Thirty normal healthy female SD rats weighing 220-250 g were purchased from Beijing Vital River Laboratory Animal Technology Co., Ltd. All rats were provided free access to food and water, maintained under a 12-h light/dark cycle, and randomly assigned to 4 groups: N (n = 6), M (n = 8), NAAC (n = 8), and EA (n = 8). All rats except the N group underwent vaginal balloon dilatation and bilateral oophorectomy to replicate the SUI model (15). Briefly, after anesthesia (sodium pentobarbital, 30 mg/kg, i.p.), an 8-Fr latex Foley catheter was inserted into the rat’s vagina, and 4.0 ml of sterile saline was slowly pushed into the balloon to expand the vagina. The catheter was sutured and fixed, and a 120 g weight was suspended in the drooping injection port. The catheter was removed after 4 hours of traction, and both ovaries were excised with a midline incision after establishing the model.

In the experiment, two urodynamic measurements were made (to reduce the discomfort of the rats, the measurements were made after anesthesia): one week after modeling and one day after the end of treatment. After emptying the bladder, a 0.7 mm epidural catheter was inserted 2 cm into the rat bladder before fixing it. One end was connected to a microinjection pump through a three-way connector to the bladder, and sterile saline was injected at a speed of 0.3 ml/min. One end was connected to a urodynamic testing instrument (Bonito urodynamic detector, Laborie Medical Technology Company of Canada). The urodynamic testing instrument was calibrated to “0”, and the test was started. When the first drop of fluid appeared at the urethral orifice, perfusion was stopped. At this time, the bladder capacity (perfusion time multiplied by perfusion speed) and intravesical pressures were recorded to determine the leak point pressure (LPP) and maximum bladder capacity (MBC). The sneeze test was carried out after the first urodynamic measurement (16). The success models (n = 17) were randomly assigned to 3 groups: the M (n = 5), NAAC (n = 6), and EA (n = 6) groups.

After the first urodynamic measurement, the N and M groups did not receive any treatment. The EA group was treated by electrical stimulation (density wave, 4/20 Hz, 20 min) at the acupoints of Shenshu (BL23, on both sides of the spinous process on the lower back) and Huiyang (BL35, positioned almost vertically underneath the periosteum approximately 5 mm lateral to the midline of the coccyx, bilateral symmetry), determined relative to their anatomical locations described in the WHO guidelines for human acupoints (17). A disposable acupuncture needle (diameter, 0.25 mm) was connected to an SDZ-V EA instrument (Shanghai, China). EA treatment was performed once a day and continued for 3 d. For the NAAC group, stimulation was at the Shenshu control point (1 cm beside Shenshu) and at the Huiyang control point (1 cm beside Huiyang). The stimulation settings were the same as in the EA group. The second urodynamic measurements were made after the above treatments. Rat stool samples were collected and stored, and 16S rRNA sequencing was performed to analyze the changes in SUI-related intestinal flora.

After the experiment, 1-2 stool samples each from 4 rats randomly selected from each group were collected using the Omega kit (Omega Bio-Tek, Norcross, GA, U.S.). For DNA extraction, a NanoDrop 2000 was used to assess the DNA purity and concentration, and agarose gel electrophoresis was used to assess DNA integrity. The primers 338F (5’-ACTCCTACGGGAGGCAGCAG-3’) and 806R (5’-GGACTACHVGGGTWTCTAAT-3’) were used for PCR amplification of the V3-V4 variable region of 16S rRNA as follows: denaturing at 95°C for 30 s, annealing at 55°C for 30 s, and extension at 72°C for 30 s for 27 cycles, and a final extension at 72°C for 10 min (PCR instrument: ABGeneAm p^® 9700). The amplification reaction consisted of 4 µl of 5* Fast Pfu buffer, 2 µl of dNTPs at 2.5 mM, 0.8 µl of primer (5 µM), 4 µl of Fast Pfu polymerase, and 10 ng of DNA template.

Samples were sequenced and analyzed on an Illumina MiSeq platform (Illumina, San Diego, USA) according to the standard protocols by Majorbio Bio-Pharm Technology Co., Ltd. (Shanghai, China). The raw reads were quality controlled with Trimmomatic software, and FLASH software was used for read assembly. Operational taxonomic units (OTUs) were identified at a 97% similarity cutoff using UPARSE (version 7.1, http://drive5.com/uparse/), and chimeric sequences were identified and removed. The taxonomy of each representative OTU sequence was analyzed by RDP Classifier (http://rdp.cme.msu.edu/) against the 16S rRNA database (Silva SSU132) using a confidence threshold of 0.7.

