| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Research Article |
Department of Urology,3 UCLA School of Medicine, Los Angeles, California 90095
Los Angeles Biomedical Research Institute at Harbor-UCLA Medical Center,4 Division of Urology, Torrance, California 90509
ABSTRACT
Tissue ossification in Peyronie disease (commonly known as Peyronie's disease [PD]), a localized fibrotic lesion within the tunica albuginea (TA) of the penis, may result from osteogenic differentiation of fibroblasts, myofibroblasts, and/or adult stem cells in the TA, and may be triggered by chronic inflammation, oxidative stress, and profibrotic factors like transforming growth factor beta 1 (TGFB1). In this study, we have investigated whether cultures of cells from normal TA and PD plaques undergo osteogenesis, express markers for stem cells, and originate other cell lineages via processes modulated by TGFB1. We found that TA and PD cells in osteogenic medium (OM) expressed osteogenic markers, alkaline phosphatase, and osteopontin and underwent calcification. PD cells, but not TA cells, formed foci in soft agar that were positive for alkaline phosphatase and calcification and expressed the mRNAs for osteoblast-specific factors pleiotrophin and periostin and bone morphogenic protein 2. Both cultures expressed stem cell marker CD34 antigen but not protein tyrosine phosphatase, receptor type c. TA and PD cells expressed smooth-muscle cell markers smoothelin and transgelin. None of the cultures underwent adipogenesis in adipogenic medium. Incubation with TGFB1 increased osteogenesis and myofibroblast differentiation and reduced CD34 antigen expression in both cultures. TA and PD cells modulated the differentiation of the multipotent C3H 10T(1/2) cells in dual cultures, into osteoblasts and myofibroblasts. In conclusion, both TA and PD cultures contain cells, presumably stem cells, that undergo osteogenic and myofibroblast differentiation, and may induce these processes by paracrine interactions. This may explain progression of fibrosis in the PD plaque and its eventual calcification.
male reproductive tract, male sexual function, penis
INTRODUCTION
Peyronie disease (commonly known as Peyronie's disease [PD]), a condition affecting from 5% to 9% of men, is characterized by the development of a fibrotic localized plaque in the tunica albuginea (TA) of the penis. The plaque is assumed to arise from abnormal wound healing after trauma during coitus and can lead to a deformation of the erect penis, localized pain, and erectile dysfunction [1, 2]. In both humans and in animal models of PD, such trauma to the TA leads to the accumulation of fibrin, production of reactive oxygen species (ROS), and the expression of other profibrotic factors, mainly transforming growth factor beta 1 (TGFB1) and plasminogen activator inhibitor type 1 (SERPINE1, formerly PAI-1), causing increased deposition and disorganization of collagen fibers in the TA [3–10].
One of the key cellular events in abnormal wound healing and tissue fibrosis, in general, is the differentiation into myofibroblasts of either the fibroblasts per se or putative multipotent stem cells not yet identified, triggered by oxidative stress and TGFB1 [11, 12]. Myofibroblasts, which share the phenotype of both the fibroblast and smooth-muscle cell, originate during wound healing to help repair the wound and they normally disappear by apoptosis after the injury is repaired. If these cells persist, they would continue to actively produce collagen, and this may lead to a localized scar. Besides wound healing, myofibroblasts have been shown to play a critical role in the development of tissue fibrosis [13–15], and in certain organs, such as in the liver, myofibroblasts originate from intermediate progenitors (stellate cells) [15]. In the penis, myofibroblasts are virtually absent in normal human TA, but abundant in the PD plaque of both men and experimentally induced animal models of PD [6–9].
The progression of the PD plaque may eventually lead to calcification or ossification in 15–25% of patients [2, 3, 16–18]. To date, the mechanism of this process remains unknown. However, by analogy to certain processes of ectopic ossification that have been better characterized in other tissues, as in arterial wall calcification [19, 20] or in some ossified traumatic lesions in the skeletal muscle [21], it may be assumed that ossification of the PD plaque is a late step in the progression of fibrosis, caused by the generation of osteoblasts, from either progenitor cells not yet identified or possibly from multipotent cells. Ossification and calcification of the TA is highly prevalent in non-PD patients with end-stage renal disease and may present rarely in the penile corpora cavernosa [22, 23] and in other urogenital organs, such as the normal prostate and seminal vesicles [24–26], and more characteristically, in primary tumors of the testis and ovary [27, 28]. In some of these cases, the associated calcinosis may be a trigger of the cell-differentiation processes occurring in the target tissues.
Tissue ossification outside the bone appears to be mediated via the activation of the same gene families and pathways operating in osteoblast differentiation in the skeleton, namely, bone morphogenic proteins (BMPs), osteoblast-specific factors POSTN, RUNX2, MSX2, SP7, NOG, and others [29–33]. These proteins in turn modulate the expression of more downstream osteogenesis protein markers, such as alkaline phosphatase (ALPL), osteopontin (SPP1), and osteocalcin (BGLAP). However, there seem to be differences in the abnormal ossification of the vascular wall as compared with bone formation in the skeleton [19, 20].
One of the main questions is whether the two sequential processes in PD progression, fibrosis and ossification, are elicited by the differentiation into myofibroblasts and osteoblasts of separate progenitor fibroblasts, or of a common multipotent cell in the TA that, in addition, may initiate other cell lineages. The existence of adult or postnatal stem cells with multipotent capability is not restricted to the bone marrow because, in the last few years, it has become evident that they are present in a number of organs and tissues, such as skeletal muscle, pancreas, central nervous system, connective tissue, liver, heart, and others [34–36]. These cells appear to constitute a reservoir that, on environmental cues, mostly of paracrine origin, can elicit differentiation into several types of lineages, particularly during tissue regeneration after injury. TGFB1, the key factor in fibrosis, is one potential paracrine modulator of stem cell commitment because it is known to stimulate the conversion of fibroblasts into myofibroblasts [11, 12, 37], inhibit myogenesis [37], and promote osteogenesis [38].
