Bacterial Lipopolysaccharides: a Remarkable Family of Bioactive Macroamphiphiles?
Chapter
69
CHRISTIAN R. H. RAETZ
Lipopolysaccharides (LPSs) are immunogenic glycolipids that make up the outer surface of the outer membranes of gram-negative bacteria, including Escherichia coli and Salmonella enterica (3, 145, 182, 193, 226, 241, 257). The elucidation of the biosynthesis, secretion, and function of LPSs remains one of the great challenges of prokaryotic biochemistry (226, 257). Because of their large size, amphiphilic composition, and microheterogeneity (M r, ∼2,000 to 50,000), it is difficult to determine the complete covalent structures and conformations of LPSs (3, 145, 182, 226, 241, 257). At least 50 genes are required for the assembly of LPS (257).
LPS consists of three covalently linked domains (Fig. 1 to 3): (i) lipid A (endotoxin), which functions as the hydrophobic anchor for LPS in the outer membrane and is the bioactive component responsible for some of the pathophysiology associated with severe gram-negative infections (145, 182, 226, 227, 242, 243); (ii) the core region, a phosphorylated nonrepeating oligosaccharide required for the function of the outer membrane as a barrier to antibiotics (193, 226, 241, 257); and (iii) the O-antigen polymer, an immunogenic repeating oligosaccharide of 1 to 40 units that varies greatly from strain to strain (3, 236, 241). O antigen is not normally synthesized by E. coli K-12 (269).
In the past 10 years, enormous progress has been made in determining the exact chemical structure (145, 182, 226, 241, 242) and enzymatic synthesis (226, 227) of LPS. Much has been learned about the molecular genetics of LPS assembly (226, 227, 236, 257), setting the stage for studies of LPS function in bacteria. The structural requirements for the bioactivity of lipid A-like molecules as immunostimulants and endotoxins have been defined (227, 243). Some of the proteins in animal cells that bind lipid A and initiate signal transduction have been identified (90, 145, 227, 231, 243, 316). The pathophysiological consequences of LPS administration to animals, especially the role of cytokines, are better understood (145, 180, 181, 192, 207). Given this progress, it is disappointing that effective therapies for LPS-induced complications of gram-negative sepsis in humans have not emerged (180, 192, 270). However, the most innovative approaches, such as use of synthetic lipid A analogs that block the actions of endotoxins (34) or chemicals that inhibit LPS-triggered signal transduction (139, 199), have yet to be tested in septic patients.
This chapter presents a critical account of what is known about LPS biochemistry and molecular biology in E. coli K-12 and S. enterica serovar Typhimurium LT2 (hereafter referred to by the former, more commonly used designation, Salmonella typhimurium LT2). Many of the conclusions pertain to diverse gram-negative bacteria, including common human pathogens, photosynthetic organisms, and strains of Rhizobium. Several books (3, 145, 182, 241) and in-depth reviews (150, 158, 226, 237, 238, 257) concerning LPS should be consulted for additional information. Recent minireviews dealing with selected aspect of LPS biology (180, 231, 236, 242, 289, 290) and chemistry (227, 243) are also available.
† This review is dedicated to R. F. W. Rätz and Eugene P. Kennedy.
Clinical isolates of enteric gram-negative bacteria usually possess an intact O antigen and are termed smooth because of their colony morphology (158, 238, 241, 257). Strains lacking O antigen, including essentially all laboratory strains of E. coli K-12 (269), are described as rough (158, 238, 241, 257). Smooth and rough LPSs (including most commercially available materials) are isolated from cells by either phenol-water (309) or phenol-chloroform-petroleum ether (66) extraction. Both smooth and rough LPSs can be further fractionated by sodium dodecyl sulfate (SDS)-gel electrophoresis (78, 144, 257). This procedure is especially useful for determining the number and distribution of O-antigen repeats (Fig. 3) (78, 144), which are displayed as a ladder of bands after silver staining or autoradiography. Gel electrophoresis of LPSs lacking O antigen results in partial separation of molecular species differing in the number of sugar units that are present in the core (Fig. 3) (144, 257). The individual bands that are resolved by gel electrophoresis are not necessarily homogeneous (257).
Most carbohydrate analyses and mass spectrometry of intact LPSs have used relatively crude materials extracted from cells with phenol (3, 31, 32, 241). Novel structural insights might be gained by doing such chemical analyses on LPS bands purified by SDS-gel electrophoresis.
Because of its amphiphilic nature, microheterogeneity, and limited solubility, it is difficult to carry out high-resolution nuclear magnetic resonance (NMR) studies on intact LPS structures larger than the Re chemotype (22, 271, 273). In addition, LPS molecules form aggregated structures in both polar and nonpolar solvents (241). The physical properties of LPS aggregates, though important for biological activity, are not extensively characterized (94, 145, 241, 277). Because of the molecular heterogeneity and composition of LPS, it is also not yet possible to obtain high-quality crystals (123).
The hydrophobic anchor of LPS, termed lipid A (Fig. 1 to 4), is released from LPS by acid hydrolysis (29, 33, 221). Lipid A obtained in this manner resembles common glycerophospholipids in some of its physical and chromatographic properties (70, 221, 223). However, glycerol is not a component of lipid A (29, 221, 223), and glycerophospholipids are not bioactive endotoxins (T. A. Garrett and C. R. H. Raetz, unpublished data).
Lipid A possesses a backbone composed of a β,1'-6-linked disaccharide of glucosamine (182, 226, 243). As shown in Fig. 3 and 4A, the glucosamine disaccharide of lipid A is acylated, phosphorylated, and further glycosylated (182, 226, 243). The pattern of acyl chain substitution shown in Fig. 4A is the predominant one in E. coli K-12, but variations on this theme occur in related organisms (116, 117, 222). Free lipid A does not exist as such in cells (70). The minimal LPS (Re chemotype; Fig. 3) required for growth of E. coli (158, 257) consists of lipid A glycosylated with two 3-deoxy-d-manno-octulosonic acid (Kdo) residues (Fig. 4A). The latter is an acidic eight-carbon sugar, the biosynthesis of which is essential for growth (73, 85, 239). Minor, nonessential species of Re LPS in E. coli K-12 may be decorated with additional phosphate-containing substituents or l-rhamnose (Fig. 3) (103, 104, 105, 273).
There are ∼2 × 106 lipid A moieties and ∼2 × 107 glycerophospholipids in a cell of E. coli (70, 225). The acyl chains attached to lipid A are usually shorter than those on glycerophospholipids (225). All acyl chains that are directly attached to the glucosamine disaccharide of lipid A are hydroxylated, whereas the acyl chains of glycerophospholipids are not (225, 228). The acyl moieties of lipid A do not usually contain double bonds (116, 117, 313).
In contrast to common glycerophospholipids (Fig. 4B), which possess only one or two asymmetric carbon atoms (225), Re LPS contains 24 chiral centers. This structural complexity greatly increases the possibilities for specific recognition by proteins. The conservation of the 24 chiral centers contributed by Kdo and lipid A in the LPSs of most gram-negative bacteria may account for the fact that mammalian cells have evolved the ability to respond to picomolar concentrations of LPS (145, 182, 231). The minimal Re LPS (Fig. 3 and 4A) required for bacterial growth possesses the full biological activity characteristic of endotoxins isolated from wild-type bacteria (182, 243). Synthetic lipid A also displays bioactivity almost comparable to that of Re LPS in most systems (65, 153, 182, 243). However, slight structural modifications can convert lipid A from a potent, cytokine-inducing agonist into an endotoxin antagonist (34, 35, 79, 136, 153, 278).
The LPS molecules of E. coli K-12 contain 6 to 10 core sugars in addition to the two Kdos (Fig. 3 and 5A) (103, 104, 105, 226, 257). The core domain of E. coli K-12 LPS is a branched, nonrepeating oligosaccharide (103, 104, 105, 226, 243, 257). The current proposed structure, which is one of several alternative forms (61, 257), is shown in Fig. 5A. A major difficulty with studies of the core is that it is heterogeneous in composition (103, 104, 105, 226, 243, 257). The core contains unusual sugars, such as Kdo and l-glycero-d-manno-heptose, and phosphorylated substituents (Fig. 5A). Mild acidic conditions (100°C, pH 4.5 or below) are used to cleave the Kdo-lipid A linkage prior to core isolation for structural analysis (29, 70, 221, 223). These hydrolysis conditions can alter Kdo and some labile phosphorylated substituents. Enzymes that cleave the Kdo-lipid A linkage would facilitate further progress on the elucidation of core structure, but such enzymes have not been discovered. Chemically synthesized standards of the full core are also lacking. Consequently, all proposed core structures, including that of Fig. 5A, must be considered tentative.
Despite structural uncertainties, mutants of E. coli K-12 and other gram-negative bacteria in which the core is truncated are available (158, 257). Strains with truncated cores are able to grow but are frequently resistant to bacteriophages (147), such as Felix O (158, 257). Such shortened cores, indicated by the two-letter code in Fig. 3, are termed chemotypes. Strains with truncated cores usually harbor mutations in the genes of the rfa cluster (Fig. 6 and 7) (158, 257). The latter includes most of the structural genes for the putative glycosyltransferases that assemble the core (Fig. 5A) (158, 257).
Selections based on bacteriophage resistance do not yield mutants lacking Kdo and lipid A, since these components of LPS are essential for growth (70, 73, 85, 239). Strains lacking the phosphate-containing substituents on heptose or lacking the heptose region altogether (Fig. 2, 3, and 5A) are hypersensitive to antibiotics and detergents (deep rough phenotype), but they do grow (158, 209, 257). An intact heptose region displaying the right phosphate moieties allows the outer membrane to act as a barrier to many antibiotics (193, 209, 227, 292). Phosphate residues on heptose are presumed to bind divalent cations (226, 257). This may serve to cross-link neighboring LPS molecules and maintain the structural integrity of the outer membrane. Washing cells with EDTA removes divalent cations and renders the outer membrane permeable to antibiotics (143, 226, 257).
The absence of the heptose/phosphate domain also leads to a profound deficiency of many outer membrane proteins, especially the porins (2, 135, 257). The molecular basis for this remarkable phenomenon is not well understood (257). It may involve a reduced rate of assembly of newly made outer membrane porins to stable trimers (137). The way in which LPS interacts with outer membrane proteins requires further study.
Laboratory strains of E. coli K-12 do not contain O antigen, have the Ra chemotype (Fig. 3) (257, 269), and usually harbor an insertion mutation in the last rfb gene (orf11) encoding a rhamnosyltransferase (Fig. 5B and 8) (269). This accounts for the absence of O-antigen production. When a functional copy of this gene is restored, E. coli K-12 is capable of synthesizing and exporting a polymeric O-antigen pentasaccharide (269). The structure is a variant of O antigen, O-16 (Fig. 5B). As with other O antigens, a nonrandom distribution of repeating units, ranging from 1 to 40, is synthesized (11, 12, 77, 78, 269).
O-antigen structures vary considerably in gram-negative bacteria (3, 110, 236, 237, 241). At least 160 different chemical arrangements of O antigens have been identified in E. coli alone (110, 236, 237). The presence of O antigen on the cell surface may enhance virulence under some circumstances and increases the resistance of bacterial cells to complement-mediated killing (245). The X-ray structure of a typical O-antigen repeat bound to a specific antibody has recently been solved (45).
Reeves and coworkers (237, 269) have sequenced the rfb clusters of E. coli K-12 (Fig. 6 and 8) and several other gram-negative bacteria. The rfb genes code for most of the unique enzymes required for O-antigen assembly, including the biosynthesis of 6-deoxy- and 3,6-dideoxyhexoses (when present) (150, 237, 257, 269). The important studies of Reeves and coworkers have shown that rfb clusters (Fig. 8) sequenced from different strains of Salmonella and E. coli are highly divergent (236, 237, 269). Many rfb genes have strikingly lower GC contents than the other genes of the strains from which they were isolated (237). The rfb genes also differ in GC content among themselves within the same gram-negative bacterium (237). Reeves argues that individual rfb genes or portions of rfb gene clusters may have been transferred between species of bacteria during evolution (236, 237). Transfer of foreign rfb clusters on hybrid plasmids into E. coli K-12 can indeed result in the biosynthesis and export of heterologous O-antigen units (1, 61, 236, 274).
In 1979, Nishijima and I (195) discovered lipid X, an unanticipated monosaccharide precursor of lipid A (195, 281) that accumulates in certain phosphatidylglycerol-deficient mutants of E. coli. The determination in 1983 of the structure of lipid X as a 2,3-diacylglucosamine 1-phosphate in which both acyl chains are R-3-hydroxymyristate (Fig. 4B) was crucial for the elucidation of lipid A biosynthesis (4, 25, 227, 233, 281). The structure of lipid X also provided independent evidence for the sites of fatty acid attachment to lipid A (107, 244, 279). Given that lipid A is relatively conserved among different gram-negative bacteria, most features of the E. coli lipid A pathway are also conserved (311). Many of the genes coding for the enzymes of lipid A biosynthesis in E. coli have now been identified (227), facilitating purification and mechanistic studies.
The biosynthesis of UDP-GlcNAc plays an important role in the assembly of bacterial surface polymers. The formation of UDP-GlcNAc is catalyzed by a peculiar bifunctional enzyme (encoded by glmU), the existence of which was not recognized until recently (171). UDP-GlcNAc resides at a complex branch point in E. coli K-12 (Fig. 9) that initiates the formation of lipid A (70), peptidoglycan (208), O antigen (269), and enterobacterial common antigen (169). Lipid A and peptidoglycan are absolutely essential for growth under all conditions. The numbers of glucosamine residues in lipid A and peptidoglycan (∼4 × 106 in each) are similar (70, 208). O antigen and enterobacterial common antigen are not required for growth under laboratory conditions but presumably confer selective advantages. The partitioning of UDP-GlcNAc may be regulated.
As shown in Fig. 9, the addition of phosphoenolpyruvate to UDP-GlcNAc (leading to peptidoglycan) occurs on the hexose 3-OH. The antibiotic phosphomycin blocks peptidoglycan synthesis at this early stage (114, 161).
The first enzyme required for lipid A biosynthesis is a cytoplasmic acyltransferase that also recognizes the hexose 3-OH of UDP-GlcNAc (Fig. 9 and 10) (5, 6). The E. coli acyltransferase is highly selective for R-3-hydroxymyristate (5, 6). This specificity is consistent with the acyl chain composition of E. coli lipid A (Fig. 4A) (116, 222). The UDP-GlcNAc acyltransferase differs from the relatively nonspecific glycerol 3-phosphate acyltransferase of E. coli (82, 256) in its absolute requirements for acyl carrier protein (ACP) thioesters and for the R-3-OH moiety. Its high specificity for acyl chain length (14 carbons) is also unusual (5, 6). UDP-GlcNAc acyltransferase is encoded by the lpxA gene (40, 70), which maps near min 4 on the chromosome (Fig. 6 and 7). The conditional lethality of E. coli mutants defective in UDP-GlcNAc O-acyltransferase demonstrates that lipid A is essential for cell viability (70). The properties of mutants in lpxA prove that the proposed pathway (Fig. 10 to 12) is not a minor one (70). The rapid loss under nonpermissive conditions of cell viability of strains bearing lpxA mutations demonstrates that the early steps of the pathway are excellent targets for the development of novel antibacterial agents (70). Furthermore, even partial inhibition of lipid A synthesis causes profound hypersensitivity to other antibiotics, like erythromycin, that are normally excluded by the outer membrane (295).
In contrast to enzymes of glycerophospholipid synthesis, which are generally membrane bound (225, 228, 262), all three of the early enzymes of lipid A biosynthesis (Fig. 10) are cytoplasmic (4, 5, 6). Crystals of UDP-GlcNAc-3-O-acyltransferase suitable for X-ray diffraction at a resolution of 2.6 Å (0.26 nm) have recently been obtained (213). This important development is certain to provide fundamental new insights into lipid-protein recognition and presents a major opportunity for drug design (see above). No other acyltransferase of phospholipid metabolism has ever been obtained in crystalline form (125, 262). How R-3-hydroxymyristate is kept out of glycerophospholipids is uncertain (225, 226).
Like UDP-GlcNAc, R-3-hydroxymyristoyl-ACP is located at a metabolic branch point. If it is not used for lipid A biosynthesis, R-3-hydroxymyristoyl-ACP can be elongated to palmitoyl-ACP, a key precursor of glycerophospholipids (176, 225). We have found that the temperature-sensitive phenotype of certain point mutations in lpxA can be suppressed by point mutations in the gene immediately upstream of lpxA, designated fabZ (Fig. 7) (176). The fabZ gene codes for a dehydrase that acts on R-3-hydroxymyristoyl-ACP (176). This reaction is the immediate next step in the conversion of R-3-hydroxymyristoyl-ACP to palmitate (176). Deficiency of the dehydrase is hypothesized to increase the pool of R-3-hydroxymyristoyl-ACP, enhancing the ability of cells to generate more lipid A in the setting of reduced lpxA function (176).
The equilibrium constant for the acylation of UDP-GlcNAc (∼0.01) is surprisingly unfavorable (5), suggesting that R-3-hydroxymyristoyl-ACP (a thioester) may be more stable than UDP-3-O-(R-3-hydroxymyristoyl)-GlcNAc (an oxygen ester). The chemical basis for this anomaly is unclear. The biological implication of this finding is that the deacetylation of the product, UDP-3-O-(R-3-hydroxymyristoyl)-GlcNAc (Fig. 10), is the first committed reaction of lipid A biosynthesis (5, 7). Accordingly, the deacetylase should be a site for regulation. Inhibition of lipid A biosynthesis in certain mutants can indeed cause a 5- to 10-fold increase in deacetylase specific activity (5). This response can be attributed to the presence of more enzyme in such cells, as judged by immunological criteria (C. R. H. Raetz, unpublished data).
The deacetylase is encoded by the envA (lpxC) gene near min 2 on the E. coli chromosome (Fig. 6) (318). envA (lpxC) was first described by Normark and coworkers in 1969 (197, 198) as a mutation that causes antibiotic hypersensitivity and a delay in cell separation following septum formation. envA (lpxC) was sequenced in 1987 (13) prior to our discovery of its function in lipid A biosynthesis. It is not homologous to any other known gene (13). The deacetylase has been purified to homogeneity (318). Like lpxA, envA (lpxC) is essential for the growth of E. coli (13). The relationship of lipid A biosynthesis to cell division requires further investigation.
Following deacetylation, the N-linked R-3-hydroxymyristate is incorporated to generate the liponucleotide, UDP-2,3-diacylglucosamine (Fig. 10) (124). Again, R-3-hydroxymyristoyl ACP is required as the cosubstrate. Neither R-3-hydroxymyristoyl coenzyme A nor myristoyl-ACP is recognized (6, 124). The N-acyltransferase is the product of the firA (lpxD) gene (52, 124), which maps within the same complex operon as lpxA near min 4 (Fig. 6 and 7) (40, 44, 287). The firA mutation was first identified in 1970 on the basis of its ability to reverse the rifampin resistance conferred by mutations in the β subunit of RNA polymerase (10). firA is the equivalent of the ssc gene of S. typhimurium (291, 294).
The lpxA and lpxD genes show statistically significant sequence homology (124), consistent with their related functions (Fig. 10). They each contain multiple copies of a hexapeptide repeat (termed an isoleucine patch), suggesting a novel secondary structure (53, 291). X-ray diffraction studies of the lpxA gene product suggest that these acyltransferases may be trimers (213) and that the repeats form a novel, left-handed β-helix (C. R. H. Raetz and S. L. Roderick, unpublished data).
UDP-2,3-diacylglucosamine (Fig. 10 and 11) is the immediate precursor of the distal (nonreducing) sugar of lipid A (4, 224, 233). Some of it is also cleaved at the pyrophosphate bond to generate 2,3-diacylglucosamine-1-phosphate (lipid X) (Fig. 11) (4, 25, 233). Wild-type cells contain about 100 and 1,000 molecules of UDP-2,3-diacylglucosamine and lipid X, respectively (25). The presence of low levels of these metabolites in normal cells explains why they were not discovered in early studies of bacterial lipid composition (225).
