Identification and functional activity of matrix-remodeling associated 5 (MXRA5) in benign hyperplastic prostate

Introduction

Benign prostatic hyperplasia (BPH) is one of the most common pathological diseases states in aging males [1], affecting approximately 50% of men in their fifties and 80% for men in their eighties. Lower urinary tract symptoms (LUTS) secondary to BPH greatly impact male quality of life [2, 3]. Despite the high prevalence, the pathogenesis of BPH still remains unclear. In recent decades, many hypotheses have been proposed including altered androgen-estrogen ratio, stromal-epithelial interactions, growth factors (Gfs) and cytokines, inflammation and autoimmune system, stem cells, and metabolic syndrome including obesity and diabetes, to likely play important roles in the onset and progression of BPH [4, 5]. Additionally, genetics, including a number of genes, contribute to the pathogenesis of this disease state. Indeed, there exists a familial form of BPH. Preliminary studies demonstrate DNA mutations [6], DNA hypomethylation [7], and abnormalities of nuclear matrix protein expression [8], along with miscellaneous genetic polymorphisms [9], and abnormal expression of the Wilms tumor gene (WT1) [10] in human BPH. Consequently, detailed analysis of the relevant pivotal genes is extremely beneficial. The establishment of the cDNA microarray technique provides the possibility to evaluate a huge quantity of genes. Indeed. Luo, et al. [11] have used 9 BPH specimens from men with extensive hyperplasia and 12 histologically normal prostates from radical prostatectomy specimens. The differentially expressed genes (DEG) from their microarray analysis were classified into four groups: (1) growth factors and their binding proteins, (2) hydrolases, proteases, and protease inhibitors, (3) stress response enzymes and (4) ECM molecules. Middleton, et al. [12] also used histologically normal prostate samples from patients undergoing radical prostatectomy as controls and demonstrated the stromal signaling molecules bone morphogenetic protein 5 (BMP5) and CXC chemokine ligand 13 (CXCL13) are enriched in BPH specimens while estrogen regulated pathways were depleted. In addition, they also performed single-cell RNA-Seq to verify their findings and whole-exome sequencing to uncovered somatic single-nucleotide variants in BPH. Different from the above two studies, Prakash, et al. [13] used prostate tissue from young donors as normal prostate, which also identified a genetic signature separating BPH from normal prostate. In our current study, we conducted a microarray analysis of five human BPH tissues and three normal prostate tissues and bioinformatics analysis of the DEGs was performed. Subsequently, we focused on the biologic functions of matrix-remodeling associated 5 (MXRA5) protein which is one of the most significantly altered DEGs.

Recent evidence suggests that BPH development involves accumulation of mesenchymal-like cells derived from the prostatic epithelium via epithelial-mesenchymal transition (EMT) [14–16]. The extracellular matrix (ECM) is a collection of extracellular molecules secreted by support cells that provides structural and biochemical support to the surrounding cells [17]. It not only serves as an external environment for the physiological activities of cells, but also supports the intercellular signal transduction pathways involved in multiple pathological and physiological processes. Moreover, androgens, estrogens and the ECM are demonstrated to have interactive effects [18]. Androgen and estrogen receptors regulate active peptides in tissues and stimulate cells to produce different components of the ECM. The ECM also interacts with hormones and growth factors to increase the sensitivity of prostate cells to androgen growth factors [19].

MXRA5 is a protein found in humans that is encoded by the MXRA5 gene. This protein contains 7 leucine-rich repeats and 12 immunoglobulin-like C2-type domains related to perlecan with a molecular weight of 312 kDa and is expressed in primates but not in rat or mouse, playing important roles in cell adhesion and matrix remodeling [20]. MXRA5 mRNA is up-regulated in human chronic ischemic myocardium and in fibroblast primary cultures from centenarians [21]. In addition to ECM remodeling, MXRA5 has been found to be involved in tumorigenesis. Somatic mutations of MXRA5 are observed in patients with non-small cell lung cancer (NSCLC) [22], MXRA5 is also aberrantly expressed in colorectal cancer (CRC) tissues, serving as a critical biomarker for the diagnosis and prevention of CRC [23]. The expression of MXRA5 in hyperplastic prostate has never been studied. However, MXRA5 has been recently found to be closely associated with the mitogen-activated protein kinases (MAPKs) pathways. It was demonstrated that the MAPK pathway plays a key role in the regulation of cell proliferation, survival, differentiation and apoptosis and is involved in signal transduction from the cell membrane to nucleus in response to a wide range of stimuli. Indeed, BPH could be a result of the imbalance between cell proliferation and cell apoptosis. Therefore, we postulate that MXRA5 is an important ECM factor involved in the remodelling of prostate tissue. In our current study, we conducted a microarray analysis comparing human BPH tissues to normal prostate tissues. Subsequently, we investigated the expression of both MXRA5 mRNA and protein in human hyperplastic prostate tissues. We further silenced and overexpressed the MXRA5 gene in human prostatic cells to identify those genes involved in the MAPK pathway.