The pathophysiology of calcification and ossification in the PD fibrotic plaque has so far remained virtually unexplored because it is the late stage of development in a non-life-threatening localized plaque that can be surgically removed. However, it does have potential implications for understanding a much more significant problem, vascular calcification, a highly prevalent complication of diabetes and aging and predictor of cardiovascular morbidity and mortality [19, 20]. The penis is an extension of the vascular tree, and the TA may be considered, in terms of its interaction with the corporal smooth-muscle tissue, as a sort of functional equivalent to the adventitia/media cross talk in the arterial wall. Indeed, the arterial adventitia also harbors fibroblasts, myofibroblasts, and putative stem cells involved in ossification [39].
In the present work we have investigated whether fibroblasts cultured from both the normal human TA and the PD plaque can undergo differentiation into osteoblasts, in addition to their known ability to develop into myofibroblasts, and can even evolve into other cell lineages, in a process modulated by TGFB1. We have also studied whether TA and PD cells can paracrinely modulate the differentiation of an established mesenchymal multipotent cell line into cell lineages similar to those originated from the PD cultures by themselves and whether stem cells are present in these TA and PD cultures that may explain their multiple differentiation capability.
MATERIALS AND METHODS
Incubation of Fibroblast Cultures from Normal TA and Fibrotic PD Plaque
Fibroblast primary cultures were obtained from fragments of human TA from non-PD patients (n = 3), undergoing penile prosthesis surgery, and from plaque tissue isolated from PD patients (n = 4) who underwent a surgical procedure to treat this condition [6, 7, 40, 41]. Written informed consent was obtained under institutional review board approval. Tissues were washed in Hanks solution, minced in a fibroblast growth medium-2 (FGM; Cambrex Inc., Walkersville, MD) and 20% fetal bovine serum, and plated onto a 25-cm2 culture flask per specimen. Fragments were left undisturbed until attachment for about 1 wk, and once the monolayer was starting to develop, they were removed. Medium with 10% serum was then changed once a week, and when cells achieved approximately 80% confluence (3–4 wk), they were trypsinized and split onto three 10-cm plates. Cells were allowed to grow again to 80% confluence, with medium changed twice a week. The cells collected from this passage were considered as passage 1. Successive passages were performed at 1/3 split ratio and experiments were performed with cells from passages 5 on. The purity of these cultures was established by immunocytochemistry for the fibroblast marker vimentin that showed 100% staining as described [6].
After tyrosinylation and centrifugation, the cell pellet from one plate was suspended and plated at 25–35% confluence on 8-well removable chamber plates (for immunocytochemistry), 12-well plates (for protein homogenates), or 6-well plates (for RNA isolation) and allowed to grow in either FGM, osteogenic medium (OM), or adipogenic medium (AM), supplemented with 10% fetal bovine serum [42]. In some cases, the incubations for immunocytochemistry were performed separately from those for protein homogenates or RNA isolation. OM consisted of Dulbecco modified essential medium (DMEM) with 0.1 µM dexamethasone, 50 µM ascorbate-2 phosphate, 10 mM ß-glycerophosphate. AM consisted of DMEM with 0.5 mM isobutylmethyl xanthine (IBMX), 1 µM dexamethasone, 10 µM insulin, for the indicated periods (usually 2 or 4 wk). The standard DMEM was used for experiments where smooth-muscle cell differentiation or cell cocultures were involved. In certain experiments, TGFB1 was added at the indicated concentrations. All experiments were duplicated or triplicated.
Clonogenic Selection of Multipotent Cells
TA and PD cells were trypsinized and subjected to the soft agar tumorigenic/stem cell selection procedure [43]. Cells were suspended at several concentrations in warm (37–39°C) 0.33% agar in FGM-10% fetal bovine serum (soft agar layer), and 0.3 ml (2000–200000 cells) were deposited in duplicate for each cell input on top of a bottom layer of 0.4 ml of 0.5% agar in the same medium that had been allowed to solidify on 24-well plates at 4°C (hard agar layer). Cultures were allowed to proceed for 3 wk, and when foci were visible, they were stained with 0.005% crystal violet in Hanks solution for 1 h, and colonies were counted. In certain cases, separate foci (n = 6) from the PD soft agar culture not subjected to staining were picked up and transferred to FGM-10% fetal bovine serum and cultured on T-25 flasks. After colony amplification and trypsinization, each clone was either cultured on 8-well removable chamber plates, 12-well plates, or 6-well plates, as above.
Dual Cultures of PD Cells and a Multipotent Cell Line
The mouse mesenchymal embryonic cell line C3H 10T(1/2) [44, 45] was cultured in 25 cm2 flasks with DMEM-10% fetal bovine serum, the medium routinely employed for these cells. At approximately 60% confluence, cells received azacytidine at 20 µM or were left untreated, and then incubated for 3 days. Cells were washed, medium replenished without azacytidine and then submitted to dual culture [46] by plating them in triplicate onto the bottom outer chamber (cluster) of the Costar Transwell Permeable Support (12-µm pore size) 12-well plates at 80% confluence. The top inner chamber (transwell or insert) received the TA and PD cells at 40% confluence or no cells (controls). Medium was either OM or FGM in two separate series. Incubation proceeded for 2 wk and, at completion, the target cells on the triplicate wells of the bottom chamber were trypsinized and pooled and seeded onto 8-well removable chamber plates or 12-well plates at 60–100% confluence and allowed to continue for 1 day.
