Hypothesis / aims of study
Pelvic Organ Prolapse (POP) is a disorder that occurs when the musculature of the pelvic floor weakens, resulting in pathologic descent of pelvic organs into the vaginal canal. While risk factors for POP have been suggested such as genetic predisposition, childbirth, obesity, and advancing age, the etiology of this condition remains largely unknown. Females have an 11% lifetime risk of POP which can result in urinary, bowel, and sexual dysfunction that can significantly impair a woman’s quality of life. Sphingosine-1-phosphate (S1P) is a bioactive lysosphingolipid with countless metabolic functions including regulation of cell proliferation and smooth muscle contractility. S1P is generated by the phosphorylation of sphingosine by sphingosine kinase, which exists as two major isoforms (SPHK1 & SPHK2). This phosphorylation step is reversible via an enzyme known as sphingosine-1-phosphate phosphatase that also exists as two major isoforms (SPP1 & SPP2). Therefore, alterations in the relative expression of these 4 enzymes would be expected to play a major role in cell regulation. We hypothesized that the terminal step of the S1P pathway, namely the reversible conversion of sphingosine-to-S1P, is present in the female vaginal wall and further that the relative expressions of SPHK and SPP isoforms may be altered in vaginal tissue, thereby leading to decreased vaginal wall stability and subsequent POP.
Study design, materials and methods
Full-thickness anterior vaginal wall tissue samples were obtained from women undergoing surgery under an approved IRB protocol with written informed consent for tissue acquisition and use of tissue. "CONTROL" samples (n=5) were obtained from women undergoing routine hysterectomy with no history of POP. "POP" samples (n=5) were obtained from women with POP undergoing reconstructive surgical repair. After removal, the sample were immediately brought back to the lab, where samples were cleaned of loose tissue and adventitia and were then separated into 2 layers under a dissecting microscope. One layer contained the epithelium and lamina propria (termed epithelial side) and the other layer contained the fibromuscular layer (termed smooth muscle side). Patients with a history of pelvic radiation, previous pelvic reconstructive surgery (for both control and POP groups), or connective tissue disorders such as Marfan syndrome or Ehler-Danlos syndrome were excluded from the study.
Total protein extracts were prepared from all samples while submerged under liquid nitrogen using a SPEX freezer/mill 6775. The frozen tissue samples were pulverized into a fine powder and then homogenized with a polytron handheld homogenizer in a protein extraction buffer that contained 20% glycerol, 0.5 M Tris buffer, 200 mM PMSF, 1% SDS, and protease inhibitor mix (Halt TM Protease inhibitor single-use cocktail).
Extracts were then loaded onto no-stain SDS-PAGE gels and separated via electrophoresis. Exposure of the gels to 1 minute of ultraviolet light activated the trihalo compounds embedded in the gel and allowed visualization of all protein bands in each lane that was captured using a ChemiDoc Imager. The gels were then scanned using Image Lab 6.0 software and the area under the curves of the protein bands profile integrated in order to determine total protein loaded per sample. As the trihalo compound and short time ultra violet light activation does not alter antigenicity of the target proteins, we directly blotted these exact same gels onto polyvinylidene difluoride membranes in only 3 minutes using a Trans-Blot Turbo Blotting System. We blocked the membranes in 5% non-fat dried milk for 30 minutes and then added our primary antibodies (in Tris-buffered saline) and incubated overnight at 4C. Membranes were then washed and incubated at room temperature for 1 hour with fluorescently tagged secondary antibodies. Target protein expression was then quantified using the ChemiDoc Imager and normalized to total protein loaded in each lane. The percentage change in target protein expression was then determined.
In addition, using the same Western blotting procedure described above we sought to determine the expression of our target proteins in various types of other muscle (cardiac and skeletal muscle) as well as bladder smooth muscle using human extracts of each of these. For these blots the same amount of total protein extract (12.5 μg) was added for each muscle extract and blotting was performed using the same primary and secondary antibodies as above.
We were able to identify a single band running at the appropriate expected molecular weight of 43 kDa for the SPHK1 isoform in vaginal samples from both the epithelial and smooth muscle side extracts. After normalizing to total protein in the gel it was determined that SPHK1 was expressed approximately 46% and 40% lower in smooth muscle side and epithelial side extracts, respectively, in POP samples compared to controls. We were also able to identify SPHK1 isoform expression in cardiac and skeletal muscle, but the presence of these enzymes in human bladder samples was inconclusive. Using the same exact smooth muscle and epithelial side samples under identical conditions, we could not identify significant expression of the SPHK2 isoform in either control or POP samples (either in smooth muscle side or epithelial side samples).
We were also able to identify a single band running at the appropriate expected molecular weight of 49 kDa for the SPP1 isoform in samples from smooth muscle side extracts. After normalizing to total protein in the gel it was determined that the SPP1 isoform was expressed approximately 13% lower in smooth muscle side extracts in POP samples compared to controls. We could not identify SPP1 in epithelial samples. Again, we were also able to identify SPP1 isoform expression in cardiac and skeletal muscle, but the presence of these enzymes in human bladder samples was inconclusive.
Interpretation of results
Our novel data shows that SPHK1 is the predominant sphingosine kinase isoform in human anterior vaginal smooth muscle side and epithelial side tissue samples and that SPHK1 levels are decreased dramatically in women with POP compared to controls. Similarly, we demonstrated that SPP1 is the predominant sphingosine-1-phosphate phosphatase isoform in human anterior vaginal smooth muscle side extracts, but we could not detect SPP1 in epithelial side extracts. SPP1 levels were only approximately 13% lower in smooth muscle side extracts of women with POP compared to controls
The larger decrease in SPHK1 vs SPP1 expression would suggest a shift toward a higher sphingosine-to-S1P ratio and thus a shift toward apoptosis and cell cycle arrest and overall decreased vaginal wall stability. These findings suggest a POP preventative treatment strategy of stabilizing SPHK1 levels in the vaginal smooth muscle.
Our study utilized human samples making our findings translationally relevant. A limitation of our data is that only 5 control and 5 POP human samples were available. More samples will be analyzed in the future and SPP2 when available.