroblast YAP/TAZ intensifies fibrotic injury.
Furthermore, canonical transient receptor potential channel 6 (TRPC6), a nonselective cation channel, has been linked to TGF-β 1-induced fibroblast-to-myofibroblast transition in renal interstitial fibroblasts.TRPC6 expression is regulated by TGF-β 1 and is mediated through the p38/YAP pathway in renal interstitial fibroblast NRK-49F cells.NRK-49F Trpc6 inhibition by siRNA or through pharmalogical intervention reduces expression levels of αsmooth muscle actin (α-SMA) and collagen I, both classical markers of myofibroblasts.Jiang et al. demonstrate that TGF-β 1 exposure causes phosphorylation of p38 and Yes-associated protein, leading to translocation into the nuclei.Inhibition p38/YAP phosphorylation decreases TRPC6 and α-SMA expression, indicating a key role for TRPC6 in TGF-β 1 stimulated fibroblast to myofibroblast transition that is also downstream of the p38/YAP pathway (Jiang et al., 2022).

Pericyte to myofibroblast
Studies utilizing pathway and gene expression analysis implicate miR-132 in pericyte transformation under TGF-β/Smad2/ Smad3 control, networking with both cell proliferation-related signaling, STAT3/ERK and Foxo3/p300.The role of miR-132 becomes particularly evident in FoxD1-GC; Z/Red-mice subjected to unilateral ureteral obstruction, where pericytes undergo a remarkable 21-fold upregulation of miR-132 during the transition from pericytes to myofibroblasts.This surge in miR-132 coincides with heightened myofibroblast transformation and an increase in collagen deposition, ultimately contributing to kidney fibrosis.When miR-132 is deliberately silenced, not only does collagen deposition decrease, but myofibroblast proliferation is attenuated.However, in the context of TGF-β control over pericyte-tomyofibroblast transformation, it's worth noting that while TGF-β stimulation in pericyte cultures typically triggers myofibroblast differentiation without significant proliferation, this is not the case downstream of miR-132 activation (Ren and Duffield, 2013), suggesting a more intricate regulatory role for this microRNA in the transformation process.
Zinc finger E-box binding homeobox 2 (ZEB2), a promoter region with binding sites for FOXD1, exerts inhibitory control over both transcription factor, Zeb2 and decorin (DCN).An intriguing aspect of DCN, a proteoglycan, is that it sequesters TGF-β, thereby neutralizing its binding capability and subsequent signaling cascade and it acts as an antagonist to SMAD signaling, particularly during nephron development.Therefore, it is plausible that Zeb2 may ultimately regulate upstream of SMAD which would explain why when inhibited in Zeb2 cKO mice there is differentiation of Foxd1positive pericytes into myofibroblasts, ultimately contributing to the progression of renal fibrosis (Fetting et al., 2014;Kumar et al., 2023).
In a comprehensive study investigating the impact of inflammation-induced upregulation of vitamin D receptors (VDR) in the kidney, a significant link was established between VDR activators and the conversion of pericytes into myofibroblasts.This study examined the influence of VDR activators in the context of hypoxia or in the presence of TGF-β 1, both of which initiate pericyte transformation in a HIF-1α-dependent manner, primarily through the Smad2 signaling pathway.Notably, the introduction of paricalcitol, a ligand targeting VDRs, resulted in a marked inhibition of pericyte transformation.Intervention coincided with a concurrent reduction in the expression of key markers, TGF-β 1, α-SMA, and PDGFR β in pericytes (El-Atifi et al., 2015).This study not only suggests a VDR link to pericyte transdifferentiation but also highlights the pivotal role of hypoxia in orchestrating HIF-1α, which integrates multiple signaling networks such as the TGF-β/Smad, Notch, and NF-κB pathways in the induction of kidney fibrosis (Liu et al., 2017;Lim et al., 2021).