Quantitative Immunocytochemical Determination of Cell Markers
At completion of incubations, chambers were removed from the eight-well plates and cells were fixed for 20 min in 3% buffered formalin at room temperature [6, 7]. For myofibroblast identification, endogenous peroxidase was quenched and cells blocked with normal goat serum and incubated with a monoclonal primary antibody against ACTA2 protein (actin, alpha 2 smooth muscle, aorta) (Sigma Immunohistology Kit, Sigma Chemical Co, St. Louis, MO), followed by a biotinylated secondary antibody with ExtrAvidin-conjugated peroxidase and 3-amino-9-ethylcarbazol (AEC) chromogen, processing according to the manufacturer's instruction. Negative controls omitted the first antibodies or were replaced by IgG isotype at the same concentration of the first antibodies. Fixed cells were also subjected to alkaline phosphatase enzyme activity detection using nitro blue tetrazolium (NBT) [47] or to adipocyte identification with 0.3% Oil Red O (44). For detection of calcification, cells were subjected to the von Kossa method [42]. Counterstaining was done with Mayer hematoxylin for ACTA2 or with nuclear fast red (alkaline phosphatase) or omitted. All the slides were mounted with Aqua Mount (Lerner, Pittsburgh, PA).
Immunoreactivity or staining intensity was quantified by computerized densitometry using the ImagePro 401 software program (Media Cybernetics, Silver Spring, MD) coupled to an Olympus BHS microscope equipped with a Spot RT digital camera [4–9]. The number of positive cells was counted in a computerized grid against the total number of cells determined by counterstaining, and results were expressed as a percentage of positive cells over total cells. In other cases, the integrated optical density (IOD) was obtained by measuring the density per object and multiplying it by the respective area. Five fields were measured per well (20x total magnification or as indicated) on duplicate wells and duplicate series.
Identification of Cell Markers by Western Blot
Cell extracts were obtained by adding to each well in the 12-well plate 0.4 ml of a buffer containing 1% sodium dodecylsulphate (SDS), 20 mM HEPES (pH 7.2), 0.5 mM EDTA, 1 mM dithiothreitol, and protease inhibitors (3 µM leupeptin, 1 µM pepstatin A, 1 mM phenylmethyl sulfonyl fluoride) preheated at 95°C, and scrapping off the cell extracts. Suspensions were heated again at 95°C for 5 min and clarified by centrifugation at 15000 x g for 15 min, estimating protein concentration in the supernatant [4–9]. Equal amounts of protein (20–30 µg) were run on polyacrylamide gels (acrylamide concentrations as indicated in each case), and submitted to Western blot immunodetection with antibodies against the following proteins: ACTA2 (actin, alpha 2 smooth muscle, aorta; as above, for myofibroblasts and smooth-muscle cells), SMTN (smoothelin; monoclonal, Abcam, Cambridge, MA, exclusively for smooth-muscle cells), SPP1 (osteopontin; polyclonal, Alexis Biochemicals, Carlsbad, CA, for osteogenesis), ALPL (alkaline phosphatase; polyclonal, USBiological, Swamscott, MA, for osteogenesis), and CD34 antigen and PTPRC (protein tyrosine phosphatase, receptor type c, previously referred to as CD45; both polyclonal, Santa Cruz Biotechnology, Santa Cruz, CA, for stem cells), followed by a luminol reaction. Negative controls were performed without primary antibody. Gels were reprobed for GAPDH (glyceraldehyde-3-phosphate dehydrogenase; monoclonal, Chemicon, Temecula, CA, a housekeeping gene). Quantitative determinations were performed by densitometry of the corresponding band intensities, normalized against the respective intensities for GAPDH.
Identification of Cell Markers by Reverse Transcription-Polymerase Chain Reaction
Total RNA was isolated from the human TA and PD tissues, from their respective fibroblast cultures by the Trizol procedure (Gibco BRL, Gaithesburg, MD). RNA was then submitted (1 µg) to reverse transcription (RT) [7, 8, 40, 41] using Superscript II RNase H– reverse transcriptase (Gibco BRL) and random primers (0.25 µg). This was followed by polymerase chain reaction (PCR) with the respective gene specific primers: a) PTN (pleiotrophin, previously referred to as OSF-1) on nt 39 (forward; AAATCCCGCCAAGAGAGCCC) and nt 420 (reverse; GCACACA CACACTCCACTGC) of the respective cDNA (GenBank # 52946); encompassing a 382-bp band; b) POSTN (periostin, previously referred to as OSF-2) on nt 1001 (forward; GATAGGATGTGACGGTGACA) and nt 1504 (reverse; GATTTCTCTGCTGGCTTGAT) of the respective cDNA (GenGank # NM_006475), generating a 485-bp band; c) BMP2 (bone morphogenic protein 2), on nt 504 (forward; TTGGCCTGAAACAGAGACCC) and nt 882 (reverse; TGTCCAAAAGTCTGGTCACGGG) of the respective cDNA (GenBank # NM_001200), generating a 400-bp band; d) TAGLN (transgelin, previously referred to as SM22), on nt 681 (forward; TGGCTGAAGAATGGCGTGAT) and nt 974 (reverse; GTCCCTGCGCTTTCTTCATA) of the respective cDNA (GenBank # NM_ 001001522), generating a band of 313 bp. PCR products obtained after 32 cycles at 94°C for 1 min denaturing, 60°C for 1 min annealing, and 72°C for 1 min extension, were separated by electrophoresis on 1% agarose gels and stained with ethidium bromide. For densitometry, normalization was performed against the GAPDH PCR fragment generated with primers on nt 75 (forward; CATGGGGAAGGTGAAGGTCG) and nt 1083 (reverse; TTACTCCTTGGAGGCCATG) (GenBank # XM006959), generating a 1009-bp band in the same PCR reaction. The number of cycles and concentration of primers were determined in preliminary experiments to maintain amplification of each PCR product within the linear range.