Data were analyzed with GraphPad Prism 6.0. One–way analysis of variance (ANOVA) was used to assess the difference between groups, followed by pairwise comparison using the least significant difference (LSD) test. P<0.05 indicated that the difference was statistically significant. The measured data are expressed as the mean ± SEM. The rarefaction curves and alpha diversity indices referring to community richness (Chao1) and community diversity (Shannon) were based on the OTU information. The species composition is represented in a histogram. Diversity among groups was shown by the phylogenetic beta diversities calculated by partial least squares discriminant analysis (PLS-DA) with unweighted UniFrac distances. ANOSIM/Adonis analysis was used to determine whether the grouping was meaningful by testing if the difference between groups was significantly greater than those within groups. Based on a normalized relative abundance matrix, the linear discriminant analysis (LDA) effect size (LEfSe) method based on the Kruskal–Wallis test was used to identify potential biomarkers. The LDA threshold was 2.0, and the significance α was 0.05.

The effect of EA on LPP in rats with SUI is shown in Figures 1A, B. Before treatment (Figure 1A), compared with that of the N group (78.83 ± 1.40), the LPPs of the M (55.40 ± 1.29), NAAC (59.83 ± 2.13), and EA groups (62.00 ± 2.50) decreased (P <0.05). After treatment (Figure 1B), the LPP of the M group (58.20 ± 2.15) was significantly lower than that of the N group (78.17 ± 2.27) (P <0.05), and the LPP of the rats with SUI increased significantly after EA treatment (73.50 ± 2.32) (P <0.05), while there was no significant difference between the NAAC group (62.17 ± 1.54) and the M group (P =0.607>0.05). Similarly, the effect of EA on MBC in rats with SUI is shown in Figures 1C, D. Before treatment (Figure 1C), compared with that of the N group (1.21 ± 0.11), the MBCs of the M (0.76 ± 0.07), NAAC (0.60 ± 0.06), and EA groups (0.81 ± 0.10) decreased (P <0.05). After treatment (Figure 1D), the MBC of the M group (1.08 ± 0.05) was significantly lower than that of the N group (1.58 ± 0.24) (P <0.05), and the MBC of the rats with SUI increased significantly after EA treatment (1.50 ± 0.08) (P <0.05), while there was no significant difference between the NAAC (1.11 ± 0.06) group and the M group (P =1.000>0.05).

MiSeq high-throughput sequencing was used to analyze 4 groups of rat stool samples. The alpha exponential rarefaction curve of the Chao1 index (Figure 2A) and the Shannon index (Figure 2B) showed clear asymptotes, which indicated near-complete sampling of the community. Violin plots of the Chao index (Figure 2C)

and the Shannon index (Figure 2D) exhibited no significant differences among the four groups. According to ANOSIM/Adonis analysis (R = 0.4721, P = 0.001) and PLS-DA (Figure 2E), there was a significant difference in the species community composition among the groups.

Figure 3A shows the OTU clustering results detailing the relative abundances of the bacterial community at the genus level in each group of rat stool samples. We found that the dominant bacteria in the N group and the EA group were the same, namely, Bacteroides and norank_f_Muribaculaceae. However, the mean proportions were different. Figure 3B shows the top 15 species with the most abundant expression at the genus level, in which the relative abundance of blautia is statistically different among groups (*0.01< P < =0.05, **0.001< P < =0.01), The comparison of the relative abundance of blautia in rats of each group is shown in Figure 3C. Blautia accounted for 6.68%, 0.37%, 1.36% and 2.52% of the microbial communities of the N group, M group, NAAC group and EA group,respectively, indicating that the content of Blautia increased after EA treatment.Electroacupuncture stimulation can significantly increase or decrease the relative abundance of blautia in Sui rats, but electroacupuncture control group can not reverse this change.Further analysis of 15 different bacteria in the N and M groups at different taxonomic levels by LEfSe (Figure 4) revealed that the relative abundances of the Allobaculum and A2 genera in the M group increased, while the relative abundances of Bacteroides, Bacteroidaceae, Blautia, f_norank_o_Rhodospirillales, norank_o_Rhodospirillales Anaerostipes, g_Ruminiclostridium_1, g_Candidatus_ Soleaferrea, g_GCA_900066225, Erysipelatoclostridium, Holdemania, Burkholderiaceae, and g_Marvinbryantia decreased (LDA = 2.0, P<0.05).

Spearman correlation analysis of a heatmap showed the relationship between microbial flora and urodynamics (Figure 5). The bacterial genera positively correlated with MBC were the Lachnospiraceae NK4A136 group and Ruminococcaceae_UCG-014. The bacterial genera positively corelated with LPP were Blautia, Ruminococcus_torques_group and Prevotellaceae_Ga6A1_group, while Parabacteroides, norank_f_Lachnospiraceae and Coprococcus_2 were negatively correlated with LPP. The results of Spearman correlation analysis for the N, M, and EA groups are shown in Table 1. The treatment of SUI rats in the EA group was positively correlated with the relative abundance of the genera Blautia and Prevotellaceae_NK3B31_group in terms of the MBC, LPP and sneezing experiments and negatively correlated with the relative abundance of Parabacteroides.