Lipid X (Fig. 11) is the direct precursor of the proximal (reducing end) sugar of lipid A (224, 233). Disaccharide formation involves the condensation of lipid X with another molecule of UDP-2,3-diacylglucosamine in a reaction catalyzed by the disaccharide synthase (224, 233). The 2,3-diacylglucosamine portion of UDP-2,3-diacylglucosamine is transferred to the 6 position of lipid X, generating the β,1'-6 linkage found in all lipid A molecules (224, 233). The disaccharide synthase has been cloned, overexpressed, and purified to homogeneity (44, 224). It can be used to synthesize lipid A substructures and analogs (259, 296). The disaccharide synthase cannot condense two molecules of UDP-2,3-diacylglucosamine directly (224). Consequently, the UDP-2,3-diacylglucosamine-specific pyrophosphatase is also a critical component of the system. The disaccharide synthase is encoded by the lpxB gene (43, 44). It is part of the same complex operon (Fig. 6 and 7) as are lpxD and lpxA (40, 124, 287). The gene encoding the pyrophosphatase is unknown. A second pyrophosphatase encoded by the cdh gene (not shown) also cleaves UDP-2,3-diacylglucosamine (25, 26, 106). However, deletion of the cdh gene has no obvious effects on LPS biogenesis (26, 106).
Lipid X (148), and probably all other monosaccharide precursors of lipid A, forms micelles in aqueous solution, whereas lipid A disaccharides generally assemble into bilayers (102, 123). The tendency of these precursors to form micelles may explain why detergents are not required for the activities of the early enzymes of the pathway, including the disaccharide synthase (5, 6, 7, 124, 224). All enzymes beyond the disaccharide synthase do require the presence of detergents for catalytic activity (22, 24, 233).
Lipid X can be isolated in gram quantities from conditional mutants altered in lpxB (224), as it accumulates to levels that are ∼500-fold higher than wild-type levels (25). Lipid X accumulation exacerbates the conditional lethality of certain point mutations in the pgsA gene (Fig. 6) that are ∼30% reduced in the ability to make phosphatidylglycerol (194, 195). The molecular basis for the interaction between phosphatidylglycerol and lipid A biosynthesis remains obscure. Nevertheless, it was the basis for our discovery of lipid X (194, 195).
The disaccharide synthase present in crude extracts of E. coli migrates as a large aggregate in sucrose velocity gradients (173, 229). The aerobic glycerol 3-phosphate dehydrogenase (glpD gene product) appears to be associated with the same aggregates as the disaccharide synthase (173). The biological significance of these new findings may provide insights into the interactions of the lipid A and glycerophospholipid pathways (195).
A specific, membrane-bound kinase (234) incorporates the 4'-monophosphate, generating the key intermediate, lipid IVA (Fig. 12). Large amounts of lipid IVA accumulate in conditional mutants that are defective in Kdo biosynthesis (230, 239, 272) or in cells treated with CMP-Kdo synthase inhibitors (Fig. 13) (73, 85). These results provide independent evidence for the role of lipid IVA as an intermediate in the pathway and as the acceptor of Kdo residues in living cells. Inhibition of Kdo biosynthesis results in an arrest of cell growth, but cell viability is not lost because the cells can recover when the inhibition is relieved (75, 201, 203, 240). It is not entirely clear whether growth is inhibited by the accumulation of lipid IVA in the inner membrane or by the absence of mature LPS in the outer membrane.
Lipid IVA possesses some of the bioactivity of endotoxins (79, 230, 264). In mouse cells it is an endotoxin-like agonist (79, 230, 264), but in human cells it is an endotoxin antagonist (79, 136, 153). Radiolabeled lipid IVA can be prepared with the 4'-kinase (88, 234). The kinase product, [4'-32P]lipid IVA, is extremely useful for detecting endotoxin-binding proteins on animal cells (86, 87) and for studying the late steps of lipid A biosynthesis (Fig. 12) (22, 24).
The gene that codes for the 4'-kinase is not known. The kinase is unstable under assay conditions when ATP concentrations are low (88, 234). It has not been purified. The phosphorylation of the 4' position should be thermodynamically favorable. Overexpression and purification of the 4'-kinase might be useful for synthetic purposes. Many disaccharide 1-phosphate analogs are already available as potential kinase substrates (259, 296). Access to novel disaccharide 1,4'-bisphosphates that could be generated with the 4'-kinase might be important, because some compounds of this kind are already known to be potent endotoxin antagonists (see below) (34, 79, 126, 136, 153, 278).
Rhizobium leguminosarum lipid A lacks a 4'-phosphate moiety (17), as do lipid A’s of many anaerobic bacteria (200, 305). R. leguminosarum does possess a 4'-kinase (216). Apparently, R. leguminosarum first generates the same phosphorylated intermediates that are found in E. coli (Fig. 12) but then selectively 4'-dephosphorylates Kdo2-lipid IVA in the later stages of LPS assembly (see below) (215a).
CMP-Kdo is an unusual labile nucleotide (15, 72, 133). Its half-life in water is ∼10 min under typical assay conditions (15, 133). CMP-Kdo must be generated in situ from CTP and Kdo by excess CMP-Kdo synthase when used as a substrate for Kdo transferase (15, 22). The enzymology of CMP-Kdo biosynthesis (Fig. 13) is well characterized (74, 133). Several mutants and inhibitors that support the scheme shown in Fig. 13 are also available (73, 85, 239).
Re LPS (Fig. 4A) contains two distinct Kdo residues (182, 241, 271). These are incorporated by an unusual bifunctional transferase (Fig. 11) encoded by the kdtA gene (15, 38). The following evidence supports the existence of a bifunctional Kdo transferase. (i) Purification of the membrane-associated Kdo transferase to homogeneity does not resolve the two components of the Kdo transferase reaction (15). (ii) The kdtA gene maps near min 81 in E. coli (Fig. 6) just clockwise of the rfa cluster (38). Overproduction of this polypeptide of 425 amino acids causes large increases in the specific activities of both the first and second Kdo transferase reactions in cell extracts (15, 38).
The purified Kdo transferase is a useful catalytic reagent for the addition of Kdo disaccharides to lipid A-like molecules (15). The presence of the 4'-phosphate moiety is an absolute requirement for Kdo transfer (15, 22). The Kdo transferase of E. coli recognizes lipid A disaccharide bisphosphate substrates with a specificity (15, 22) resembling that of lipid A-recognizing proteins in animal systems (145). However, the kdtA gene shares no obvious sequence homology with known mammalian proteins (38, 145) or with the lipid A-binding protein that functions as an inhibitor of lipid A activation of the Limulus clotting cascade (101). The X-ray structure of the Limulus lipid A inhibitory protein has recently been solved at 1.5 Å (0.15 nm), providing a novel paradigm for lipid A-recognition by proteins (101).
A gene (gseA) from Chlamydia trachomatis with significant sequence homology to kdtA has been described (14). The C. trachomatis Kdo transferase encodes a single polypeptide that attaches three Kdos to lipid IVA (14). Its expression in enteric gram-negative bacteria results in the appearance of a third Kdo residue on a subset of LPS molecules (14, 191). E. coli mutants in which the chromosomal copy of kdtA has been inactivated by transposon insertion can be rescued by expression of gseA on a plasmid (Raetz, unpublished).
Antibodies directed against the third Kdo residue can be used to detect its presence on the surface of E. coli cells expressing the gseA gene (191). Such antibodies react specifically with many Chlamydia strains and originally served to define the genus-specific epitope of chlamydiae (14, 191). These antibodies are used to diagnose chlamydial infections.
Extracts of Haemophilus influenzae contain a monofunctional Kdo transferase (310). This finding is consistent with the core sugar composition of Haemophilus strains (97, 170). It will be important to determine whether a monofunctional Kdo transferase can support the growth of E. coli mutants lacking the kdtA gene.
In the last steps of Re LPS formation, laurate and myristate residues are incorporated into the distal unit of lipid A (Fig. 12). This process generates the acyloxyacyl moieties (24), a conserved feature of lipid A molecules from diverse sources (67, 182, 241). Like the UDP-GlcNAc acyltransferases, the "late" acyltransferases can utilize only thioesters of ACP as substrates (24). However, the equilibrium constants estimated for the reactions catalyzed by the late acyltransferases (>2) are reasonable for S-to-O acyl transfers (Raetz, unpublished). The late acyltransferases display at least 1,000-fold selectivity for Kdo2-lipid IVA (Fig. 12) over lipid IVA as the acyl acceptor (24). This property accounts for the fact that lipid IVA (not lipid A) accumulates in Kdo-deficient Salmonella mutants (230, 239, 272). Similarly, lipid IVA accumulates in strains of Salmonella and most other gram-negative bacteria treated with CMP-Kdo synthase inhibitors (Fig. 13) (73, 75, 85). Pseudomonas aeruginosa is an exception because fully acylated lipid A builds up when CMP-Kdo formation is inhibited (76). P. aeruginosa extracts contain an unusual lauroyltransferase that recognizes both lipid IVA and Kdo2-lipid IVA (177).
Two E. coli genes encoding enzymes involved in Kdo-dependent late acylation (24) have recently been identified (37). This was accomplished by assaying lysates of cells infected with the individual hybrid λ bacteriophages of the Kohara library (132). If a gene of interest is present on a given DNA insert, approximately 10-fold overproduction of enzymatic activity is observed in the lysate (as shown with known hybrid λ bacteriophages of the collection bearing kdtA). In this way, we identified htrB (Fig. 6) (118, 121) as the gene that codes for the Kdo-dependent lauroyltransferase (Fig. 12) (37). The msbB gene (which suppresses the temperature-sensitive growth of strains bearing htrB mutations when introduced on high-copy-number plasmids) (120) encodes a second late acyltransferase that normally functions after laurate has been added to Kdo2-lipid IVA (24, 37). The htrB and msbB genes display significant sequence homology to each other but not to other known genes (120). The htrB-encoded enzyme displays ∼10-fold selectivity for lauroyl-ACP over myristoyl-ACP (24, 37). The msbB-encoded enzyme functions well with both lauroyl-ACP and myristoyl-ACP under the conditions examined (37).
The htrB gene was described previously by Karow et al. as the locus of a transposon-induced mutation that causes temperature-sensitive growth on nutrient media (118, 121). Conditions that reduce the growth rate at elevated temperatures (such as minimal media) restore the ability of htrB-deficient cells to divide (119, 121). Given our findings, it must now be determined why laurate-deficient lipid A is incompatible with rapid growth on rich media above 32°C.
The msbB gene (Fig. 6) presumably rescues the growth of htrB-deficient cells at 42°C by providing alternative enzymatic machinery for the generation of acyloxyacyl moieties (37). Transposon-induced mutations in msbB alone have no obvious phenotype (120). Fatty acid analyses of lipid A isolated from htrB-deficient mutants show a significant reduction in laurate content (119). Fatty acid analyses of lipid A from htrB-deficient strains rescued by msbB have not been described.
A second E. coli gene (msbA; Fig. 6) that suppresses htrB mutations when introduced on low- or high-copy-number plasmids was described by Karow and Georgopoulos (122). The msbA gene displays striking sequence similarity to proteins involved in secretion of drugs, such as the Mdr proteins of animal cells (122). This provocative finding could now be interpreted to mean that lipid A precursors lacking laurate are toxic to E. coli above 32°C and that enhanced export of these precursors mediated by extra copies of msbA permits rapid growth at elevated temperatures. The possibility that multiple copies of msbA can rescue mutants unable to generate Kdo (230, 239, 272) must also now be examined.
It will be especially interesting to determine whether MsbA is part of the normal machinery responsible for the secretion of LPS, an important aspect of the LPS story about which very little is known (201, 226). Karow and Georgopoulos have demonstrated that the msbA gene is essential for the growth of E. coli at all temperatures (122). Recent evidence in animal systems indicates that the Mdr-2 protein functions to transport phosphatidylcholine across the bile canalicular membranes of hepatocytes (100, 206, 255, 268). Mice lacking functional Mdr-2 contain no phosphatidylcholine in the bile (206, 255, 268). Given the precedent for the involvement of Mdr proteins in phospholipid transport in animal systems, it will be necessary to investigate the possible function of MsbA in lipid A translocation in E. coli.
A different reaction for the generation of acyloxyacyl moieties has been described for palmitate transfer to lipid X (21, 280). In this case, a palmitate residue from the 1 position of a glycerophospholipid is transferred to R-3-hydroxy function of the N-linked R-3-hydroxymyristoyl moiety (21, 280) (Fig. 14). Both lipid IVA and Kdo2-lipid IVA are alternative palmitate acceptors. The importance of this transesterification is unknown. Mature E. coli K-12 lipid A contains very little palmitate (116). However, other strains of E. coli as well as Salmonella strains do contain significant lipid A molecular species (116) or precursors bearing palmitate (21, 95, 230, 272, 280), especially under conditions of limited lipid A biosynthesis. The gene encoding the palmitoyltransferase is unknown. Its overexpression might provide another way to bypass the need for htrB at elevated temperatures.
In certain E. coli (Fig. 5A) and Salmonella strains (Fig. 15A), lipid A is derivatized with additional phosphate, phosphoethanolamine, and/or 4-aminoarabinose moieties (182, 230, 241). Polymyxin-resistant Salmonella mutants contain higher levels of 4-aminoarabinose than do wild-type cells (96). Although the pmrA locus that is altered in polymyxin-resistant strains has been cloned and sequenced, its exact function is unknown (250). Sequencing suggests that the pmrA locus encodes a two-component regulatory system like the OmpR-EnvZ system (250). Polymyxin resistance is also observed in strains overexpressing a small protein of unknown function encoded by the pmrD gene (249). The latter type of resistance requires the presence of a functional pmrA locus, suggesting some kind of interaction (249).
The enzymatic reactions for the incorporation of lipid A "decorations," like 4-aminoarabinose, are not known (182, 230, 241). In general, these modifications reduce the overall negative charge of the lipid A domain (182, 241). The long-standing controversy concerning the locations of the ethanolamine pyrophosphate and the 4-aminoarabinose residues on lipid A of Salmonella strains (182, 230, 241, 272) needs to be resolved. Our studies demonstrate unequivocally (230, 272) that the ethanolamine pyrophosphate attached to certain lipid A precursors that accumulate in Kdo-deficient mutants is located at position 4'. All other workers have claimed that ethanolamine pyrophosphate is attached to the 1 position of lipid A (182, 241). Furthermore, our data show that 4-aminoarabinose is linked to position 1, not to position 4' (272).
The possibility of additional enzymes catalyzing the removal or remodeling of acyl chains attached to lipid A has not been explored. Lipid A disaccharides lacking one of the usual O-linked hydroxy fatty acids (Fig. 4A) do exist in some organisms (116). However, the disaccharide synthase functions several orders of magnitude better with monosaccharide precursors containing two acyl chains rather than one (224). Underacylated lipid A species could arise late in the pathway by action of a lipid A-specific lipase(s).
The observed regulation of the deacetylase (lpxC gene product) (5) implies that cells have mechanisms for sensing their lipid A content and adjusting the rate of lipid A synthesis accordingly. Nothing is known about the biochemistry of deacetylase regulation. Similarly, the significance of the complex operon near min 4 (Fig. 6 and 7) encoding genes of glycerophospholipid metabolism, lipid A biogenesis, and DNA replication remains uncertain (40, 44, 124, 287, 288). Its transcription involves several distinct promoters spaced along the length of the operon (287).
The enzymes that incorporate core sugars and other substituents beyond the Kdo disaccharide (Fig. 5A and 15A) require considerable further investigation. The following problems have hindered progress in this field. (i) The structures of the substrates utilized by these enzymes (271, 273) and the products generated by them (266) are often not fully characterized. (ii) Assays that are dependent on time and protein concentration are not well developed (238, 266). (iii) Synthetic substrates are not available, reflecting the fact that the structure of the core is complex and not fully established (Fig. 5A and 15A).
The published work on the enzymatic synthesis of the heptose region of the core (Fig. 5A and 15A) (41, 55, 134, 266) illustrates some current limitations. The rfaD, rfaC, and rfaF genes (Fig. 5A and 7) are believed to encode the ADP-heptose epimerase (Fig. 16) (41, 55, 134), the first heptosyltransferase (Fig. 17) (266), and the second heptosyltransferase (Fig. 17), respectively (266). Studies of the epimerase (41, 55, 134) and of the heptosyltransferases (266) have not used synthetic ADP-l (or d)-glycero-d-manno-heptoses as substrates. All published investigations of ADP-heptose enzymology (Fig. 16 and 17) have relied on incompletely characterized ADP-heptose preparations obtained from cells (41, 134). Although the structure of the heptosyl acceptor, Kdo2-lipid IVA (Fig. 17), was validated (22), the reaction products generated by the heptosyltransferases could be obtained only in radiochemical amounts, insufficient for linkage analyses, mass spectrometry, or NMR spectroscopy (266).
Similar limitations pertain to the important early observations that Rd1 or Rc LPSs isolated from strains defective in UDP-glucose or UDP-galactose synthesis could serve as acceptors of radioactive Glc, Gal, or GlcNAc when incubated with appropriate sugar nucleotides and crude cell extracts (202, 204, 238, 254, 257, 314). The putative Salmonella glycosyltransferases encoded by the rfaG, rfaB, rfaI, rfaJ, and rfaK genes (Fig. 7 and 15A) were first detected in this manner (202, 204, 238, 254, 257, 314). Several of these enzymes were purified and reconstituted in monolayer systems (59, 187, 251). Unfortunately, synthetic LPS substrates were not available (202, 204, 238, 254, 257, 314). Since the exact covalent structures of the Rd1 and Rc LPSs used as acceptors were uncertain to begin with (238, 257, 271, 273), the structures of the products generated by these enzymes could not be fully confirmed.
Very early in vitro studies of the phosphorylation of LPS preparations extracted from rfaP-deficient strains (Fig. 7) suffered from the same shortcomings (183, 184, 238, 257). An additional problem with the rfaP gene is the fact that it is not homologous to any known kinase (257) and appears to be required for the addition of more than one core substituent (Fig. 5A and 15A) (257). Despite the interesting antibiotic-hypersensitive phenotype associated with rfaP mutations, the exact biochemical functioning of the rfaP gene product remains elusive (209, 257). The only rfa gene that displays some homology to kinases is rfaY (Raetz, unpublished).
Future progress with the biochemistry of the core will require more aggressive use of the methods of carbohydrate structure elucidation and synthetic chemistry. An enzymatic method for cleaving the Kdo-lipid A linkage will have to be found so that the structures of the core oligosaccharides present in various substrates and products can be evaluated in water (as is normally done for NMR studies of oligosaccharides) without interference from the amphiphilic lipid A anchor. Only then can the enzymology of core assembly be worked out in an unequivocal manner. Novel biochemical insights may well emerge, given that the functions of at least six of the rfa genes (Fig. 7) are not altogether obvious from their DNA sequences (see below) (257).
In the past 5 years, significant progress has been made in isolating and sequencing the genes required for core assembly (9, 129, 155, 210, 212, 214, 252, 257, 258, 266, 267) (Fig. 5A, 6, 7, and 15A). The rfa genes of E. coli K-12 and S. typhimurium LT2, in which the core region is slightly different (compare Fig. 5A and 15A), have been examined (257). Despite the limitations of the available enzymatic assays, it does seem likely that many of the rfa genes have been assigned their proper biochemical functions (Fig. 5A, 7, and 15A). The detailed evidence supporting these tentative assignments is found in a thorough recent review by Schnaitman and Klena (257).
As shown in Fig. 7, the rfa region of E. coli consists of 14 genes organized into two major operons that are transcribed in opposite, convergent directions (9, 252, 258). The proposed functions of each rfa gene product are indicated in Fig. 7. These should be compared with the tentative core structure in Fig. 5A. A small operon containing the KDO transferase gene (kdtA) and orf18 is situated immediately clockwise of the larger rfa operon (36, 252). The transcription of the latter is initiated in the interval between rfaQ and kdtA (36).