Statistical Analysis
Values were expressed as mean ± SEM. The normality distribution of the data were established using the Wilk-Shapiro test, and the outcome measures between two groups were compared by the paired t-test. Multiple comparisons among different groups were analyzed by a single-factor ANOVA, followed by post hoc comparisons with the Student-Neuman Keuls test, according to the Graph Pad prism V. 30. Differences among groups were considered significant at P < 0.05.
RESULTS
Fibroblast Cultures from the Normal and Fibrotic TA Contain Osteogenic Progenitor Cells
Cells cultured from the normal TA (Fig. 1A) or the fibrotic PD plaque (Fig. 1B) cultured in regular FGM showed osteogenic transformation at a relatively early period (2 wk), as evidenced, without counterstain, by dispersed zones of faint blue staining for ALPL activity, an early marker of osteogenesis, which were abundant and intense in the PD cells. In OM, a medium that is designed to stimulate osteogenic differentiation, the staining intensified considerably in the PD culture, and less in the TA cells, with blue cells growing in concentric circles, and much larger positive zones than in TA. Some areas in the culture were negative for alkaline phosphatase activity. This suggested that progenitor cells able to undergo osteogenic differentiation were present in the PD culture and to a lesser extent in the TA culture.
|
When cultures were allowed to proceed for 4 wk in OM, cells underwent calcification in certain areas, as shown by Von Kossa staining (Fig. 1A and B). The staining was visible in the TA cells, but was, as with alkaline phosphatase, less intense than in PD cells, and also restricted to smaller patches throughout the monolayer. A higher magnification in those areas in the PD culture with fewer positive cells allowed detecting some discrete Ca2+ deposits inside the cells (not shown).
To investigate whether this osteogenic conversion occurred in stem cells or cells with multipotent capacity, TA and PD cultures in FGM were subjected to the soft agar clonogenic assay, a system used to select proliferating hematopoietic stem cells or neoplastic cells by their survival under conditions that eliminate most other types of cells [43]. TA and PD cultures in OM were seeded on the soft agar layer and, after 3 wk, scattered foci composed of >200 cells were visible with the naked eye in the PD wells, but not in the TA wells, where less than 10–20 cells grouped sporadically. When the PD foci were picked up and cultured on monolayer in OM for 2 wk, they showed a considerable intensification of ossification as shown by alkaline phosphatase activity, and the concentric growth pattern expanded over the well at 4 wk. In several instances, the typical mineralized nodules could be seen, with alkaline phosphatase staining at the borders (Fig. 1C, both panels).
The presence of progenitor osteogenic cells in the PD cultures and their enrichment by growth in soft agar was confirmed by RT-PCR of RNA isolated after 4 wk of growth in six selected clones (Fig. 2). The mRNAs for the osteogenic markers PTN1, POSTN, and BMP2 (all positive in RNA from fetal human osteoblasts, not shown), were expressed in both FGM and OM, but to a higher extent in the latter medium.
|
Osteogenic Commitment of TA Fibroblast Cultures Is Stimulated by TGFB1, and These Cultures Express Stem Cell Markers
TGFB1 is assumed to be involved in osteogenesis [38, 48], and it is an important factor in fibroblast differentiation in vivo in the PD plaque and in fibrosis [3–7, 9], as well as in the modulation of stem cell lineage in other tissues and cell cultures [49]. Therefore, we tested whether this factor intensifies the osteogenic transformation in OM of TA and PD plaque fibroblasts, using new cultures obtained from other patient specimens. Figure 3 (left panels) shows that, in FGM without TGFB1 supplementation, there is basal expression of the 70-kDa ALPL protein alkaline phosphatase band detected by Western blot in extracts of TA and PD cells at 2 wk, in agreement with Figure 1. This was not substantially increased by prolonging the incubation to 4 wk. The 70-kDa band was intense in the control fetal osteoblast cell (FOB) extract. Surprisingly, the upregulation of alkaline phosphatase activity by switching cells to OM, seen in Figure 1, was not observed at the level of alkaline phosphatase protein expression by Western blot. Adding TGFB1 (5 ng/ml) to the OM stimulated the levels of this osteogenic marker, particularly in PD cells after 4 wk.
|
We examined next the levels of the SPP1 protein (osteopontin), a negative regulator of osteogenesis, that is posttranslationally modified to generate bands of differing sizes [50]. The 66-kDa band was only faintly expressed in fetal osteoblasts and was very low in TA and PD cells in both FGM and OM in a 2-wk incubation (Fig. 3 right panels). However, a smaller band (33 kDa), absent in the fetal osteoblasts, was much more expressed in both cultures and both media at 2 wk. The 66-kDa SPP1 band was intensified at 4 wk, particularly in OM. TGFB1 (5 ng/ml) virtually did not affect expression of the 66-kDa band in OM at any period or type of culture, but did reduce the levels of the 33-kDa band in TA and PD cells at both periods. This suggests that an early reduction of the osteogenesis inhibitor allows the initial stimulation of osteogenesis and that TGFB1 downregulates SPP1 to stimulate osteogenesis.
