Innate Immunity

Erythrocytes have been suggested to bind chemokines, thus limiting their effective concentration in the circulation. We investigated binding of the chemokines IL-8 and RANTES to erythrocytes following endotoxin challenge. Endotoxin caused a transient increase of IL-8 in plasma, and a gradual increase of erythrocyte-bound IL-8, remaining elevated for at least 24 h in vivo. Binding of IL-8 to erythrocytes was confirmed by incubation of freshly isolated erythrocytes with endotoxin stimulated plasma. Incubation for 1 h led to a decrease in plasma concentrations of IL-8 with a concomitant increase in erythrocyte-bound IL-8 levels. Erythrocyte-bound RANTES was induced early following endotoxin infusion and our data may suggest competition between the two chemokines for the same binding site on erythrocytes in vivo. In conclusion, in humans, both IL-8 and RANTES are induced by endotoxin and are, at least in part, cleared from the plasma by binding to erythrocytes.

We conducted a prospective study of a chromogenic LAL assay in 346 patients with sepsis syndrome, as defined by a modification of the Bone criteria, and 131 healthy volunteers at eight member centers of the Academic Medical College Consortium (AMCC). We identified patients with endotoxemia (> 0.40 EU/ml) by measuring LAL-reactive material in whole blood, extracted by the Tamura nitric acid method, with the chromogenic LAL (Pyrochrome®) assay. The mean result in sepsis patients with detectable endotoxemia (n = 241) was 1.07 ± 1.57 EU/ml, and the mean result in 131 volunteers was 0.151 ± 0.113 EU/ml, with 73% of the volunteers' results falling below the detectable limit. The average incidence of endotoxemia in sepsis patients was 33%, but varied 2.7-fold among the clinical centers (range 16-44%). Assay results were repeatable when samples tested frozen at the clinical sites were compared to results on frozen samples tested at Associates of Cape Cod, Inc. (ACC). Multiple samples were obtained from 40 patients at 18-24 h interval(s). Fourteen multidraw patients (35%) were endotoxemic at one or more draw(s) and eight of these patients had two or more draws with endotoxin levels > 1.0 EU/ml. The presence of sulfa drugs gave false positive results in two patient samples. A positive LAL test did not correlate with culture-proven bacterial infection and did not significantly correlate with mortality. There was a correlation (P = 0.014) between a patient having a positive LAL test and the presence of a fungal infection when mixed fungal and bacterial infections were included. There was no correlation with a positive LAL test when only a fungal infection was present (P = 0.425) or when only a fungal and a Gram-positive infection was present (P = 0.087).

Plasma levels of endotoxin, TNFα, interleukin-1β (IL-1β), and interleukin-6 (IL-6) as well as the endotoxin-induced release of these cytokines from whole blood were determined in 20 patients after elective aseptic surgery and eight patients suffering from septic disease. Blood samples were drawn before surgery as well as once daily over 7 days postoperatively. Endotoxin plasma levels increased on the day of the operation from 0.025 ± 0.005 EU/ml to 0.099 ± 0.022 EU/ml and remained elevated for 1 day. No changes were noted in the plasma levels of TNFα and IL-1. Levels of IL-6 increased to 198 ± 52 pg/ml on the day of the operation, while the release of TNFα and IL-6 markedly declined after whole blood was stimulated with endotoxin. The difference between basal and stimulated TNF values amounted to 505 ± 129 pg/ml preoperatively and 132 ± 41 pg/ml after the operation and 297 ± 28 pg/ml and 137 ± 43 pg/ml for IL-6. From the second day of the postoperative period onwards, the endotoxin-induced release of IL-6 was higher than the preoperative values. The results of whole blood stimulation in patients with sepsis were similar to those achieved directly after elective surgery. The cytokine release of TNFα, IL-1β, and IL-6 was reduced. We conclude that, during the early aseptic acute phase reaction, the changes in the functioning of monocytes resemble - qualitatively - those occurring during septic disease, and that this may be related to endotoxemia originating in the gastrointestinal tract.

Damage to the intestinal mucosal barrier sustained during the compensated stages of shock may result in bacterial translocation and systemic accumulation of toxins, leading to sepsis and multiple organ failure. Neutrophils are known to be an important component of endotoxin and cytokine-induced tissue injury. The purpose of this study, therefore, was to test the hypothesis that neutrophil infiltration and MPO production accompany endotoxin and TNFα induced intestinal mucosal dysfunction.

To evaluate mucosal dysfunction, jejunal blood-to-lumen clearances of chromium 51-labeled ethylenediaminetetraacetate (⁵¹Cr-EDTA) were measured in cats administered fetal calf serum (controls), endotoxin, TNFα, or polyclonal anti-TNFα antibody followed by endotoxin. Neutrophil accumulation in the jejunal mucosa was evaluated by histologic enumeration and measurement of jejunal myeloperoxidase (MPO) activity.