The larger rfa operon that begins with rfaQ (Fig. 7) is regulated by the rfaH gene (19, 62, 215). The latter maps outside the rfa region (Fig. 6) and also regulates the expression of the tra genes (62, 215, 257). The RfaH protein appears to function by an antitermination mechanism (62). The smaller rfa operon, beginning with rfaD, encodes four genes (Fig. 7) and is under heat shock control (232). The physiological significance of rfa operon regulation is unclear (257).
It is possible to create viable strains with deletions extending from the middle of rfaD to rfaG (Fig. 7) (257). Such mutants have a deep rough phenotype, as is observed with other heptose- or heptose phosphate-deficient organisms (257). Deep rough mutants may accumulate second-site suppressors with continued growth under laboratory conditions, reducing the severity of their antibiotic hypersensitivity (209, 257). Conversely, Raina and Georgopoulos (232) have found that cells with insertion mutations in rfaD (also designated htrM) are temperature sensitive for growth on rich media (the htr phenotype), similar to what they observed with insertions in the gene coding for the Kdo-dependent lauroyltransferase (htrB) (Fig. 12) (232). Perhaps the heptose domain is required for growth at elevated temperatures in the genetic backgrounds used by Raina and Georgopoulos (232) because of other, as yet uncharacterized mutations.
Of the many genes in the rfa region, only kdtA (the Kdo transferase) and rfaL (the putative O-antigen ligase) have membrane-spanning domains (38, 155, 257). The former has 1 predicted membrane-spanning region, whereas the latter possesses at least 10 (38, 155, 257). The other rfa genes do not encode long hydrophobic sequences, consistent with their proposed functioning as cytoplasmic or peripheral membrane enzymes (257). The entire core domain is probably assembled on the cytoplasmic side of the inner membrane prior to translocation to the periplasmic surface and O-antigen attachment (see below). The possibility that the rfa gene products function as a complex with each other or with other proteins has not been explored. Assembly of the entire core is not necessary for translocation, as Re LPS is exported at normal rates (201).
The rfaL gene can be deleted (257). Accordingly, the rfaL gene product is unlikely to be part of a general core-lipid A export system, the functioning of which would be essential for growth. However, a direct demonstration that purified RfaL has O-antigen ligase activity is lacking.
The roles of the rfaQ, -S, -Y, -Z, and -K genes in E. coli and Salmonella strains are uncertain (257). The rfaL and rfaK genes near the center of the rfa region in E. coli K-12 and S. typhimurium LT2 show the least amino acid sequence conservation (less than 20% identity) (129, 155). Most other rfa genes show at least 70% identity at the amino acid level when E. coli K-12 and S. typhimurium LT2 are compared (257).
Within the E. coli K-12 rfa region, the rfaQ gene (Fig. 7) displays some sequence similarity to rfaC and rfaF (210). Accordingly, rfaQ might code for an additional heptosyltransferase that attaches one of the substoichiometric heptose residues of the core (Fig. 5A and 15A) (210, 266). In S. typhimurium LT2, the rfaK gene is thought to code for an enzyme that attaches GlcNAc to the outer core (Fig. 15A) (129, 155, 257). The addition of GlcNAc to the outer core is apparently required for subsequent O-antigen attachment (128, 257). The sequence of E. coli K-12 rfaK is not very similar to that of S. typhimurium LT2 rfaK (129, 155). Recent studies suggest that the cores of some strains of E. coli K-12 do not contain this GlcNAc residue (61, 104). Its role in O-antigen attachment in E. coli is uncertain (257).
The possible functions of rfaS, rfaY, and rfaZ (128, 129, 130, 131, 214) are discussed in detail by Schnaitman and Klena (257), who base their interpretations on the effects of mutations in these genes on LPS electrophoretic mobility. Klena et al. (128, 130, 131, 257) suggest that rfaS and rfaZ may be required for the formation of an alternative core, termed lipooligosaccharide, the exact structure of which is unknown. Lipooligosaccharide is proposed to be incapable of accepting the O-antigen polymer (128, 130, 131, 257). It will not be possible to determine the exact biochemical functions of rfaS and rfaZ without better correlation between mutations and structural alterations of the core.
Several genes that participate in core assembly map outside the rfa cluster (Fig. 6). Additional genes of this kind may be found as novel features of core structure emerge. The locations of the regulatory gene rfaH, which controls the rfaQ operon (19, 62, 215), and the rfaE gene, which is required for ADP-heptose generation (Fig. 16) (266), are shown in Fig. 6. Deletion of the pss gene (48) (encoding phosphatidylserine synthase) (Fig. 6) results in LPS lacking ethanolamine substituents (Fig. 5A) (257), since pss insertion mutations eliminate all phosphatidylethanolamine from the envelope (228). Mutations in the tolC gene (Fig. 6) result in an increased amount of ethanolamine pyrophosphate associated with the inner heptose (Fig. 5A) (257). Schnaitman and Klena suggest that the tolC gene product normally trims ethanolamine phosphate from this site, leaving behind a free monophosphate residue with a high affinity for divalent cations (257). Strains bearing tolC mutations display antibiotic hypersensitivity that is similar to that of heptose-deficient mutants (257), and they fail to secrete hemolysin (298), but unlike heptose-deficient strains, they are not deficient in outer membrane proteins (257).
The enzymes that generate the minimal Kdo2-lipid A structure required for growth perform their functions in a defined linear sequence (Fig. 10, 11, 12, 13). Each enzyme can only act on the product generated by the preceding enzyme in the pathway. For instance, Kdo transferase does not work prior to 4'-kinase (15), and late acylation does not occur prior to Kdo addition (24). Such a precise arrangement results in the generation of a single, homogeneous product at each stage (Fig. 10 to 12). One exception appears to be the Kdo-independent palmitoyltransferase (Fig. 14) that can function at any stage after the formation of lipid X (21; Raetz, unpublished). However, the palmitoyltransferase may be an outer membrane enzyme (196; Raetz, unpublished) and may have no access to Kdo2-lipid IVA under ordinary circumstances.
The order of core domain biosynthesis is less rigid. The major sugars of the core (Fig. 5A) are incorporated sequentially, but other core substituents are incorporated independently of one another. For instance, heptose can be transferred either to Kdo2-lipid IVA or Kdo2-lipid A, independent of late acylation (17) (24). Under conditions of limited fatty acid synthesis, the entire core domain can be added to Kdo2-lipid IVA (297). Similarly, the outer core can be completed whether or not the phosphorylated substituents of the heptose region have been attached (98, 209). Three of the enzymes that can act independently of each other on Kdo2-lipid IVA in extracts of E. coli (Raetz, unpublished) are shown in Fig. 17.
Alternative enzymatic processing of Kdo2-lipid IVA has been uncovered in recent studies of core and lipid A biosynthesis in extracts of R. leguminosarum (23; Raetz, unpublished). As shown in Fig. 18, R. leguminosarum lipid A differs significantly from that of E. coli. These differences include the absence of the 4'- and 1-phosphate residues, the presence of an acylated aminogluconic acid moiety instead of the usual glucosamine 1-phosphate residue at the reducing end, the presence of very long acyl chains (C28) instead of laurate or myristate, and the presence of mannose instead of heptose attached to the inner Kdo (not shown in Fig. 18) (17, 18).
Despite these structural differences, we have shown that R. leguminosarum generates Kdo2-lipid IVA by the same linear pathway as found in E. coli (Fig. 10, 11, 12, 13) (216). However, the combination of enzymes that is available to process Kdo2-lipid IVA in R. leguminosarum (Fig. 19) is entirely different from that in E. coli (Fig. 17). R. leguminosarum extracts contain two unique phosphatases (Fig. 19) that act on Kdo2-lipid IVA (23, 215a). Kdo2-lipid IVA is also an acceptor for mannose residues from GDP-mannose in extracts of R. leguminosarum but not E. coli (23). Lastly, R. leguminosarum extracts do not contain the Kdo-dependent lauroyltransferases, but they do possess a novel, high-molecular-weight acyl donor (Fig. 19) of unknown structure (23). The latter may be involved in the biogenesis of the C28 acyl chains found in rhizobia (Fig. 18) (23). It will be interesting to determine whether the combination of enzymes that process Kdo2-lipid IVA in R. leguminosarum can replace those that normally act on Kdo2-lipid IVA in E. coli, and vice versa. It will also be important to determine whether the unique structure of R. leguminosarum lipid A plays a role in the biology of symbiosis.
An important feature of combinatorial enzymology is its ability to generate molecular diversity. As methods for analyzing the core region improve, it will be interesting to examine the complete structures of cores isolated from cells growing under different physiological conditions or in infected animals. The ability of enzymes of core assembly to function in parallel rather than in sequence may provide a mechanism for creating structural diversity on the bacterial cell surface. This flexibility might increase the chances for survival under adverse circumstances or in the presence of immune surveillance. The phenomenon of phase variation in organisms like H. influenzae provides a further genetic mechanism for switching on alternative enzymes for generating core diversity (306).
The broad outlines of O-antigen structure, biosynthesis, and immunology have been known for over 20 years (3, 67, 110, 158, 238, 241, 242, 257). The recent sequencing of cloned rfb clusters (Fig. 8) from several sources has nevertheless provided important new insights (236, 237). The notion that individual rfb genes or gene clusters were acquired by interspecific transfer is especially intriguing (236). Perhaps other bacterial surface carbohydrates evolved by similar mechanisms.
Facile access to all of the rfb genes is necessary for incisive studies of O-antigen biochemistry. The availability of key enzymes for the construction of O-antigen oligosaccharides (158, 226, 238, 257) should facilitate the preparation of defined substrates for better in vitro studies of O-antigen polymerization (115, 246, 248, 302) and ligation (158, 238). The enzymology of polymerization has received very little attention since its discovery 30 years ago (115, 158, 238, 246, 248, 302). In vitro ligation has not been described at all (158, 238), but mutants defective in the rfaL gene do accumulate polymeric O antigen attached to bactoprenol pyrophosphate in vivo (186).
Recent enzymatic studies of 3,6-dideoxyhexose formation have provided new insights into the functions of the rfb gene products (150). The biosynthesis and structures of the common 6-deoxy- and 3,6-dideoxyhexoses are illustrated in Fig. 20 and 21 (150). The enzymes that generate dTDP-rhamnose (163) are encoded by the first four open reading frames (BDAC) of many rfb operons (Fig. 8), and the sequences of these four genes are relatively conserved (236, 237, 257, 269). Following the genes for dTDP-rhamnose assembly, one sometimes finds a cluster of genes (rfbIFGHJ in the case of group B S. typhimurium) (Fig. 8) that encode enzymes for 3,6-dideoxyhexose formation (compare Fig. 8 and 21) (111, 236, 257, 269). Since the structures of dideoxyhexoses are quite variable (150, 236), the DNAs from this region are correspondingly variable in different gram-negative bacteria (111, 236, 237, 269). E. coli K-12 is very different from S. typhimurium LT2, as it does not make any 3,6-dideoxyhexoses, even when its ability to synthesize its own O antigen (Fig. 5B) is restored by introduction of a functional orf11 (Fig. 8) (269). The rfb cluster of E. coli K-12 does not possess any DNA sequences related to the 3,6-dideoxyhexose pathway (269). Its rfb cluster, like that of many other gram-negative bacteria studied so far, is therefore somewhat shorter than that of S. typhimurium LT2 (Fig. 8) (111, 236, 237, 269).
Liu and coworkers have recently examined the enzymatic mechanisms of 3,6-dideoxyhexose formation in Yersinia pseudotuberculosis (150, 151, 174, 175, 283, 284, 285, 301). Many of the genes in this system (which generates ascarylose as its primary 3,6-dideoxyhexose) are related to those of salmonellae (150, 285). With the development of better enzymatic assays, Liu and coworkers have been able to assign unequivocally the genes encoding the rfbH and rfbI functions (150, 285) (Fig. 8) among the open reading frames sequenced by Reeves and coworkers (236). The unusual role of pyridoxamine 5'-monophosphate (Fig. 21) in the deoxygenation and reduction of the hexose C-3 position (catalyzed by the rfbH and rfbI gene products, respectively) is especially intriguing (283, 284). The dehydrase (rfbH) has an iron-sulfur cluster cofactor in addition to pyridoxamine 5'-monophosphate (150, 283, 284, 285, 301), and the reductase (rfbI) is a [2Fe-2S]-containing flavoprotein (151, 174, 175). The FeS clusters are essential for enzymatic activity (150). The reduction of the 3 position apparently involves one-electron transfers to the pyridoxamine 5'-monophosphate–sugar adduct (Fig. 21), generating novel radical intermediates (not shown) (174, 284).
The middle of many rfb clusters contains a hydrophobic open reading frame (rfbX) that codes for a protein with ∼12 membrane-spanning segments (Fig. 8) (111, 236, 257, 269). Despite their similar hydropathy profiles, the sequences of rfbX genes from different organisms are not conserved. The function of rfbX is unknown (111, 236, 257, 269). Its hydropathy profile suggests a role in transport. In S. enterica C1, rfbX is at the distal end of the rfb gene cluster (236, 237).
The distal ends of many rfb operons (Fig. 8) contain genes encoding enzymes required for the formation of certain sugar nucleotides (such as GDP-mannose) (236, 269). The distal rfb region may also code for glycosyltransferases that are involved in the cycle of O-antigen oligosaccharide assembly (Fig. 22) (111, 236, 257, 269). Since O antigens vary considerably in oligosaccharide composition and configuration (Fig. 5B and 15B), the distal rfb genes are correspondingly variable (236), as shown for E. coli K-12 and S. typhimurium LT2 in Fig. 8 (111, 269). In S. enterica C1,the presumptive glycosyltransferases are in the middle of the rfb cluster (236, 237).
The O-antigen polymer is synthesized independently of lipid A and core (158, 184, 201, 238). The cycle of reactions shown in Fig. 22 illustrates the sequence of glycosylations that are thought to give rise to a single unit of S. typhimurium group B O antigen (Fig. 15) and related species (115, 149, 246, 248, 302). None of these enzymes has been purified. The products generated by these enzymes in vitro have not been characterized by NMR spectroscopy or mass spectrometry. Consequently, some aspects of structure and mechanism require further investigation.
The enzymology of O-antigen biogenesis has not been investigated at all in E. coli K-12, given that the capacity of E. coli K-12 to synthesize O antigen was not recognized until very recently (269).
The O-antigen cycle is initiated by the transfer of galactose 1-phosphate from UDP-galactose to bactoprenol phosphate (Fig. 22) (205). In S. typhimurium LT2 and all other strains in which O-antigen synthesis starts with galactose, this reversible reaction is catalyzed by the rfbP gene product (Fig. 8), a membrane-associated enzyme that may also play a role in O-antigen export (see below) (299). The product, bactoprenol pyrophosphate-galactose, is the acceptor for three sequential glycosylations, catalyzed by sugar transferases encoded by the rfbN, rfbU, and rfbV genes (Fig. 8) (149, 236). Incorporation of the partial acetyl and glucosyl substituents of S. typhimurium group B O antigen (Fig. 15B) occurs after polymerization (158, 257). The genes responsible for the latter modifications (oafA and oafR) map outside the rfb cluster (Fig. 6 and 15B) (158, 257).
In E. coli, the first sugar of the O-antigen repeat is GlcNAc (Fig. 5B) (257, 269). The complete cycle of O-antigen enzymes has not yet been studied in E. coli. However, it is clear that the first step of O-antigen synthesis in E. coli is the reversible transfer of GlcNAc 1-phosphate from UDP-GlcNAc to bactoprenol phosphate (Fig. 9), analogous to the transfer of galactose 1-phosphate to bactoprenol phosphate in S. typhimurium (257). The GlcNAc 1-phosphate transferase is encoded by the rfe gene (Fig. 6), which maps outside the rfb cluster (169, 257). The rfe gene product is also required for the assembly of enterobacterial common antigen (Fig. 9) (169).
An important feature of O-antigen polymerization is that the newly generated repeat is incorporated at the reducing end of the growing chain (246, 247). This mechanism is illustrated with boldface in Fig. 22. Peptidoglycan grows in a similar manner (208). Growth at the reducing end has the advantage that the membrane-bound polymerase (encoded by rfc) does not have to search for the nonreducing terminus of the growing O-antigen polymer. The latter may be extended away from the membrane and the active site of the polymerase. In E. coli K-12 (269) and most other bacteria examined so far (236, 237), the rfc gene is part of the rfb cluster (Fig. 8), whereas in S. typhimurium LT2 it is not (42, 236). All O-antigen polymerases are hydrophobic proteins with at least 12 putative membrane-spanning regions (42, 178, 236, 269).
When LPSs bearing O-antigen polymers are fractionated by SDS-polyacrylamide gel electrophoresis, one observes a bimodal distribution of repeats (77, 78, 257). LPS molecules bearing a single O-antigen unit are relatively abundant (77, 78, 257). As more units are added, the relative amounts of the corresponding LPSs that are observed on gels rapidly decline (77, 78, 257). However, LPS species with ∼25 to 35 repeats are again relatively abundant (11, 12, 77, 78, 179, 257). Above ∼35 repeats, polymerization becomes inefficient, leading to a complete disappearance of polymers bearing more than 40 units (11, 12, 77, 78, 179, 257).
The nonrandom distribution of O-antigen units is brought about by the product of the rol (cld) gene (Fig. 6) (11, 12, 179). The latter encodes a membrane protein of unknown function that must somehow influence the catalytic efficiency of the rfc-encoded polymerase at each stage of elongation (11, 12, 77). Inactivation of the rol gene has no obvious effects on growth. However, such cells have a unimodal population of O-antigen repeats, as predicted by a simple model that assigns a single probability for the successful transfer of each O-antigen unit (11, 12, 77, 179). The physiological consequences of the rol mutation have not been examined. It may be that bacteria lacking the rol function are less capable of establishing infections. The reconstitution of the O-antigen polymerase together with the rol gene product in cell extracts has not been attempted.
The ligase that transfers O antigen to the outer core has not received much attention (158, 257) (Fig. 22). Two genes (rfaL and rfbT) were thought to be required for ligase function (158, 257). However, the rfbT gene has recently been shown by sequencing and complementation to be the same as rfbP (Fig. 8) (299). Strains bearing such rfbP(T) mutations are able to make a single O-antigen repeat, but it is not polymerized (299). In contrast, when the O-antigen ligase encoded by rfaL is defective (Fig. 22), cells accumulate polymeric O antigen on the periplasmic surfaces (see below) (186). Wang and Reeves speculate that the N-terminal domain of the rfbP protein may function as a flippase to deliver a single O-antigen unit attached to bactoprenol pyrophosphate to the periplasmic surface of the inner membrane (Fig. 23, step 1) (299). The N-terminal domain of the rfbP-encoded protein has four membrane-spanning segments, whereas the C-terminal region is the catalytic site for galactose 1-phosphate transfer (236, 299).
When the O-antigen polymerase is rendered defective by an rfc mutation, the ligase is still able to attach a single O-antigen repeat to the core (158, 189). Such strains are termed semirough (158). This interesting observation has not been reconstructed with purified proteins in vitro. The catalytic mechanism, specificity, and topography of the ligase have not been studied (158). Synthetic acceptors and defined bactoprenol-linked oligosaccharides would greatly facilitate characterization of the ligase.
Certain bacteriophages, such as P27, code for unique O-antigen polymerases that synthesize linkages different from those present in uninfected host cells (158). The phenomenon of bacteriophage-mediated O-antigen conversion was recognized over 30 years ago, but the enzymology has not been investigated. The same is true of the phenomenon of form variation (158).
Some investigators have questioned the generality of the cycle shown in Fig. 22 (109, 238). In certain gram-negative bacteria, O-antigen growth is thought to occur at the nonreducing end (109, 238).