Endothelial to mesenchymal transition
In UUO mice induced for kidney fibrosis, matrix metalloproteinase 9 (MMP-9) knockouts have reduced EndoMT, evidenced by decreased histological VE-cadherin and α-SMA colocalization compared to wildtype controls.Primary endothelial cells (MRPECs) treated with recombinant TGF-β 1 and MMP-9 inhibitor also demonstrated the same reduction in EndoMT.These changes correlated with a decrease in Notch activation and its downstream reduction of transcription factor, Hey-1, indicating a MMP-9 dependent -Notch regulation of myofibroblast formation (Zhao et al., 2017).
While ERK1/2 signaling is crucial in development and homeostasis, modest pharmacological intervention to prevent inhibition of ERK1/2 signaling may halt endothelial-tomyofibroblast transformation (Cao et al., 2019).Mice lacking global ERK1 and containing endothelial-specific ERK2 (Erk1 −/− Erk2 iEC−/− ) by two and 3 weeks of age undergo myofibroblast transition of endothelial cells as indicated by VE cadherin + αSMA + expression in the kidney.By 5 weeks of age, these mice succumb to organ failure, primarily a result of fibrosis of the heart.An increase of EndoMT was replicated through siRNA inhibition of ERK1/2 signaling of human endothelial cells (HUVECs) when compared to wildtype controls.RNA Seq analysis of these siRNA treated cells and scrambled controls indicated a gene hub that consisted of a TGF-β 2 gene driver with downstream regulation of ECM genes COL25A1, COL5A1, and CDH2 vasoactive genes EDN1 and NOS3, and fenestration gene, PVLAP (Ricard et al., 2019).
MiRNA-driven myofibroblast transformation is evident in endothelial cells, as indicated by a study highlighting the regulatory role of miR-126-3p.Endothelial derived myofibroblasts were traced by YFP under a Cdh5 promoter in mice which were induced to kidney fibrosis by unilateral ureteral obstruction.Five days post-operation, YFP + cells differentiated within the glomeruli, capillaries, and blood vessel intima, and represented 9% of αSMA + cells.Examination via RT-qPCR and in situ hybridization of kidney sections revealed a significant downregulation of miR-126-3p in fibrotic mouse kidneys compared to healthy kidneys.This downregulation was also observed in fibrotic human kidneys compared to normal kidneys.Diminished levels of miR-126-3p are evident in diverse injury contexts, including human renal IRI and myocardial infarction, indicating its prospective utility as a valuable molecular biomarker for disease monitoring and a potential target for intervention (Jordan et al., 2021).

Macrophage to myofibroblast
Studies utilizing chromatin immunoprecipitation (ChIP) techniques have identified POU Class 4 Homeobox 1 (Pou4f1) as a downstream target of Smad3 and a regulator of macrophage-tomyofibroblast transition (MMT).Microarray analysis further revealed Pou4f1 as a pivotal node in a fibrogenic gene network that promotes TGF-β 1/Smad3-driven MMT in bone-marrow derived macrophages, suggesting Pou4f1 as a potentially viable therapeutic target.Macrophage-specific inhibition of Pou4f1 signaling, both in vitro and in vivo using UUO and IRI mouse models of kidney fibrosis, effectively impedes macrophage transition to myofibroblasts and halts fibrosis progression.Notably, elevated Pou4f1 levels also show a strong correlation with increased renal injury associated with fibrosis in human diseases (Tang et al., 2020).
The TGF-β/Smad signaling pathway dynamically engages with the Wnt/β-catenin pathway.A fundamental part of the canonical Wnt signaling cascade is mediated by β-catenin and T-cell factor (TCF).Interventions that hinder β-catenin's interaction with TCF not only reduce MMT but also steer β-catenin towards interacting with Foxo1 during the progression of renal fibrosis (Vierhout et al., 2021).Intriguingly, the non-canonical Wnt signaling pathway can be set into motion through TGF-β 1 induction, where Wnt5a operates independently by elevating YAP/TAZ levels.This elevation then triggers the induction of macrophage M2 polarization, driving forward myofibroblast transformation and the fibrotic response (An et al., 2022).
In a concurrent signaling cascade, the A2B adenosine receptor, notable for its ability to initiate G protein signaling that sets in motion a multitude of intracellular processes, such as alterations in gene expression, ion channel activity, and metabolism, showcases the potential to amplify the TGF-β pathway.This augmentation, in turn, is implicated in an escalation of renal fibrosis by instigating MMT.Experimental treatments involving the use of an A2B adenosine receptor antagonist in diabetic rats revealed a notable decrease in glomerulosclerosis, reduced levels of collagen and α-SMA, and a significant reduction in MMT, indicating a promising approach for mitigating renal fibrosis (Cardenas et al., 2013;Torres et al., 2020).
Another receptor pathway-the aldosterone mineralocorticoid receptor (MR) has been shown to aide the transition of M1 macrophages to myofibroblasts upon stimulation of TGF-β.Inhibition of this pathway using the MR blocker, esaxerenone, has been demonstrated to mitigate the fibrotic phenotype and reduce macrophage-myofibroblast transition both in in vivo and in vitro settings (Qiang et al., 2022).
Lastly, analyzing the transcriptome of macrophages treated with TGF-β revealed the involvement of CXCL10 in the signaling pathway.Notably, in UUO mice, there exists a noteworthy positive correlation between the accumulation of MMT and increased levels of CXCL10.Interventions targeting CXCL10, whether through siRNA knockdown techniques, or chemical inhibition, resulted in a reduction of MMT in mouse models of kidney fibrosis, underlining the dynamic impact of CXCL10 (Feng et al., 2020a).It is worth mentioning that CXCL10 exhibits a dual molecular potential, sometimes acting as an antifibrotic agent, depending on the specific fibrotic context (Tager et al., 2004;Roman and Mutsaers, 2018;Feng et al., 2020b).

