Inhibition of DNA methylation during chronic obstructive bladder disease (COBD) improves function, pathology and expression

Abstract

Partial bladder outlet obstruction due to prostate hyperplasia or posterior urethral valves, is a widespread cause of urinary dysfunction, patient discomfort and also responsible for immense health care costs. Even after removal or relief of obstruction, the functional and pathologic aspects of obstruction remain as a chronic obstructive bladder disease (COBD). Epigenetic changes, such as DNA methylation, contribute to the persistent character of many chronic diseases, and may be altered in COBD. We tested whether candidate genes and pathways and the pathophysiology of COBD were affected by a hypomethylating agent, decitabine (DAC). COBD was created in female Sprague-Dawley rats by surgical ligation of the urethra for 6 weeks, followed by removal of the suture. Sham ligations were performed by passing the suture behind the urethra. After removal of the obstruction or sham removal, animals were randomized to DAC treatment (1 mg/kg/3-times/week intraperitoneally) or vehicle (normal saline). Bladder function was non-invasively tested using metabolic cages, both one day prior to de-obstruction at 6 weeks and prior to sacrifice at 10 weeks. Residual volume and bladder mass were measured for each bladder. Bladders were examined by immunostaining as well as qPCR. The effects of DNA methyltransferase (DNMT)-3A knockout or overexpression on smooth muscle cell (SMC) function and phenotype were also examined in bladder SMC and ex vivo culture. Residual volumes of the DAC treated group were not significantly different from the NS group. Compared to COBD NS, COBD DAC treatment helped preserve micturition volume with a significant recovery of the voiding efficiency (ratio of the maximum voided volume/maximum bladder capacity) by one third (Fig. 1, p > 0.05). Brain-derived neurotrophic factor (BDNF) variants 1 and 5 were upregulated by COBD and significantly reduced by DAC treatment. Deposition of collagen in the COBD bladder was reduced by DAC, but gross hypertrophy remained. In bladder SMC, DNMT3A overexpression led to a loss of contractile function and phenotype. In bladders, persistently altered by COBD, inhibition of DNA-methylation enhances functional recovery, unlike treatment during partial obstruction, which exacerbates obstructive pathology. The underlying mechanisms may relate to the gene expression changes in BDNF and their effects on signaling in the bladder.

Figure 1

Decitabine treatment improves discrete aspects of bladder physiology and pathology during COBD. Previously we noted that 1 mg/kg/day decitabine (DAC) on physiology during obstruction (PBO) exacerbated the bladder dysfunction. Here we examined DAC treatment in chronic obstructive bladder disease after de-obstruction (COBD). DAC treatment of COBD had either an improvement ( maximum voiding efficiency), no change (bladder mass, maximum capacity, mean voided volume) or a modest exacerbation in pathophysiology (maximum voided volume) , ^# p < 0.05, One-tailed t-test. *Two-tailed t-test, p < 0.05. **Two-tailed t-test, p < 0.01.

Figure 2

COBD increased collagen type I expression within SMC bundles, while decitabine (DAC) treatment reversed changes in ECM. (A) Collagen type I expression by immunofluorescent staining (RED) was increased and more widely distributed during COBD. (B) Analysis on Volocity 6.3 revealed that DAC treatment decreased collagen expression significantly in bladder SMC compartments (identified by WWTR1 staining in far-red-blue). (C) The SMC area (size of smooth muscle cells) and area occupied by the SMC in the detrusor fields was quantified on Volocity, by counting nuclei and the areas of the detrusor. The SMC area and the total smooth muscle/field was increased by chronic obstruction and reduced by decitabine treatment. Total collagen area was also significantly altered by COBD and DAC treatment. A ratio of the total SMC area to collagen areas in each field were calculated as well (right-most panel), but the ratio was not significantly altered. Nuclei are identified with DAPI (light blue). *p < 0.05 by 1-tailed t-test, **p < 0.05 by 2-tailed t-test, ***p < 0.01 by 2-tailed t-test.

Figure 3

Expression of obstruction- and DAC-responsive genes is altered during COBD. qPCR was performed in COBD bladder tissues to determine expression of genes which we previously found to be restricted by DNA methylation during obstruction in contrast to after de-obstruction. BDNF variants 1 and 5 (exons VI and IV) shown here increased significantly with COBD and decreased with DAC treatment in COBD. CTGF was increased with COBD and remained upregulated despite DAC treatment. MECP2, which is a regulator of BDNF, was significantly downregulated with DAC, though during COBD it only tended towards upregulation. We also examined genes regulating DNA de-methylation, APOBEC2 and TET2. APOBEC2 was upregulated with COBD and COBD + DAC. TET2 showed a downregulation with COBD and DAC treatment vs. either sham or COBD alone. Expression of other genes implicated in COBD in the bladder were also examined in Supplemental Fig. S4, but none of these genes showed expression patterns concordant with BDNF mRNA expression. *p < 0.05.

Figure 4

DNMT3A alters contractile function of bladder SMC. (A) Overexpression of DNMT3A + GFP (DNMT3A OE + GFP plasmid) vs. GFP plasmid alone was performed by nucleofection in human bladder SMC. Anti-DNMT3A (Magenta-far-red) detected a strong overexpression of DNMT3A in cells with GFP co-expression, representative image shown. All images at same magnification. White bar = 23 microns. (B) In parallel cultures, cells were plated onto collagen gels at 3 × 10⁴ cells/mL in 24 well plates and allowed to grow 1 day prior to releasing gels from wells to allow for attachment and contraction. DNMT3A overexpression significantly decreased gel contraction, Overexpression was confirmed in parallel cultures. Representative photomicrographs shown. *p < 0.002.

Figure 5

Activation of ERK1/2 is associated with persistent and decitabine (DAC)-resistant features of COBD. (A) Localization of activated ERK was performed by quantifying phospho-ERK1/2 (red-Cy3) in the SMC bundles (positive for BDNF, purple-far-red, as in previous work), and compared to the nuclear staining by DAPI (light blue), on Volocity. (B) Staining for Phospho-ERK was most intense in the nucleus of COBD cells. Nuclear phospho-ERK staining of interstitial cells and SMC was increased by DAC treatment. (C) Primary human bladder SMC maintained in 2% growth media were treated with vehicle or 2 μM DAC for 48 h to examine if signaling was altered by DAC treatment. Phospho-S6 (purple-Far-red) and phospho-ERK (red-Cy3) were detected by immunofluorescence, which showed a reduction of Phospho-S6 and an increase in phospho-ERK1/2. *p < 0.05, **p < 0.01, ***p < 0.005.

Figure 6

DNMT knockout is associated with altered smooth muscle cell (SMC) phenotype. Bladder SMC targeted AAV6-cre-Tomato (vs. AAV6-GFP) leads to a reduction in DNMT3a/3b (Supplemental Fig. S2). This decrease is associated with a reduction in far-red staining for desmin and smooth muscle myosin (Myh11). Representative photomicrographs shown at the same magnification for each antibody set. White bar for myosin = 12 microns. White bar for desmin staining = 90 microns.