Cerclage prevents ascending intrauterine infection in pregnant mice

Abstract

The treatment for cervical insufficiency is cerclage surgery. Although cerclage is a common therapy for prevention of preterm birth, there is no consensus about its mechanism of efficacy. Previous investigators have hypothesized that cerclage prevents preterm birth by improving the cervical barrier to ascending infection. However, this hypothesis is difficult to study in human pregnancy. In a mouse model of ascending infection, we hypothesized that a cerclage improves the cervical barrier leading to decreased ascending intrauterine infection and inflammation. An abdominal cerclage was studied because a vaginal cerclage is not feasible in mice. To perform an abdominal cerclage, laparotomy was performed on timed, pregnant C57BL/6 mice on gestational day 10 (E10). A 6-0 silk suture was placed around the cervix just below the junction of the 2 uterine horns. Sham controls received the same surgery, but no cerclage was placed. To track ascending infection nonpathogenic E coli K12 was genetically modified to express bioluminescence. On E15, bioluminescent E coli K12 (20 μL of 1×109 bacteria) was inoculated into the vagina. Whole-body bioluminescence imaging was performed 0.5 hours and 24 hours after inoculation. To assess intrauterine inflammation, pathogenic E coli K1 was used. On E15, bacterial inoculums of E coli K1 (20 μL of 1×104 bacteria) were vaginally administered. Samples of uterus, placenta, and fetal membranes were collected 24 hours after inoculation. Gene expression of inflammation-related proteins was compared between 3 groups: (1) sham control surgery+ inoculation of phosphate-buffered saline (PBS), (2) sham control surgery+ inoculation of E coli K1, and (3) cerclage surgery+ inoculation of E coli K1. Abdominal cerclage was well tolerated. No cases of preterm birth were seen following abdominal cerclage. Whole-body bioluminescent imaging performed 0.5 hours post inoculation showed a strong luminescence signal in the vaginal region of mice in both control and experimental groups indicating successful bacteria inoculation. Twenty-four hours after inoculation, bioluminescent signal was seen ascending into the uterine horns in all control mice. However, in mice with abdominal cerclages, no bioluminescent signal was seen after 24 hours. When the reproductive tissues were imaged separately in control mice, strong bioluminescence signal was detected in the placenta, fetal membranes, and uterus. Gene expression studies showed that cerclage significantly decreased the expression of inflammatory proteins induced by E coli K1 in the uterus, placenta, and fetal membranes. In this mouse model of ascending intrauterine infection, abdominal cerclage prevented ascending infection of E coli. In addition, abdominal cerclage prevented expression of inflammatory cytokines in the uterus, placenta, and membranes of mice. The study provides evidence for a potential mechanism of cerclage success in human pregnancy.