The important studies of Osborn and coworkers (159, 160, 167, 168, 186, 201) have provided considerable evidence for the existence of periplasmic intermediates in LPS biogenesis. In strains of S. typhimurium unable to complete the biosynthesis of the core-lipid A domain (or in mutants defective in the rfaL gene), polymerized O antigens linked to bactoprenol pyrophosphate accumulate on the periplasmic surface of the inner membrane (167, 186). This material can be chased into mature LPS when core-lipid A synthesis is reinitiated (167). The data support the hypothesis that the active sites of the O-antigen ligase and O-antigen polymerase are located on the periplasmic surface of the inner membrane (167, 186, 201, 226). All enzymes of core-lipid A and O-antigen tetrasaccharide assembly are thought to be cytoplasmic or associated with the inner membrane. Transporters that export sugar nucleotides to the periplasm have not been detected. Accordingly, core-lipid A and single O-antigen units attached to bactoprenol pyrophosphate appear to flip independently from the cytoplasmic to the periplasmic surface of the inner membrane prior to polymerization and ligation (Fig. 23, steps 1 and 2) (167, 186, 201, 226). Additional carriers may be required to move the completed LPS through the periplasm and to the outer surface of the outer membrane (Fig. 23, steps 3 and 4) (186, 201, 226). Alternative mechanisms for LPS export have been discussed previously (201, 226), but the limited available evidence is most consistent with the model shown in Fig. 23.
The export of LPS appears to be permissive with respect to variations in O-antigen sugar composition. For example, it is possible to express foreign O antigens in E. coli K-12 and in other gram-negative bacteria by introduction of appropriate hybrid plasmids (1, 274, 275). Alternatively, the minimal Re LPS (Kdo2-lipid A) required for growth is also exported efficiently (201), presumably via steps 2 to 4 shown in Fig. 23.
Several key events in LPS biogenesis (Fig. 23) require the proton motive force. These include (i) the transposition of newly made core-lipid A from the cytoplasmic to the periplasmic surface of the inner membrane (159, 168) and (ii) the initial step of O-antigen formation (galactose 1-phosphate transfer from UDP-galactose to bactoprenol phosphate) in living cells (160). The latter may be related to the rfbP(T) phenomenon described above (299). Transport of core-lipid A from inner to outer membranes also requires ATP (159). Newly made LPS supposedly first appears on the outer surface of the outer membrane over putative zones of adhesion (185, 201).
In short-term incubations of S. typhimurium with 1 mM dinitrophenol, there is little or no inhibition of de novo core-lipid A biosynthesis, showing that the membrane potential is not required for the enzymatic synthesis of core-lipid A, provided that cellular ATP levels are not depleted (159, 160, 168). If the translocation of core-lipid A to the periplasm is mediated by a membrane protein (Fig. 23, step 2), it should be possible to identify mutants specifically defective in core-lipid A export. As yet, there is no direct biochemical or genetic evidence for such a transport protein. However, the possible clue provided by the msbA multicopy suppressor of temperature-sensitive mutants defective in laurate addition to lipid A (htrB) (118, 122) needs to be examined further. Transporters that have recently been identified in other polysaccharide-secreting systems also may provide relevant paradigms (211, 317).
Future progress with LPS export will require the development of new biochemical assays for measuring the rate of LPS secretion directly and the isolation of new mutants specifically blocked at each step of the export pathway. The recent identification of mutations in the genes that are involved in the final stages of Re LPS assembly, such as kdtA, htrB, and msbB (37), may also lead to identification of LPS transporters, since phenotypes associated with late biosynthetic mutations may be suppressed by compensatory mutations in the export machinery.
LPS molecules can dissociate from the surfaces of gram-negative bacteria (164). During severe gram-negative infections, macrophages and endothelial cells are greatly stimulated by the released LPS, resulting in excess production of cytokines and inflammatory mediators (145, 192, 207). These products of macrophage activation, not LPS itself, are thought to be responsible for the complications of gram-negative sepsis (145, 192, 207). Certain antibiotics may actually enhance the release of LPS from bacteria (108, 217). Synthetic lipid A closely mimics the effects of LPS in animal models (65, 145).
Considerable progress has been made recently in defining the initial steps in the interaction of lipid A with animal cells (263, 289, 290, 315, 316) (Fig. 24). The lipid A anchor of LPS can bind to an acute-phase serum protein (LBP), which is produced in hepatocytes in response to cytokines (83). LBP displays significant sequence homology to cholesterol ester transfer protein (57). LBP can function as a lipid transfer protein that delivers lipid A to CD14 (84, 91), although in some situations LPS-LBP complexes appear to interact with membrane-associated CD14 (71). CD14 is a surface protein found on macrophages and other cells (63, 261, 265). A soluble form of CD14 is also present in serum (64). CD14 had been identified as an epitope and cloned prior to the discovery of its involvement in lipid A recognition (63, 261, 265).
The cell activation scheme shown in Fig. 24 was proposed on the basis of the neutralizing effects of certain antibodies directed against CD14 on diverse LPS responses (162, 165, 316). To observe the maximal potency of LPS as an activator of macrophages in cell culture (263, 286) and as a toxin in animal models (69), LBP is also required. Certain cells, such as 70Z/3 pre-B lymphocytes, lack CD14 (140), and while they can respond to "high" levels (nanomolar) of lipid A (140, 264), they are capable of responding to very "low" levels (picomolar) when transfected with CD14 (140, 142). However, it is not entirely certain whether the same signal transduction pathways are employed in the presence and absence of CD14 (see below).
Interestingly, even Chinese hamster ovary (CHO) cells can respond to LPS when transfected with CD14 (49, 50, 80). Not only is arachidonate released from phospholipids when LPS is added to the growth medium of such transfectants (80), but also NF-κB is activated at low LPS concentrations (49, 50). Normal untransfected CHO cells display no apparent responses to LPS. These findings establish CHO cells as a convenient somatic cell model system for studies of endotoxin biology, and they support the general importance of the CD14-LBP system in cell activation.
As shown in Fig. 24, CD14 is a 55-kDa glycoprotein that is bound to the cell surface by a phosphatidylinositol anchor (265). The latter is not critical to CD14 action, since recombinant CD14 that is anchored to the membrane by a hydrophobic C-terminal peptide is also functional (141). In endothelial cell lines, soluble CD14 can even substitute for membrane-bound CD14 (64, 220, 235). The N-terminal domain of CD14 (residues 1 to 152) appears to contain the LPS binding site and is sufficient for cell activation (113, 293). Residues 57 to 64 may form a part of the actual LPS binding site (112).
CD14 is believed to direct LPS to another, as yet unknown transmembrane protein capable of generating a key intracellular signal(s) (Fig. 24) (50, 126, 127, 142). In macrophages, the signals produced by the putative protein downstream of CD14 stimulate transcription of mRNAs encoding cytokines, such as tumor necrosis factor and interleukin-1 (16, 54, 145). LPS-induced enzymes, like NO synthase, are also expressed because of transcriptional activation (154, 308). Excessive production of multiple cytokines and other nonprotein mediators, including NO and platelet-activating factor, is responsible for the symptoms of endotoxin-induced shock (16, 54, 145, 181, 192). Lipid A itself is not directly toxic to animals or to animal cells in culture. CHO cells lacking CD14 can grow in the presence of 100 μg of LPS per ml (49; Raetz, unpublished). Transgenic mice lacking CD14 are orders of magnitude more resistant to LPS than are normal mice (S. A. Goyert, paper presented at the International Business Communications Meeting "Endotoxemia and Sepsis," Philadelphia, Pa., 1984).
The notion that phosphatidylinositol-linked proteins can interact both with specific ligands and with additional transmembrane proteins to generate intracellular signals is not without precedent. The receptor for interleukin-6 (276) displays features that are very similar to the scheme of Fig. 24.
The relevant intracellular signals resulting from LPS stimulation have recently been shown to include rapid activation of specific protein tyrosine phosphorylations (303, 304). In macrophages and macrophage tumor cells, two isoforms of mitogen-activated protein (MAP) kinase are activated within minutes by phosphorylation on tyrosine (92, 303, 304). In addition, a MAP kinase homolog of ∼38,000 kDa is phosphorylated on tyrosine after exposure of the cells to lipid A or LPS (92, 303, 304). In CD14-transfected 70Z/3 cells, only the protein with a molecular mass of ∼38,000 kDa is phosphorylated (92, 142). The activation of NF-κB and the transcription of light chains in 70Z/3 cells following LPS exposure are somewhat slower (253).
Han et al. have isolated and cloned the gene encoding p38 (90). p38 displays significant homology to the well-characterized MAP kinases, but it is nevertheless a member of a distinctly different kinase family (68, 90). The sequence of p38 is homologous to that of a kinase in yeast cells (encoded by the HOG gene) required for growth at high osmolarity (30, 68, 90). Expression of murine p38 in yeast cells can restore the ability of yeast mutants lacking HOG to grow at high osmolarity (68, 90). The recognition of a yeast homolog of p38 also raises the intriguing question of whether yeast cells can respond to bacterial endotoxins in certain situations.
The enzymes that may be responsible for the tyrosine phosphorylation of p38 following exposure of animal cells to LPS have recently been identified (51). Novel isoforms of MAP kinase kinase (designated MKK3 and MKK4) can phosphorylate p38 at its dual phosphorylation sites, Thr-180 and Tyr-182 (51). MKK3 appears to be selective for p38, whereas MKK4 also phosphorylates the JNK subgroup of MAP kinases (51).
The human homolog of murine p38 has also recently been cloned (139). This was accomplished by determining the protein target of specific chemicals that block the ability of human THP cells to produce tumor necrosis factor in response to LPS (139). These small molecules, which have potential as novel anti-inflammatory agents, presumably block the enzymatic activity of p38 (a serine/threonine kinase) (139). The relevant substrates for p38 are still unknown. The latter might include IκB-α, a cytosolic protein inhibitor of the NF-κB family of transcription factors. IκB must be removed from NF-κB by a combination of phosphorylation and proteolysis prior to translocation of active NF-κB into the nucleus (20, 99, 146). The promoters of genes that are transcribed in response to lipid A usually contain NF-κB binding sequences (154, 253).
It is probable that several other components downstream of CD14 but upstream of MKK3 and p38 remain to be discovered in the lipid A-mediated signal transduction cascade (Fig. 24). Identification of the sensor that specifically recognizes lipid A-like molecules as either agonists or antagonists (34, 79, 126, 127, 278) would be of special interest. CD14 is not entirely specific for lipid A, since it also binds to lipoteichoic acids (219).
In 70Z/3 cells that are not transfected with CD14, it is still possible to achieve full activation of κ-chain transcription and surface immunoglobulin assembly, but higher LPS concentrations (∼100 nM) are required (142, 264). Under these conditions, there is no detectable tyrosine phosphorylation of p38 (142). This important observation is not entirely consistent with the direct involvement of p38 phosphorylation in LPS-triggered activation of κ-chain transcription but rather implies that p38 phosphorylation may be a separate LPS-activated event. Perhaps lipid A-mediated signal transduction that is independent of CD14 involves entirely different steps that converge beyond p38. In 70Z/3 cells, both CD14-dependent and CD14-independent activation of κ-chain transcription are blocked by endotoxin antagonists, such as lipid A from Rhodobacter sphaeroides (Fig. 25) (142). Tyrosine phosphorylation of p38 in CD14-transfected cells is also blocked by endotoxin antagonists, suggesting that the antagonists function upstream of p38 (142).
Another powerful approach to the identification of additional signal-transducing proteins would be afforded by the cloning of the genes that are defective in somatic cell mutants that are selectively unresponsive to LPS. Such mutants have been isolated from 70Z/3 cells (157) and from macrophage tumor cells (93). Lastly, the identification of the gene that is altered in LPS-resistant C3H/HeJ mice would provide important new insights (190).
Lipid A and LPS are bound and internalized by cells in culture that express the macrophage scavenger receptor (86) (not shown in Fig. 24). Inhibition of uptake mediated by the scavenger receptor does not prevent stimulation of cytokine synthesis (86), indicating that the scavenger receptor is not involved in lipid A-mediated signal transduction. Expression of CD14 can also increase the ability of cells to take up LPS bound to LBP (71), but this phenomenon can be blocked by certain antibodies without inhibition of LPS-activated signal transduction (127). Following endocytosis, some cells are capable of dephosphorylating (89) or deacylating (60, 188) lipid A. Such modifications reduce the potency of lipid A as an agonist. Selective removal of the normal fatty acyl chains from lipid A actually gives rise to molecules that function as endotoxin antagonists in human cells (60, 79, 127, 136).
Albumin has a considerable capacity for binding LPS and other molecules derived from bacterial cell surfaces (58). This phenomenon can interfere with attempts to identify lipid A receptors by direct biochemical methods. Binding of lipid A by albumin does not play a direct role in signal transduction.
The intermediates of the lipid A pathway (Fig. 10 to 12) have been evaluated for the ability to function as endotoxin mimetics or antagonists (79, 136, 153). Monosaccharide precursors, when properly purified, display no agonist activity (138; Raetz, unpublished), but they do possess slight endotoxin-antagonizing activity in some systems (46, 80a, 218). Underacylated disaccharide bisphosphate precursors such as lipid IVA (Fig. 12) are relatively effective lipid A (endotoxin) antagonists in human cells (79, 136, 153), but they are recognized as endotoxin-like agonists by murine lines (79, 264).
Some naturally occurring lipid A’s, such as that from R. sphaeroides (Fig. 25), also function as antagonists of the action of E. coli endotoxin on animal cells (34, 35, 79, 278). These derivatives resemble lipid IVA in that they are somewhat underacylated compared with mature E. coli lipid A (Fig. 25). R. sphaeroides lipid A is an antagonist of E. coli endotoxin in both human and murine cells (79). The remarkable ability of different animal cell lines to distinguish rather similar lipid A-like molecules provides additional evidence for specificity in lipid A recognition. As noted above, the specificity-conferring factor appears to reside downstream of CD14 (which cannot distinguish between the lipid A analogs shown in Fig. 25) (50, 127), but it must be upstream of p38 (Fig. 24), since tyrosine phosphorylation of p38 is blocked by lipid A antagonists (142).
Potent, synthetic lipid A-like molecules that function as general endotoxin antagonists in human cells (Fig. 25) have recently been reported by Christ and coworkers (34, 35). These remarkable compounds are able to block almost completely the effects of a low-dose endotoxin challenge in humans (27, 28). The LPS antagonists have not yet been evaluated in septic patients. Whether these compounds work as competitive inhibitors of LPS binding, as inverse agonists, or as activators of alternative, negative signalling pathways is not known. Perhaps they rapidly induce a state of LPS tolerance (166), an interesting phenomenon that requires further study at the molecular level (152).
Complications of gram-negative infections are a common cause of death in debilitated patients (192, 205). On the basis of current knowledge of the pathophysiology of LPS-stimulated mediator production (Fig. 24), three general approaches are being explored (47). The first involves the sequestration of lipid A with antibodies (300) or with a recombinant bactericidal/permeability- increasing protein normally produced in neutrophils (81, 260, 307). Clinical trials with the former have been disappointing (47, 300), probably because the available antibodies were not sufficiently potent to block the effects of endotoxins in vivo. Bactericidal/permeability-increasing protein, which appears promising in animal models (145) and during low-dose endotoxin administration to normal individuals (S. J. H. van Deventer, paper presented at the International Business Communications Meeting "Endotoxemia and Sepsis," Philadelphia, Pa., 1994), has not yet been tested in septic patients. The second approach involves blocking of the action of key cytokines, as with antibodies directed against tumor necrosis factor (39) or with a protein antagonist that binds to the interleukin-1 receptor (8, 54, 282). Despite initial promise, phase III trials in patients have failed to confirm good efficacy (282). Perhaps it is insufficient to block one arm of the inflammatory response, given the multitude of cytokines and mediators that are induced by LPS (145). The third, newest approach involves lipid A antagonists, such as the analogs of R. sphaeroides lipid A (34, 35, 79, 278) (Fig. 25), to block the initial activation of macrophages (Fig. 24). Blocking the initial event(s) should prevent the production of all mediators induced by LPS while not completely inactivating the inflammatory response. Lipid A antagonists have not yet been tested in septic patients, but such antagonists are very effective in normal volunteers challenged with low doses of intravenous endotoxins (27, 28).
In addition to these approaches, compounds that block tyrosine phosphorylation (199) or that inhibit p38 directly (139) will have to be examined more thoroughly in relevant animal models of endotoxemia and in clinical situations. Pharmaceutical applications of the enormous progress that has occurred since the elucidation of the structure of lipid A in 1983 (244) are still in their infancy.
Future progress with bacterial LPSs will require elucidation of the structural biology, secretion, and functions of LPSs in bacteria. The physical properties of purified lipid A-like molecules and their interactions with proteins have yet to be investigated in detail. New chemical methods for synthesizing defined cores and lipid A-like molecules are needed. Although many key genes of lipid A, core, and O-antigen biosynthesis have now been identified (Fig. 5, 6, 7, 8, 9, 10, 11, 12, 13, 15, 16, 17, and 20, 21, 22), important components, including the genes for the 4'-kinase, possible regulatory factors, and putative transporters required for the export of LPS from its site of biosynthesis on the cytoplasmic membrane to the outer membrane (Fig. 23), remain to be characterized. Novel inhibitors of lipid A biosynthesis and tighter conditional mutants would also facilitate functional studies in bacteria. Genetic analyses of second-site suppressors and of protein-protein interactions are likely to be fruitful. Modification of the structure of lipid A in living cells of E. coli by expression of genes coding for enzymes that act on lipid A in other systems (like rhizobia) might shed some light on lipid A function. Perhaps bacterial cells could be made to grow without any lipid A in the presence of appropriate second-site suppressors or under stabilizing conditions.
Lastly, the complex interactions of lipid A with the signal transduction cascades of animal cells require considerable further work. It may be that lipid A functions by mimicking a minor, as yet undiscovered lipid mediator of animal cell signal transduction. The identification of a mammalian lipid A equivalent might have profound implications for the role of lipid A in the evolution of the immune system.
I thank all present and past members of my laboratory for their contributions to this story. I thank Kathryn Brozek, Teresa Garrett, and Peter Reeves for their thoughtful comments on the manuscript. Past research funding was from NIH and Merck & Co., Inc. Current studies at Duke University are supported by start-up funds from the North Carolina Biotechnology Center and by NIH grants GM-51310 and GM-51796.
References
1. al-Hendy, A., P. Toivanen, and M. Skurnik. 1991. Expression cloning of Yersinia enterocolitica O:3 rfb gene cluster in Escherichia coli K-12. Microb. Pathog. 10:47–59.
2. Ames, G.-F., E. N. Spudich, and H. Nikaido. 1974. Protein composition of the outer membrane of Salmonella typhimurium: effect of lipopolysaccharide mutations. J. Bacteriol. 117:406–416.
3. Anderson, L., and F. M. Unger (ed.). 1983. Bacterial Lipopolysaccharides. ACS Symp. Ser., vol. 231. American Chemical Society, Washington, D.C.
4. Anderson, M. S., C. E. Bulawa, and C. R. H. Raetz. 1985. The biosynthesis of gram-negative endotoxin: formation of lipid A precursors from UDP-GlcNAc in extracts of Escherichia coli. J. Biol. Chem. 260:15536–15541.
5. Anderson, M. S., H. S. Bull, S. M. Galloway, T. M. Kelly, S. Mohan, K. Radika, and C. R. H. Raetz. 1993. UDP-N-acetylglucosamine acyltransferase of Escherichia coli: the first step of endotoxin biosynthesis is thermodynamically unfavorable. J. Biol. Chem. 268:19858–19865.
6. Anderson, M. S., and C. R. H. Raetz. 1987. Biosynthesis of lipid A precursors in Escherichia coli: a cytoplasmic acyltransferase that converts UDP-N-acetylglucosamine to UDP-3-O-(R-3-hydroxymyristoyl)-N-acetylglucosamine. J. Biol. Chem. 262:5159–5169.
7. Anderson, M. S., A. D. Robertson, I. Macher, and C. R. H. Raetz. 1988. Biosynthesis of lipid A in Escherichia coli: identification of UDP-3-O-(R-3-hydroxymyristoyl)-α-d-glucosamine as a precuosor of UDP-N 2-O 3-bis-(R-3-hydroxymyristoyl)-α-d-glucosamine. Biochemistry 27:1908–1917.