Results

To explore the biological mechanisms of BPH, we performed mRNA expression profiling of three normal prostates and five BPH tissues. A total of 198 DEGs were identified between the two groups (Figure 1A). This microarray data has already been submitted to the GEO database (serial number: GSE119195) and is the first microarray data on BPH that has been submitted to this database. MXRA5 is one of the highly altered DEGs and was of interest to us for further study. Based on the KEGG database, DEGs related pathways were also analyzed. As shown in Figure 1B, a number of pathways were found to be upregulated, including cell adhesion, Wnt signaling, and the Rap1 signaling, with the MAPK signaling pathway being most significantly altered (p < 0.01).

MXRA5 expression was further demonstrated in the Oncomine database with mRNA levels in BPH stroma being increased by 4.5 folds compared to controls (p = 0.013) (Figure 2A). For BPH and normal samples (n = 15) harvested at our institute, MXRA5 was found consistently upregulated over 2-fold both at the transcriptional and translational level (p < 0.01) (Figure 2B–2D)

Additionally, immunofluorescence staining demonstrated MXRA5 was predominantly localized in the stromal compartment of human prostate with lesser staining observed in the epithelium (Figure 3A). An overall increased MXRA5 staining was also observed in the enlarged prostate when compared with normal prostates using immunofluorescence microscopy (Figure 3B). Negative controls omitting the primary antibody failed to stain (Figure 3C) and positive controls using human renal cortex tissue showed a strong immune positivity (Figure 3D). Similarly, immunohistology demonstrated MXRA5 was present predominantly in stromal cells (Figure 4A) but to a much lesser degree in epithelial cells (Figure 4B).

To create a cell model of MXRA5 deficiency, 3 distinct MXRA5-target-specific-siRNAs (si-MXRA5s) were transfected in to WPMY-1 cells. After 48 h, the knockdown efficiency was validated by qRT-PCR (Figure 5B) and Western blot analysis (Figure 5C, 5D). si-MXRA5-3 exhibited an inhibitory efficiency over 80% and was chosen for subsequent experimentation. Immunofluorescence staining showed the high expression of MXRA5 protein was strongly downregulated in si-MXRA5-3 transfected WPMY-1 cells (Figure 5A). Cell apoptosis and cell cycle stage were further analyzed for these transfected cells. A significant cell cycle arrest at the G0/G1 phase was determined with flow cytometry analysis (Figure 6A, 6B). Immunofluorescence staining showed that the MXRA5-siRNA group exhibited considerably less Ki-67 positive cells than the siRNA-control group (Figure 6C). Moreover, an MTT assay indicated that knockdown of MXRA5 restrained WPMY-1 proliferation drastically (Figure 6D). Additionally, expression proteins involved in G0/G1 phase regulation (cyclin A1/2, cyclin D1 and CDK2/4) was strongly decreased in MXRA5 silenced stromal cells (Figure 7A, 7B). However, no obvious effect on apoptosis was observed in si-MXRA5-3 transfected WPMY-1 cells (Figure 6E, 6F) as revealed by flow cytometry analysis. Finally, Western blot demonstrated that the protein levels of phospho-p38 MAPK (p-p38 MAPK) and p-ERK1/2 were significantly reduced with p38 MAPK and ERK1/2 unchanged following MXRA5-siRNA transfection (Figure 7C, 7D).