The demonstration of osteogenic differentiation in the TA and PD cultures posed the question as to whether this is due to the presence of stem cells in the culture that can initiate several cell lineages. As a preliminary approach to this issue that would explain the ability of cells to form colonies in soft agar as shown in Figure 1, the expression of two stem cell markers, protein tyrosine phosphatase, receptor type c (PTPRC) and CD34 antigen, was examined by Western blot. PTPRC (previously referred to as CD45) was negative (not shown), but the 97 kDa CD34 antigen protein was strongly expressed in TA and PD cells in FGM at 2 wk, suggesting the presence of stem cells (Fig. 4, top). This assumption was supported by the fact that further incubation (4 wk), which allows for more osteogenic transformation, led to downregulation of expression of CD34 antigen, and that incubation of these cultures in OM, reduced it even more. When TGFB1 was added to increase differentiation, there was downregulation of CD34 antigen expression at 2 and, more considerably, at 4 wk, at which time the stem cell marker reached negligible levels in the PD cells. The identity of the smaller size CD34 antigen band seen particularly at 4 wk in the TA cells, but also in PD cells, is unknown. This suggests that osteogenic differentiation occurs with depletion of CD34 antigen+/PTPRC– stem cells.
|
Tunical and PD Cultures Can Also Differentiate into Myofibroblasts and Smooth-Muscle Cells in a Process Modulated by TGFB1 but Do Not Undergo Adipogenesis
The possible presence of stem cells, or multipotent cells, was supported by the fact that OM stimulated the early appearance of another cell lineage, the myofibroblast, which our previous work had shown to occur in vivo and in vitro in FGM [6]. OM intensified the early expression (2 wk) of the myofibroblast cell marker, ACTA2 (actin, alpha 2, smooth muscle, 43-kDa band), rather considerably in PD cells, and a longer time (4 wk) upregulated further this process (Fig. 5). Incubation with TGFB1 considerably stimulated ACTA2 protein expression in both TA and PD cells in OM. Although FGM was much less effective than OM in inducing myofibroblast differentiation, addition of TGFB1 enhanced ACTA2 expression in this medium if sufficient time (4 wk) was allowed to elapse. In fact, in PD cells at this period, ACTA2 expression reached about the same levels irrespective of whether FGM or OM was used.
|
Although ACTA2 protein expression as stated above may indicate mainly the differentiation of cultures previously identified as fibroblasts because of vimentin (VIM) protein expression [6, 51], into myofibroblasts, it may also suggest the generation of another lineage, smooth-muscle cells. This, so far, has not been demonstrated with TA or PD cultures, or with myofibroblasts from other tissues, and would further support the concept that stem cells are present. The detection in TA and PD cell extracts of the 59-kDa band of the smoothelin (SMTN) protein (Fig. 6A), a late marker of contractile smooth-muscle cells that is absent in myofibroblasts [52], suggests that the TA cells can evolve into smooth-muscle cells with prolonged incubation in FGM and OM. TGFB1 protein moderately stimulated expression in this experiment. The 50-kDa band corresponds to SMTN seen in visceral smooth-muscle cells, whereas the very faint 50-kDa band was expressed in the control vascular smooth-muscle cells. This indicates that a small fraction of the ACTA2 expression detected in Figure 5 may actually pertain to smooth-muscle cells, in addition to the one in myofibroblasts. Another marker of smooth-muscle cells was examined, TAGLN (transgelin, previously known as SM22), which is also expressed in some types of myofibroblasts. In this case, we used RT-PCR that showed the expected 313-bp band in FGM in both TA and PD, both at 2 and 4 wk (Fig. 5B). The differentiation also occurred in DMEM, a medium routinely used for smooth-muscle cell culture.
|
Because the detection of markers for three different cell lineages (osteoblasts, myofibroblasts, smooth-muscle cells) evolving with time of incubation and differing according to the type of medium and the response to TGFB1 supported the view that this may originate from putative CD34 antigen+/PTPRC– stem cells, we examined whether a fourth cell type, the adipocyte, may be also initiated. Adipogenesis is a process relatively common in many cell cultures and usually occurs at much earlier periods (1 wk after azacytidine or switch to nonstandard medium). TA cells were incubated in FGM, OM, DMEM, and AM and observed under the microscope without staining for the accumulation of fat droplets, or subjected to staining with Oil Red O that detects adipocytes. Even after 4 wk, virtually no clearly identifiable adipocytes were found, although some cells appeared to contain a few droplets or vacuoles (not shown). Because a positive control, the mouse C3H 10T(1/2) cell line, fibroblast-like cells of embryonic and mesenchymal origin [45], underwent a strong adipogenic conversion upon azacytidine induction (20–50% of the population), even in FGM or OM, we interpreted these results as indicating that the TA cells do not undergo adipogenesis.
The Comparison of Cell Lineage Marker Levels Suggests that the Differentiation of TA and PD Culture Is Intrinsically Similar but with Some Quantitative Differences
The 4-wk incubations and subsequent Western blot analysis of lineage markers presented above were replicated 2–3 additional times in OM and FGM, and quantitative densitometry of band intensities was applied to all experiments, correcting by the respective housekeeping protein (GAPDH) band intensity to compensate for variations in protein loading or transfer, and finally normalized for the expected variations in cell culture responsiveness and immunodetection conditions by dividing by the respective basal value in the reference medium (FGM).
