Extracellular matrix and Hippo signaling as therapeutic targets of antifibrotic compounds for uterine fibroids

Abstract

Background: Uterine fibroids are highly prevalent, collagen-rich, mechanically stiff, fibrotic tumors for which new therapeutic options are needed. Increased extracellular matrix (ECM) stiffness activates mechanical signaling and Hippo/YAP promoting fibroid growth, but no prior studies have tested either as a therapeutic target. We tested the hypothesis that injection of a purified form of collagenase Clostridium histolyticum (CCH) that selectively digests type I and type III collagens would alter ECM stiffness, Hippo signaling, and selectively reduce fibroid cell growth. We also used two FDA-approved drugs, verteporfin and nintedanib, to elucidate the role of Hippo/YAP signaling in uterine fibroid and myometrial cells.

Methods: The clinical trial was registered (NCT02889848). Stiffness of samples was measured by rheometry. Protein expression in surgical samples was analyzed via immunofluorescence. Protein and gene expression in uterine fibroid or myometrial cell lines were measured by real time PCR and western blot, and immunofluorescence.

Results: Injection of CCH at high doses (0.1-0.2 mg/cm3 ) into fibroids resulted in a 46% reduction in stiffness in injected fibroids compared to controls after 60 days. Levels of the cell proliferation marker proliferative cell nuclear antigen (PCNA) were decreased in fibroids 60 days after injection at high doses of CCH. Key Hippo signaling factors, specifically the transcriptionally inactive phosphorylated YAP (p-YAP), was increased at high CCH doses, supporting the role of YAP in fibroid growth. Furthermore, inhibition of YAP via verteporfin (YAP inhibitor) decreased cell proliferation, gene and protein expression of key factors promoting fibrosis and mechanotransduction in fibroid cells. Additionally, the anti-fibrotic drug, nintedanib, inhibited YAP and showed anti-fibrotic effects.

Conclusions: This is the first report that in vivo injection of collagenase into uterine fibroids led to a reduction in Hippo/YAP signaling and crucial genes and pathways involved in fibroid growth. These results indicate that targeting ECM stiffness and Hippo signaling might be an effective strategy for uterine fibroids.

Keywords: Hippo signaling; collagenase; extracellular matrix; nintedanib; uterine fibroids; verteporfin.

Conflict of interest statement

James H. Segars is or was a PI on research sponsored by Bayer, Abbvie, Biospecifics, Allergan, Inc., and Myovant. James H. Segars serves on boards of the Society for Reproductive Investigation, and the American Gynecological and Obstetrical Society. Md Soriful Islam, Friederike L. Jayes, Bhuchitra Singh, Sadia Afrin, Joshua T. Brennan, and Phyllis C. Leppert have nothing to disclose. Mostafa A. Borahay serves as Advisory Board member for Myovant Sciences.

© 2021 The Authors. Clinical and Translational Medicine published by John Wiley & Sons Australia, Ltd on behalf of Shanghai Institute of Clinical Bioinformatics.

