Rough Mutant Research Articles

Mouse monoclonal antibodies reactive with J5 lipopolysaccharide exhibit extensive serological cross-reactivity with a variety of gram-negative bacteria.

We describe two mouse monoclonal antibodies reactive with lipopolysaccharide derived from the J5 mutant of Escherichia coli O111:B4. These antibodies react with purified lipopolysaccharide derived from rough mutants of E. coli and Salmonella typhimurium and also with lipopolysaccharide associated with both smooth- and rough-phenotype, gram-negative bacteria. Both antibodies appear to bind determinants present in the lipopolysaccharide core region, and this reactivity is inhibited in the presence of polymyxin B. Although their patterns of reactivity differ, both antibodies exhibit extensive serological cross-reactivity with a variety of gram-negative bacteria. Reagents of this type should prove useful in animal models to delineate the requisite affinity, epitope specificity, immunoglobulin class, etc., needed for the prevention and treatment of gram-negative bacteremia.

Brucellacidal activity of human and bovine polymorphonuclear leukocyte granule extracts against smooth and rough strains of Brucella abortus.

The microbicidal activities of freeze-thaw and high-salt extracts of human and bovine polymorphonuclear leukocyte (PMN) granules were tested against a smooth intermediate strain (45/0) and a rough strain (45/20) of Brucella abortus which differ in virulence and survival within PMNs. Freeze-thaw extracts of human PMN granules were more brucellacidal than high-salt extracts when supplemented with hydrogen peroxide (H2O2) and potassium iodide (KI), whereas the opposite was found with freeze-thaw and high-salt extracts of bovine PMN granules. There was no oxygen-independent killing of either the smooth or rough strain of B. abortus by amounts of granule extracts which caused 100% killing of a deep rough mutant (Re) of Salmonella typhimurium. The oxygen-dependent brucellacidal activity of granule extracts was dependent on concentrations of myeloperoxidase (MPO) units, H2O2, and KI. Maximal brucellacidal activity was observed at pH 5.5 to 6.0. The smooth strain, 45/0, was more resistant to oxygen-dependent killing by granule extracts than was the rough strain, 45/20. Granule extracts were more brucellacidal than purified MPO at equivalent levels of MPO enzyme units, suggesting that at least one other reaction enhances killing by the MPO-H2O2-I- system.

Monoclonal antibodies specific for Escherichia coli J5 lipopolysaccharide: cross-reaction with other gram-negative bacterial species.

Four monoclonal antibodies against Escherichia coli J5 were studied. Each of these monoclonal antibodies reacted with purified lipopolysaccharides from E. coli J5, the deep rough mutant Salmonella minnesota Re595, Agrobacterium tumefaciens, and Pseudomonas aeruginosa PAO1 as well as with the purified lipid A of P. aeruginosa. Enzyme-linked immunosorbent assays using the outer membranes from a variety of gram-negative bacteria demonstrated that these lipid A-specific monoclonal antibodies interacted with between 84 and 97% of the gram-negative bacterial species tested. One of the monoclonal antibodies, 5E4, was shown to interact with 34 of the 35 outer membrane or lipopolysaccharide antigens tested. Immunoenzymatic staining of Western electrophoretic blots of separated P. aeruginosa outer membrane components was used to demonstrate that antibody 5E4 interacted with a similar fast-migrating band, corresponding to rough lipopolysaccharide, from all 17 serotype strains and all 14 clinical isolates of P. aeruginosa. Similarly, iodinated goat anti-mouse immunoglobulin was used to detect the binding of monoclonal antibody 8A1 to a fast-migrating band on Western electrophoretic blots of purified lipopolysaccharides from Klebsiella pneumoniae and both smooth and rough strains of E. coli, Salmonella typhimurium, and S. minnesota. These results suggest considerable conservation of single antigenic sites in the lipid A of gram-negative bacteria.

Heterogeneity of molecular size and antigenic expression within lipooligosaccharides of individual strains of Neisseria gonorrhoeae and Neisseria meningitidis.