As MXRA5 is poorly expressed in prostate epithelial cells, plasmid was transfected into BPH-1 cells to over-express the MXRA5 protein. After 48h, efficiency of over-expression was determined by RT-PCR (Figure 8B) and Western blot analysis (Figure 8C, 8D) and MXRA5 was upregulated both at mRNA and protein levels. Immunofluorescence staining also showed the expression of MXRA5 protein was strongly upregulated in MXRA5-specific plasmid transfected BPH-1 cells (Figure 8A). Cell apoptosis and cycle were further analyzed for these transfected cells. In contrast to MXRA5 silenced stromal cells, the percentage of MXRA5 transfected BPH-1 cells at the G0/G1 phase was decreased while the percentage of cells at the G2 phase was increased (Figure 9A, Figure 9B). Immunofluorescence staining revealed that the MXRA5 group exhibited considerably more Ki-67 positive cells than the vector control group (Figure 9C). Moreover, MTT assay indicated that over-expression of MXRA5 significantly increased BPH-1 proliferation (Figure 9D). Additionally, proteins involved in G0/G1 phase regulation (cyclin A1/2, cyclin D1 and CDK2/4) were strongly upregulated in MXRA5 over-expression epithelial cells (Figure 10A, 10B). However, no obvious effect on apoptosis was also observed in MXRA5-specific plasmid transfected BPH-1 cells (Figure 9E, 9F) as revealed by flow cytometry analysis. Finally, Western blot demonstrated that the protein levels of phospho-p38 MAPK (p-p38 MAPK) and p-ERK1/2 were significantly increased with p38 MAPK and ERK1/2 unchanged following MXRA5-specific plasmid transfection (Figure 10A, 10C).

Discussion

In the current study, we performed a transcriptome analysis comparing BPH tissues versus normal prostatic samples. A total of 198 DEGs were identified with MXRA5 markedly upregulated. Along with these DEGs, a number of signaling pathways were altered in the prostate. Based on the Oncomine database, the MAPK pathway was the most significantly changed. The expression, localization and functional activity of MXRA5 were further elucidated using human hyperplastic prostate tissues and cultured human prostate cell lines. Our novel data demonstrated that the upregulation of MXRA5 in the enlarged prostate could contribute to the development of BPH via increasing cell proliferation instead of inhibiting cell apoptosis via phosphorylating ERK1/2 and p38 of the MAPK pathway.

BPH is a quite common condition in aging males with the incidence continuing to rise due to an increasing aging population worldwide. α 1-adrenoceptor antagonists [4] and 5α-reductase inhibitors [5] are efficacious nonsurgical treatments for BPH. However, around 30% of patients still require surgery. Moreover, the mechanisms of BPH are not completely understood. Thus, it is necessary to obtain a central knowledge of the molecular mechanisms involved in the etiology of BPH to discover new therapeutic strategies. As microarray analysis could precisely provide us differentially expressed genes and related biological functions, in this current study we conducted microarray analysis of 5 human prostatic hyperplasia tissues and 3 normal prostatic tissues from young donors. The microarray data were submitted to the GEO database and a serial number (GSE119195) was assigned. Until now, there has been no published microarray data available in the public database concerning benign prostate tissues.

Many of the prostatic stromal-epithelial interactions that are observed during normal development of prostate and BPH may be mediated by soluble growth factors or by the ECM, which itself has growth factor–like properties [24, 25]. Our microarray analysis identified hundreds of DEGs with MXRA5 being one of the most markedly changed, which is a matrix-remodeling related gene of interest to us. Indeed, MXRA5 mRNA levels in BPH stroma were increased by 4.5 fold when analyzing data from the Oncomine database. The increased expression of MXRA5 was further demonstrated via RT-PCR and Western blot assays using prostate tissues from our institute. Both mRNA and protein levels were significantly upregulated in human hyperplasia prostatic tissues when compared with normal ones. MXRA5 has been reported to be expressed in primates, marsupials, some mammals, birds and fish, but not in rat or mouse [26]. Indeed, we did not observe MXRA5 staining in rat prostate (data not shown). Our present study also demonstrated that MXRA5 was mainly localized in the stroma of human prostate [27].

As our understanding of stromal-epithelial cell relationships in the prostate increases, it is possible that therapies may be designed to induce regression of established BPH by modulating one class of stromal cell excretory protein (i.e., ECM, such as MXRA5). In turn, these proteins regulate epithelial cell differentiation. The present study created a MXRA5 deficiency cell model. As MXRA5 was mainly localized in stroma, cultured stromal cells (WPMY-1) were transfected with siRNA to knockdown MXRA5. The silencing of MXRA5 ameliorated WPMY-1 proliferation drastically and a significant cell cycle arrest at G0/G1 phase was observed. However, knockdown of MXRA5 had no obvious effect on apoptosis in WPMY-1 cells. We also co-cultured siMXRA5 transfected WPMY-1 cells with BPH-1 cells with no inhibitory effect observed on epithelial cells (data not shown). But when a MXRA5 overexpressed epithelial cell model was created, enhanced BPH-1 cell proliferation and accelerated cell cycle from G0/G1 phase to G2 phase were observed. MXRA5 has been reported to be a novel biomarker which is upregulated in non–small cell lung carcinoma [22, 28] and colorectal cancer [23]. Therefore, MXRA5 could modulate both prostate stromal cell and epithelial cell proliferation.