Figures

FIGURE 1

Collagenase Clostridium histolyticum (CCH) injection regulates stiffness of extracellular matrix and cell proliferation as well as Hippo signaling. Uterine fibroid tissues were injected with CCH at a fixed dose (1.16 mg) (n = 3), and tissues were collected after 24–48 h (Group 1). In group 2, fibroids were injected with CCH at 0.05 mg/cm 3 (dose 1) (n = 3), 0.1 mg/cm 3 (dose 2) (n = 3), and 0.2 mg/cm 3 (dose 3) (n = 3) and were collected after 60–90 days. One sample was morcellated and was not able to be tested. Therefore, we excluded that sample from the analysis. (A) Quantification of mechanical stiffness in control and CCH‐treated fibroid tissues. Injection of CCH at fixed dose (1.16 mg) reduced stiffness by 21% (p = 0.500) (a). In contrast, CCH injection at low dose (0.05 mg/cm 3 ) increased stiffness by 22% (p = 0.630) (b). Of note, stiffness of fibroid tissue was significantly reduced (46%) by CCH injection at the two highest doses (0.1–0.2 mg/cm 3 ) (p = 0.031) (c). Data are presented as means ± SEM of CCH‐treated or control fibroids (these are considered as biological replicates). Statistical significance was determined using Wilcoxon matched‐pairs signed rank test and set a p < 0.05. *p < 0.05. (B) Levels of the proliferation marker, proliferative cell nuclear antigen (PCNA), in CCH‐treated fibroids. CCH injection increased PCNA expression in two out of two subjects in 24–48 h treatment group and decreased in five of six subjects in 60–90 days at highest doses (values below red bar) (p < 0.05). The numbers below the X‐bars (such as 007–019) indicate subject (or fibroid) number. (C) Quantification of the Hippo signaling factor phosphorylated YAP (p‐YAP) in control (open bars) or CCH‐treated fibroid samples (black bars). The levels of p‐YAP were normalized with YAP/TAZ (total). The p‐YAP expression was increased in 27% (three out of 11) CCH‐injected fibroids (black bars), compared to control samples (open bars). Notably, CCH at doses 2–3 increased p‐YAP expression in 50% (three out six) (values below red bar) CCH‐treated fibroid tissues (p < 0.05). Data are presented as mean ± SD of five images (these are considered as technical replicates) (20X) from CCH treated fibroid or control adjacent fibroid. (D) Representative immunofluorescence images of p‐YAP expression in control and CCH treated fibroids. Phospho‐YAP immunostaining is indicated as red, DAPI is indicated as blue, and α‐SMA is indicated as green. Scale bar = 50 μM. Negative controls were performed by omitting the first or the secondary antibody. (E) Schematic representation of mechanotransduction and growth factor signaling in fibroid growth. In fibroid cells, extracellular matrix (ECM) initiates mechanotransduction process by interacting with integrin receptors (ITGA6, ITGB1) and activates cytoplasmic kinase cascades (such as FAK, AKAP13). This leads to activate downstream signaling pathways (such as ERK 1/2) that regulates transcription of genes involved in cell growth and fibrosis. Profibrotic growth factor activin A binds to its receptors that activates SMAD2/3 signaling pathway and induces fibrotic gene expression. In fibroid tumors the Hippo target “YAP” is in the nucleus. In this state, Hippo signaling is OFF, the MST1/2 and LATS1/2 kinases are inactive, YAP and TAZ are dephosphorylated, and translocate into the nucleus where they are in complex with TEAD, inducing transcription of genes. (F) Regulation of multiple critical targets of fibroid growth by antifibrotic compounds through inactivation of YAP pathway. CCH‐treated fibroid shows a reduction in stiffness of ECM (46% reduction), reduction in staining intensity of PCNA (cell proliferation marker) (six out of 11 subjects) at high doses. There is an increase in phosphorylated YAP at high dose of CCH, suggesting the inactivation of YAP. The inhibitor of YAP, verteporfin inhibits series of targets of fibrosis and proliferation, at least in part, by inhibition of YAP, as well as ERK 1/2 and SMAD 2/3 signaling pathways. The antifibrotic drug, nintedanib also inhibits YAP in fibroid cells that leads to changes in the expression of genes involved in cell growth and fibrosis. These suggests that these three drugs might be used as alone or as a combination therapy for fibroid treatment

FIGURE 2

Verteporfin (VP) shows YAP‐specific actions in fibroid and myometrial cells. (A and B) The expression and localization of YAP/TAZ in myometrium and fibroid tissues. YAP/TAZ expression levels were quantified in myometrial and fibroid tissues (A) or cells (B). Both myometrium and fibroids expressed YAP/TAZ (nuclear) but intensity was higher in fibroids. Results are presented as mean ± SEM. For tissues, data were calculated from 10 immunofluorescence images of two patients. For cells, data were calculated from three independent experiments (n = 9) (20X). Scale bar = 50 μM. (C and D) VP reduced fibroid cell proliferation. MTS assays were performed after treatment with verteporfin for 24 h. The differential effect in the viability of fibroid cells was observed after VP treatment compared to myometrial cells (n = 4) (C). Results are presented as mean ± SEM. Data were calculated from four independent biologic experimental replicates. VP (1 μM) reduced protein (n = 5) expression of proliferative gene PCNA (proliferative cell nuclear antigen) in fibroid cells (D). (E‐G) YAP‐specific actions of VP in fibroid and myometrial cells. The reduced protein expression of non‐phospho‐YAP (active) in fibroid and myometrial cells after VP treatment (VP) for 24 h was compared to untreated (NT) control (n = 3) (E). The house keeping protein β‐actin was used to normalize non‐phospho‐YAP levels. The YAP‐responsive genes connective tissue growth factor (CTGF) (F) and CYR61 (cysteine‐rich angiogenic inducer 61) (G) were highly expressed in fibroid, compared to myometrial cells. VP significantly decreased mRNA levels in both fibroid and myometrial cells (n = 3). Each graph represents experimental data with means ± SD. Data were calculated from 3–5 independent biologic experimental replicates. Statistical significance was determined using the Mann Whitney U test and set at p < 0.05. *p < 0.05, **p < 0.01, ***p < 0.001. Abbreviation: NT, no treatment (control)

FIGURE 3

Verteporfin (VP) reduces key targets of fibrosis in fibroid cells. (A‐C) VP reduced extracellular matrix (ECM) expression in fibroid cells. Human fibroid and myometrial cells were treated with VP at 1 μM for 24 h. The mRNA and protein expression was measured using real time qPCR and western blot, respectively. (A and B) Fibronectin was highly expressed in fibroid compared to myometrial cells. VP decreased mRNA (n = 3) (A) and protein (n = 4) (B) levels in both fibroid and myometrial cells. (C) Versican (VCAN) was also highly expressed in fibroid compared to myometrial cells. VP significantly reduced mRNA expression in both fibroid and myometrial cells (n = 3) (C). (D‐K) VP reduced activin A, its receptors, and downstream targets in fibroid cells. Uterine fibroid cells highly expressed activin A (INHBA) (D) and its receptors, ACVR2A (E), ACVR2B (F), and ACVR1B (G) which were downregulated by VP treatment (n = 3). VP also decreased SMAD2 (H) and p‐SMAD2 (I) levels in fibroid cells (n = 3). The downstream target of activin A, PAI‐1 (plasminogen activator inhibitor‐1) mRNA (J) and protein (K) levels were reduced by verteporfin in fibroid and myometrial cells (n = 3). Results are expressed as mean ± SD. Data were calculated from 3–4 independent biologic experimental replicates. Statistical significance was determined using the Mann Whitney U test and set at p < 0.05. *p < 0.05, **p < 0.01, ***p < 0.001. Abbreviations: NT, no treatment (control)

