Materials and Methods
This study was conducted with the approval of the Institutional Animal Care and Use Committee (IACUC) at Virginia Tech. A total of sixteen adult female Long-Evans rats, aged between 70-89 days, were used for this study. After euthanizing each rat via decapitation, the entire vagina was immediately isolated. During the dissection, the vagina was continuously kept moist with Krebs-Ringer Bicarbonate Buffer. The vaginal canal was opened into a flat specimen by placing a cut along the urethra. It was then trimmed to be approximately square with sides parallel to the longitudinal direction (LD) and the circumferential direction (CD) of the vagina (Figure 1). Two small hooks (size 18 dry fly hook, Mustad) that were connected via a silk thread were placed on each side of the specimen.
The specimen was then secured to a custom-built biaxial testing machine described in detail elsewhere (Figure 2(a))55. The linear actuators (T-NA08A25, Zaber Technologies, Inc.) of the biaxial system had a maximum travel length of 25 mm and micro-step size resolution of 0.048 µm while the two load cells (FSH02663, Futek Advanced Sensor Technology, Inc) had a maximum load capacity of 50 g and accuracy of +/- 0.1%.
Once mounted on the biaxial testing machine, each specimen was submerged in a bath of Krebs-Ringer bicarbonate buffer with calcium (2.00 mM). In order to establish the initial configuration more consistently across different specimens, each specimen was stretched with a displacement rate of 0.05 mm/s until a pre-load of approximately 4 mN was recorded in both axial loading directions. It was then allowed to relax for 30 minutes before being re-stretched at 0.05 mm/s displacement rate until a 4 mN preload was reached in both directions. In this configuration, a picture of each specimen was quickly taken using a CMOS camera (DCC1545M, Thorlabs) equipped with lens (59-871, Edmund Optics) to measure the specimen dimensions using ImageJ (NIH, Bethesda, MD). Specifically, the average of two distances between the hooks that were placed directly opposite of each other along the LD or CD was considered to be the initial side-length of the specimen in that direction (Table 1). In this initial configuration, the specimen was then electrically stimulated in the LD and CD using a high-power pulse stimulator (701C, Aurora Scientific, Inc.) through custom made stainless steel electrodes (Figure 2(b)). The electrodes were designed such that one plate could be positioned directly above the specimen and the other plate directly below the specimen. This design ensured that the direction of the electric current would be perpendicular to both the LD and CD and applied through the specimen thickness. The EFS consisted of one train of 2 ms pulses delivered at 70 Hz for 0.8 s, with 700 mA of current intensity. These EFS parameters were selected since they produced maximum forces in preliminary tests.
Each specimen was then stretched simultaneously in each direction by increasing the length of the specimen by 2% of its initial LD and CD lengths for 21 consecutive increments to a final stretch of 42%. The specimen was held at each stretch for 90 seconds and, at the end of each holding period, the forces in the LD and CD were measured. These forces were the “passive forces” of the specimen, i.e. the forces that resulted from the application of the stretches prior to the EFS. Note that these are not fully passive because the smooth muscle was not chemically passivated. After each holding period, the specimen was electrically stimulated using the same methods and parameters presented above and the contraction forces in the LD and CD were measured. At each stretch, the “active force” in the LD or CD was defined as the difference between the maximum force reached via EFS and the passive force at that stretch. A schematic of this testing protocol is presented in Figure 3(a)-(b).
After electrically stimulating the specimen multiple times, the specimen was unloaded until the load reached about 4 mN and it was then allowed to relax for 5 minutes. While recording force data, the testing solution was removed and replaced with Krebs-Ringer bicarbonate buffer with calcium (2.00 mM) and a high concentration of potassium chloride (124 mM) (Figure 3(c)) 21. The specimen remained in the bath until the force in the LD and CD reached a maximum value and began to decrease, which typically took 1-2 minutes. The “active force” due to the KCl in the LD or CD was calculated as the difference between the maximum force reached with the KCl stimulation and the passive force recorded before the KCl solution was added (Figure 3(c)). Specimen thickness was measured immediately after testing in 5 different locations through the use of a CCD laser displacement sensor (LK-G82, Keyence, Inc.) with an accuracy of 7.5 µm (Table 1). It was assumed that changes in thickness during biaxial testing were negligible. Cross sectional area was measured as the average of distances between hooks on the same side multiplied by specimen thickness. Passive and active force data in the LD or CD were divided by the cross-sectional area that was perpendicular to the LD or CD, respectively, to obtain passive (nominal) and active stress data.
Maximum active stresses induced by EFS or KCl, stretches at which the maximum EFS-induced active stresses were achieved, and maximum EFS-induced active stresses normalized by KCl induced active stresses were compared between the LD and CD using Wilcoxon signed-rank tests. Similarly, in each direction, the maximum EFS-induced and KCl-induced active stresses were compared. Differences in passive stresses and EFS-induced active stresses were evaluated with 2-way ANOVA using direction and stretch as factors. Statistical analysis was performed with GraphPad Prism 6.0, and differences were considered significant when p<0.05.
The mean passive stress-stretch data collected from n=16 specimens in both loading directions are shown in Figure 4. The rat vaginal tissue exhibits the typical non-linear stress-stretch behavior of soft biological tissues. The 2-way ANOVA revealed that direction (p<0.001) and stretch (p<0.0001) significantly affected the passive force. As expected, the mean passive force increased as the stretch increases. More interestingly, the tissue resulted to be anisotropic, being, on average, stiffer in the CD than in the LD.