8. Arend, W. P. 1991. Interleukin 1 receptor antagonist: a new member of the interleukin 1 family. J. Clin. Invest. 88:1445–1451.
9. Austin, E. A., J. F. Graves, L. A. Hite, C. T. Parker, and C. A. Schnaitman. 1990. Genetic analysis of lipopolysaccharide core biosynthesis by Escherichia coli: insertion mutagenesis of the rfa locus. J. Bacteriol. 172:5312–5325.
10. Babinet, C. 1970. A mutation that affects the resistance of E. coli to rifampicin, p. 37–45. In Proceedings of the First International Lepetit Colloquium.
11. Bastin, D. A., G. Stevenson, P. K. Brown, A. Haase, and P. R. Reeves. 1993. Repeat unit polysaccharides of bacteria: a model for polymerization resembling that of ribosomes and fatty acid synthetase, with a novel mechanism for determining chain length. Mol. Microbiol. 7:725–734.
12. Batchelor, R. A., P. Alifano, E. Biffali, S. I. Hull, and R. A. Hull. 1992. Nucleotide sequences of the genes regulating O-polysaccharide antigen chain length (rol) from Escherichia coli and Salmonella typhimurium. J. Bacteriol. 174:5228–5236.
13. Beall, B., and J. Lutkenhaus. 1987. Sequence analysis, transcriptional organization, and insertional mutagenesis of the envA gene of Escherichia coli. J. Bacteriol. 169:5408–5415.
14. Belunis, C. J., K. E. Mdluli, C. R. H. Raetz, and F. E. Nano. 1992. A novel 3-deoxy-d-manno-octulosonic acid transferase from Chlamydia trachomatis required for expression of the genus-specific epitope. J. Biol. Chem. 267:18702–18707.
15. Belunis, C. J., and C. R. H. Raetz. 1992. Biosynthesis of endotoxins: purification and catalytic properties of 3-deoxy-d-manno-octulosonic acid transferase from Escherichia coli. J. Biol. Chem. 267:9988–9997.
16. Beutler, B., and A. Cerami. 1988. Tumor necrosis, cachexia, shock, and inflammation: a common mediator. Annu. Rev. Biochem. 57:505–518.
17. Bhat, U. R., L. S. Forsberg, and R. W. Carlson. 1994. The structure of the lipid A component of Rhizobium leguminosarum bv. phaseoli lipopolysaccharide. A unique, non-phosphorylated lipid A containing 2-amino-2-deoxy-gluconate, galacturonate, and glucosamine. J. Biol. Chem. 269:14402–14410.
18. Bhat, U. R., B. S. Krishnaiah, and R. W. Carlson. 1991. Re-examination of the structures of the lipopolysaccharide core oligosaccharides from Rhizobium leguminosarum biovar phaseoli. Carbohydr. Res. 220:219–227.
19. Brazas, R., E. Davie, A. Farewell, and L. I. Rothfield. 1991. Transcriptional organization of the rfaGBIJ locus of Salmonella typhimurium. J. Bacteriol. 173:6168–6173.
20. Brown, K., S. Gerstberger, L. Carlson, G. Franzoso, and U. Siebenlist. 1995. Control of IκB-α proteolysis by site-specific, signal-induced phosphorylation. Science 267:1485–1488.
21. Brozek, K. A., C. E. Bulawa, and C. R. H. Raetz. 1987. Biosynthesis of lipid A precursors in Escherichia coli: a membrane bound enzyme that transfers a palmitoyl residue from a glycerophospholipid to lipid X. J. Biol. Chem. 262:5170–5179.
22. Brozek, K. A., K. Hosaka, A. D. Robertson, and C. R. H. Raetz. 1989. Biosynthesis of lipopolysaccharide in Escherichia coli: cytoplasmic enzymes that attach 3-deoxy-d-manno-octulosonic acid to lipid A. J. Biol. Chem. 264:6956–6966.
23. Brozek, K. A., J. L. Kadrmas, and C. R. H. Raetz. 1995. Lipid A biosynthesis in Rhizobium: novel phosphatases, glycosyl transferases, and an unusual acyl transferase. FASEB J. 9:A1376.
24. Brozek, K. A., and C. R. H. Raetz. 1990. Biosynthesis of lipid A in Escherichia coli: acyl carrier protein-dependent incorporation of laurate and myristate. J. Biol. Chem. 265:15410–15417.
25. Bulawa, C. E., and C. R. H. Raetz. 1984. The biosynthesis of gram-negative endotoxin: identification and function of UDP-2, 3-diacylglucosamine in Escherichia coli. J. Biol. Chem. 259:4846–4851.
26. Bulawa, C. E., and C. R. H. Raetz. 1984. Isolation and characterization of Escherichia coli strains defective in CDP-diglyceride hydrolase. J. Biol. Chem. 259:11257–11264.
27. Bunnell, E., M. Lynn, J. E. Parillo, K. Habet, L. T. Friedhoff, and S. L. Rogers. 1995. Effect of E5531 on systemic responses to endotoxin in healthy volunteers. Crit. Care Med. Suppl. 23:A147.
28. Bunnell, E., A. Neumann, M. Lynn, L. T. Friedhoff, S. L. Rogers, K. Habet, and J. E. Parillo. 1995. E5531, an endotoxin antagonist, blocks the hyperdynamic and depressant cardiovascular effects of endotoxin in healthy subjects. Crit. Care Med. Suppl. 23:A151.
29. Burton, A. J., and H. E. Carter. 1964. Purification and characterization of the lipid A component of the lipopolysaccharides from Escherichia coli. Biochemistry 3:411–418.
30. Cano, E., and L. C. Mahadevan. 1995. Parallel signal processing among mammalian MAPKs. Trends Biochem. Sci. 20:117–122.
31. Caroff, M., C. Deprun, and D. Karibian. 1993. 252Cf plasma desorption mass spectrometry applied to underivatized rough-type endotoxin preparations. J. Biol. Chem. 268:12321–12324.
32. Caroff, M., C. Deprun, D. Karibian, and L. Szabó. 1991. Analysis of unmodified endotoxin preparations by 252Cf plasma desorption mass spectrometry. J. Biol. Chem. 266:18543–18549.
33. Caroff, M., A. Tacken, and L. Szabó. 1988. Detergent-accelerated hydrolysis of bacterial endotoxins and determination of the anomeric configuration of the glycosyl phosphate present in the "isolated lipid A" fragment of the B. pertussis endotoxin. Carbohydr. Res. 175:273–282.
34. Christ, W. J., O. Asano, A. L. Robidoux, M. Perez, Y. Wang, G. R. Dubuc, W. E. Gavin, L. D. Hawkins, P. D. McGuinness, M. A. Mullarkey, M. D. Lewis, Y. Kishi, T. Kawata, J. R. Bristol, J. R. Rose, D. P. Rossignol, S. Kobayashi, I. Hishinuma, A. Kimura, N. Asakawa, K. Katayama, and I. Yamatsu. 1995. E5531, a pure endotoxin antagonist of extraordinary potency: chemistry and biology. Science 268:80–83.
35. Christ, W. J., P. D. McGuinness, O. Asano, Y. Wang, M. A. Mullarkey, M. Perez, L. D. Hawkins, T. A. Blythe, G. R. Dubuc, and A. L. Robidoux. 1994. Total synthesis of the proposed structure of Rhodobacter sphaeroides lipid A resulting in the synthesis of a new potent lipopolysaccharide antagonist. J. Am. Chem. Soc. 116:3637–3638.
36. Clementz, T. 1992. The gene coding for 3-deoxy-d-manno-octulosonic acid transferase and the rfaQ gene are transcribed from divergently arranged promoters in Escherichia coli. J. Bacteriol. 174: 7750–7756.
37. Clementz, T., J. Bednarski, and C. R. H. Raetz. 1995. Escherichia coli genes encoding Kdo dependent acyltransferases that incorporate laurate and myristate into lipid A. FASEB J. 9:A1311.
38. Clementz, T., and C. R. H. Raetz. 1991. A gene coding for 3-deoxy-d-manno-octulosonic acid transferase in Escherichia coli: identification, mapping, cloning, and sequencing. J. Biol. Chem. 266:9687–9696.
39. Cohen, J., A. R. Exley, W. Buurman, R. Owen, G. Hanson, J. Lumley, J. M. Aulakh, M. Bodmer, A. Riddle, S. Stephens, and M. Perry. 1990. Monoclonal antibody to TNF in severe septic shock. Lancet 335:1275–1277.
40. Coleman, J., and C. R. H. Raetz. 1988. First committed step of lipid A biosynthesis in Escherichia coli: sequence of the lpxA gene. J. Bacteriol. 170:1268–1274.
41. Coleman, W. G., Jr. 1983. The rfaD gene codes for ADP-l-glycero-d-manno-heptose-6-epimerase. J. Biol. Chem. 258:1985–1990.
42. Collins, L. V., and J. Hackett. 1991. Molecular cloning, characterization, and nucleotide sequence of the rfc gene, which encodes an O-antigen polymerase of Salmonella typhimurium. J. Bacteriol. 173:2521–2529.
43. Crowell, D. N., M. S. Anderson, and C. R. H. Raetz. 1986. Molecular cloning of the genes for lipid A disaccharide synthase and UDP-N-acetylglucosamine acyltransferase in Escherichia coli. J. Bacteriol. 168:152–159.
44. Crowell, D. N., W. S. Reznikoff, and C. R. H. Raetz. 1987. Nucelotide sequence of the Escherichia coli gene for lipid A disaccharide synthase. J. Bacteriol. 169:5727–5734.
45. Cygler, M., D. R. Rose, and D. R. Bundle. 1992. Recognition of a cell surface oligosaccharide of pathogenic Salmonella by an antibody fab fragment. Science 253:442–445.
46. Danner, R. L., K. A. Joiner, and J. E. Parillo. 1987. Inhibition of endotoxin-induced priming of human neutrophils by lipid X and 3-aza-lipid X. J. Clin. Invest. 80:605–612.
47. Davis, J. 1993. New approaches to septic shock. SCRIP 1793:22–23.
48. DeChavigny, A., P. N. Heacock, and W. Dowhan. 1991. Sequence and inactivation of the pss gene of Escherichia coli. J. Biol. Chem. 266:5323–5332.
49. Delude, R. L., M. J. Fenton, R. Savedra, Jr., P.-Y. Perera, S. N. Vogel, R. Thieringer, and D. T. Golenbock. 1994. CD14-mediated translocation of nuclear factor-κB induced by lipopolysaccharide does not require tyrosine kinase activity. J. Biol. Chem. 269:22253–22260.
50. Delude, R. L., R. Savedra, Jr., H. Zhao, R. Thieringer, S. Yamamoto, M. J. Fenton, and D. T. Golenbock. CD14 enhances cellular responses to endotoxin without imparting ligand-specific recognition. Proc. Natl. Acad. Sci. USA, in press.
51. Dérijard, B., J. Raingeaud, T. Barrett, I.-H. Wu, J. Han, R. J. Ulevitch, and R. J. Davis. 1995. Independent human MAP kinase signal transduction pathways defined by MEK and MKK isoforms. Science 267:682–685.
52. Dicker, I. B., and S. Seetharam. 1991. Cloning and nucleotide sequence of the firA gene and the firA200 allele from Escherichia coli. J. Bacteriol. 173:334–344.
53. Dicker, I. B., and S. Seetharam. 1992. What is known about the structure and function of the Escherichia coli protein FirA? Mol. Microbiol. 6:817–823.
54. Dinarello, C. A. 1991. Interleukin-1 and interleukin-1 antagonism. Blood 77:1627–1652.
55. Ding, L., B. L. Seto, S. A. Ahmed, and W. G. Coleman, Jr. 1994. Purification and properties of the Escherichia coli K-12 NAD-dependent nucleotide diphosphosugar epimerase, ADP-l-glycero-d-mannoheptose 6-epimerase. J. Biol. Chem. 269:24384–24390.
56. Dotson, G. D., P. Nanjappan, M. D. Reily, and R. W. Woodard. 1993. Stereochemistry of 3-deoxyoctulosonate 8-phosphate synthase. Biochemistry 32:12392–12397.
57. Drayna, D., A. S. Jarnagin, J. McLean, W. Henzel, W. Kohr, C. Fielding, and R. Lawn. 1987. Cloning and sequencing of human cholesteryl ester transfer protein cDNA. Nature (London) 327:632–634.
58. Dziarski, R. 1994. Cell-bound albumin is the 70-kDa peptidoglycan-, lipopolysaccharide-, and lipoteichoic acid-binding protein on lymphocytes and macrophages. J. Biol. Chem. 269:20431–20436.
59. Endo, A., and L. I. Rothfield. 1969. Studies of a phospholipid-requiring bacterial enzyme. II. The role of phospholipid in the uridine diphosphate galactose: lipopolysaccharide α-3-galactosyltransferase reaction. Biochemistry 8:3508–3515.
60. Erwin, A. L., and R. S. Munford. 1992. Processing of LPS by phagocytes, p. 405–434. In D. C. Morrison and J. L. Ryan (ed.), Bacterial Endotoxic Lipopolysaccharides, vol. I. Molecular Biochemistry and Cellular Biology. CRC Press, Boca Raton, Fla.
61. Fält, I. C., E. Schweda, A. Weintraub, S. Sturm, K. N. Timmis, and A. A. Lindberg. 1993. Expression of Shigella dysenteriae type 1 lipopolysaccharide repeating units in Escherichia coli-K-12-Shigella dysenteriae type 1 hybrids. Eur. J. Biochem. 213:573–581.
62. Farewell, A., R. Brazas, E. Davie, J. Mason, and L. I. Rothfield. 1991. Suppression of the abnormal phenotype of Salmonella typhimurium rfaH mutants by mutations in the gene for transcription termination factor Rho. J. Bacteriol. 173:5188–5193.
63. Ferrero, E., C.-H. Hsieh, U. Francke, and S. M. Goyert. 1990. CD14 is a member of the family of leucine-rich proteins and is encoded by a gene syntenic with multiple receptor genes. J. Immunol. 145:331–336.
64. Frey, E. A., D. S. Miller, T. G. Jahr, A. Sundan, V. Bazil, T. Espevik, B. B. Finlay, and S. D. Wright. 1992. Soluble CD14 participates in the response of cells to lipopolysaccharide. J. Exp. Med. 176:1665–1671.
65. Galanos, C., O. Lüderitz, E. T. Rietschel, O. Westphal, H. Brade, L. Brade, M. Freudenberg, U. Schade, M. Imoto, H. Yoshimura, S. Kusumoto, and T. Shiba. 1985. Synthetic and natural Escherichia coli free lipid A express identical endotoxic activities. Eur. J. Biochem. 148:1–5.
66. Galanos, C., O. Lüderitz, and O. Westphal. 1969. A new method for the extraction of R lipopolysaccharides. Eur. J. Biochem. 9:245–249.
67. Galanos, C., E. T. Rietschel, O. Lüderitz, and O. Westphal. 1977. Newer aspects of the chemistry and biology of bacterial lipopolysaccharides, with special reference to their lipid A component, p. 239–335. In T. W. Goodwin (ed.), International Review of Biochemistry: Biochemistry of Lipids II, vol. 14. University Park Press, Baltimore.
68. Galcheva-Gargova, Z., B. Dérijard, I.-H. Wu, and R. J. Davis. 1994. An osmosensing signal transduction pathway in mammalian cells. Science 265:806–808.
69. Gallay, P., D. Heumann, D. LeRoy, C. Barras, and M.-P. Glauser. 1994. Mode of action of anti-lipopolysaccharide-binding protein antibodies for prevention of endotoxemic shock in mice. Proc. Natl. Acad. Sci. USA 91:7922–7926.
70. Galloway, S. M., and C. R. H. Raetz. 1990. A mutant of Escherichia coli defective in the first step of endotoxin biosynthesis. J. Biol. Chem. 265:6394–6402.
71. Gegner, J. A., R. J. Ulevitch, and P. S. Tobias. 1995. Lipopolysaccharide (LPS) signal transduction and clearance: dual roles for LPS binding protein and membrane CD14. J. Biol. Chem. 270:5320–5325.
72. Ghalambor, M. A., and E. C. Heath. 1966. The biosynthesis of cell wall lipopolysaccharide in Escherichia coli. IV. Purification and properties of cytidine monophosphate 3-deoxy-d-manno-octulosonate synthetase. J. Biol. Chem. 241:3216–3221.
73. Goldman, R., W. Kohlbrenner, P. Lartey, and A. Pernet. 1987. Antibacterial agents specifically inhibiting lipopolysaccharide synthesis. Nature (London) 329:162–164.
74. Goldman, R. C., T. J. Bolling, W. E. Kohlbrenner, Y. Kim, and J. L. Fox. 1986. Primary structure of CTP:CMP-3-deoxy-d-manno-octulosonate cytidylyltransferase from Escherichia coli. J. Biol. Chem. 261:15831–15835.
75. Goldman, R. C., C. C. Doran, and J. O. Capobianco. 1988. Analysis of lipolysaccharide synthesis in Salmonella typhimurium and Escherichia coli using agents which block incorporation of KDO. J. Bacteriol. 170:2185–2192.
76. Goldman, R. C., C. C. Doran, S. K. Kadam, and J. O. Capobianco. 1988. Lipid A precursor from Pseudomonas aeruginosa is completely acylated prior to addition of 3-deoxy-d-manno-octulosonate. J. Biol. Chem. 263:5217–5223.
77. Goldman, R. C., and F. Hunt. 1990. Mechanism of O-antigen distribution in lipopolysaccharide. J. Bacteriol. 172:5352–5359.
78. Goldman, R. C., and L. Leive. 1980. Heterogeneity of antigenic side-chain length in lipopolysaccharide from Escherichia coli O111 and Salmonella typhimurium LT2. Eur. J. Biochem. 107:145–153.
79. Golenbock, D. T., R. Y. Hampton, N. Qureshi, K. Takayama, and C. R. H. Raetz. 1991. Lipid A-like molecules that antagonize the effects of endotoxins on human monocytes. J. Biol. Chem. 266:19490–19498.
80. Golenbock, D. T., Y. Liu, F. H. Milham, M. W. Freeman, and R. A. Zoeller. 1993. Surface expression of human CD14 in Chinese hamster ovary fibroblasts imparts macrophage-like responsiveness to bacterial endotoxins. J. Biol. Chem. 268:22055–22059.
80a. Golenbock, D. T., J. A. Will, C. R. H. Raetz, and R. A. Proctor. 1987. Lipid X ameliorates pulmonary hypertension and protects sheep from death due to endotoxin. Infect. Immun. 55:2471–2476.
81. Gray, P. W., G. Flaggs, S. R. Leong, R. J. Gumina, J. Weiss, C. E. Ooi, and P. Elsbach. 1989. Cloning of the cDNA of a human neutrophil bactericidal protein. Structural and functional correlations. J. Biol. Chem. 264:9505–9509.
82. Green, P. R., A. H. Merrill, Jr., and R. M. Bell. 1981. Membrane phospholipid synthesis in Escherichia coli. Purification, reconstitution, and characterization of sn-glycerol-3-phosphate acyltransferase. J. Biol. Chem. 256:11151–11159.
83. Grube, B. J., C. G. Cochane, R. D. Ye, C. E. Green, M. E. McPhail, R. J. Ulevitch, and P. S. Tobias. 1994. Lipopolysaccharide binding protein expression in primary human hepatocytes and HepG2 hepatoma cells. J. Biol. Chem. 269:8477–8482.
84. Haliman, E., H. S. Lichenstein, M. M. Wurfel, D. S. Miller, D. A. Johnson, M. Kelley, L. A. Busse, M. M. Zukowski, and S. D. Wright. 1994. Lipopolysaccharide (LPS)-binding protein accelerates the binding of LPS to CD14. J. Exp. Med. 179:269–277.