Figure 7 shows that, unexpectedly, OM was not more effective than FGM in stimulating osteogenic differentiation of TA and PD cells, and was less effective stimulating expression for SMTN and ACTA2 proteins. However, TGFB1 did considerably and significantly stimulate ALPL protein expression as a marker of osteoblast differentiation in TA and particularly in PD, in agreement with the experiment represented in Figure 3, whereas the mean for SPP1 (osteopontin) protein exhibited only a moderate trend to downregulation rather than the more marked reduction of Figure 3. The decrease of CD34 antigen protein by TGFB1 in TA and particularly in PD, presented in Figure 4, was confirmed statistically. The same occurred with the very considerable and significant induction of ACTA2 in TA by the growth factor, thus agreeing with Figure 4, whereas the mean increase in PD was much less pronounced. Finally, the differentiation into smooth muscle cells of both the TA and PD cells seen in Figure 5 was confirmed, although the observed stimulation by TGFB1 was not reproducible and effects were not significant.
|
PD Fibroblast Cultures Can Paracrinely Induce Osteogenesis and Myofibroblast Differentiation of Other Mesenchymal Multipotent Cells While Inhibiting Adipogenesis
To determine whether the osteogenic and myofibroblast differentiation of PD cells may involve the secretion of factors acting paracrinely on cells to induce lineage commitment, we investigated the effects of coculturing these cells with the target C3H 10T(1/2) cell line. These target cells, on incubation with azacytidine as a demethylating agent, undergo differentiation into several lineages, mainly myotubes, adipocytes, chondrocytes, and osteoblasts, whereas in the absence of azacytidine, the differentiation is low or does not occur [45]. Azacytidine-induced and untreated C3H 10T(1/2) cells were plated on the bottom outer compartment of a dual chamber system with a porous separation membrane not allowing cell-cell contact between compartments, and TA and PD cells were seeded in the top inner compartment. Control cultures of C3H 10T(1/2) cells were run without cells on top. This device allows mutual paracrine effects to be exerted between the top and bottom cultures. The cultures were then trypsinized, centrifuged, resuspended, seeded for 1 day in OM onto removable chamber plates, and subjected to immunocytochemistry and histochemistry. The first row of micrographs in Figure 8 show the considerable modulation exerted by TA and PD cells on the phenotype of the target differentiating C3H 10T(1/2) cells in OM. The latter azacytidine-treated cells maintained in monoculture for 2 wk and not trypsinized underwent what appears to be an aborted or artifactual osteogenesis, with many cells staining intensively for ALPL protein, but appearing rounded, without the osteoblast morphology and loaded with fat droplets, suggesting concurrent adipogenesis (not shown). However, when cells were subjected to trypsinization, followed by centrifugation that eliminates adipocytes, and were replated for 1 day, only the true non-fat-loaded osteoblasts and the undifferentiated cells attached. Under these conditions, some azacytidine-treated target cells (C), and a few in the case of untreated target cells (not shown), showed alkaline phosphatase activity as judged by the blue color, but when TA or PD cells were present, the number of cells in the azacytidine-treated cultures was increased, and, in the case of at least the PD cells, the intensity of the staining was also higher. The same occurred with the untreated cultures (not shown).
|
Alternative removable chambers from the experiment described above were stained for alkaline phosphatase to determine the paracrine effects exerted by TA and PD on myofibroblast/smooth-muscle differentiation. The bottom row of micrographs in Figure 8 show that the differentiating target azacytidine-treated C3H 10T(1/2) cells became mostly positive for ACTA2 protein expression, indicating the fibroblast/myofibroblast transition (C). When these same target cells were not subjected to adipocyte depletion by trypsinization and sedimentation, they contained numerous ACTA2+ degenerating fat-loaded cells that lacked the myofibroblast morphology (not shown). A distinctive difference was detected between the paracrine effect exerted by the TA cells that did not affect or stimulated ACTA2 protein expression, and PD cells that considerably reduced it. Both TA and PD cells drove the target C3H 10T(1/2) cells to a true myofibroblast morphology with the typical actin filaments.
When the target azacytidine-treated control cells (no cells on top) that were directly plated for 2 wk and not subjected to trypsinization and centrifugation were treated with Oil Red O, the fat droplets were intensively stained in red (not shown). However, virtually no staining was observed in the control target cells (no cells on top) or in the target cells cocultured with TA and PD cells, where adipocytes had been eliminated (not shown). This confirms the above results with ALPL and ACTA2 protein staining and extends the previous observations in monocultures of PD cells (see above) that were unable to undergo adipogenesis, even in AM. Essentially the same paracrine modulation of osteogenesis and myofibroblast differentiation of the azacytidine-treated and untreated target cells was observed in FGM (not shown), thus indicating that the additives in OM are not essential or may be produced by the cultures. Similar treatments with human penile corpora cavernosa smooth-muscle cells on top in the dual culture in OM did not show any significant modulation of the target azacytidine-treated cells, thus showing the specificity of the response (not shown).
Figure 8 graphs summarize the data obtained by quantitative image analysis from duplicate experiments in terms of the expression of the respective marker in the target azacytidine-treated and untreated C3H 10T(1/2) cells in OM. The graphs show that, once all the abnormal fat-loaded positive cells seen in direct monocultures of the target cells were eliminated by trypsinization and sedimentation at completion of the dual cultures, the paracrine modulation exerted by both the TA and PD cells could be cleanly ascertained because, as seen above, only true osteoblasts and myofibroblasts remained plated.
In monocultures of target cells not treated with azacytidine, the expression of alkaline phosphatase was seen in only 5% of the plated cells, but azacytidine, as expected, increased them to 20%. Dual culture with either TA or PD considerably and significantly increased the osteogenic differentiation, reaching in some cases 6–7% of the cells. The intensity per cell was also increased in the untreated dual cultures with TA and PD and in the azacytidine-treated target cells in the presence of PD cells. A similar positive modulation of untreated target cells by coculturing with TA cells was seen for ACTA2 staining, reaching close to 95% of positive cells versus 55% in the monocultures, and also intensifying the IOD/cell, thus indicating a strong stimulation of myofibroblast differentiation. The morphology showed the typical actin filaments. In contrast with osteogenesis, PD cells considerably reduced the number of ACTA2+ cells and the IOD/cell in both types of target cells (azacytidine-treated and untreated).