Force vs. time data collected during the first, middle, and last EFSs corresponding to stretch values of L= L, =1.2 L, and =1.42 L are shown for a representative specimen in Figure 5. After the EFS was applied (t=1 s in Figure 5(a), (b), and (c)) the vagina contracted producing gradual increases in forces in the LD and CD until the stimulation ended (t=1.8 s in Figure 5(a), (b), and (c)) and, at that time, the forces returned to their passive values. At L=1 L, the specimen was pre-loaded so that the initial passive forces acting along each direction were nearly identical, around 4 mN (Figure 5(a)). As the stretch increased, the difference between the passive forces along the CD and LD increased with higher passive forces in the CD than in the LD (Figure 5(b)-(c)). The activeforces were typically higher in the CD than in the LD at each stretch value. The differences of both active and passive forces in the two directions demonstrate the anisotropy of the vaginal tissue.Figure 6. The 2-way ANOVA revealed that the stretch (p<0.0001) and direction (p<0.0001) factors significantly affected the active stress. EFS-induced active stresses were higher in the CD than in the LD, increasing with stretch until the maximum values were reached. These stresses then decreased even though the stretch continued to increase. The maximum active stress induced by EFS was about 60% higher in the CD than in the LD (p<0.001) The maximum active stress along the LD was measured at a stretch of about 1.32 on average, while the maximum active stress along the CD was obtained at a stretch of around 1.25 on average (Figure 7). This difference was found to be statistically significant (p<0.01).
In Figure 8, the forces along the LD and the CD for a representative specimen that was stimulated using KCl are reported. During the first few seconds of KCl stimulation, the forces in both the LD and CD quickly increased and decreased due to the removal of the first solution and the addition of the KCl solution. Once the KCl solution was added, the forces in each direction gradually increased until plateauing after around 100 s.
On average, the KCl stimulation caused 60% higher active stresses in the LD than in the CD (p<0.01). The mean active stresses caused by the KCl stimulation were 3.46 kPa in the LD, and 2.15 kPa in the CD (Error! Reference source not found.9). When comparing the mean maximum active force induced by EFS to the mean active force induced by the KCl, no differences were found in the CD (p=0.404). On the other hand, in the LD, the mean active force was much higher in response to the KCl stimulation than to in response to EFS (p<0.001). All statistically significant differences found with respect to stimulation methods and direction are noted in Figure 9. Finally, the mean maximum EFS- induced active stress normalized by the KCl-induced active stress was significantly higher in the CD than the LD (Figure 10).
In this study, we presented the first characterization of the biaxial passive and active mechanical properties of the rat vagina. The mechanical properties of this reproductive organ are crucial to the proper function of the entire female pelvic floor. Indeed, alterations of these properties are implicated in the development of pelvic floor disorders such as sexual dysfunction, urinary incontinence, and prolapse. The vagina is a tubular organ that, in vivo, is primarily loaded along the LD and CD, and smooth muscle fibers within the muscularis layer of the vagina are mainly oriented along these directions. The vagina is widely assumed to be anisotropic, but very few tests have sought to quantify this behavior. For these reasons, we performed planar biaxial tests, rather than the commonly used uniaxial tests, to quantify the mechanical behavior of the vagina. During mechanical testing, we stimulated the vaginal tissue through nerve fibers using EFS and through direct membrane depolarization using KCl. Our findings demonstrated that vaginal tissue is highly anisotropic: the tissue in the CD is significantly stiffer than in the LD in the passive state (Figure 4)and EFS-induced active state (Figures 6). However, the vaginal tissue generated higher stress in the LD when activated via KCl (Figure 9).
Previous studies have sought to quantify the anisotropy of vaginal tissue in the passive state by comparing results of uniaxial tests collected from tissue strips that were cut along the LD and CD 39, 43, 45, and, very recently, through inflation-extension testing 41. The results from these studies have been conflicting, though. Pena et al. found that the vagina was stiffer in the LD than in the CD 39. Rubod et al. reported that the anisotropy of the vaginal tissue was strain-dependent: there were no significant differences in stiffness at strains below 50%, but the tissue in the LD became stiffer at higher strains 43, 45. Similarly, Robison et al. also found the vagina to be stiffer in the LD at high strains, but found it stiffer in the CD at low strains 41. In this study, we found that, onaverage, the rat vagina was significantly stiffer in the CD at strains between 0 and 42% (Figure 4). However, it should be noted that not every specimen followed this trend, as there were some specimens that exhibited higher stresses in the LD than in the CD for the same stretches. Our anisotropy findings align with those of Robison et al.  but additional testing at higher strains would need to be performed to potentially observe the reported strain-dependent anisotropy of vaginal tissue. The stress values reached during our testing were lower than previously reported stresses at similar strains 40. This is likely due to our incremental protocol that allowed the specimens to relax significantly throughout the test.
The mechanical properties of the vagina in the active state have been exclusively studied via uniaxial tests. In these tests, either longitudinal 4-6, 19, 27, 34 or circumferential 16, 21, 47, 54 strips of vaginal tissue were tested, and the active forces generated by various stimulation methods were recorded. Oh et al. 35 tested both longitudinal and circumferential strips of vaginal tissue via uniaxial tests and found no significant differences in the active responses in the LD and CD 35. This is in contrast with the results of our biaxial tests as we found that, in response to EFS, the vagina contracted significantly more in the CD than in the LD as demonstrated by the higher active stresses in the CD (Figures 5, 6 and 9). Moreover, in response to KCl, the vagina contracted more in the LD than in the CD, generating higher active stresses in the LD (Figures 8 and 9). Before normalizing by cross sectional area, we found that our testing yielded higher, but comparable, EFS-induced active forces than tests by van Helden 54; this was expected since we used the entire vagina rather than segments. We found that the magnitudes of the active forces normalized by specimen volume generated by KCl in our tests were similar to those found in uniaxial testing .
Materials and Methods
GET THE COMPLETE PROJECT
Anisotropy of Passive and Active Rat Vagina under Biaxial Loading