85. Hammond, S. M., A. Claesson, A. M. Jansson, L. G. Larsson, B. G. Pring, C. M. Town, and B. Ekström. 1987. A new class of synthetic antibacterials acting on lipopolysaccharide biosynthesis. Nature (London) 327:730–732.
86. Hampton, R. Y., D. T. Golenbock, M. Penman, M. Krieger, and C. R. H. Raetz. 1991. Recognition and plasma clearance of endotoxin by scavenger receptors. Nature (London) 352:342–344.
87. Hampton, R. Y., D. T. Golenbock, and C. R. H. Raetz. 1988. Lipid A binding sites in membranes of macrophage tumor cells. J. Biol. Chem. 263:14802–14807.
88. Hampton, R. Y., and C. R. H. Raetz. 1992. Lipid A 4'-kinase of Escherichia coli: enzyme assay and preparation of 4'-32P-labeled probes of high specific radioactivity. Methods Enzymol. 209:466–475.
89. Hampton, R. Y., and C. R. H. Raetz. 1991. Macrophage catabolism of lipid A is regulated by endotoxin stimulation. J. Biol. Chem. 266:19499–19509.
90. Han, J., J. D. Lee, L. Bibbs, and R. J. Ulevitch. 1994. A MAP kinase targeted by endotoxin and hyperosmolarity in mammalian cells. Science 265:808–811.
91. Han, J., J. C. Mathison, R. J. Ulevitch, and P. S. Tobias. 1994. Lipopolysaccharide (LPS) binding protein, truncated at Ile-97, binds LPS but does not transfer LPS to CD 14. J. Biol. Chem. 269:8172–8175.
92. Han, J. H., J.-D. Lee, P. S. Tobias, and R. J. Ulevitch. 1993. Endotoxin induces rapid protein tyrosine phosphorylation in 70Z/3 cells expressing CD14. J. Biol. Chem. 268:25009–25014.
93. Hara-Kuge, S., F. Amano, M. Nishijima, and Y. Akamatsu. 1990. Isolation of a lipopolysaccharide (LPS)-resistant mutant, with defective LPS binding, of cultured macrophage-like cells. J. Biol. Chem. 265:6606–6610.
94. Hayter, J. B., M. Rivera, and E. J. McGroarty. 1987. Neutron scattering analysis of bacterial lipopolysaccharide phase structure: changes at high pH. J. Biol. Chem. 262:5100–5105.
95. Helander, I. M., L. Hirvas, J. Tuominen, and M. Vaara. 1992. Preferential synthesis of heptaacyl lipopolysaccharide by the ssc permeability mutant of Salmonella typhimurium. Eur. J. Biochem. 212:363–369.
96. Helander, I. M., I. Kilpeläinen, and M. Vaara. 1994. Increased substitution of phosphate groups in lipopolysaccharides and lipid A of the polymyxin-resistant pmrA mutants of Salmonella typhimurium: a 31P-NMR study. Mol. Microbiol. 11:481–487.
97. Helander, I. M., R. B. Lindner, H. Brade, K. Altmann, A. A. Lindberg, E. T. Rietschel, and U. Zähringer. 1988. Chemical structure of the lipopolysaccharide of Haemophilus influenzae strain I-69 Rd-/b+: description of a novel deep-rough mutant. Eur. J. Biochem. 177:483–492.
98. Helander, I. M., M. Vaara, S. Sukupolvi, M. Rhen, S. Saarela, U. Zähringer, and H. Mäkelä. 1989. RfaP mutants of Salmonella typhimurium. Eur. J. Biochem. 185:541–546.
99. Henkel, T., T. Machleidt, I. Alkalay, M. Krönke, Y. Ben-Neriah, and P. A. Baeuerle. 1993. Rapid proteolysis of IκB-α is necessary for activation of transcription factor NF-κB. Nature (London) 365:182–185.
100. Higgins, C. F., and M. M. Gottesman. 1992. Is the multidrug transporter a flippase? Trends Biochem. Sci. 17:18–19.
101. Hoess, A., S. Watson, G. R. Siber, and R. Liddington. 1993. Crystal structure of an endotoxin-neutralizing protein from the horseshoe crab, Limulus anti-LPS factor, at 1.5 Å resolution. EMBO J. 12:3351–3356.
102. Hofer, M., R. Y. Hampton, C. R. H. Raetz, and H. Yu. 1991. Aggregation behavior of lipid IVA in aqueous solutions at physiological pH. 1) Simple buffer solutions. Chem. Phys. Lipids 59:167–181.
103. Holst, O., and H. Brade. 1990. Isolation and identification of 3-deoxy-5-O-α-l-rhamnopyranosyl-d-manno-2-octulopyranosonate from the inner core region of the lipopolysaccharide of Escherichia coli K-12. Carbohydr. Res. 207:327–331.
104. Holst, O., and H. Brade. 1992. Chemical structure of the core region of lipopolysaccharides, p. 135–170. In D. C. Morrison and J. L. Ryan (ed.), Bacterial Endotoxic Lipopolysaccharides, vol. I. Molecular Biochemistry and Cellular Biology. CRC Press, Boca Raton, Fla.
105. Holst, O., E. Röhrscheidt-Andrejewski, H. Brade, and D. Charon. 1990. Isolation and characterization of 3-deoxy-d-manno-2-octulopyranosonate 7-(2-aminoethyl phosphate) from the inner core region of Escherichia coli K-12 and Salmonella minnesota lipopolysaccharides. Carbohydr. Res. 204:93–102.
106. Icho, T., C. E. Bulawa, and C. R. H. Raetz. 1985. Molecular cloning and sequencing of the gene for CDP-diglyceride hydrolase in Escherichia coli. J. Biol. Chem. 260:12092–12098.
107. Imoto, M., S. Kusumoto, T. Shiba, H. Naoki, T. Iwashita, E. T. Rietschel, H.-W. Wollenweber, C. Galanos, and O. Lüderitz. 1983. Chemical structure of E. coli lipid A: linkage site of acyl groups in the disaccharide backbone. Tetrahedron Lett. 24:4017–4020.
108. Jackson, J. J., and H. Kropp. 1992. β-Lactam antibiotic-induced release of free endotoxin: in vitro comparison of penicillin-binding protein (PBP) 2-specific imipenem and PBP 3-specific ceftazidime. J. Infect. Dis. 165:1033–1041.
109. Jann, K., G. Goldemann, C. Weisgerber, C. Wolff-Ullisch, and S. Kanegasaki. 1982. Biosynthesis of the O9 antigen of Escherichia coli. Initial reaction and overall mechanism. Eur. J. Biochem. 127:157–164.
110. Jann, K., and B. Jann. 1984. Structure and biosynthesis of O-antigens, p. 138–166. In E. T. Rietschel (ed.), Handbook of Endotoxin, vol. I. Chemistry of Endotoxin. Elsevier/North-Holland Biomedical Press, Amsterdam.
111. Jiang, X. M., B. Neal, F. Santiago, S. J. Lee, L. K. Romana, and P. R. Reeves. 1991. Structure and sequence of the rfb (O antigen) gene cluster of Salmonella serovar typhimurium (strain LT2). Mol. Microbiol. 5:695–713.
112. Juan, T. S.-C., E. Hailman, M. J. Kelley, L. A. Busse, E. Davy, C. J. Empig, L. O. Narhi, S. D. Wright, and H. S. Lichenstein. 1995. Identification of a lipopolysaccharide binding domain in CD14 between amino acids 57 and 64. J. Biol. Chem. 270:5219–5224.
113. Juan, T. S.-C., M. J. Kelley, D. A. Johnson, L. A. Busse, E. Hailman, S. D. Wright, and H. S. Lichenstein. 1995. Soluble CD14 truncated at amino acid 152 binds lipopolysaccharide (LPS) and enables cellular response to LPS. J. Biol. Chem. 270:1382–1387.
114. Kahan, F. M., J. S. Kahan, P. J. Cassidy, and H. Kropp. 1974. Mechanism of action of fosfomycin. Ann. N. Y. Acad. Sci. 235:364–385.
115. Kanegasaki, S., and A. Wright. 1970. Mechanism of polymerization of the Salmonella O antigen: utilization of lipid-linked intermediates. Proc. Natl. Acad. Sci. USA 67:951–958.
116. Karibian, D., C. Deprun, and M. Caroff. 1993. Comparison of lipids A of several Salmonella and Escherichia strains by 252Cf plasma desorption mass spectrometry. J. Bacteriol. 175:2988–2993.
117. Karibian, D., C. Deprun, L. Szabó, Y. Le Beyee, and M. Caroff. 1991. 252Cf plasma desorption mass spectrometry applied to the analysis of endotoxin lipid A preparations. Int. J. Mass Spectrom. Ion Processes 111:273–286.
118. Karow, M., O. Fayet, A. Cegielska, T. Ziegelhoffer, and C. Georgopoulos. 1991. Isolation and characterization of the Escherichia coli htrB gene, whose product is essential for bacterial viability above 33oC in rich media. J. Bacteriol. 173:741–750.
119. Karow, M., O. Fayet, and C. Georgopoulos. 1992. The lethal phenotype caused by null mutations in the Escherichia coli htrB gene is suppressed by mutations in the accBC operon, encoding two subunits of acetyl coenzyme A carboxylase. J. Bacteriol. 174:7407–7418.
120. Karow, M., and C. Georgopoulos. 1992. Isolation and characterization of the Escherichia coli msbB gene, a multicopy suppressor of null mutations in the high-temperature requirement gene htrB. J. Bacteriol. 174:702–710.
121. Karow, M., and C. Georgopoulos. 1991. Sequencing, mutational analysis, and transcriptional regulation of the Escherichia coli htrB gene. Mol. Microbiol. 5:2285–2292.
122. Karow, M., and C. Georgopoulos. 1993. The essential Escherichia coli msbA gene, a multicopy suppressor of null mutations in the htrB gene, is related to the universally conserved family of ATP-dependent translocators. Mol. Microbiol. 7:69–79.
123. Kastowsky, M., T. Gutberlet, and H. Bradaczek. 1992. Molecular modelling of the three-dimensional structure and conformational flexibility of bacterial lipopolysaccharide. J. Bacteriol. 174:4798–4806.
124. Kelly, T. M., S. A. Stachula, C. R. H. Raetz, and M. S. Anderson. 1993. The firA gene of Escherichia coli encodes UDP-3-O-(R-3-hydroxymyristoyl)-α-d-glucosamine N-acyltransferase: the third step of endotoxin biosynthesis. J. Biol. Chem. 268:19866–19874.
125. Kent, C. 1995. Eucaryotic phospholipid synthesis. Annu. Rev. Biochem. 64:315–343.
126. Kitchens, R. L., and R. S. Munford. 1995. Enzymatically deacylated lipopolysaccharide (dLPS) can antagonize LPS at multiple sites in the LPS recognition pathway. J. Biol. Chem. 270:9904–9910.
127. Kitchens, R. L., R. J. Ulevitch, and R. S. Munford. 1992. Lipopolysaccharide (LPS) partial structures inhibit responses to LPS in a human macrophage cell line without inhibiting LPS uptake by a CD14-mediated pathway. J. Exp. Med. 176:485–494.
128. Klena, J. D., R. S. Ashford, and C. A. Schnaitman. 1992. Role of Escherichia coli K-12 rfa genes and the rfp gene of Shigella dysenteriae 1 in the generation of lipopolysaccharide core heterogeneity and attachment of O antigen. J. Bacteriol. 174:7297–7303.
129. Klena, J. D., E. Pradel, and C. A. Schnaitman. 1992. Comparison of the lipopolysaccharide biosynthesis genes rfaK, rfaL, rfaY, and rfaZ of Escherichia coli K-12 and Salmonella typhimurium. J. Bacteriol. 174:4746–4752.
130. Klena, J. D., E. Pradel, and C. A. Schnaitman. 1993. The rfaS gene which is involved in production of a rough form of lipopolysaccharide core in Escherichia coli K-12 is absent in Salmonella typhimurium. J. Bacteriol. 175:1524–1527.
131. Klena, J. D., and C. A. Schnaitman. 1994. Genes for TDP-rhamnose synthesis affect the pattern of lipopolysaccharide heterogeneity in Escherichia coli K-12. J. Bacteriol. 176:4003–4010.
132. Kohara, Y., K. Akiyama, and K. Isono. 1987. The physical map of the whole Escherichia coli chromosome: application of a new strategy for rapid analysis and sorting of a large genomic library. Cell 50:495–508.
133. Kohlbrenner, W. E., and S. W. Fesik. 1985. Determination of the anomeric specificity of the Escherichia coli CTP:CMP-3-deoxy-d-manno-octulosonate cytidylyltransferase by 13C NMR spectroscopy. J. Biol. Chem. 260:14695–14700.
134. Kontrohr, T., and B. Kocsis. 1981. Isolation of adenosine 5'-diphosphate-d-glycero-d-mannoheptose. An intermediate in lipopolysaccharide biosynthesis of Shigella sonnei. J. Biol. Chem. 256:7715–7718.
135. Koplow, J., and H. Goldfine. 1974. Alterations in the outer membrane of the cell envelope of heptose-deficient mutants of Escherichia coli. J. Bacteriol. 117:527–543.
136. Kovach, N. L., E. Yee, R. S. Munford, C. R. H. Raetz, and J. M. Harlan. 1990. Lipid IVA inhibits synthesis and release of tumor necrosis factor induced by lipopolysaccharide in whole human blood ex vivo. J. Exp. Med. 172:78–84.
137. Laird, M. W., A. W. Kloser, and R. Misra. 1994. Assembly of LamB and OmpF in deep rough lipopolysaccharide mutants of Escherichia coli. J. Bacteriol. 176:2259–2264.
138. Lam, C., J. Hildebrandt, E. Schütze, B. Rosenwirth, R. A. Proctor, E. Liehl, and P. Stütz. 1991. Immunostimulatory, but not antiendotoxin, activity of lipid X is due to small amounts of contaminating N,O-acylated disaccharide-1-phosphate: in vitro and in vivo reevaluation of the biological activity of synthetic lipid X. Infect. Immun. 59:2351–2358.
139. Lee, J. C., J. T. Laydon, P. C. McDonnell, T. F. Gallagher, R. S. Kuma, D. Green, D. McNulty, M. J. Blumenthal, J. R. Heys, S. W. Landvatter, J. E. Strickler, M. M. McLaughlin, I. R. Siemens, S. M. Fisher, G. P. Livi, J. R. White, J. L. Adams, and P. R. Young. 1994. A protein kinase involved in the regulation of inflammatory cytokine biosynthesis. Nature (London) 372:739–746.
140. Lee, J.-D., K. Kato, P. S. Tobias, T. N. Kirkland, and R. J. Ulevitch. 1992. Transfection of CD14 into 70Z/3 cells dramatically enhances the sensitivity to complexes of lipopolysaccharide (LPS) and LPS binding protein. J. Exp. Med. 175:1697–1705.
141. Lee, J.-D., V. Kravchenko, T. N. Kirkland, J. Han, N. Mackman, A. Moriarty, D. Leturcq, P. S. Tobias, and R. J. Ulevitch. 1993. Glycosyl-phosphatidylinositol-anchored or integral membrane forms of CD14 mediate identical cellular responses to endotoxin. Proc. Natl. Acad. Sci. USA 90:9930–9934.
142. LeGrand, C. B., and R. Thieringer. 1994. CD14-dependent induction of protein tyrosine phosphorylation by lipopolysaccharide in murine B-lymphoma cells. Biochim. Biophys. Acta 1223:36–46.
143. Leive, L. 1974. The barrier function of the gram-negative envelope. Ann. N. Y. Acad. Sci. 235:109–127.
144. Lesse, A. J., A. A. Campagnari, W. E. Bittner, and M. A. Apicella. 1990. Increased resolution of lipopolysaccharides and lipooligosaccharides utilizing tricine-sodium dodecyl sulfate polyacrylamide gel electrophoresis. J. Immunol. Methods 126:109–117.
145. Levin, J., C. R. Alving, R. S. Munford, and P. L. Stütz (ed.). 1993. Endotoxin Research Series, vol 2. Bacterial Endotoxin: Recognition and Effector Mechanisms. Excerpta Medica, Amsterdam.
146. Lin, Y.-C., K. Brown, and U. Siebenlist. 1995. Activation of NF-κB requires proteolysis of the inhibitor IκB-α: signal-induced phosphorylation of IκB-1α alone does not release NF-κB. Proc. Natl. Acad. Sci. USA 92:552–556.
147. Lindberg, A. A. 1977. Bacterial surface carbohydrates and bacteriophage absorption, p. 289–356. In I. Sutherland (ed.), Surface Carbohydrates of the Procaryotic Cell. Academic Press, New York.
148. Lipka, G., R. A. Demel, and H. Hauser. 1988. Phase behaviour of lipid X. Chem. Phys. Lipids 48:267–280.
149. Liu, D., A. M. Haase, L. Lindqvist, A. A. Lindberg, and P. R. Reeves. 1993. The glycosyl transferases of O antigen biosynthesis in S. enterica: identification and characterization of transferase genes of groups B, C2, and E1. J. Bacteriol. 175:3408–3413.
150. Liu, H. W., and J. S. Thorson. 1994. Pathways and mechanisms in the biogenesis of novel deoxysugars by bacteria. Annu. Rev. Microbiol. 48:233–256.
151. Lo, S. F., V. P. Miller, Y. Lei, J. S. Thorson, H.-W. Liu, and J. L. Schottel. 1994. CDP-6-deoxy-Δ3,4-glucoseen reductase from Yersinia pseudotuberculosis: enzyme purification and characterization of the cloned gene. J. Bacteriol. 176:460–468.
152. Löms Ziegler-Heitbrock, H. W., A. Wedel, W. Schraut, M. Ströbel, P. Wendelgass, T. Sternsdorf, P. A. Bäuerle, J. G. Haas, and G. Riethmüller. 1994. Tolerance to lipopolysaccharide involves mobilization of nuclear factor κB with predominance of p50 homodimers. J. Biol. Chem. 269:17001–17004.
153. Loppnow, H., H. Brade, I. Dürrbaum, C. A. Dinarello, S. Kusumoto, E. T. Rietschel, and H. D. Flad. 1989. IL-1 induction capacity of defined lipopolysaccharide partial structures. J. Immunol. 142:3229–3238.
154. Lowenstein, C. J., E. W. Alley, P. Raval, A. M. Snowman, S. H. Snyder, S. W. Russell, and W. J. Murphy. 1993. Macrophage nitric oxide synthase gene: two upstream regions mediate induction by interferon γ and lipopolysaccharide. Proc. Natl. Acad. Sci. USA 90:9730–9734.
155. MacLachlan, P. R., S. K. Kadam, and K. E. Sanderson. 1991. Cloning, characterization, and DNA sequence of the rfaLK region for lipopolysaccharide synthesis in Salmonella typhimurium LT-2. J. Bacteriol. 173:7151–7163.
156. Magnuson, K., S. Jackowski, C. O. Rock, and J. E. Cronan, Jr. 1993. Regulation of fatty acid biosynthesis in Escherichia coli. Microbiol. Rev. 57:522–542.
157. Mains, P. E., and C. H. Sibley. 1983. LPS-nonresponsive variants of mouse B cell lymphoma, 70Z/3: isolation and characterization. Somatic Cell Genet. 9:699–720.
158. Mäkelä, P. H., and B. A. D. Stocker. 1984. Genetics of lipopolysaccharide, p. 59–137. In E. T. Rietschel (ed.), Handbook of Endotoxin, vol. I. Chemistry of Endotoxin. Elsevier/North-Holland Biomedical Press, Amsterdam.
159. Marino, P. A., K. A. Phan, and M. J. Osborn. 1985. Energy dependence of lipopolysaccharide translocation in Salmonella typhimurium. J. Biol. Chem. 260:14965–14970.
160. Marino, P. M., B. C. McGrath, and M. J. Osborn. 1991. Energy dependence of O-antigen synthesis in Salmonella typhimurium. J. Bacteriol. 173:3128–3133.
