Aberrant Transforming Growth Factor‐β Activation Recruits Mesenchymal Stem Cells During Prostatic Hyperplasia

Abstract Benign prostatic hyperplasia (BPH) is the overgrowth of prostate tissues with high prevalence in older men. BPH pathogenesis is not completely understood, but it is believed to be a result of de novo overgrowth of prostatic stroma. In this study, we show that aberrant activation of transforming growth factor‐β (TGF‐β) mobilizes mesenchymal/stromal stem cells (MSCs) in circulating blood, which are recruited for the prostatic stromal hyperplasia. Elevated levels of active TGF‐β were observed in both a phenylephrine‐induced prostatic hyperplasia mouse model and human BPH tissues. Nestin lineage tracing revealed that 39.6% ± 6.3% of fibroblasts and 73.3% ± 4.2% smooth muscle cells were derived from nestin+ cells in Nestin‐Cre, Rosa26‐YFPflox/+mice. Nestin+ MSCs were increased in the prostatic hyperplasia mice. Our parabiosis experiment demonstrate that nestin+ MSCs were mobilized and recruited to the prostatic stroma of wild‐type mice and gave rise to the fibroblasts. Moreover, injection of a TGF‐β neutralizing antibody (1D11) inhibits mobilization of MSCs, their recruitment to the prostatic stroma and hyperplasia. Importantly, knockout of TβRII in nestin+ cell lineage ameliorated stromal hyperplasia. Thus, elevated levels of TGF‐β‐induced mobilization and recruitment of MSCs to the reactive stroma resulting in overgrowth of prostate tissues in BPH and, thus, inhibition of TGF‐β activity could be a potential therapy for BPH. Stem Cells Translational Medicine 2017;6:394–404


INTRODUCTION
Benign prostatic hyperplasia (BPH) is a nonmalignant overgrowth of the prostatic tissue within the transition zone affecting approximately 50% of men aged 51-60 years and 90% by the ninth decade of life [1]. BPH is associated lower urinary tract symptoms (LUTS), including storage and voiding symptoms that can progress to urinary retention, bladder dysfunction, and renal impairment if left untreated or treated ineffectively. Despite its prevalence, the exact etiology of BPH remains poorly understood. As a result, current therapies including a-blockers and 5-a reductase inhibitors [2] are not completely effective [3]. In addition, these treatments could lead to vascular side effects [4] and sexual dysfunction [5]. Surgical intervention is reserved for patients who have failed medical therapy or present with moderate-to-severe LUTS.
BPH is characterized by the increased stromal compartment of the prostate with age. The stromal-to-epithelial ratio is relatively constant at 2:1 from birth to 40 years of age in normal glands increasing to 5:1 in BPH [6]. Embryonic reawakening [7,8], imbalance between androgen and estrogen [9], chronic inflammation [10,11], stem cell defects [12], and epithelial-stromal interactions [13,14] have been implicated in the pathogenesis of BPH. Investigation of the morphological analogies between the fetal prostatic stroma and the BPH stroma nodules showed a similar developmental profile: from immature mesenchyme toward fibroblastic stroma, culminating in predominantly smooth muscle [8]. Androgen/androgen receptor (AR) signaling plays an important role in BPH development by influencing the stromal cell paracrine factors that can regulate the adjacent epithelial growth and differentiation [13][14][15]. Inversely, epithelial cells contribute to the increased stromal cell population via epithelialmesenchymal transition (EMT) [16,17]. Therefore, BPH is a stromal enlargement disease likely associated with reawakening of embryonic inductive properties.
High levels of transforming growth factor-b (TGF-b) activities are often observed in the pathogenesis of BPH [18]. Induction of fibroplasia and collagenous micronodules was observed with TGF-b1 overexpression in transgenic mice [19]. Furthermore, the expression of TGF-b receptor II protein is associated with increased prostatic gland volume and lymphocyte infiltration in BPH patients [20]. Among the three members of the TGF-b subfamily, TGF-b1 is predominantly expressed in the prostate [21][22][23]. TGF-b is synthesized and deposited into the extracellular matrix (ECM) as a latent complex with mature TGF-b bound to latency-associated peptide (LAP). Active TGF-b can be released from the complex by cleavage of the LAP [24]. Temporo-spatial activation of latent TGF-b in the matrix recruits stem/progenitor cells for remodeling and repair in various tissues such as bone [25][26][27], vasculature [28,29], and lung [30]. Increasing evidence suggests that TGF-b constitutes integral components in the crosstalk between stem cells and their microenvironment [31].
Mesenchymal/stromal stem cells (MSCs), also known as multipotent stromal cells, have the capacity of self-renewal and differentiation into various cell types, thereby supporting tissue growth and regeneration [32,33]. Nestin, detected originally in neuronal stem cells, represents one of the class VI intermediate filament protein. Recently, nestin has been used as a marker for bone marrow derived MSCs that have both self-renewal and multilineage differentiation capacity [28,34]. The existence of MSCs in prostate has been verified in different studies [12,35,36]. In the present study, we used several animal models including phenylephrine (PE)-induced prostatic hyperplasia mice, a parabiosis mice model by Nestin-Cre, Rosa26-YFP flox/+ mice surgically sutured with wild-type littermates and conditional knockout Nestin-Cre-ER, Tgfbr2 flox/flox mice to investigate whether nestin + MSCs are mobilized in peripheral circulation, recruited to the prostatic stroma and give rise to the fibroblasts during prostatic hyperplasia. We found that systemic blockade of TGF-b activity or conditional deletion of the TGF-b receptor II (Tgfbr2) in nestin + cell lineage attenuated stromal hyperplasia with potential as an effective therapy for human BPH.

