Search

N-Acetylmuramyl-L-alanyl-D-isoglutamine (muramyl dipeptide [MDP]) injected intraperitoneally significantly increased the number of cells entering the peritoneal cavity of guinea pigs primed with liquid paraffin or thioglycollate. There was a close relationship between peritoneal polymorphonuclear leukocyte (PMN) accumulation and the uptake of glucosamine by macrophages in guinea pigs treated with a variety of bacterial cell surface components such as cell wall peptidoglycan subunits and bacterial or synthetic lipid A. The PMN accumulation was also facilitated by the intraperitoneal transfer of the peritoneal macrophages that had been stimulated by MDP in vitro. Furthermore, cell-free lavage fluids taken from the peritoneum of MDP-treated guinea pigs also initiated the influx of PMNs when introduced into the peritoneal cavities of liquid paraffin-pretreated guinea pigs. These results suggest that a soluble factor which attracts neutrophils is produced by MDP-treated macrophages. Partial characterization of the factor is described.

The ability of synthetic lipid A analogs to induce prostaglandin synthesis in macrophages was compared with that of native lipopolysaccharide. The synthetic preparations comprised monomeric or dimeric derivatives of D-glucosamine with different patterns of substitution by phosphate and tetradecanoic, (R)-3-hydroxytetradecanoic, and (R)-3-tetradecanoyloxytetradecanoic acid. All of these preparations are structurally distinct from native lipid A (principally regarding the position of fatty acid substitution) and hence have been previously shown to be endotoxically inactive in many biological tests. It was found that many of these synthetic samples exhibit strong activity in inducing prostaglandin E2 and prostaglandin F2 alpha, with some of them having activity comparable to that of lipopolysaccharide. Experiments with macrophages of C3H/HeJ mice enabled us to differentiate between endotoxin-specific (in the case of dimeric preparations) and endotoxin-nonspecific (for monomeric preparations) mechanisms for the induction of prostaglandins. These results indicate that there is a difference in the mechanism of induction of prostaglandin synthesis between monomeric lipid A's and dimeric or native lipid A structures.

A synthetic adjuvant, N-acetylmuramyl-L-alanyl-D-isoglutamine (MDP), failed to produce polyarthritis with a wide dose range in a water-in-oil emulsion of mineral oil such as liquid paraffin, Drakeol, or heavy mineral oil. MDP, however, produced moderate to severe arthritis with almost 100% incidence in a water-in-oil emulsion made up of Difco incomplete adjuvant, which consists of Bavol F as an oil vehicle and Arlacel A as an emulsifier. N-acetylmuramyl-L-alanyl-L-isoglutamine did not produce arthritis, whereas 4,6-diacetyl-MDP produced the disease. Bacterial peptidoglycans, such as disaccharide peptides which were N-acetylglucosaminyl-N-acetylmuramyl-L-alanyl-D-isoglutaminyl-meso-diaminopimelyl-D-alanine and N-acetylglucosaminyl-6,o-acetyl-N-acetylmuramyl-L-alanyl-D-isoglutaminyl-meso-diaminopimelyl-D-alanine, also produced polyarthritis with low incidence in Difco oil but not in the other mineral oils described above. MDP and bacterial disaccharide peptides were able to produce the disease even in the latter mineral oil only when the concentration of Arlacel A was increased from 15% to 20 to 30% in the oil. We concluded that one of the minimal essential structures responsible for development of this disease is MDP, although the role of the oil vehicle remained uncertain, and there is no direct correlation between granulona formation and arthritogenicity of MDP.

Thirteen 6-O-acyl-N-acetylmuramyl-L-alanyl-D-isoglutamines (6-O-acyl-MDPs), including four inactive D-isoasparagine and L-isoglutamine analogs, were tested for their pyrogenicity and immunopotentiating activity to stimulate primary humoral and cellular immune responses in guinea pigs to a model protein antigen, ovalbumin, when administered in various vehicles. Among them, derivatives whose muramic acid residue was substituted by alpha-branched (and beta-hydroxylated) higher fatty acids at the carbon-6 position, especially 6-O-(2-tetradecylhexadecanoyl)-MDP (B3O-MDP) and, to a lesser extent, 6-O-(3-hydroxy-2-docosylhexacosanoyl)-MDP (BH48-MDP) and its L-serine analog [BH48-MDP(L-Ser)], were found to exert strong adjuvant activity in both the induction of delayed-type hypersensitivity and the stimulation of circulating precipitating antibody levels when combined with nonirritating vehicles (liposomes, squalene-in-water emulsion, and phosphate-buffered saline). These vehicles did not efficiently support the adjuvant activity of MDP, the parent molecule of the above lipophilic derivatives. Pyrogenicity tests showed that introduction of alpha-branched higher fatty acid groups but not of straight, long-chain fatty acids at the 6-position of the muramic acid residue resulted in marked decrease of the pyrogenicity inherent to MDP via intravenous administration.