Discussion
It is important to recognize that often the mechanism driving alterations in myofibroblast cells involves paracrine cross-talk.Tubular cells, when damaged during renal fibrosis, transition to a secretory phenotype and generate fibrogenic agents such as sonic hedgehog (Shh), Wnt ligands, and TGF-β, as well as IL-6, monocyte chemotactic protein-1, TNF-α, and other inflammatory cytokines.These molecules contribute to the activation and transdifferentiation of myofibroblasts.Often, this paracrine interaction is facilitated through the production and transmission of mRNA exosomes containing factors such as TGF-β, delivered to neighboring fibroblasts (Mack and Yanagita, 2015).An example of paracrine communication at the cellular level can be observed with putative endothelial progenitor cells (pEPCs) or their microvesicles, which are believed to secrete factors that inhibit TGF-β regulation, effectively reducing pericyte-to-myofibroblast transition.Following pEPC treatment, UUOinjured mice not only exhibit reduced expression of α-SMA and Collagen IV but also experience decreased myofibroblast accumulation and mitigation of renal fibrosis (Schrimpf and Duffield, 2011;Smith et al., 2012;Yang et al., 2019).
Since the hallmark of successful renal fibrosis resolution in disease models includes a decrease in myofibroblasts (Sun and Kisseleva, 2015), comprehending myofibroblast generation dynamics is key in identifying therapeutic targets in renal fibrotic diseases.In order to realistically carry out myofibroblast transdifferentiation inhibition it is reasonable to either alter the secretory phenotype of kidney resident cells or to block circulating recruitment into the renal interstitium of bone marrow derived macrophages fated for MMT (Kok et al., 2014;Meng et al., 2014;Tampe and Zeisberg, 2014).Targeting the TGF-β canonical or non-canonical signaling pathways within the renal interstitium could be achieved through a hydrogel depot containing inhibitors such as microRNA-21-5p/miR-132 antagonists and CXCL10 inhibitors (Adler et al., 2010;Qin et al., 2011) and may present a feasible method to inhibit local activation and proliferation of myofibroblasts.
Expanding the search for targets outside of the classical TGFβ/Smad pathway is crucial for advancing antifibrotic therapies.Currently, ACE-Is and ARBs dominate therapeutic options for renal fibrosis (Liu and Zhuang, 2019).While therapies addressing leukocyte recruitment, soluble factors, and matrix protein production are valuable, strategically targeting myofibroblasts at the developmental stage emerges as a promising approach to control excessive extracellular matrix deposition and impede fibrosis progression (Tampe and Zeisberg, 2014).

TABLE 1
Emerging targets outside classical TGF-β signaling in renal fibrosis.