Background
Benign prostatic hyperplasia can, independent of prostate size, cause bladder outlet obstruction (BOO) in elderly males termed benign prostatic obstruction (BPO). BPO is a dynamic chronic process accompanied by lower urinary tract symptoms (LUTS), including storage symptoms such as frequency, urgency and nocturia and voiding symptoms such as weak stream, delayed or intermittent voiding and incomplete emptying. Moderate to severe LUTS are found in approximately 25% of men aged 40 to 49 and in 50% aged 70 to 79 [
1]. After exhausting less invasive medical treatment options, patients are offered surgical treatment, for example transurethral resection of the prostate (TURP) to de-obstruct the bladder outlet. Although voiding parameters significantly improve, 20 to 40% of patients continue to experience at least some bothersome LUTS [
2,
3], inciting further research into the factors contributing to BPO-induced bladder dysfunction. Specifically, identifying the men at risk of irreparable bladder damage due to BPO and optimal timing of de-obstruction surgery are paramount to avoid loss of bladder function. Limited evidence from human studies and animal models, summarized in a recent report [
4], supports the notion that BPO gradually progresses from inflammation to hypertrophy to fibrosis [
5]. BPO-induced bladder remodeling includes initial bladder hypertrophy during the compensated stage characterized by the increased detrusor contractility / pressure during voiding, and can be accompanied by detrusor overactivity (DO). This ultimately can lead to loss of bladder function (detrusor underactivity) [
4].
In humans, the advancement of BPO-induced bladder remodelling is impossible to monitor, because several years can pass between the onset of symptoms and presentation in clinic. Nevertheless, it is possible to establish a correlation between the urodynamic phenotypes of BPO-induced lower urinary tract dysfunction (LUTD) and the molecular alterations in the bladder, as we have recently shown in a comprehensive study of human bladder biopsies, obtained before TURP [
6]. The overall number of gene expression changes increased progressively: It was the lowest in BPO with detrusor overactivity and the highest in underactive decompensated bladders [
6] compared to controls without LUTD. Animal models of LUTD cannot replicate the chronic gradual longitudinal changes seen in human disease. To mimic human bladder outlet obstruction in rodents, the urethra is loosely ligated creating a partial bladder outlet obstruction (pBOO). In contrast to humans this causes acute obstruction resulting in a significant initial inflammatory impact immediately after surgery. There are other species-specific differences in immune processes, for example, the TNF-alpha-induced changes in human BPO are not observed in mouse pBOO, indicating a different pathophysiological mechanism of organ remodelling [
7]. Similarly, the acute partial obstruction induced in animals results in bladder stretch which is rarely encountered in humans and may affect the results generated by this model.
Alternatively, it is possible to monitor both the functional and molecular changes in the bladder after surgical de-obstruction, and this was done in a number of animal studies, mostly in larger rodents such as rats [
8,
9], guinea pigs [
10] and rabbits [
11]. It was noted that despite the overall improvement of the micturition parameters, complete restoration of bladder function did not occur [
12], and many molecular changes persisted after the relief of obstruction in animals [
13]. Similarly, in a large proportion of human patients, followed after TURP, removal of obstruction improved symptom scores and flow rate [
14], but did not completely reverse the LUTD evident by persistent DO [
15] and low or inadequate detrusor contractility [
16].
Understanding the molecular alterations in the human bladder caused by BPO and persisting after the relief of obstruction is indispensable for defining the “point of no return”, when the organ deterioration becomes irreversible. This should help identify new therapeutic options, including correct timing of de-obstruction surgery. As a follow-up of our earlier study, which revealed molecular networks, hubs of signalling, and biomarkers in BPO-induced bladder dysfunction in men with defined functional phenotypes [
6], we now report a comprehensive molecular and urodynamic characterization of the bladders in men with BPO before TURP and 3 months after the relief of obstruction. We performed an integrated transcriptome and proteome analysis of the bladder biopsies in the two patient groups with a significant difference in the voiding detrusor pressure (PdetQmax), and delineated the molecular classifiers of each group, pointing at the different pre-TURP bladder status. The gene expression follow-up 3 months after surgery sheds light on the processes, contributing to the recovery of bladder function and transcription factors, involved in the regulation of bladder remodelling in BPO.