Synthetic lipid A analogs (compounds 404 through 406) were examined for their immunopharmacological activities. These compounds had two amide-bound and two ester-bound (R)-3-hydroxytetradecanoyl groups at the C-2 and C-2' and the C-3 and C-3' positions, respectively, of beta (1-3)glucosamine disaccharide. In all of the in vitro assays, these synthetic compounds exhibited high activities comparable to those of a reference lipid A prepared from Escherichia coli O8:K27 Re-mutant strain F515. The compounds activated the clotting enzyme cascade of the horseshoe crab, activated the human complement via the classical pathway, caused polyclonal B-cell activation, stimulated the phagocytosis of sheep erythrocytes by murine peritoneal macrophages, and enhanced the migration of human polymorphonuclear leukocytes. They also increased the thymidine uptake of splenocytes of BALB/c nu/nu and C3H/HeN mice but not those of C3H/HeJ (a nonresponder to lipopolysaccharide). A dephosphorylated derivative, compound 403, was barely active in all of the above assays except for the enhancement of polymorphonuclear leukocyte migration. However, compounds 404 through 406 were feeble in pyrogenicity and could not prepare the local Shwartzman reaction, although they were very lethal to galactosamine-loaded mice. Therefore, synthetic lipid A analogs described here were fully immunopharmacologically active in in vitro assays, but all of them were far less active than natural E. coli F515 lipid A regarding the biological activities characteristic of endotoxic lipopolysaccharides and lipid A's. The high lethal toxicity of compound 406 (1,4'-bisphosphate) to the galactosamine-loaded mice may not reflect its real toxicity to normal mice. In all activities examined, compound 406 was quite comparable to a biosynthetic lipid A precursor, a natural counterpart of compound 406. The immunopharmacological activities of these newly synthesized lipid A analogs, especially compound 406, were much stronger than those of compounds that had been synthesized earlier by using the originally proposed model of the lipid A structure. The findings described in this report justify the acylation pattern of a disaccharide backbone of lipid A, revised on the basis of recent analytical studies. The low in vivo endotoxic activities of the present lipid A analogs are most probably due to the fact that the kinds of acyl groups were different from those of the complete lipid A from E. coli, although there were no differences in the acylation positions on the disaccharide backbone.

The experimental system utilized in investigating the correlation between the chemical structures of muramyl peptides and their protective activities in the sepsis type of systemic infections caused by Escherichia coli was applied in evaluating the enhancement of resistance to infection induced by 32 synthetic glycopeptide analogs, including 6-O-acyl derivatives and 1-alpha-O-benzyl derivatives of muramyl dipeptide (N-acetyl muramyl-L-alanyl-D-isoglutamine). In assessing the 6-O-acyl derivatives of muramyl dipeptide, we found that the degree of protective activity was attributable to the kinds of fatty acids introduced. Acylation of the 6-hydroxy group on the muramic acid moiety in muramyl dipeptide with natural mycolic acid or a synthetic fatty acid possessing either an alpha-branched or an alpha-branched, beta-hydroxylated group resulted in a decrease in or a disappearance of the protective activity of muramyl dipeptide. Acylation with a normal fatty acid or an iso fatty acid resulted in a retention or enhancement of muramyl dipeptide activity. The activity of acylated derivatives containing linear fatty acids was stimulated by increasing the chain length up to 18 carbon atoms. The highest degree of protective activity occurred with the derivatives acylated with straight-chain fatty acids, particularly with the derivatives acylated with palmitic acid and arachidic acid. Benzylation of the 1-hydroxy group of muramyl dipeptide resulted in a decrease in or a loss of protective activity.

Muramyl dipeptide (MDP), a part of bacterial cell wall peptidoglycans, was active as adjuvant and stimulated the reticuloendothelial system (RES) of mice to increase its phagocytic function. A series of analogs of MDP was tested for their adjuvant activity and RES-stimulating activity. Sex differences were observed in the adjuvant activity and RES-stimulating capacity of some MDP analogs. A stereochemically highly specific structure required for MDP to exert adjuvant activity was also required for its RES-stimulating activity. Based on this close correlation among the structure, adjuvant activity, and RES-stimulating capacity of MDP, we infer that macrophages may play an important role in the expression of adjuvant activity of MDP.

