We have developed a functional assay to study the inflammatory capacity of plasma collected from patients with severe gram-negative septic shock. In this assay, elutriation-purified, cryo-preserved human monocytes from one healthy donor are combined with plasma from patients with severe persistent septic shock for 5 h. Subsequently, the plasma is removed, medium added, and procoagulant activity (PCA) and secretion of tumor necrosis factor alpha (TNF-alpha) and interleukin 6 (IL-6) measured after 18-h incubation. Plasma from 10 patients (6 died) infected with Neisseria meningitidis previously shown to contain high levels of native lipopolysaccharide (LPS) (median 2,700 pg/ml), TNF-alpha, IL-6, IL-8, and complement activation products, had a low net spontaneous inflammatory capacity on the monocytes. The median levels of PCA, TNF-alpha, and IL-6 were 5, 0, and 4%, respectively, of the monocyte activities induced by normal plasma boosted with purified N. meningitidis (Nm)-LPS (2,500 pg/ml; net LPS-boosted capacity, 100%). The levels of PCA, TNF-alpha, and IL-6 obtained with plasma from shock patients were not different from those induced by plasma from 10 meningococcal patients without shock or with plasma from healthy persons. Boosting shock plasma with 2,500 pg/ml Nm-LPS had little effect on the monocyte activities since the median values of PCA, TNF-alpha, and IL-6 revealed a minimal increase from 5, 0, and 4% to 9, 2, and 6%, respectively. The shock plasmas revealed a strong LPS-inhibitory capacity that was largely absent in plasmas from 10 meningococcal patients without shock since the median levels of PCA, TNF-alpha, and IL-6 increased from 5, 0, and 0% to 135, 51, and 73%, respectively, after boosting with 2,500 pg/ml Nm-LPS. The LPS-inhibitory capacity was closely associated with the levels of IL-10. The median levels of IL-10 were 19,000 pg/ml in nine shock patients vs. 22 pg/ml in nine nonshock patients with systemic meningococcal disease. Removal of native IL-10 by immunoprecipitation restored the capacity of plasmas to induce monocyte activation either by native LPS or by boosting with Nm-LPS. IL-4 and TGF-beta were not detected in shock plasmas. In 24 patients with detectable meningococcal LPS ( > 10 pg/ml, 0.1 endotoxin units/ml), the levels of IL-10 were correlated to the levels of LPS (r = 0.79, P < 0.001). IL-10 declined from initiation of antibiotic therapy and paralleled the levels of native LPS. Decreasing levels of IL-10 in serially collected shock plasmas were directly related to increasing monocyte responsiveness after Nm-LPS boosting. These results suggest that IL-10 plays a major role in containing activation of monocytes and possibly other LPS-responsive cells during overwhelming meningococcemia.