DISCUSSION
In this work, we provide the first evidence that the PD fibrotic plaque tissue, as well as the normal penile tunica albuginea, harbors progenitor cells that, in culture, can differentiate into different lineages, namely osteogenic cells and possibly smooth-muscle cells, in addition to the known ability of some of these cells to differentiate into myofibroblasts, the key cells in the development of fibrosis [1, 2, 4–9]. A fraction of cells in these cultures are also potentially tumorigenic or stem cells because, in the case of the PD cultures, they can grow in soft agar and, after this selection, the stem cells still undergo osteogenesis and calcification. In addition, both TA and PD cultures express stem cell markers. The differentiation of these progenitor cells from the fibrotic plaque and the normal tunica into putative osteoblasts and well-characterized myofibroblasts is stimulated by the main profibrotic factor, TGFB1 [1, 2]. Therefore, there does not seem to be a major intrinsic difference between TA and PD cells in terms of their ability to differentiate. However, both in vivo and in vitro, PD cells may be more active in osteogenesis because of either an autocrine loop, the presence of unknown progenitors in a more advanced stage of differentiation, and/or a more intense paracrine modulation by factors such as TGFB1. In fact, we have shown that both the TA and PD cultures can paracrinely stimulate and even elicit the osteogenic and myofibroblast lineage transformation of a mesenchymal multipotent cell line, the C3H 10T(1/2) cells.
In addition to the already defined role of these tunical fibroblast cultures in generating myofibroblasts, a process that is assumed to mimic the one occurring in the development of the fibrotic PD plaque in the TA of the penis, these cells (already present in the normal TA) are probably responsible for the plaque ossification observed during its fibrotic progression in some patients with PD [1, 2, 17, 18]. In addition, because both TA and PD cells at 4 wk express low levels of smooth-muscle cell markers, the tunical fibroblasts may speculatively constitute in vivo a reservoir to replace smooth-muscle cells in the corpora cavernosa that may be damaged or lost during certain processes, e.g., aging and diabetes [53, 54]. The fact that some cells in the PD fibroblast cultures can be selected by soft agar and cells expressing the stem cell marker CD34 antigen have been identified in TA and PD cultures, allowed us to hypothesize that multipotent cells may be present. Whether these putative stem cells [34–36], or multipotent progenitor cells, originate into the different lineages we have observed, still remains to be determined by cell cloning or dual-marker immunostaining.
The ability of the TA and PD cells to undergo osteogenesis has been demonstrated in this work by the detection of alkaline phosphatase activity and ALPL protein, the expression of some osteogenic genes (BMP2, PTN, POSTN, SPP1), evidence of calcification, presence of nodules similar to the mineralized nodules occurring in osteoblast cultures, and the morphological appearance of some areas in the monolayer that resembles the concentric disposition of ossifying cells [30, 38]. These observations taken collectively suggest that the cells exhibiting these features are true osteoblasts, but further characterization is needed to conclusively prove this assumption. However, it is clear that the cultures from the human penile TA can mimic the process that occurs in vivo in about 15–20% of the PD plaques, (e.g., a condition consisting of their evolution into very hard and bone-like tissues). Ultrasound and x-rays have shown that calcification is relatively common and in one recent report reached 27% of the palpable, ultrasound-positive plaques, a prevalence similar to the one described in a previous microscopic study [17, 18].
Our observations showing that TGFB1 stimulated in vitro osteogenesis and particularly that both TA and PD cells paracrinely modulate the osteoblast phenotype (e.g., alkaline phosphatase/morphology) in multipotent C3H 10T(1/2) cells may suggest that plaque ossification proceeds through osteogenic differentiation during fibrosis progression of progenitor cells dormant in the normal TA. This differentiation may be activated by chronic inflammation, oxidative stress, and fibrosis, that would upregulate TGFB1, BMP2, BMP4 (bone morphogenic protein 4), MSX2 (msh homeo box homolog 2), and other cytokines and proteins [20, 29–31, 33, 48]. In other words, progenitor or stem cells do exist in the normal TA that may be identical to those found in PD, but their differentiation into lineages associated with fibrosis and ossification may be dependent on their activation by the more intense secretion of these cytokines as fibrosis proceeds.
We have previously shown that myofibroblast differentiation is a feature that defines the PD plaque [1, 2, 6, 7], similar to what occurs in other fibrosis and in abnormal wound healing [11–15]. Here we have confirmed in vitro the dependence of this process on TGFB1 levels, which is in agreement with the induction of the PD-like plaque in the rat models of PD by either direct injection of TGFB1 into the tunica albuginea or its upregulation caused by a similar injection of fibrin, a factor that would be produced from the putative extravasation of fibrinogen into the tunica albuginea, subsequent to trauma to the erect penis [4, 8, 9]. TGFB1 is a major factor in the appearance of myofibroblasts during normal wound healing, cells that share the phenotype of a smooth muscle and a fibroblast, which actively produce extracellular matrix and contractile force, and later disappear by apoptosis. It is not known, not just in PD, but also in any other tissue, whether myofibroblasts derive by differentiation of a regular fibroblast or from a progenitor or stem cell not yet differentiated from the surrounding fibroblast population.
More surprising is our finding that the TA and PD cultures may differentiate into smooth-muscle cells, as judged from the expression of a specific smooth-muscle marker, SMTN, that is not expressed in myofibroblasts [52], and another one, transgelin (TAGLN), that is less specific. This occurs in regular DMEM, used for smooth-muscle cells culture, upon passage (it is not present initially) and also in FGM and OM. In contrast with osteoblast and myofibroblast differentiation, the expression of smooth-muscle markers does not seem to be affected by TGFB1. This suggests that ossification/fibrosis and smooth-muscle generation are independent processes, e.g., that the latter may not proceed via a myofibroblast intermediate; otherwise, TGFB1 would stimulate it because they considerably increase myofibroblast number.