161. Marquardt, J. L., E. D. Brown, W. S. Lane, T. M. Haley, Y. Ichikawa, C. H. Wong, and C. T. Walsh. 1994. Kinetics, stoichiometry, and identification of the reactive thiolate in the inactivation of UDP-GlcNAc enolpyruvoyl transferase by the antibiotic fosfomycin. Biochemistry 33:10646–10651.
162. Martin, T. R., J. C. Mathison, P. S. Tobias, D. J. Leturcq, A. M. Moriarty, R. J. Maunder, and R. J. Ulevitch. 1992. Lipopolysaccharide binding protein enhances the responsiveness of alveolar macrophages to bacterial lipopolysaccharide. Implications for cytokine production in normal and injured lungs. J. Clin. Invest. 90:2209–2219.
163. Marumo, K., L. Lindqvist, N. Verma, A. Weintraub, P. R. Reeves, and A. A. Lindberg. 1992. Enzymatic synthesis and isolation of thymidine diphosphate-6-deoxy-d-xylo-4-hexulose and thymidine diphosphate-l-rhamnose. Production using cloned gene products and separation by HPLC. Eur. J. Biochem. 204:539–545.
164. Marvin, H. J. P., M. B. A. ter Best, and B. Witholt. 1989. Release of outer membrane fragments from wild-type Escherichia coli and several Escherichia coli LPS mutants by EDTA and heat shock treatments. J. Bacteriol. 171:5262–5267.
165. Mathison, J., E. Wolfson, S. Steinemann, P. Tobias, and R. J. Ulevitch. 1993. Lipopolysaccharide (LPS) recognition in macrophages. Participation of LPS-binding protein and CD14 in LPS-induced adaptation in rabbit peritoneal exudate macrophages. J. Clin. Invest. 92:2053–2059.
166. McCall, C. E., L. M. Grosso-Wilmoth, K. LaRue, R. N. Guzman, and S. L. Cousart. 1993. Tolerance to endotoxin-induced expression of the interleukin-1β gene in blood neutrophils of humans with the sepsis syndrome. J. Clin. Invest. 91:853–861.
167. McGrath, B. C., and M. J. Osborn. 1991. Evidence for energy-dependent transport of core lipopolysaccharide across the inner membrane of Salmonella typhimurium. J. Bacteriol. 173:649–654.
168. McGrath, B. C., and M. J. Osborn. 1991. Localization of the terminal steps of O-antigen synthesis in Salmonella typhimurium. J. Bacteriol. 173:3134–3137.
169. Meier-Dieter, U. K. Barr, R. Starman, L. Hatch, and P. D. Rick. 1992. Nucleotide sequence of the Escherichia coli rfe gene involved in the synthesis of enterobacterial common antigen. Molecular cloning of the rfe-rff gene cluster. J. Biol. Chem. 267:746–753.
170. Melaugh, W., N. J. Phillips, A. A. Campagnari, M. V. Tullius, and B. W. Gibson. 1994. Structure of the major oligosaccharide from the lipooligosaccharide of Haemophilus ducreyi strain 35000 and evidence for additional glycoforms. Biochemistry 33:13070–13078.
171. Mengin-Lecreulx, D., and J. van Heijenoort. 1994. Copurification of glucosamine-1-phosphate acetyltransferase and N-acetylglucosamine-1-phosphate uridyltransferase activities of Escherichia coli: characterization of the glmU gene product as a bifunctional enzyme catalyzing two subsequent steps in the pathway for UDP-N-acetylglucosamine synthesis. J. Bacteriol. 176:5788–5795.
172. Mengin-Lecreulx, D., B. Flouret, and J. van Heijenoort. 1983. Pool levels of UDP-N-acetylglucosamine and UDP-N-acetylglucosamine-enolpyruvate in Escherichia coli and correlation with peptidoglycan synthesis. J. Bacteriol. 154:1284–1290.
173. Milla, M., and C. R. H. Raetz. 1995. Interaction of Escherichia coli lipid A disaccharide synthase (LpxB) with the aerobic glycerol-3-phosphate dehydrogenase (GlpD). FASEB J. 9:A1310.
174. Miller, V. P., and H.-W. Liu. 1992. Biosynthesis of 3, 6-dideoxyhexoses: new evidence supporting a radical mechanism for C-3 deoxygenation. J. Am. Chem. Soc. 114:1880–1881.
175. Miller, V. P., J. S. Thorson, O. Ploux, S. F. Lo, and H.-W. Liu. 1993. Cofactor characterization and mechanistic studies of CDP-6-deoxy-Δ3,4-glucoseen reductase: exploration into a novel enzymatic C-O bond cleavage event. Biochemistry 32:11934–11942.
176. Mohan, S., T. M. Kelly, S. S. Eveland, C. R. H. Raetz, and M. S. Anderson. 1994. An Escherichia coli gene (fabZ) encoding R-3-hydroxymyristoyl acyl carrier protein dehydrase. Relation to fabA and suppression of mutations in lipid A biosynthesis. J. Biol. Chem. 269:32896–32903.
177. Mohan, S., and C. R. H. Raetz. 1994. Endotoxin biosynthesis in Pseudomonas aeruginosa: enzymatic incorporation of laurate before 3-deoxy-d-manno-2-octulosonate. J. Bacteriol. 176:6944–6951.
178. Morona, R., M. Mavris, A. Fallarino, and P. A. Manning. 1994. Characterization of the rfc region of Shigella flexneri. J. Bacteriol. 176:733–747.
179. Morona, R., L. van den Bosch, and P. A. Manning. 1995. Molecular, genetic, and topological characterization of O-antigen chain length regulation in Shigella flexneri. J. Bacteriol. 177:1059–1068.
180. Morrison, D. C., C. A. Dinarello, R. S. Munford, C. Natanson, R. Danner, M. Pollack, J. J. Spitzer, R. J. Ulevitch, S. N. Vogel, and E. McSweegan. 1994. Current status of bacterial endotoxins. ASM News 60:479–484.
181. Morrison, D. C., and J. L. Ryan. 1987. Endotoxins and disease mechanisms. Annu. Rev. Med. 38:417–432.
182. Morrison, D. C., and J. L. Ryan (ed.). 1992. Bacterial Endotoxic Lipopolysaccharides, vol. I. Molecular Biochemistry and Cellular Biology. CRC Press, Boca Raton, Fla.
183. Mühlradt, P. 1969. Biosynthesis of Salmonella lipopolysaccharide. The in vitro transfer of phosphate to the heptose moiety of the core. Eur. J. Biochem. 11:241–248.
184. Mühlradt, P. 1970. Biosynthesis of Salmonella lipopolysaccharide. Studies of the transfer of glucose, galactose, and phosphate to the core in a cell-free system. Eur. J. Biochem. 18:20–27.
185. Mühlradt, P. F., J. Menzel, J. R. Golecki, and V. Speth. 1973. Outer membrane of Salmonella: site of export of newly synthesized lipopolysaccharide on the surface. Eur. J. Biochem. 35:471–481.
186. Mulford, C. A., and M. J. Osborn. 1983. An intermediate step in translocation of lipopolysaccharide to the outer membrane of Salmonella typhimurium. Proc. Natl. Acad. Sci. USA 80:1159–1163.
187. Müller, E., A. Hinckley, and L. I. Rothfield. 1972. Studies of phospholipid-requiring bacterial enzymes. III. Purification and properties of uridine diphosphate glucose: lipopolysaccharide glucosyltransferase I. J. Biol. Chem. 247:2614–2622.
188. Munford, R. S., and C. L. Hall. 1986. Detoxification of bacterial lipopolysaccharides (endotoxins) by a human neutrophil enzyme. Science 234:203–205.
189. Naide, Y., H. Nikaido, M. P. H., R. G. Wilkinson, and B. A. D. Stocker. 1965. Semirough strains of Salmonella. Proc. Natl. Acad. Sci. USA 53:147–153.
190. Nakano, M., and H. Shinomiya. 1992. The Lps mutational defect in C3H/HeJ mice, p. 311–328. In D. C. Morrison and J. L. Ryan (ed.), Bacterial Endotoxic Lipopolysaccharides, vol. I. Molecular Biochemistry and Cellular Biology. CRC Press, Boca Raton, Fla.
191. Nano, F. E., and H. D. Caldwell. 1985. Expression of the genus-specific lipopolysaccharide epitope in Escherichia coli. Science 228:742–744.
192. Natanson, C., W. D. Hoffman, A. F. Suffredini, P. Q. Eichacker, and R. L. Danner. 1994. Selected treatment strategies for septic shock based on proposed mechanisms of pathogenesis. Ann. Intern. Med. 120:771–783.
193. Nikaido, H., and M. Vaara. 1987. Outer membrane, p. 7–22. In F. C. Neidhardt, J. L. Ingraham, K. B. Low, B. Magasanik, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology, vol. 1. American Society for Microbiology, Washington, D.C.
194. Nishijima, M., C. E. Bulawa, and C. R. H. Raetz. 1981. Two interacting mutations causing temperature-sensitive phosphatidylglycerol synthesis in Escherichia coli membranes. J. Bacteriol. 145:113–121.
195. Nishijima, M., and C. R. H. Raetz. 1979. Membrane lipid biogenesis in Escherichia coli: identification of genetic loci for phosphatidylglycerophosphate synthetase and construction of mutants lacking phosphatidylglycerol. J. Biol. Chem. 254:7837–7844.
196. Nishijima, M., and C. R. H. Raetz. 1981. Characterization of two membrane-associated glycolipids from an Escherichia coli mutant deficient in phosphatidylglycerol. J. Biol. Chem. 256:10690–10696.
197. Normark, S. 1970. Genetics of a chain forming mutant of Escherichia coli: transduction and dominance of the envA gene mediating increased penetration to some antibacterial agents. Genet. Res. (Cambridge) 16:63–78.
198. Normark, S., H. G. Boman, and E. Matsson. 1969. Mutant of Escherichia coli with anomalous cell division and ability to decrease episomally and chromosomally mediated resistance to ampicillin and several other antibiotics. J. Bacteriol. 97:1334–1342.
199. Novogrodsky, A., A. Vanichkin, M. Patya, A. Gazit, N. Osherov, and A. Levitzki. 1994. Prevention of lipopolysaccharide-induced lethal toxicity by tyrosine kinase inhibitors. Science 264:1319–1322.
200. Ogawa, T. 1993. Chemical structure of lipid A from Porphyromonas (Bacteroides) gingivalis lipopolysaccharide. FEBS Lett. 332:197–201.
201. Osborn, M. J. 1979. Biosynthesis and assembly of the lipopolysaccharide of the outer membrane, p. 15–34. In M. Inouye (ed.), Bacterial Outer Membranes. John Wiley & Sons, New York.
202. Osborn, M. J., and L. D’Ari. 1964. Enzymatic incorporation of N-acetylglucosamine into cell wall lipopolysaccharide in a mutant strain of Salmonella typhimurium. Biochem. Biophys. Res. Commun. 16:568–575.
203. Osborn, M. J., P. D. Rick, and N. S. Rasmussen. 1980. Mechanism of assembly of the outer membrane of Salmonella typhimurium. Translocation and integration of an incomplete mutant lipid A into the outer membrane. J. Biol. Chem. 255:4246–4251.
204. Osborn, M. J., S. M. Rosen, L. I. Rothfield, and B. L. Horecker. 1962. Biosynthesis of bacterial lipopolysaccharide. I. Enzymatic incorporation of galactose in a mutant strain of Salmonella. Proc. Natl. Acad. Sci. USA 48:1831–1838.
205. Osborn, M. J., and R. T. Tze-Yen. 1968. Biosynthesis of bacterial lipopolysaccharide. VII. Enzymatic formation of the first intermediate in biosynthesis of the O-antigen of Salmonella typhimurium. J. Biol. Chem. 243:5145–5152.
206. Oude Elferink, R. P. J., R. Ottenhoff, M. van Wijland, J. J. M. Smit, A. H. Schinkel, and A. K. Groen. 1995. Regulation of biliary lipid secretion by the mdr2 P-glycoprotein in the mouse. J. Clin. Invest. 95:31–38.
207. Parillo, J. E. 1993. Pathogenic mechanisms of septic shock. N. Engl. J. Med. 328:1471–1477.
208. Park, J. T. 1987. Murein synthesis, p. 663–671. In F. C. Neidhardt, J. L. Ingraham, K. B. Low, B. Magasanik, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology, vol. 1. American Society for Microbiology, Washington, D.C.
209. Parker, C. T., A. W. Kloser, C. A. Schnaitman, M. A. Stein, S. Gottesman, and B. W. Gibson. 1992. Role of rfaG and rfaP genes in determining the lipopolysaccharide core structure and cell surface properties of Escherichia coli K-12. J. Bacteriol. 174:2525–2538.
210. Parker, C. T., E. Pradel, and C. A. Schnaitman. 1992. Identification and sequence of the lipopolysaccharide core biosynthetic genes rfaQ, rfaP, and rfaG of Escherichia coli K-12. J. Bacteriol. 174:930–934.
211. Pavelka Jr., M. S., S. F. Hayes, and R. P. Silver. 1994. Characterization of KpsT, the ATP-binding component of the ABC-transporter involved with the export of capsular polysialic acid in Escherichia coli K1. J. Biol. Chem. 269:20149–20158.
212. Pegues, J. C., L. Chen, A. W. Gordon, L. Ding, and W. G. Coleman. 1990. Cloning, expression and characterization of the Escherichia coli K-12 rfaD gene. J. Bacteriol. 172:4652–4660.
213. Pfitzner, U., C. R. H. Raetz, and S. L. Roderick. 1995. Crystallization of UDP-N-acetylglucosamine O-acyltransferase from Escherichia coli. Proteins: Structure, Function, Genetics 22:191–192.
214. Pradel, E., C. T. Parker, and C. A. Schnaitman. 1992. The structure of the rfaB, rfaI, rfaJ, and rfaS genes of Escherichia coli K-12 and their roles in the assembly of the lipopolysaccharide core. J. Bacteriol. 174:4736–4745.
215. Pradel, E., and C. A. Schnaitman. 1991. Effect of rfaH (sfrB) and temperature on the expression of rfa genes of Escherichia coli K-12. J. Bacteriol. 173:6428–6431.
215a. Price, N. P. J., B. Jeyaretnam, R. W. Carlson, J. L. Kadrmas, C. R. H. Raetz, and K. A. Brozek. 1995. Lipid A biosynthesis in Rhizobium leguminosarum: role of a 2-keto-3-deoxyoctulosonate-activated 4'-phosphatase. Proc. Natl. Acad. Sci USA 92:7352–7356.
216. Price, N. P. J., T. M. Kelly, C. R. H. Raetz, and R. W. Carlson. 1994. Biosynthesis of a structurally novel lipid A in Rhizobium leguminosarum: identification and characterization of six metabolic steps leading from UDP-GlcNAc to (Kdo)2-lipid IVA. J. Bacteriol. 176:4646–4655.
217. Prins, J. M., S. J. H. van Deventer, E. J. Kuijper, and P. Speelman. 1994. Clinical relevance of antibiotic-induced endotoxin release. Antimicrob. Agents Chemother. 38:1211–1218.
218. Proctor, R. A., J. A. Will, K. E. Burhop, and C. R. H. Raetz. 1986. Protection of mice against lethal endotoxemia by a monosaccharide lipid A precursor. Infect. Immun. 52:905–907.
219. Pugin, J., D. Heumann, A. Tomasz, V. V. Kravchenko, Y. Akamatsu, M. Nishijima, M. P. Glauser, P. S. Tobias, and R. J. Ulevitch. 1994. CD14 is a pattern recognition receptor. Immunity 1:509–516.
220. Pugin, J., C. Schürer-Maly, D. Leturcq, A. Moriarty, R. J. Ulevitch, and P. S. Tobias. 1993. Lipopolysaccharide activation of human endothelial and epithelial cells is mediated by lipopolysaccharide-binding protein and soluble CD14. Proc. Natl. Acad. Sci. USA 90:2744–2748.
221. Qureshi, N., R. J. Cotter, and K. Takayama. 1986. Application of fast atom bombardment mass spectrometry and nuclear magnetic resonance on the structural analysis of purified lipid A. J. Microbiol. Methods 5:65–77.
222. Qureshi, N., K. Takayama, P. Mascagni, J. Honovich, R. Wong, and R. J. Cotter. 1988. Complete structural determination of lipopolysaccharide obtained from deep rough mutant of Escherichia coli. J. Biol. Chem. 263:11971–11976.
223. Qureshi, N., K. Takayama, and E. Ribi. 1982. Purification and structural determination of non-toxic lipid A obtained from the lipopolysaccharide of Salmonella typhimurium. J. Biol. Chem. 257:11808–11815.
224. Radika, K., and C. R. H. Raetz. 1988. Purification and properties of lipid A disaccharide synthase of Escherichia coli. J. Biol. Chem. 263:14859–14867.
225. Raetz, C. R. H. 1986. Molecular genetics of membrane phospholipid synthesis. Annu. Rev. Genet. 20:253–295.
226. Raetz, C. R. H. 1990. Biochemistry of endotoxins. Annu. Rev. Biochem. 59:129–170.
227. Raetz, C. R. H. 1993. Bacterial endotoxins: extraordinary lipids that activate eucaryotic signal transduction. J. Bacteriol. 175:5745–5753.
228. Raetz, C. R. H., and W. Dowhan. 1990. Biosynthesis and function of phospholipids in Escherichia coli. J. Biol. Chem. 265:1235–1238.
229. Raetz, C. R. H., and E. P. Kennedy. 1972. The association of phosphatidylserine synthetase with ribosomes in extracts of Escherichia coli. J. Biol. Chem. 247:2008–2014.
230. Raetz, C. R. H., S. Purcell, M. V. Meyer, N. Qureshi, and K. Takayama. 1985. Isolation and characterization of eight lipid A precursors from a 3-deoxy-d-manno-octulosonic acid-deficient mutant of Salmonella typhimurium. J. Biol. Chem. 260:16080–16088.
231. Raetz, C. R. H., R. J. Ulevitch, S. D. Wright, C. H. Sibley, A. Ding, and C. F. Nathan. 1991. Gram-negative endotoxin: an extraordinary lipid with profound effects on eukaryotic signal transduction. FASEB J. 5:2652–2660.
232. Raina, S., and C. Georgopoulos. 1991. The htrM gene, whose product is essential for Escherichia coli viability only at elevated temperatures, is identical to the rfaD gene. Nucleic Acids Res. 19:3811–3819.
233. Ray, B. L., G. Painter, and C. R. H. Raetz. 1984. The biosynthesis of gram-negative endotoxin: formation of lipid A disaccharides from monosaccharide precursors in extracts of Escherichia coli. J. Biol. Chem. 259:4852–4859.
234. Ray, B. L., and C. R. H. Raetz. 1987. The biosynthesis of gram-negative endotoxin: a novel kinase in Escherichia coli membranes that incorporates the 4' phosphate of lipid A. J. Biol. Chem. 262:1122–1128.
235. Read, M. A., S. R. Cordle, R. A. Veach, C. D. Carlisle, and J. Hawiger. 1993. Cell-free pool of CD14 mediates activation of transcription fractor NF-κB by lipopolysaccharide in human endothelial cells. Proc. Natl. Acad. Sci. USA 90:9887–9891.
236. Reeves, P. 1993. Evolution of Salmonella O antigen variation by interspecific gene transfer on a large scale. Trends Genet. 9:17–22.
237. Reeves, P. 1994. Biosynthesis and assembly of lipopolysaccharide, p. 281–314. In A. Neuberger and L. L. M. van Deenen (ed.), Bacterial Cell Wall: New Comprehensive Biochemistry, vol. 27. Elsevier Science Publishers, New York.
238. Rick, P. D. 1987. Lipopolysaccharide biosynthesis, p. 648–662. In F. Neidhardt, J. L. Ingraham, K. B. Low, B. Magasanik, M. Schaechter, and H. E. Umbarger (ed.), Escherichia coli and Salmonella typhimurium: Cellular and Molecular Biology, vol. 1. American Society for Microbiology, Washington, D.C.