Human Specimens
BPH tissue specimens were collected from 32 patients with BPH that were undergoing transurethral resection of the prostate. As controls, normal prostate tissues were obtained from 18 bladder tumor patients younger than 40 undergoing radical cystoprostatectomy. No other inclusion or exclusion criteria were used other than tissue quality after thawing. Our study was performed according to the ethical standards of the 1975 Declaration of Helsinki, as revised in 2008, and was approved by the Central South University Ethic Committee. Informed consent was obtained from all the participating patients.
Phenylephrine induction of ventral prostate hyperplasia procedures were performed on 6-week-old mice with the methods previously described by Marinese et al. [37]. As the ventral lobe is often considered as the common site for prostatic hyperplasia, the ventral prostates were harvested and analyzed. For the time course experiments, PE-induced mice and the saline-treated controls were sacrificed at 0, 1, 2, or 4 weeks after initial injection. For the treatment with TGF-b neutralizing antibody 1D11 (5 mg/kg 21 ; Sanofi Genzyme, Cambridge, MA, https://www.sanofigenzyme. com) or the equivalent dose of control antibody 13C4 were injected intraperitoneally three times per week starting with PE injection.
We determined the genotype of the mice by polymerase chain reaction analyses of genomic DNA isolated from mouse tails using the following primer sequences: No statistical method was used to predetermine sample size. Male mice were randomly assigned to both control and testing groups, each typically containing three to five animals.
(control) mouse was surgically joined to a wild-type littermate. Briefly, the mice were anesthetized, and longitudinal skin incisions were performed from the elbow to the knee joint of each mouse. The elbow and knee joints were attached by a surgical suture, and then the dorsal and ventral skin was stitched by a continuous 5-0 Vicryl (Ethicon, Somerville, NJ, http://www.ethicon. com) suture. Each parabiotic pair was housed in a clean cage with moistened food pellets on the floor to decrease the movement of reaching for food while adjusting to parabiotic existence. After 2 weeks, shared blood circulation between the mice was confirmed by injection of Evans blue dye (Sigma-Aldrich, St. Louis, MO, https://www.sigmaaldrich.com). The parabiotic wild-type partner was injected with either PE (10 mg/kg 21 /day 21 ) or saline for 4 weeks before the mice were sacrificed. Ventral prostates from the mouse of each parabiotic pair were harvested for analysis.
All animals were maintained in the Animal Facility of the Johns Hopkins University School of Medicine. The experimental protocols for both species were reviewed and approved by the Institutional Animal Care and Use Committee of the Johns Hopkins University.