Discussion
BPO induces significant remodelling in the human urinary bladder, which demonstrates a very similar reaction to increased outlet resistance as the heart subjected to pressure overload [
34]. Based on the observations made in animal models of pBOO, BOO is a chronic gradually progressive disease. Hypertrophy is the bladder’s initial response which most likely advances further until the final decompensatory stage, loss of contractility [
35]. On the molecular level, these processes are characterised by hypoxia and inflammatory response, which induce organ fibrosis [
5]. Indirect evidence collected in humans also points to the fact that BPO is a progressive disease, which can be delayed by pharmacological treatment with alpha-1 adrenergic antagonists and/or 5-alpha reductase inhibitors [
36,
37]. In humans, monitoring urodynamic changes in the obstructed bladder after TURP allows the assessment of functional recovery, although drawing conclusions about the morphological alterations in the affected organ is difficult because a very limited number of studies offer relevant follow-up information. Increased bladder pressure and a reduced flow rate are the physiological parameters seen to be improved by therapeutic measures including surgical de-obstruction. Ultrasound measurements of the bladder or detrusor wall thickness, indicative of muscle hypertrophy, have been proposed to non-invasively monitor bladder remodelling during BPO, with significant differences observed between obstructed and non-obstructed patients [
38,
39]. Bladder wall thickness was decreased one month after TURP surgery, indicating a recovery trend after de-obstruction [
40]. However, there were no symptomatic or urodynamic gains from de-obstruction in men with BPO and detrusor underactivity [
41], implying that the timing of surgery is crucial for the outcome. Indeed, our previous study of the molecular changes in bladder dome biopsies from patients with different urodynamic phenotypes [
6] showed profound gene expression changes in BPO-induced detrusor underactivity leading to the loss of contractility.
This study investigated the changes in cell signalling processes within BPO-affected bladders before and after de-obstruction. The analysis focused on examining the transcriptomes and proteomes of bladder dome biopsies collected from men who experienced urodynamically confirmed functional improvement following TURP. Age-matched patients with BPO without DO were divided into two groups based on the PdetQmax values recorded by UDI before de-obstruction: high and medium pressure (HP and MP) groups. PdetQmax was the only statistically significant parameter, separating the HP and MP groups (Fig.
1), however, the MP group had a slightly higher residual volume (RV) and a considerably lower BCI, although the differenece in BCI did not reach statistical significance. Three months after de-obstruction surgery, the voiding parameters PdetQmax, Qmax and RV were significantly improved in both groups, without any significant inter-group difference in the values after TURP.
The small number of patients per group (n = 3) was a limitation of this study, which did not allow certain observed trends in the UDI parameters to reach statistical significance. The mean age of the control patients (n = 6) was lower than in patients with BPO, because due to the increasing age-related prevalence of BPO in the male population, it was impossible to recruit truly age-matched controls without any LUTS.
The overall number of gene expression changes in both groups compared to controls without LUTS showed an increased number of DEGs in the MP group before TURP, and comprehensive bioinformatics analysis revealed 10 molecular classifiers (CYP1B1, TIPARP, AREG, FOXE1, CYSRT1, PLAAT2, OVOL1, MYBPC1, CDX2, CYP1A1) reliably differentiating between the HP and MP groups before TURP in PCA. The proteins encoded by these genes contribute to oxidative homeostasis (CYP1B1, CYP1A1, CYSRT1), are transcription factors (CDX2, FOXE1 which targets TGF-beta, EGF/TGF-family member AREG, OVOL1), are involved in muscle contraction (MYBPC1) and immune function (TIPARP). Comparison of bladder transcriptomes before and after TURP in the HP group revealed 12 markers (CXCL13, BHMT, EGR3, CCL19, CCL21, NR4A3, CRTAC1, SAA1, UPK2, NR4A1, CCL18, UPK1A), discriminating the samples collected at two time-points from the same patients, indicating that de-obstruction induced significant alterations in the gene expression profiles of the affected bladders. Interestingly, here in addition to transcription factors (NR4A1, BHMT, EGR3, NR4A3) and inflammatory markers (CXCL13, CCL19, CCL21, CCL18 and SAA1) we discovered two uroplakin genes (UPK2 and UPK1A). Further analysis revealed that these genes were significantly down-regulated before TURP in the HP group, compared to controls and the MP group. Their expression was improved by de-obstruction but did not reach control levels. A combination of the 22 markers, used in PCA, revealed that before TURP the MP group was highly different from both the HP group and the controls. Although de-obstruction did not completely restore gene expression in the HP and MP groups, the resulting profiles in the “after” samples were more similar to each other and closer to the controls. Thus, our unbiased bioinformatics analysis of the whole transcriptomes revealed a partial normalization of gene expression, in line with functional improvement observed by UDI.