239. Rick, P. D., L. W.-M. Fung, C. Ho, and M. J. Osborn. 1977. Lipid A mutants of Salmonella typhimurium. Purification and characterization of a lipid A precursor produced by a mutant in 3-deoxy-d-mannooctulosonate-8-phosphate synthetase. J. Biol. Chem. 252:4904–4912.
240. Rick, P. D., and M. J. Osborn. 1977. Lipid A mutants of Salmonella typhimurium. Characterization of conditional lethal mutants in 3-deoxy-d-mannooctulosonate-8-phosphate synthetase. J. Biol. Chem. 252:4895–4903.
241. Rietschel, E. T. (ed.). 1984. Handbook of Endotoxin, vol. I. Chemistry of Endotoxin. Elsevier/North-Holland Biomedical Press, Amsterdam.
242. Rietschel, E. T., and H. Brade. 1992. Bacterial endotoxins. Sci. Am. 267:54–61.
243. Rietschel, E. T., T. Kirikae, F. U. Schade, U. Mamat, G. Schmidt, H. Loppnow, A. J. Ulmer, U. Zähringer, U. Seydel, F. Di Padova, M. Schreier, and H. Brade. 1994. Bacterial endotoxin: molecular relationships of structure to activity and function. FASEB J. 8:217–225.
244. Rietschel, E. T., Z. Sidorczyk, U. Zähringer, H.-W. Wollenweber, and O. Lüderitz. 1983. Analysis of the primary structure of lipid A. ACS Symp. Ser. 231:214.
245. Roantree, R. J. 1967. Salmonella O antigen and virulence. Annu. Rev. Microbiol. 21:443–466.
246. Robbins, P. W., D. Bray, M. Dankert, and A. Wright. 1967. Direction of chain growth in polysaccharide synthesis. Science 158:1536–1542.
247. Robbins, P. W., and A. Wright. 1971. Biosynthesis of O-antigens, p. 351–368. In G. Weinbaum, S. Kadis, and S. J. Ajl (ed.), Microbial Toxins, vol. 4. Bacterial Endotoxins. Academic Press, New York.
248. Robbins, P. W., A. Wright, and M. Dankert. 1966. Polysaccharide biosynthesis. J. Gen. Physiol. 49:331–346.
249. Roland, K. L., C. R. Esther, and J. K. Spitznagel. 1994. Isolation and characterization of a gene, pmrD, from Salmonella typhimurium that confers resistance to polymyxin when expressed in multiple copies. J. Bacteriol. 176:3589–3597.
250. Roland, K. L., L. E. Martin, C. R. Esther, and J. K. Spitznagel. 1993. Spontaneous pmrA mutants of Salmonella typhimurium LT2 define a new two-component regulatory system with a possible role in virulence. J. Bacteriol. 175:4154–4164.
251. Romeo, D., A. Hinckley, and L. I. Rothfield. 1970. Reconstitution of a functional membrane enzyme system in a macromolecular film. II. Formation of a functional ternary film of lipopolysaccharide, phospholipid, and transferase enzyme. J. Mol. Biol. 53:491–501.
252. Roncero, C., and M. Casabadan. 1992. Genetic analysis of the genes involved in the synthesis of lipopolysaccharide core in E. coli K-12: three operons in the rfa locus. J. Bacteriol. 174:3250–3260.
253. Rooney, J. W., D. W. Emery, and C. H. Sibley. 1990. Slow response variant of the B lymphoma 70Z/3 defective in lipopolysaccharide activation of NF-κB. Immunogenetics 31:65–72.
254. Rothfield, L. I., M. J. Osborn, and B. L. Horecker. 1964. Biosynthesis of bacterial lipopolysaccharide. II. Incorporation of glucose and galactose catalyzed by particulate and soluble enzymes in Salmonella. J. Biol. Chem. 239:2788–2795.
255. Ruetz, S., and P. Gros. 1994. Phosphatidylcholine translocase: a physiological role for the mdr2 gene. Cell 77:1071–1081.
256. Scheidler, M. A., and R. M. Bell. 1992. Glycerolphosphate acyltransferase from Escherichia coli. Methods Enzymol. 209:55–63.
257. Schnaitman, C. A., and J. D. Klena. 1993. Genetics of lipopolysaccharide biosynthesis in enteric bacteria. Microbiol. Rev. 57:655–682.
258. Schnaitman, C. A., C. T. Parker, E. Pradel, J. D. Klena, N. Pearson, K. E. Sanderson, and P. R. MacLachlan. 1991. Physical map of the rfa locus of Escherichia coli K-12 and Salmonella typhimurium. J. Bacteriol. 173:7410–7411.
259. Scholz, D., K. Bednarik, G. Ehn, W. Neruda, E. Janzek, H. Loibner, K. Briner, and A. Vasella. 1992. Enzymatic synthesis and comparative biological evaluation of a phosphonate analogue of the lipid A precursor. J. Med. Chem. 35:2070–2074.
260. Scott, R. W., C. G. Wilde, J. C. Lane, J. L. Snable, and M. N. Marra. 1993. Antimicrobial and antiendotoxin activities of bactericidal/permeability-increasing protein in vitro and in vivo, p. 373–378. In J. Levin, C. R. Alving, R. S. Munford, and P. L. Stütz (ed.), Endotoxin Research Series, vol. 2. Bacterial Endotoxin: Recognition and Effector Mechanisms. Excerpta Medica, Amsterdam.
261. Setocuchi, M., N. Nasu, S. Yoshida, Y. Higuchi, S. Akizuki, and S. Yamamoto. 1989. Mouse and human CD14 (myeloid cell-specific leucine-rich glycoprotein) primary structure deduced from cDNA clones. Biochim. Biophys. Acta 1008:213–222.
262. Shibuya, I. 1992. Metabolic regulations and biological functions of phospholipids in Escherichia coli. Prog. Lipid Res. 31:245–299.
263. Shumann, R. R., S. R. Leong, G. W. Flaggs, P. W. Gray, S. D. Wright, J. C. Mathison, P. S. Tobias, and R. J. Ulevitch. 1990. Structure and function of lipopolysaccharide binding protein. Science 249:1429–1431.
264. Sibley, C. H., A. Terry, and C. R. H. Raetz. 1988. Induction of κ-light chain synthesis in 70Z/3 B cell lymphoma cells by chemically defined lipid A precursors. J. Biol. Chem. 263:5098–5103.
265. Simmons, D. L., S. Tan, D. G. Tenen, A. Nicholson-Weller, and B. Seed. 1989. Monocyte antigen CD14 is a phospholipid anchored membrane protein. Blood 73:284–289.
266. Sirisena, D. M., K. A. Brozek, P. R. MacLachlan, K. E. Sanderson, and C. R. H. Raetz. 1992. The rfaC gene of Salmonella typhimurium: cloning, sequencing and enzymatic function in heptose transfer to lipopolysaccharide. J. Biol. Chem. 267:18874–18884.
267. Sirisena, D. M., P. R. MacLachlan, S. L. Liu, A. Hessel, and K. E. Sanderson. 1994. Molecular analysis of the rfaD gene, for heptose synthesis, and the rfaF gene, for heptose transfer, in lipopolysaccharide synthesis in Salmonella typhimurium. J. Bacteriol. 176:2379–2385.
268. Smit, J. J. M., A. H. Schinkel, R. P. J. Oude Elferink, A. K. Groen, E. Wagenaar, L. van Deemter, C. A. A. M. Mol, R. Ottenhoff, N. M. T. van der Lugt, M. A. van Roon, M. A. van der Valk, G. J. A. Offerhaus, A. J. M. Berns, and P. Borst. 1993. Homozygous disruption of the murine mdr2 P-glycoprotein gene leads to a complete absence of phospholipid from bile and to liver disease. Cell 75:451–462.
269. Stevenson, G., B. Neal, D. Liu, M. Hobbs, N. H. Packer, M. Batley, J. W. Redmond, L. Lindquist, and P. Reeves. 1994. Structure of the O antigen of Escherichia coli K-12 and the sequence of its rfb gene cluster. J. Bacteriol. 176:4144–4156.
270. Stone, R. 1994. Search for sepsis drugs goes on despite past failures. Science 264:365–367.
271. Strain, S. M., and I. M. Armitage. 1985. Selective detection of 3-deoxymannooctulosonic acid in lipopolysaccharides by spin-echo nuclear magnetic resonance. J. Biol. Chem. 260:12974–12977.
272. Strain, S. M., I. M. Armitage, L. Anderson, K. Takayama, N. Quershi, and C. R. H. Raetz. 1985. Location of polar substituents and fatty acyl chains on lipid A precursors from a 3-deoxy-d-manno-octulosonic acid-deficient mutant of Salmonella typhimurium: studies by 1H, 13C and 31P nuclear magnetic resonance. J. Biol. Chem. 260:16089–16098.
273. Strain, S. M., S. W. Fesik, and I. M. Armitage. 1983. Structure and metal binding properties of lipopolysaccharides from heptoseless mutants of E. coli studied by 13C and 31P nuclear magnetic resonance. J. Biol. Chem. 258:13466–13477.
274. Sturm, S., and K. N. Timmis. 1986. Cloning of the rfb gene region of Shigella dysenteriae 1 and construction of an rfb-rfp gene cassette for the development of lipopolysaccharide-based live anti-dysentery vaccines. Microb. Pathog. 1:289–297.
275. Sugiyama, T., N. Kido, T. Komatsu, M. Ohta, and N. Kato. 1991. Expression of the cloned Escherichia coli 09 rfb gene in various mutant strains of Salmonella typhimurium. J. Bacteriol. 173:55–58.
276. Taga, T., and T. Kishimoto. 1992. Cytokine receptors and signal transduction. FASEB J. 6:3387–3396.
277. Takayama, K., D. H. Mitchell, Z. Z. Din, P. L. C. Mukerjee, and D. L. Coleman. 1994. Monomeric Re lipopolysaccharide from Escherichia coli is more active than the aggregated form in the Limulus amebocyte lysate assay and in inducing Egr-1 mRNA in murine peritoneal macrophages. J. Biol. Chem. 269:2241–2244.
278. Takayama, K., N. Qureshi, B. Beutler, and T. N. Kirkland. 1989. Diphosphoryl lipid A from Rhodopseudomonas sphaeroides ATCC 17023 blocks induction of cachectin in macrophages by lipopolysaccharide. Infect. Immun. 57:1336–1338.
279. Takayama, K., N. Qureshi, and P. Mascagni. 1983. Complete structure of lipid A from the lipopolysaccharides of the heptoseless mutant of Salmonella typhimurium. J. Biol. Chem. 258:12801–12803.
280. Takayama, K., N. Qureshi, P. Mascagni, L. Anderson, and C. R. H. Raetz. 1983. Glucosamine-derived phospholipids in Escherichia coli: structure and chemical modification of a triacyl GlcN-1-P found in a phosphatidylglycerol-deficient mutant. J. Biol. Chem. 258:14245–14252.
281. Takayama, K., N. Qureshi, P. Mascagni, M. A. Nashed, L. Anderson, and C. R. H. Raetz. 1983. Fatty acyl derivatives of glucosamine 1-phosphate in Escherichia coli and their relation to lipid A: complete structure of a diacyl GlcN-1-P found in a phosphatidylglycerol-deficient mutant. J. Biol. Chem. 258:7379–7385.
282. Thayer, A. 1993. Synergen stock plunges after drug trial setback. Chem. Eng. News 71:7.
283. Thorson, J. S., and H.-W. Liu. 1993. Characterization of the first PMP-dependent iron-sulfur containing enzyme which is essential for the biosynthesis of 3, 6-dideoxyhexoses. J. Am. Chem. Soc. 115:7539–7540.
284. Thorson, J. S., and H.-W. Liu. 1993. Coenzyme B6 as a redox cofactor—a new role for an old coenzyme? J. Am. Chem. Soc. 115:12177–12178.
285. Thorson, J. S., S. F. Lo, O. Ploux, X. He, and H.-W. Liu. 1994. Studies of the biosynthesis of 3,6-dideoxyhexoses: molecular cloning and characterization of the asc (ascarylose) region from Yersinia pseudotuberculosis serogroup VA. J. Bacteriol. 176:5483–5493.
286. Tobias, P. S., K. Soldau, and R. J. Ulevitch. 1986. Isolation of a lipopolysaccharide-binding acute phase reactant from rabbit serum. J. Exp. Med. 164:777–793.
287. Tomasiewicz, H. G. 1990. Ph.D. thesis. University of Colorado, Denver.
288. Tomasiewicz, H. G., and C. S. McHenry. 1987. Sequence analysis of the dnaE gene of Escherichia coli. J. Bacteriol. 169:5735–5744.
289. Ulevitch, R. J. 1993. Recognition of bacterial endotoxins by receptor-dependent mechanisms. Adv. Immunol. 53:267–289.
290. Ulevitch, R. J., and P. S. Tobias. 1994. Recognition of endotoxin by cells leading to transmembrane signaling. Curr. Opin. Immunol. 6:125–130.
291. Vaara, M. 1992. Eight bacterial proteins, including UDP-N-acetylglucosamine acyltransferase (LpxA) and three other transferases of Escherichia coli, consist of a six-residue periodicity theme. FEMS Microbiol. Lett. 97:249–254.
292. Vaara, M. 1993. Antibiotic-supersusceptible mutants of Escherichia coli and Salmonella typhimurium. Antimicrob. Agents Chemother. 37:2255–2260.
293. Viriyakosol, S., and T. N. Kirkland. 1995. A region of human CD14 required for lipopolysaccharide binding. J. Biol. Chem. 270:361–368.
294. Vuorio, R., L. Hirvas, and V. M. 1991. The Ssc protein of enteric bacteria has significant homology to the acyltransferase LpxA of lipid A biosynthesis, and to three acetyltransferases. FEBS Lett. 292:90–94.
295. Vuorio, R., and M. Vaara. 1992. The lipid A biosynthesis mutation lpxA2 of Escherichia coli results in drastic antibiotic supersusceptibility. Antimicrob. Agents Chemother. 37:354–356.
296. Vyplel, M., D. Scholz, I. Macher, K. Schindelmaier, and E. Schütze. 1991. C-glycosidic analogues of lipid A and lipid X. J. Med. Chem. 34:2759–2767.
297. Walenga, R. W., and M. J. Osborn. 1980. Biosynthesis of lipid A. Formation of acyl-deficient lipopolysaccharides in Salmonella typhimurium and Escherichia coli. J. Biol. Chem. 255:4257–4263.
298. Wandersman, C., and S. Létoffé. 1993. The study of vancomycin resistant mutants reveals a role of lipopolysaccharide in the secretion of Escherichia coli α-hemolysin and Erwinia chrysanthemi proteases. Mol. Microbiol. 7:141–150.
299. Wang, L., and P. R. Reeves. 1994. Involvement of the galactosyl-1-phosphate transferase encoded by the Salmonella enterica rfbP gene in O-antigen subunit processing. J. Bacteriol. 176:4348–4356.
300. Warren, H. S., R. L. Danner, and R. S. Munford. 1992. Anti-endotoxin monoclonal antibodies. N. Engl. J. Med. 326:1153–1157.
301. Weigel, T. M., L.-D. Liu, and H.-W. Liu. 1992. Mechanistic studies of the biosynthesis of 3,6-dideoxyhexoses in Yersinia pseudotuberculosis: purification and characterization of CDP-4-keto-6-deoxy-d-glucose-3-dehydrase. Biochemistry 31:2129–2139.
302. Weiner, I. M., T. Higuchi, L. I. Rothfield, M. Saltmarsh-Andrew, M. J. Osborn, and B. L. Horecker. 1965. Biosynthesis of bacterial lipopolysaccharide. V. Lipid-linked intermediates in the biosynthesis of the O-antigen groups of Salmonella typhimurium. Proc. Natl. Acad. Sci. USA 54:228–235.
303. Weinstein, S. L., M. R. Gold, and A. L. DeFranco. 1991. Bacterial lipopolysaccharide stimulates protein tyrosine phosphorylation in macrophages. Proc. Natl. Acad. Sci. USA 88:4148–4152.
304. Weinstein, S. L., J. S. Sanghera, K. Lemke, A. L. DeFranco, and S. L. Pelech. 1992. Bacterial lipopolysaccharide induces tyrosine phosphorylation and activation of mitogen-activated protein kinases in macrophages. J. Biol. Chem. 267:14955–14962.
305. Weintraub, A., U. Zähringer, H.-W. Wollenweber, U. Seydel, and E. T. Rietschel. 1989. Structural characterization of the lipid A component of Bacteroides fragilis strain NCTC 9343 lipopolysaccharide. Eur. J. Biochem. 183:425–432.
306. Weiser, J. N., J. M. Love, and E. R. Moxon. 1989. The molecular mechanism of phase variation of H. influenzae lipopolysaccharide. Cell 59:657–665.
307. Weiss, J., P. Elsbach, C. Shu, J. Castillo, L. Grinna, A. Horwitz, and G. Theofan. 1992. Human bactericidal/permeability-increasing protein and a recombinant NH2-terminal fragment cause killing of serum-resistant gram-negative bacteria in whole blood and inhibit tumor necrosis factor release induced by the bacteria. J. Clin. Invest. 90:1122–1130.
308. Weisz, A., S. Oguchi, L. Cicatiello, and H. Esumi. 1994. Dual mechanism for the control of inducible-type NO synthase gene expression in macrophages during activation by interferon-γ and bacterial lipopolysaccharide. J. Biol. Chem. 269:8324–8333.
309. Westphal, O., O. Lüderitz, and F. Bister. 1952. Über die Extraktion von Bakterien mit Phenol/Wasser. Z. Naturforsch. Teil B 7:148–155.
310. White, K. A., and C. R. H. Raetz. 1995. Characterization of a mono-functional Kdo transferase in extracts of Haemophilus influenzae. FASEB J. 9:A1376.
311. Williamson, J. M., M. S. Anderson, and C. R. H. Raetz. 1991. Acyl-acyl carrier protein specificity of UDP-GlcNAc acyltransferases from gram-negative bacteria: relationship to lipid A structure. J. Bacteriol. 173:3591–3596.
312. Woisetschläger, M., and G. Högenauer. 1987. The kdsA gene coding for 3-deoxy-d-manno-octulosonic acid 8-phosphate synthetase is part of an operon in Escherichia coli. Mol. Gen. Genet. 207:369–373.
313. Wollenweber, H.-W., S. Schlecht, O. Lüderitz, and E. T. Rietschel. 1983. Fatty acids in lipopolysaccharides of Salmonella species grown at low temperatures. Eur. J. Biochem. 130:167–171.
314. Wollin, R., E. S. Creeger, L. I. Rothfield, B. A. D. Stocker, and A. A. Lindberg. 1983. Salmonella typhimurium mutants defective in UDP-d-galactose:lipopolysaccharide α1, 6-d-galactosyltransferase. J. Biol. Chem. 258:3769–3774.
315. Wright, S. D. 1991. Multiple receptors for endotoxin. Curr. Opin. Immunol. 3:83–90.
316. Wright, S. D., R. A. Ramos, P. S. Tobias, R. J. Ulevitch, and J. C. Mathison. 1990. CD14: a receptor for complexes of lipopolysaccharide (LPS) and LPS binding protein. Science 249:1431–1433.
317. Wunder, D. E., W. Aaronson, S. F. Hayes, J. M. Bliss, and R. P. Silver. 1994. Nucleotide sequence and mutational analysis of the gene encoding KpsD, a periplasmic protein involved in transport of polysialic acid in Escherichia coli K1. J. Bacteriol. 176:4025–4033.
318. Young, K., L. L. Silver, D. Bramhill, C. A. Caceres, S. A. Stachula, S. E. Shelly, C. R. H. Raetz, and M. S. Anderson. 1993. The second step of lipid A biosynthesis, UDP-3-O-acyl-GlcNAc deacetylase is encoded by the pleiotropic permeability/cell division gene envA of Escherichia coli. FASEB J. 7:A1268.
Chapter69