Enzyme-Linked Immunosorbent Assay of TGF-b1 in Plasma
We detected the concentration of active TGF-b1 in plasma of mice with an enzyme-linked immunosorbent assay (ELISA) development kit (R&D Systems, Minneapolis, MN, https://www. rndsystems.com) according to the manufacturer's instructions.

Isolation and Culture of Murine GFP-Labeled Prostatic Nestin + Cells
To obtain prostatic nestin + cells, ventral prostate isolated from 6-week-old male Nestin-GFP transgenic mice after euthanization were carefully dissected under sterile conditions. Subsequently, the procured tissues were submerged in 5% low melting point agarose (Thermo Fisher) for protection from digestion. After solidification of the agarose, prostatic cells were isolated by collagenase in a-minimal essential medium (a-MEM; Corning Cellgro, Manassas, VA, http://www.cellgro.com). The digestion was stopped by addition of an equal volume of fetal bovine serum (FBS; Atlanta Biologicals, Flowery Branch, GA, https://www. atlantabio.com). The cell suspension was passed through a 70-mm nylon cell strainer (BD Biosciences, San Jose, CA, http:// www.bdbiosciences.com), washed three times with a-MEM. Cells aliquots were incubated for 60 minutes at 4°C with FITC-conjugated GFP antibody and perCP-conjugated CD45 antibody. Acquisition was performed on a fluorescence-activated cell sorting (FACS) Aria model (BD Biosciences), and analysis was performed with a FACS DIVE software version 6.1.3 (BD Biosciences). To isolate individual clonal strains, the sorted GFP + CD45cells were passed consecutively through 16-and 20-gauge needles to obtain single cell suspensions. For single colony forming, 1 3 10 3 cells were cultured in 100-mm dishes with a-MEM containing penicillin (100 U/ml; Sigma-Aldrich), streptomycin sulfate (100 mg/ml; Sigma-Aldrich), and 20% FBS at 37°C in a 5% CO 2 humidified incubator. Individual, well-separated colonies were selected with cloning cylinders and the number of cells was individually expanded by passaging and confirmed by flow cytometry.

Statistical Analysis
Data are presented as mean 6 SD and were analyzed using twotailed Student's t tests for comparisons between two groups, oneway analysis of variance (ANOVA) with Bonferroni post hoc test for multiple comparisons. For all experiments, p , .05 was considered to be significant and was indicated by a single asterisk; p , .01 was indicated by a double asterisk. No statistical method was used to predetermine the sample size. The experiments were randomized. The investigators were not blinded to allocation during experiments and outcome assessment.

Elevated TGF-b Activity in Prostatic Stromal Hyperplasia
To delineate the potential role of TGF-b signaling in prostatic stromal hyperplasia, we generated a PE-induced mouse model [37,40]. Mice induced with PE showed stromal compartment hyperplasia with concomitant papillary in folding and/or piling up of epithelial cells after 4 weeks of injection relative to saline-treated mice (Fig. 1A). Collagen deposition and stromal content were significantly increased from 2 to 4 weeks after initial PE injection with Masson's trichrome staining (Fig. 1B). In parallel, the number of vimentin + fibroblasts were also increased (Fig. 1C). Ventral prostate sections were immunostained for phosphorylated Smad2/3 (pSmad2/3) and the results demonstrated that pSmad2/3levels were elevated in the nuclei of both epithelial and stromal cells continuously from1week after initial PE injection (Fig. 1D). To examine whether active TGF-b level in circulation is changed, we measured the concentration of active TGF-b1 in blood. In PEinduced mice, active TGF-b1was significantly higher at 1 week and peaked at 4 weeks relative to the level of week 0 (Fig. 1E). These results indicate that TGF-b signaling was elevated with reactive stromal component in PE-induced mice. We then examined whether high level of active TGF-b is directly associated with prostate stroma augmentation in human BPH tissues. A constant increase of the relative and absolute amount of stromal component was observed in all human BPH specimens stained by H&E ( Fig. 2A). Masson's trichrome staining showed far more stromal compartment with large quantity of blue-staining collagen fibers mixed with red-staining smooth muscle cells in BPH relative to normal prostate tissue (Fig. 2B). Vimentin + area was also more extensive in BPH (Fig. 2C). Again, pSmad2/3 + cells were dramatically increased in both the epithelium and stromal cells in BPH samples, whereas fewer pSmad2/3 + cells were loosely scattered in normal prostate (Fig. 2D). Thus, high levels of TGF-b activity are associated with stromal expansion in the pathogenesis of human BPH.