Thirteen acylated and phosphorylated derivatives of beta-1,6-linked glucosamine disaccharide (lipid A analogs), which were synthesized after the structural model of Salmonella-type lipid A, and seven similar derivatives of glucosamine monosaccharide (lipid A-related compounds) were studied for their immunobiological activities. These included mitogenicity and polyclonal B cell activation enhancement of migration of monocytes and polymorphonuclear leukocytes derived from human peripheral blood, stimulation of guinea pig peritoneal macrophages, activation of human complement, and stimulation of serum antibody production and induction of delayed-type hypersensitivity against ovalbumin in guinea pigs. Comparisons were made with lipid A, RE-glycolipid, lipopolysaccharide of natural sources, and a well-known synthetic adjuvant, N-acetylmuramyl-L-alanyl-D-isoglutamine. Some of the lipid A analogs were found to manifest the mitogenic, polyclonal B cell-activating macrophage-stimulating, complement-activating, and immunostimulating activities, although the observed activities were generally far less than those of natural products in intensity and efficiency. Other immunobiological effects exhibited by most of the synthetic lipid A analogs were the enhancement of migration of monocytes and polymorphonuclear leukocytes. It is premature to draw definite conclusions on structure-activity relationships, since a few compounds which were active in some assay systems were scarcely active in other assays. However, an indisputable fact was that beta-1,6-glucosamine disaccharide 1 alpha,4'-diphosphate, which carries two amide-bound (R)-3-hydroxytetradecanoyl and three ester-bound tetradecanoyl residues, and thus has the structure most closely resembling natural lipid A among test compounds in this study, was definitely active in all of the present assay systems. However, its potency was generally much less than natural products. Some of glucosamine monosaccharide derivatives, especially N-(R)-3-[(R)-3-hydroxytetradecanoyloxy]tetradecanoyl glucosamine, also exerted all of the in vitro activities described above. This fact suggests that a glucosamine disaccharide structure may not necessarily be a prerequisite as far as the in vitro immunobiological activities tested are concerned.

In the capillary tube migration system a synthetic muramyl dipeptide (MDP; N-acetylmuramyl-L-alanyl-D-isoglutamine), a part of bacterial cell wall peptidoglycans, inhibited the migration of peritoneal exudate macrophages from normal guinea pigs or rats. The migration inhibition was also caused by some MDP-containing peptidoglycan fragments from cell walls of Lactobacillus plantarum and Staphylococcus epidermidis. The migration inhibition could not be explained on the basis of macrophage migration inhibitory factor. A stereochemically highly specific structure of MDP required for its adjuvant activity was also required for the macrophage migration inhibition. These findings suggest that MDP and MDP-containing cell wall fragments may activate macrophages and that this activation may be important in the exertion of their adjuvant activity.

Chemically synthesized lipid A analogs were investigated for several endotoxic activities, including pyrogenicity, lethal toxicity, anticomplement activity, and the capacity to gelate Limulus amoebocyte lysate in comparison to natural lipid A. The synthetic preparations contained D-glucosamine or D-glucosamine-beta-1,6-D-glucosamine disaccharide substituted by ester- and amide-bound hydroxylated or non-hydroxylated fatty acids and by phosphate groups in different combinations. Some preparations which were insoluble in water were succinylated and thus rendered more soluble. Strong biphasic pyrogenic responses with a maximal increase in body temperature of 1 to 2 degrees C were obtained with 50 micrograms/kg doses of 3 disaccharide preparations of 15 tested. With two preparations (50 micrograms/kg) moderate pyrogenicity with monophasic fever curves and a maximal temperature increase of about 0.6 degrees C was obtained. Lethal toxicity tests were carried out in galactosamine-sensitized mice. Of 15 synthetic preparations, 4 exhibited lethal toxicity under these conditions. The effective doses of the lipid A analogs in both in vivo tests were, however, several hundred times higher than those of bacterial lipid A. For the activities in vivo, hydroxyacyl residues seemed to be important. Anticomplement activity was demonstrable in seven preparations, one of which expressed an activity comparable to that of lipid A. Preparations containing non-hydroxylated fatty acids seemed to be most active in this test. None of the synthetic preparations was found to exhibit gelation activity for Limulus amoebocyte lysate when tested in doses up to 0.4 micrograms, whereas bacterial free lipid A was active in doses of about 2 pg. None of the monosaccharide derivatives exhibited any of these activities.

Bacterial cell walls, water-soluble fragments of the wall peptidoglycan, N-acetylmuramyl-L-alanyl-D-isoglutamine (MDP), and 6-O-acyl derivatives of MDP were examined for migration-stimulating activity on human peripheral blood monocytes by using a multiwell chemotaxis assembly. Cell walls isolated from 11 bacterial species caused a definite increase in monocyte migration, but the walls of Micrococcus lysodeikticus were scarely active. The migration-enhancing activity of Staphylococcus epidermidis cell walls was retained by a monomer as well as a polymer of disaccharide peptides which were prepared by digestion of the peptidoglycan with enzymes. It was finally revealed that the migration of monocytes was enhanced by MDP. 6-O-Octadecanoyl-MDP, 6-O-(2-tetradecylhexadecanoyl)-MDP, and 6-O-(3-hydroxy-2-docosylhexacosanoyl)-N-acetylmuramyl-L-seryl-D-isoglutamine were active, but to a lesser extent. A checkerboard assay demonstrated that the increased monocyte migration caused by S. epidermidis cell walls was directed toward a positive stimulus (chemotaxis).

