Search

Recently published studies showed that human hemoglobin (Hb) forms stable complexes with bacterial endotoxin (lipopolysaccharide, LPS), and that these complexes have much greater biological activity than LPS alone. LPS in LPS-Hb complexes has greater filterability, is less turbid, and has a lower sedimentation velocity compared to LPS alone, changes consistent with smaller LPS size. To further characterize the physical changes in LPS induced by Hb, we examined the ultrastructure of LPS alone and LPS-Hb complexes by negative staining. We studied two rough (Re) LPS and one smooth LPS, before and after incubation with purified human HbA0. Salmonella minnesota (Re) 595 LPS was visibly turbid, and was composed mostly of ribbon-like and mesh-like structures. In addition, this LPS contained large irregular membrane-like masses, and a few smaller discoidal particles. After incubation with Hb, none of the large mesh-like, ribbon-like or membrane-like structures remained; the resultant LPS particles were discoidal particles of 5-20 nm. Escherichia coli (Re) LPS alone was comprised predominantly of ribbon-like structures; after incubation with Hb, this LPS was comprised of circular membrane-like sheets (predominantly 40-200 nm). Smooth E. coli 055:B5 LPS alone contained heterogeneous particles of 10-100 nm. Most of the largest of these particles were disintegrated by the addition of Hb. The time course of Hb-induced LPS disaggregation was studied with S. minnesota LPS because the great turbidity of this LPS allowed us to monitor its disaggregation spectroscopically. Turbidity was decreased by 50% within approximately 5 h, and was not measurable by 25 h. We propose that the toxic domains of LPS (i.e. lipid A) are cryptic in native LPS aggregates and become exposed upon binding with Hb, thereby resulting in increased biological activity of LPS-Hb complexes compared to LPS alone.

The protein product of the early response gene, Egr-1, is required for macrophage differentiation. Egr-1 mRNA is induced by lipopolysaccharide within minutes in murine peritoneal macrophages through an effect on the rate of transcription of Egr-1. Therefore, we undertook a series of studies to characterize Egr-1 protein induced by LPS and to define enhancer elements in the Egr-1 promoter which mediate the effect of LPS in murine macrophages. Salmonella minnesota Re 595 LPS (10 ng/ml or higher) induced Egr-1 protein of 80 kDa within 45 min in thioglycollate-elicited peritoneal macrophages. The promoter region of Egr-1 contains six serum response elements (SRE) which each contain a core CArG box. CAT activity in LPStreated macrophages transfected with Egr-1-CAT constructs containing the two most upstream 5' SREs was increased 4.7-fold whereas CAT activity in macrophages transfected with constructs lacking these two most upstream SREs was not affected by LPS. LPS induced CAT activity in macrophages transfected with constructs containing tetramers of each of the two most 5' SREs upstream of a heterologous promoter linked to CAT. Mutation of the CArG box abolished the effect of LPS. These findings suggest that the CArG box elements within the two most upstream SREs may mediate the effect of LPS on transcription of Egr-1.

The phenol-chloroform-petroleum ether (PCP) extraction is the method of choice for isolating LPS chemotypes from rough mutant bacteria. However, we have observed that a high percentage of PCP-purified LPS preparations contain trace levels of protein contaminants, and these protein contaminants may exhibit endotoxin-like activity. To obtain protein-free rough LPS, a modified phenol-water extraction procedure was developed for use as a final step to follow PCP extraction. Using this procedure, Salmonella minnesota Ra, Rc, Rd and Re LPS were recovered in yields of 82-100% from the original PCP-extracted preparations. Yields were determined by recovery of KDO and supported by analysis of silver stained SDS-PAGE gels. Although the original PCP-purified chemotypes were contaminated by major 41, 38.5, 29.5, 23, 17 and 14 kDa proteins (detected by Western blotting and gold staining), the repurified LPS contained no detectable protein. The repurified LPS retained all of its original bioactivity as defined by the ability to stimulate normal C3H/OuJ macrophages to secrete TNF. The importance of utilizing repurified LPS in biologic studies was illustrated by the loss of > 99% of activity when repurified LPS was used to stimulate TNF secretion by LPS 'hyporesponsive' C3H/HeJ macrophages.

The ability of rat serum to inactivate endotoxin (LPS) was assessed with the aid of the limulus amebocyte lysate assay. Following the addition of various amounts of endotoxin to normal serum the mixture was incubated for 1 hr at 37 degrees C and the residual endotoxin activity determined. One milliliter of rat serum inactivated between 5 and 10 micrograms Escherichia coli LPS per hour. Heating serum for 45 min at 56 degrees C resulted in loss of 80-90% of the LPS inhibitor (LPSI) activity. Serum from cobra venom factor (CVF)-treated rats inactivated between 0.5 and 2.5 micrograms LPS/ml serum. Serum from tolerant rats, even after heating for 45 min at 56 degrees C, inactivates between 10 and 15 micrograms LPS/ml serum/hr; decomplemented tolerant rat serum neutralizes between 5 and 10 micrograms LPS/ml serum/hr. Clearly, the tolerant rat has large quantities of LPSI activity, which does not appear to be complement. The inhibitor found in tolerant rat serum is not species specific since it inactivates Salmonella minnesota and Salmonella typhimurium endotoxins to the same degree and in the same amount as E. coli endotoxin, the agent used to induce tolerance. Both heating serum (56 degrees C) and lead acetate reduce LPSI activity.

The clinical toxicity of acetic acid-treated « Salmonella minnesota » R595 (Re) organisms was evaluated in 24 cancer patients. Bacteria were injected i.v. four times at increasing doses for a total of 6.5 μg. This therapeutic regimen was free of major side effects (one patient had fever higher than 38 °C and 10 patients complained of pruritus). Furthermore, this bacterial preparation which possesses a more exposed lipid A on its surface, exhibited immunomodulating capacities in that it normalized the inverted T helper/T suppressor ratio and enhanced natural killer activity in tumor patients. The mechanisms of the lower toxicity and immunomodulating activities of these bacteria compared to other lipid A preparations are discussed.