MSCs Are Recruited to the Stroma During Prostatic Hyperplasia
To investigate whether high levels of active TGF-b are involved in the recruitment of MSCs during prostatic hyperplasia, we used Nestin-GFP transgenic mouse in which nestin + cells were labeled with GFP. GFP + cells increased significantly with time in the prostate stroma of the PE-induced mice (Fig. 3A). Ventral prostate sections were immunostained with antibodies against CD90 and Sca1, markers for MSCs [41]. CD90 and Sca1 were colocalized with GFP in the stroma (Fig. 3B), indicating increase of MSCs. Furthermore, the percentage of GFP + CD45 2 cells in blood circulation were almost twofold at 1 week and fourfold at 2 and 4 weeks after injection of PE relative to baseline at week 0 (Fig. 3C). Similarly, in wild-type mice, the percentages of CD90 + Sca1 + CD45 2 CD11b 2 cells in blood were also increased with PE injection (Fig. 3D).
To examine whether the recruited GFP + cells in prostate stroma are multipotential MSCs, cells isolated from ventral prostate of Nestin-GFP mice were sorted by GFP + and CD45-negative  (Fig. 3E). The sorted GFP + CD45cells were assessed by colony-forming unit fibroblast (CFU-F) assay (Fig. 3F) and displayed CFU-F activity. Their multipotential was tested for fibrogenesis, osteogenesis, adipogenesis, and chondrogenesis (Fig.  3G). These results suggest that MSCs were mobilized into the peripheral blood and recruited to the stroma during prostatic hyperplasia.
To determine whether prostate stromal cells are differentiated from nestin + cells, we performed nestin-lineage tracing experiment by crossing Nestin-Cre mice with Rosa26-YFP mice. Both nestin + cells and their descendants were labeled with YFP. In 6-week-old Nestin-Cre, Rosa26-YFP flox/+ mice, double immunostaining of YFP and vimentin or a-SMA demonstrated that 39.6% 6 6.3% of fibroblasts were derived from nestin + cells, and, 73.3% 6 4.2% smooth muscle cells were of YFP + nestin origin (Fig. 3H). This shows that nestin + cells are an important source of prostate stromal cells.

Circulating Nestin + Cells Recruited to the Reactive Stroma Give Rise to Fibroblasts
To validate that recruited nestin + cells from peripheral blood are involved in prostatic hyperplasia, we generated parabiotic pairs composed of Nestin-Cre, Rosa26-YFP flox/+ mice and wild-type littermates (Para1 and Para3). The mice were treated with PE or vehicle after parabiosis for 4 weeks as shown in Figure 4A and 4B. In parallel, Nes-Cre mice were joined to their wild-type littermates (Para2) as controls of the nonspecific tissue autofluorescence. Notably, YFP + cells, which are descendants of nestin + cells derived from Nestin-Cre, Rosa26-YFP flox/+ mouse partners through circulation, were detected in the prostate stroma of wild-type partners in Para3 with PE induction, but they were significantly less in Para1 wild-type partners (Fig. 4C, 4D). Double-immunostaining of YFP and vimentin or a-SMA showed that all of the nestinlineage cells were vimentin + but not a-SMA + (Fig. 4C). YFP + CD45 2 cells in circulating blood were significantly elevated after injection of PE in Para3 relative to Para1 and Para2 control groups (Fig. 4E,  4F). Thus, nestin + cells were mobilized in blood circulation and recruited to the stroma in prostatic hyperplasia.