A study was made on the activity of various bacterial cell walls and peptidoglycans to liberate serotonin from rabbit blood platelets. All of the test cell walls or peptidoglycans prepared from 27 strains of 21 bacterial species were shown to cause a marked release of serotonin, regardless of differences in types of peptidoglycan and non-peptidoglycan moieties and in some biological properties. The assay made with the water-soluble "digests" of Staphylococcus epidermidis cell wall peptidoglycans, which were prepared by use of appropriate enzymes, revealed that a polymer of peptidoglycan subunits (a disaccharide-stempeptide) was definitely active in the release of serotonin, but a structural unit monomer was inactive. Among a variety of synthetic muramylpeptides and their 6-O-acyl derivatives, only 6-O-(3-hydroxy-2-docosylhexacosanoyl)-N-acetylmuramyl-L-alanyl-D-isoglutaminyl- L-lysyl-D-alanine was found to hold a strong serotonin-liberating activity.

Several synthetic acylated glucosamine monophosphates, with structures corresponding to the nonreducing or reducing moiety of the lipid A of the Escherichia coli or Salmonella minnesota type, and a synthetic compound corresponding to a biosynthetic disaccharide lipid A precursor (designated Ia or IVA) were examined for their endotoxic and related bioactivities in comparison with those of the synthetic and bacterial parent molecules, i.e., acylated beta(1-6)-D-glucosamine disaccharide bisphosphates. Some of the test monosaccharide compounds were definitely active in most of the in vitro assays. Their activities, except for complement activation, however, were weaker than those of the reference compounds, synthetic and bacterial acylated disaccharide bisphosphates. The differences between the test monosaccharide and disaccharide compounds were much more apparent in in vivo assays, in which the test acylated glucosamine monophosphates were scarcely active, though some test compounds exhibited weak lethal toxicity in galactosamine-loaded mice and were weakly active in pyrogenicity, immunoadjuvant activity, and possible tumor necrosis factor and alpha and beta interferon-inducing ability in Mycobacterium bovis BCG- and Propionibacterium acnes-primed mice, respectively. Mixture at an equimolar ratio of acyl glucosamine monophosphates, each of which has the structure of the reducing or nonreducing moiety of the reference disaccharide compound, did not restore the endotoxic or associated bioactivities of the corresponding parent molecules. No essential differences in bioactivity were noted between synthetic and bacterial monosaccharide compounds, i.e., lipid X, whose structure corresponds to the reducing moiety of E. coli-type lipid A.

A synthetic compound (compound 516), beta(1-6)-linked D-glucosamine disaccharide 1,4'-bisphosphate, which is acylated by (R)-3-hexadecanoyloxytetradecanoyl, (R)-3-hydroxytetradecanoyl, (R)-3-dodecanoyloxytetradecanoyl, and (R)-3-tetradecanoyloxytetradecanoyl groups at positions 2,3,2', and 3', respectively, exhibited in vitro antigenic reactivity of high specificity comparable to that of free lipid A from Salmonella minnesota R595. This was confirmed by an enzyme-linked immunosorbent assay and an enzyme-linked immunosorbent assay inhibition test with monoclonal and conventional antibodies. The results of comparative analysis performed with several synthetic lipid A analogs as well as three monosaccharide derivatives suggested that the complete structure involving both phosphate groups at the C-1 and C-4' positions and the 3-acyloxyacyl groups at the C-2, C-2', and C-3' positions of the glucosamine disaccharide are required for the expression of the serological specificity of Salmonella-type lipid A. This was deduced from the observations that compound 506, a synthetic Escherichia coli-type lipid A which has the same structure as that of compound 516, except that 3-hydroxytetradecanoyl group is substituted for an acyloxyacyl residue at the C-2 position, exhibited significantly reduced antigenic reactivity as compared with compound 516 and that the replacement by the hydrogen atom of the phosphoryl group at the C-1 position or by 3-hydroxytetradecanoyl or tetradecanoyl groups of acyl residues at the 2, 3, 2', and 3' positions of compound 516 results in a marked reduction of reactivity with monoclonal antibodies 5G and 36G. Similar results were obtained by assays with conventional rabbit antibodies, but the structural difference between compounds 516 and 506 could not be distinguished by these polyclonal antibodies. The results of cross-reactions among synthetic analogs with monoclonal antibodies 161M and 1-9M, which have been confirmed to exhibit different serological specificities from the 5G or 36G antibody, also suggested that the nature and linkage of fatty acyl residues as well as the backbone structure of lipid A play an important role in determining serological specificity of the lipid A molecule.