The analysis of the biological processes and activated pathways in HP and MP groups before and after TURP showed considerable improvement but no compete reversal of the BPO-induced bladder gene expression deterioration. Activation of the immune response processes was a prominent feature in both patient groups before TURP, but the hallmarks of inflammation and the main activated pathways were strikingly different in both groups before TURP. Complement activation was the main up-regulated process in the HP group, whereas the MP group showed more advanced signs of immune response, including significant up-regulation of TNF-driven inflammation, and concomitant IL6, IL1B and PTGS2 up-regulation. B cell activation, neutrophil migration and humoral immune response with activated cytokine production were the top BPs in MP “before” samples, while the complement activation was the top BP in HP “before” samples. These differences, together with the higher number of DEGs, might be an indication of a progressive bladder deterioration in the MP group in response to BPO. Complement activation is becoming increasingly recognized as a key contributor to the beginning sterile inflammation, when the damaged tissues release danger signals and trigger complement, which acts on a range of leukocytes to augment and bridge the innate and adaptive immune systems [
42]. Complement triggers phagocytosis [
43] and the subsequent neutrophil infiltration, observed in the MP group. Thus, the changes in immune response in the obstructed bladder might serve as an indicator of the disease progression. Likewise, our earlier study [
6] showed a steady increase of DEGs in obstructed acontractile patients (UA group) compared to those who were still able to void (BO group). After TURP there was a compensatory down-regulation of many affected processes in both groups, particularly those controlling cell division and cell cycle progression. Interestingly, the expression levels of detrusor muscle genes, which were already significantly up-regulated in the HP “before” samples but down-regulated in MP “before” samples, were elevated following TURP in both groups, accompanied by the activation of muscle- and contractility-related pathways. This indicates that de-obstruction was beneficial for bladder contractility. In the “before” MP group, we observed activated TNF-driven signalling and concomitant down-regulation of detrusor gene expression. This could be an indication of the adverse effects of bladder inflammation on smooth muscle contractility, as previously described in TNF-alpha treated SMCs in vitro [
44], and as a consequence result in the lower PdetQmax before TURP compared to the HP group, where no such processes were recorded. The down-regulated BPs of “cornification” and “intermediate filament organization” in the HP group “before” samples contain many urothelial genes, including uroplakins, all of which were significantly down-regulated. This might be indicative of urothelial dysfunction, exacerbated by high bladder pressure, in humans similar to the animal models of pBOO [
45,
46]. A previous study showed significantly lower expression of E-cadherin, and a higher number of apoptotic cells in humans with BPO [
47], confirming the adverse effects of BOO on urothelial morphology and function.
Proteome analysis indicated a significant difference in protein composition between the ‘before’ and ‘after’ TURP states in both the HP and MP groups. Overall, less DEPs were detected (195 for HP, 165 for MP before TURP compared to controls) compared to the DEGs (855 for HP, 1496 for MP before TURP compared to controls). Immune processes were highly regulated in the proteomes of BPO patients and showed partial normalization after de-obstruction. We also observed changes in the metabolic and proliferative processes, evident by alteration of mTOR and 3-phosphoinositide biosynthesis and degradation pathways after TURP.
To comprehend the impact of obstruction on the factors driving gene expression changes, we investigated the expression levels of known or predicted transcription factors (TFs) and regulators in the transcriptomes of all bladder biopsies before and after TURP. In particular, we looked for TFs which were altered in the “before” state and normalized 3 months after surgery. Only 2 TFs matched these criteria: SOX21, which was significantly up-regulated in both HP and MP groups, and NR1I3 which was specifically down-regulated in the HP group.
The SRY-Box Transcription Factor 21 (SOX21) participates in regulating cell proliferation and differentiation across various tissues [
48]. A database search was performed to identify known and predicted mRNA targets regulated by SOX21. Subsequently, using the mRNA levels of these targets we performed hierarchical clustering analysis to examine the correlation of their expression level changes with the up-regulation of SOX21 and its subsequent normalization. We identified three distinct gene clusters, denoted as Cluster 2, Cluster 4, and Cluster 5, consisting of 7, 34, and 16 genes, respectively, which were regulated in accordance with SOX21 levels before and after TURP. Genes in Cluster 2 were involved in the immune-related BPs (granulocyte chemotaxis, antimicrobial response, etc.) with thrombospondin THBS4 being a prominent signalling molecule; these genes were highly elevated in “before” HP and MP and reduced after TURP. The genes in Clusters 4 and 5 were related to the processes of cell division and chemotaxis; they were significantly elevated in the MP group, and down-regulated after de-obstruction. The proteome analysis of SOX21 targets revealed one protein cluster, also containing THBS4 and significantly down-regulated concomitant with normalization of SOX21 expression.
THBS4 is a glycoprotein mediating cell-to-cell and cell-to-matrix interactions. It is involved in cellular proliferation, migration, adhesion and attachment, inflammatory response and adaptive responses of the heart to pressure overload and in myocardial function and remodelling [
49]. THBS4 was induced after outlet obstruction in rodents, and considered to be a sensitive marker of obstruction, although its knock-out in mice did not affect bladder growth or repression of contractile markers [
50]. Here we confirm the up-regulation of THBS4 in human BPO at both protein and mRNA levels, and its normalization 3 months after de-obstruction. Importantly, THBS4 can be detected in urine as was shown in human urinary proteomics studies from healthy and diseased individuals [
51], making it a potential non-invasive BPO biomarker candidate, if its levels in the urine from men with BPO correlate with the functional impairment caused by obstruction.
Here we established a possible link between SOX21 up-regulation in obstructed bladders, and an increased cell proliferation leading to organ hypertrophy. Earlier studies showed that SOX21 suppressed differentiation of airway progenitor cells during lung development and promoted cell division [
52]. Similarly, higher concentration of SOX21 inhibited neuron formation and instead promoted progenitor maintenance [
48]. These data are in agreement with our observation that SOX21 targets in the Clusters 4 and 5 regulate cell cycling and proliferation pathways. Elucidating the mechanisms, which induce the high responsiveness of SOX21 to obstruction would be important for understanding the progressive changes caused by BPO and may represent a novel approach for the diagnosis and treatment of BPO.