Gram-negative shock in rats depends on the presence of capsulated bacteria and is modified by laparotomy.

To develop a hyperdynamic sepsis model in rats, four Escherichia coli strains were used, which differed in the presence or absence of a capsule or K antigen (K1 and K-, respectively) and/or in O serogroup (O9 and O18). Of the two clinical isolates, O9K- did not survive in rat serum, whereas O18K1 and two isogenic laboratory strains (O18K1 and O18K-) were able to resist serum bacteriolysis. Pentobarbital-anesthetized rats (n = 21) received an intravenous bolus of 10(9) bacteria. In contrast to the two noncapsulated strains, both capsulated strains induced hyperdynamic shock; arterial lactate rose from a mean value of .91 to 3.09 mmol.L-1, systemic vascular resistance dropped from 1.15 to .78 mmHg.min.mL-1, and cardiac output transiently increased from 98 to 115 mL.min-1; renal plasma flow remained at 3-4 mL.min-1, whereas glomerular filtration rate decreased from 1.3 to .7 mL.min-1. Laparotomy, which is often performed to study kidney function, completely abolished the hyperdynamic condition, while glomerular filtration rate was still decreased. We conclude that in rats, in contrast to humans, capsulated bacteria are required to induce a hyperdynamic septic shock; the hyperdynamic characteristics of the shock do not occur in animals subjected to a laparotomy.