Endotoxin and TNFα induced mucosal dysfunction in the proximal jejunum is accompanied by neutrophil accumulation in the jejunal mucosa as indicated by histologic enumeration and increased MPO activity. Pretreatment with anti-TNFα polyclonal antibody significantly decreased endotoxin-induced increases in jejunal MPO activity. The results of this study support the contribution of neutrophils to endotoxin and TNFα induced mucosal dysfunction.

Deletion homologues of the cloned Factor C cDNAs from the horseshoe crab Carcinoscorpius rotundicauda were engineered to express in Saccharomyces cerevisiae under the regulation of a galactose-inducible promoter. Expression cassettes were constructed in the vectors: pEMBLyex4 and YEpsec1 to direct, respectively, the intracellular expression, and the secretion of the protein into the culture medium using a heterologous signal sequence. The effect of insert size on the efficiency of expression and the functionality of the resulting recombinant Factor C (rFC) were studied by creating expression constructs bearing various deletion and/or hybrid fragments of Factor C. Removal of the long 5' UTR from the Factor C cDNA improved expression of the rFC. 3' Deletions of up to 84%, or internal deletions of 65% of the Factor C cDNA resulted in either the lack of detectable amounts of Factor C or loss of immunoreactivity. Depending on the construct, full length or partial rFC-related proteins were correspondingly expressed intracellularly, regardless of the vector. The rFC partitioned with the insoluble cell fraction, was solubilised with either SDS or Triton X-100, and found to be immunoreactive. The rFCs were functionally active, being able to bind Gram-negative bacterial endotoxin, provided critical regions of the endotoxin-binding domain were preserved.

This work uses a novel heart-perfusion technique to measure [³H]-lipopolysaccharide ([³H]-LPS) binding on capillary endothelium and myocyte cell membranes in Sprague-Dawley rats. Free or serum-containing Ringer-Lock buffer was infused at a rate of 1 ml/min and in the presence of 20 mM K⁺ and [ ³H]-LPS through an aortic cannula, and the effluent was collected through a catheter introduced into the right atrium cavity. The capillary endothelial lining was removed by CHAPS treatment to expose the cardiac myocyte surface. A physical model describing 1:1 binding stoichiometry of [³H]-LPS with its receptors is proposed and the mathematical equations derived allow for the calculation of binding constants (k_n), reversal constants (k_-n), dissociation constants (k_d), and residency time constants (τ). The results showed that the presence of serum in the perfusate, slowed the binding of [³H]-LPS with the endothelial lining and myocytes, but increased the residency time by 3- and 50-fold, respectively. Hence, the endothelium and myofiber may contain LPS receptors that can bind more strongly with the ligand in association with sCD14-like and LBP-like molecules in rat serum. Thus it is postulated that the affinity of LPS to its receptor subtypes is not strictly and specifically dependent on the CD14 binding profile.

Binding of pig hemoglobin (Hb) to Actinobacillus pleuropneumoniae lipopolysaccharide (LPS), either extracted or present at the surface of whole cells, was studied. After a short incubation period, pig Hb seemed to cover the bacterial cell surface and enhanced the cells' contrast when examined by transmission electron microscopy (TEM). Energy-dispersive X-ray spectroscopy analysis showed that the amount of elemental iron detected was increased when cells of A. pleuropneumoniae were incubated with pig Hb. Coating with pig Hb, however, did not interfere with the accessibility of O- and capsular antigens to antibodies on the bacterial cell surface. Binding of pig Hb and polymyxin B to lipid A of A. pleuropneumoniae was confirmed with a fluorescent probe (dansylcadaverine) displacement assay. The binding of pig Hb to extracted LPS resulted in a disaggregation of LPS as observed by TEM after negative staining. Additional evidence for a direct physical interaction between pig Hb and A. pleuropneumoniae LPS was demonstrated by a shift in the sedimentation velocity of LPS-Hb complexes determined by sucrose gradient centrifugation. Pig Hb binding to extracted LPS or to bacterial cells resulted in a significant decrease of chromogenic Limulus amebocyte lysate activation. Finally, the capacity of extracted LPS to induce NO₂ ^- in the presence of pig Hb was tested by using cell line J774 and determined by the Greiss' reaction. LPS alone induced, as expected, NO₂- production, whereas the presence of pig Hb significantly reduced NO₂ ^- production by murine macrophages. Taken together, our results indicate that binding of pig Hb affected some physical and biological properties of A. pleuropneumoniae LPS.