FIGURE 4

Verteporfin (VP) regulates of mechanotransduction in fibroid cells. Human fibroid and myometrial cells were treated with VP (1 μM) for 24 h. The mRNA and protein expressions of selected genes were measured through real time qPCR and western blot, respectively. The integrin receptors ITGB1 (A) but not ITGA6 (B) were more highly expressed in fibroid compared to myometrial cells and were reduced by VP treatment (n = 3). The cytoplasmic kinase focal adhesion kinase (FAK) (C) and phospho‐FAK (D) were also highly expressed in fibroid cells and were decreased by VP treatment (n = 3). A‐kinase anchor protein 13 (AKAP13) is a target of FAK that was expressed at higher levels in fibroid cells and likewise decreased by VP treatment (n = 3) (E). VP also reduced phospho‐ERK 1/2 (F) levels in both fibroid and myometrial cells (n = 3). Results are presented as mean ± SD. Data were calculated from three independent biologic experimental replicates. Statistical significance was determined using the Mann Whitney U test and set at p < 0.05. *p < 0.05, **p < 0.01, ***p < 0.001. Abbreviation: NT, no treatment (control)

FIGURE 5

Nintedanib (NDB) treatment reduces growth of fibroid cells and expression of genes or proteins involved in fibrosis. (A and B) Nintedanib treatment reduced proliferation of fibroid cells. Human fibroid and myometrial cells were treated with NDB (5 μM) for 24 h and 48 h or cells were treated with vehicle. An MTS assay was performed after treatment with NDB for indicated durations. The differential effect of NDB was observed in cell proliferation of fibroid cells versus myometrial cells (n = 4) (A). Results are presented as mean ± SEM. Data were calculated from four independent biologic experiments. NDB reduced PCNA protein expression in both myometrial and fibroid cells, compared to untreated controls (n = 3) (B). (C‐E) NDB treatment decreased expression of key targets of fibrosis in fibroid cells. Myometrial and fibroid cells were treated with NDB at 5 μM concentration for 24 h. The mRNA and protein expression of selected genes was measured using real time qPCR and western blot, respectively. Fibronectin (FN) was highly expressed at both mRNA (C) and protein (D) levels in fibroid than myometrial cells, and levels were reduced by NDB treatment (n = 3). Versican (VCAN) was also highly expressed in fibroid cells and levels were reduced by NDB treatment (n = 2) (E). (F‐H) NDB treatment decreased activin A and its downstream target in fibroid cells. Activin A (INHBA) was highly expressed in fibroid cells, compared to myometrial cells, and levels were reduced by nintedanib treatment (n = 3) (F). The mRNA and protein levels of PAI‐1 (plasminogen activator inhibitor‐1) were higher in fibroid, compared to myometrial cells. NDB treatment decreased PAI‐1 mRNA (n = 3) (G) and protein (n = 2) (H) levels in both fibroid and myometrial cells. Results are presented as mean ± SD. Data were calculated from 2–4 independent biologic experiments. Statistical significance was determined using the Mann Whitney U test and set at p < 0.05. *p < 0.05, **p < 0.01, ***p < 0.001. Abbreviation: NT, no treatment (vehicle)

FIGURE 6

Nintedanib (NDB) treatment alters Hippo pathway in fibroid cells. Human fibroid and myometrial cells were treated with NDB (5 μM) for 24 h. NDB treatment decreased yes‐associated protein (YAP ) (A) and slightly increased phospho‐YAP (B) in fibroid cells (n = 3). The house keeping protein β‐actin was used to normalize YAP and phospho‐YAP levels. The YAP‐responsive genes connective tissue growth factor (CTGF) (C) and CCND1 (cyclin D1) (D) levels were reduced by NDB treatment of fibroid and myometrial cells (n = 3). The Hippo kinases SAV1 (Salvador Family WW Domain Containing Protein 1) (E) was increased but MST1 (macrophage stimulating 1) (F), and LATS1 (large tumor suppressor 1) (G) levels were unchanged after NDB treatment (n = 3). Results are expressed as mean ± SD. All the data were calculated from three independent biologic experiments. Statistical significance was determined using the Mann Whitney U test and set at p < 0.05. *p < 0.05, **p < 0.01, ***p < 0.001. Abbreviation: NT, no treatment (vehicle)