Inhibition of TGF-b Signaling Attenuates Prostatic Stromal Hyperplasia
To validate that aberrant activation of TGF-b in the prostatic stroma mobilizes nestin + cells, we injected the PE-induced mice intraperitoneally with TGF-b neutralizing antibody (1D11) or an isotype-matched murine IgG1 control antibody (13C4). Immunohistological staining showed that injection of TGF-b antibody 1D11 reduced Smad2/3 + phosphorylation in the stroma (Fig.  5A). The percentage of CD90 + Sca1 + CD45 2 CD11b 2 MSCs in blood circulation was also decreased (Fig. 5B). Significant reduction of the collagen accumulation and decrease of vimentin + fibroblasts were observed in the 1D11-treated mice relative to the mice injected with 13C4 control antibody (Fig. 5C, 5D). Moreover, 4 weeks of 1D11 antibody treatment not only significantly reduced GFP + cells in the prostate stroma, but also reversed the elevated GFP + CD45 2 cells in peripheral blood of the PE-induced Nestin-GFP mice (Fig. 5E, 5F). Taken together, these results demonstrate that inhibition of TGF-b activity could prevent prostate stroma expansion by reducing recruitment of MSCs.

Knockout of Tgfbr2 in Nestin + Cells Ameliorates Prostatic Stromal Hyperplasia
To further demonstrate that local excessive activation of TGF-b recruits nestin + cells to prostate stroma during pathogenesis of prostatic hyperplasia, we generated Nestin-Cre-ER, Tgfbr2 flox/flox mice by crossing Nestin-Cre-ER with Tgfbr2 flox/flox . Tgfbr2was deleted in the nestin + cells by intraperitoneal injection of tamoxifen to the Nestin-Cre-ER, Tgfbr2 flox/flox mice every other day starting with PE injection. Immunostaining of prostatic tissue sections showed that collagen deposition (Fig. 6A) and fibroblasts accumulation (Fig. 6B) in the stroma of Tgfbr2 2/2 mice were significantly reduced relative to their wild-type littermates (Tgfbr2 +/+ ). These Tgfbr2 knockout data confirmed that increased TGF-b activity in the stroma enhances recruitment of nestin + cells for prostatic hyperplasia.