Our microarray analysis also suggested that the MAPK signaling pathway was linked with BPH. When the growth regulation and the homeostatic control between proliferative and non-proliferative behavior in the periurethral and transition zones of the prostate are disrupted, they lead to the expansion of the cell proliferating compartment and the concomitant decrease in the extent of the differentiated compartment [29]. Histologically, BPH represents hyperplasia of both the stromal and glandular components of the prostate, with stromal smooth muscle elements comprising approximately 40% of the cellular volume of the enlarged prostate [30]. This is the end result of an ensemble of different stimuli and local conditions that are influenced by androgen metabolism, stromal-epithelial interactions with autocrine–paracrine secretion of GFs and the action of cytokines produced by inflammatory cells [29, 31]. There is indirect evidence that all three MAPK cascades might be implicated in these pathological alterations by modulating the local prostatic environment to favor the progression of BPH. It is known that MAPK family members participate in regulating the cell cycle in various ways [32]. ERK mainly is involved in regulating the progression of G0/G1 to the S phase and p38 primarily regulates the G2 checkpoint [33, 34]. Interestingly, a recent study reported that the MXRA5 protein is a positive regulator of the MAPK signaling pathway in preeclampsia [20]. Similarly, the current study found that MXRA5 knockdown induced prostatic stromal cell cycle arrest at G0/G1 phase which was confirmed by downregulation of related proteins (cyclin D1 and CDK2/4) and MXRA5 over-expression significant decreased the percentage of epithelial cells at the G0/G1 phase. Consistently, our study also showed that p-p38 and p-ERK1/2 were significantly reduced with p38 MAPK and ERK1/2 unchanged following MXRA5 knockdown, p-p38 and p-ERK1/2 were significantly increased with p38 MAPK and ERK1/2 unchanged following MXRA5 overexpression, suggesting MAPK has important roles in regulating cell proliferation [35] via connecting extracellular stimuli from the cell membrane to the nucleus. Currently, the first-line therapies for BPH are based on the physiology of the prostate with α1-blockers (reduce the prostate tone) and 5α-reductase inhibitors (reduce the prostate volume). However, these treatments cannot completely prevent the progression of BPH. The present study identified genetically altered genes and related signaling pathways of BPH. Therefore, treatment based on pathophysiology of BPH could play a more important role for the future therapy for BPH during routine clinical practice.

There are several factors that may have influenced the outcome of the current study. First, the difference in age between groups may influence the outcome with normal prostate samples being obtained from young donors and BPH samples obtained from aging males. Thus, some differentially expressed genes (DEGs) identified in the present study may be attributed to different age, instead of BPH. However, Luo J, et al. [11] harvested the normal prostate samples from histologically normal regions within the radical prostatectomy specimens. They compared gene expression profile from normal samples with age ≥59 ys and BPH samples with age ≤59 ys and showed similar alterations as in their analysis with the presence of age difference (52.5 vs. 70.6, normal vs. BPH). Consistently, our current microarray analysis identified the same DEGs as theirs, such as BMP5, NELL2, IGF1. Thus, age should not have much influence on the outcomes of our study. Second, origin of tissue may also influence the outcome since different prostate zones have specific architectural and stromal feature. To avoid this interference, samples in our study were all collected from the prostate transition zone. Third, hyperplastic prostate might have cancer-related genetic changes since they were obtained from patients undergoing cystoprostatectomy for infiltrating bladder cancer. However, the prostate samples we used showed no tumor infiltration.

Collectively, this is the first study to reveal both MXRA5 gene and protein expression increased in hyperplastic prostate, while knockdown of MXRA5 inhibited and overexpression of MXRA5 enhanced prostatic stromal cell cycle. Moreover, this process may be partially connected with alteration of the MAPK signaling pathway. Our data suggests that the MXRA5-MAPK system could play important roles in the development of BPH and it could be rediscovered as new therapeutic targets for BPH. Nevertheless, there is some limitation in our study. It will be better to establish a BPH rat model to explore the expression and functional activities of MXRA5 in animal prostate. Unfortunately, rodent animals do not express MXRA5 protein. In the future, genetic mice with overexpressed MXRA5 gene could be an intriguing model to explore the in vivo function of MXRA5.

Materials and Methods