A model of bacterially induced umbilical vein spasm, relevant to fetal hypoperfusion.

From experience with placental pathology and autopsies, we hypothesize that infected amniotic fluid may produce umbilical cord and placental surface vasospasm, reduced fetal perfusion, and perinatal asphyxia, morbidity, and mortality. We tested the hypothesis by a preliminary investigation with group B beta-hemolytic streptococcus, a common cause of perinatal morbidity and mortality. Our methods included suspension of isolated human umbilical vein segments in a 30-mL organ bath with modified Krebs solution, pH 7.3-7.4, under continuous aeration by 95% O2 and 5% CO2 at 37C. The vessels were attached to an isometric transducer which was connected to a polygraph. We tested Krebs solution, sterile culture medium, group B streptococcus spent medium, and washed whole group B streptococcus bacteria. Contractility was compared with results produced by 10(-5) M serotonin. The average contraction produced by 30 microL/mL group B streptococcus spent medium was 40.8% of that produced by serotonin. Whole washed bacteria produced 8.6% of the serotonin contraction. Krebs solution and sterile culture medium did not stimulate vascular contraction. Intraperitoneal injection of group B streptococcus spent medium in mice at four times the maximal experimental dose had little effect. Boiling group B streptococcus spent medium failed to affect its vasoactive stimulatory properties. We conclude that in vitro, group B streptococcus synthesizes a heat-stable, non-protein exotoxin that causes human umbilical vein contraction.