DISCUSSION
The embryonic reawakening theory implicated a potential role of MSCs in the pathogenesis of BPH. Evidence has recently emerged to support this argument [12,35,[42][43][44]. Lin et al. [12] identified and isolated prostatic stromal cells from BPH specimens, which possessed strong proliferative potential and the ability to differentiate to myogenic, adipogenic, and osteogenic lineage. Ceder et al. [35] found a stromal stem/ progenitor cells population in both fresh and cultured human prostate tissue. This stromal population was positive for KIT, vimentin, CD133, and SCF, and negative for AC133 and Bcl-2. Brennen et al. [45] also demonstrated that the number of MSCs constitute 0.01%-1.1% of the total cells in human prostatectomy specimens. These cells expressed CD90, CD73, and CD105 in the absence of hematopoietic lineage markers using flow cytometry assay. However, studies of the physiological or pathological function of MSCs in prostate have been hampered by a lack of specific markers that permit both identification and linage tracing in vivo. Nestin is an intermediate filament protein described originally in neuronal stem cells during embryonic development and nerve injury [46]. Recently, nestin + cells have been shown to be a subpopulation of bone marrow derived MSCs and have the ability of self-renewal and multilineage differentiation [34,47,48]. Our study provides further evidence that nestin + cells isolated from mice prostate have the classic MSC properties, including CFU-F activity, fibrogenic, osteogenic, adipogenic and chondrogenic potentials. The main cell types in prostate stroma are smooth muscle cells and fibroblasts, both contributing to the synthesis of the extracellular matrix. Analysis of 6-week-old Nestin-Cre, Rosa26-YFP flox/+ mice revealed that nestin + lineage cells contribute significantly to prostate stromal formation physiologically including nearly 75% smooth muscle cells and 40% fibroblasts. Thus, nestin + MSCs are involved in prostate growth and stromal tissue remodeling.
MSCs have both profibrotic and antifibrotic effects in various tissues and organs including liver, heart, lungs, skin, and pancreas. During the past few years, an increasing number of studies have evaluated the antifibrotic potential of MSCs. Conditioned media from MSCs attenuated fibroblast proliferation and collagen synthesis in vitro. In vivo, MSCs have been shown to attenuate fibrosis in nephrectomized rats [49] as well as having antifibrotic actions in animal models of cardiac disease [50]. The propensity of MSCs migrate, proliferate and differentiate depending on the microenvironment surrounding them enhances their curative properties. The profibrotic role of MSCs finds usefulness in diabetic mice wound healing where aside producing ECM they contribute to wound healing by releasing various mediators [51]. Profibrotic actions of MSCs in mouse model of liver damage have been demonstrated [52]. MSCs migrated to sites during chronic rejection and differentiated toward a fibroblast phenotype in a rat model of chronic allograft heart rejection [53]. In murine lungs fibroblast proliferation and matrix production were induced by MSC conditioned medium [54]. In our study, we focused mainly on the recruitment of MSCs and their potential profibrotic role in BPH.
Bone marrow-derived MSCs were recruited and then incorporated into the prostate epithelia through fusion during the process of prostate regrowth [43,44]. Peng et al. [55] demonstrated sonic hedgehog signaling from prostate basal epithelial cells to adjacent smooth muscle cells in mature prostate. They performed lineage tracing in the stroma to address smooth muscle cells as largely self-sustained lineages during multiple rounds of castration-induced involution and androgen-mediated regeneration. Our parabiosis results demonstrated that nestin + MSCs recruited from circulation differentiate to interstitial fibroblasts, but not to smooth muscle cells in prostate stroma. These results suggest there is heterogeneity of origin of interstitial fibroblasts and smooth muscle cells during prostatic hyperplasia.
TGF-b signaling pathways are ubiquitous and fundamental regulators of cellular processes including proliferation, differentiation, migration, and survival, as well as physiological processes, including tissue development and wound healing [24]. The appropriate regulation of these pathways is essential at all levels, especially at the ligand level, with either a deficiency or an excess of specific TGF-b ligands resulting in human disease [31]. The critical role of TGF-b in prostatic stromal accumulation, which eventually leads to BPH, is well known [19,20,[56][57][58][59]. Aberrant activation of TGF-b canonical signaling has been observed in both rodent and human prostatic hyperplasia [18,40]. Recent studies showed that the murine pan-specific TGF-b neutralizing monoclonal antibody, 1D11, had the effect of pharmacological inhibition of TGF-b in different mouse models, including osteogenesis imperfecta [60], chronic nephropathy [61], and asthma [30]. In humans, fresolimumab, similar to TGF-b antibody 1D11 in specificity, has been used for clinical trials in patients with idiopathic pulmonary fibrosis [62], primary focal segmental glomerulosclerosis [63] and cancer [64]. Our study demonstrated that injection of 1D11 effectively reduced TGF-b signaling, fibroblast numbers, and collagen deposition in prostatic hyperplasia mice with an accompanying reduction of the nestin + cells in both prostatic stroma and blood. Knockout of Tgfbr2 in nestin + cells attenuated the stromal accumulation in PEinduced mice prostate. Clearly, upregulated TGF-b signaling contributed to the higher number of nestin + cells. The elevated levels of active TGF-b mobilized MSCs from circulation to the stroma during the process of prostatic hyperplasia. Thus, targeting TGF-b mediated recruitment of MSCs may

CONCLUSION
Activation of TGF-b plays a role in the pathogenesis of benign prostatic hyperplasia. TGF-b activation results in recruitment of MSCs and their subsequent differentiation into prostatic stromal cells. Thus, careful modulation of TGF-b activity could provide potential therapeutic options in the management of benign prostatic hyperplasia.