If this expression of smooth-muscle markers defined in vitro would be identifiable in a substantial number of well-recognized smooth-muscle cells and would also occur in vivo, the tunical fibroblasts may be envisaged as a reservoir of smooth-muscle cells to replenish their loss during oxidative stress and fibrosis. The latter are processes that overexpress or activate TGFB1 and the cell lineage cross talk between the TA and the trabecular smooth muscle may exist at some stages, as postulated for the arterial adventitia fibroblasts in relation to the media [39]. So far, there is no evidence in the literature showing that myofibroblasts can differentiate into smooth muscle and, although it has been suspected that this may be the case, the possibility of a smooth-muscle progenitor different from the myofibroblast being present in the tunical cell population cannot be excluded.
The ability of both TA and PD cells to induce a typical osteogenic differentiation of the target mesenchymal multipotent cell line C3H10T(1/2), even in the absence of azacytidine, may be due to the secretion of osteogenic factors such as BMP2 or some other BMP, like BMP4 or MSX2 [29–31, 33], some of which we found were upregulated in the PD cells in OM. An interesting result was the finding that PD cells, as opposed to TA cells, paracrinely reduced the number of both azacytidine-treated and untreated target cells that become ACTA2 positive and have the typical morphology of myofibroblasts, despite the fact that PD cells tend to express more ACTA2 protein upon incubation in both FGM and OM. This should be investigated to determine why PD cells, but not TA cells, seem to differ in paracrine and autocrine effects. On the other hand, the coincidence of ALPL or ACTA2 protein expression with fat droplets in the differentiating target cells is not unexpected, at least in the case of osteogenesis, because in general, adipogenesis and osteogenesis may overlap at a certain stage but are mutually exclusive in vitro [55]. Our procedure assures their elimination subsequent to dual culture and before replating for immunocytochemistry
A fundamental question posed by our observations is whether the induction of at least three lineages (osteoblast, myofibroblast, smooth muscle), and the inhibition of a fourth one (adipocyte) in, or by, TA and PD cell cultures is due to independent progenitor cells for each lineage, or conversely, whether all processes occur from a single adult stem cell present in vivo in the TA and the PD plaque [34–36]. The ability of some of the PD cells to grow in soft agar and the expression of the stem cell marker CD34 antigen in a fraction of the TA and PD cells would suggest that stem cells are indeed present in the normal and fibrotic (PD) tunical tissue, thus adding these sources to the large series of tissues and organs harboring stem cells. This needs further investigation based on dual immunocytochemistry and clonal selection to determine whether the progeny of a single CD34 antigen+ cell originates two or more lineages.
The ability of PD cells to form colonies on soft agar, which is virtually absent in TA cells, may imply that these stem cells (or other cells in the culture) are potentially tumorigenic. In fact, some fibroblast cell lines and primary cultures can grow in soft agar, possibly because they contain multipotent cells, as assumed above, or alternatively, due to their well-known spontaneous tumorigenicity [56]. This assumption is consistent with the occurrence of chromosomal abnormalities in PD cells that progress during culture [57], an increased S-phase, and inactivation of the antitumor gene, p53, all features of uncontrolled replication [58]. In fact, PD cells, but not the tunica cells, injected into SCID immunodeficient mice, were found to develop subcutaneous tumors [59]. Myofibroblasts may undergo uncontrolled replication and lead to myofibroblastomas, aggressive fibromatoses (like PD), and even sarcomas [60]. Because PD is not associated with penile cancer, it is possible that paracrine factors in the normal penile TA or corpora cavernosa maintain these potentially transformed cells in check, allowing osteogenesis and fibrosis but not neoplastic development.
In conclusion, cultures of human tunical and PD plaque fibroblasts can undergo and induce differentiation into multiple cell lineages and this may explain plaque ossification and fibrosis progression through paracrine modulation. Future work is needed to clarify whether a) this occurs from multiple preomitted progenitors or from true stem cells; b) this process can be replicated in vivo in a PD-like lesion in nude mice receiving implants of human PD fibroblasts; c) the main mechanism is the activation of the BMP/TGFB1/SMAD pathway and involves other members of the TGFB1 family, like GDF8 (growth differentiation factor 8, previously known as myostatin) [44]; d) the inhibition of osteogenic differentiation has therapeutic potential to prevent plaque ossification; and e) this process is related to oxidative stress and may be similar to the one occurring in arterial wall ossification [19, 20].
FOOTNOTES
1 Supported by the Eli and Edythe L. Broad Foundation, and partially from NIH grant R01DK-53069, and NIH Program grants G12RR-03026 and 5P20MD000545. 
2 Correspondence: Nestor F. Gonzalez-Cadavid, Los Angeles Biomedical Research Institute at Harbor-UCLA Medical Center, Division of Urology, Bldg. F-6, 1000 West Carson Street, Torrance, CA 90509. FAX: 310 222 1914; ncadavid{at}ucla.edu
Received: 15 February 2005.
First decision: 21 March 2005.
Accepted: 8 August 2005.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  M.G. Ferrini, I. Kovanecz, S. Sanchez, D. Vernet, H.H. Davila, J. Rajfer, and N.F. Gonzalez-Cadavid Long-Term Continuous Treatment with Sildenafil Ameliorates Aging-Related Erectile Dysfunction and the Underlying Corporal Fibrosis in the Rat Biol Reprod, May 1, 2007; 76(5): 915 - 923. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||
cell line was incubated in DMEM-10% fetal bovine serum containing either 20 µM azacytidine (+AZCT) or no addition (–AZCT) for 3 days, and then cultured for 2 wk on the lower compartment of a dual chamber containing inserts with TA or PD cells on top of the membrane separating both chambers or no cells (C), using OM in both compartments. C3H 10T