We determined the Mr of neisserial lipooligosaccharides (LOS) by using discontinuous sodium dodecyl sulfate-polyacrylamide gel electrophoresis, minimal loading concentrations, and Salmonella isogenic rough mutant LOS as Mr standards. Salmonella LOS were resolved into three components. The migration distance of each component was linearly related to its theoretical Mr (r = 0.99). Neisserial LOS also contained multiple components whose calculated Mr ranged from 3,200 to 7,100. The relative abundance of components and their MrS varied greatly among strains. Meningococcal LOS were composed almost exclusively of two closely migrating components; gonococcal LOS were more heterogeneous. LOS from a gonococcus selected for resistance to a Pseudomonas pyocin contained only a single component that was different from and of intermediate Mr among the three components of the parent strain. A monoclonal antibody directed against the meningococcal L8 LOS epitope was used to determine whether heterogeneity of antigen expression reflected Mr heterogeneity. Single components of the L8 meningococcal LOS and of the LOS of 3 of 19 gonococcal strains bound the monoclonal antibody. Gonococcal LOS components that expressed the L8 epitope were of a similar Mr (4,800). Smaller components of these same LOS did not express the epitope.

New Salmonella typhimurium mutants with altered outer membrane permeability.

We describe three new classes of Salmonella typhimurium mutants with increased sensitivity to hydrophobic agents. In contrast to many previously described mutants, the phage sensitivity pattern of these mutants did not give any indication of defective lipopolysaccharide. Furthermore, they had no detectable changes in their phospholipid or outer membrane protein composition, and their growth rate and cell morphology were normal. Class B mutants were nearly as sensitive to novobiocin, fusidic acid, erythromycin, rifampin, and clindamycin as are deep rough (heptoseless) mutants; in addition they were sensitive to methicillin, penicillin (to which heptoseless mutants are resistant), gentian violet, and anionic and cationic detergents. Class A and C mutants had less sensitive, but characteristic phenotypes. None of the three classes were sensitive to serum bactericidal action. The class B mutation mapped between map positions 7 and 11 on the S. typhimurium chromosome, and the class C mutation mapped between positions 5 and 7. The map position for the class A mutation remained undefined, but it was separate from the class B and C mutations and, like those, did not correspond to any gene loci known to participate in the synthesis of major outer membrane constituents.

Physical aspects of structure and function of membranes made from lipopolysaccharides and free lipid A

The physical structure of the lipid component of the outer membrane of Gram-negative bacteria, here mutants of Salmonella minnesota, was studied with different optical and calorimetric techniques on the bilayer and with film balance measurements on the monolayer system. Special emphasis was laid on the elucidation of the phase behavior of lipopolysaccharides and its isolated lipid component, free lipid A, differing in the length of the polysaccharide moiety. All samples exhibit the phase transition gel-liquid crystalline of the hydrocarbon chains with T c for free lipid A lying well above the growth temperature of the bacteria (37°C)m while the mutant lipopolysaccharides show values in the range 3-–37°C. The state of order of the hydrocarbon chains at 37°C is lowest for the deep rough mutant lipopolysaccharides. Thus, an explanation of the high sensitivity of the mutants Re and Rd against hydrophobic drugs is possible. The behaviour of several physical quantities, e.g. enthalpy and cooperativity of the phase transition indicates that free lipid A and, to a lower extent, also the deep rough mutant lipopolysaccharide (mR595) form inverted structures. From this a model is derived for the mechanism of incorporation of lipopolysaccharide into the outer leaflet of the outer membrane after its lipid and saccharide parts have been translocated separately from the cytoplasmic membrane.

A group of bactericidal factors conserved by vertebrates for more than 300 million years.

A complement-dependent bactericidal factor (RaRF) specific for the Ra chemotype strains of Salmonella and for the Ra-like strains of other enterobacteria has been found in sera of mice. We show here that the anti-Ra bactericidal factors are present in sera of all species, so far tested, of mammals, birds, reptiles, amphibians, and bony and cartilaginous fish. Certain properties, such as binding specificity, requirement of divalent cations for binding, and sensitivities to heat and reducing agents, of the factor in the representative animals were the same as or similar to those of the mouse RaRF. These results indicate that these factors have been conserved by vertebrates for more than 300 million years as a result of the necessity for the resistance to rough mutants of Gram-negative bacteria.

Change in the surface hydrophobicity of substrate cells during bdelloplast formation by Bdellovibrio bacteriovorus 109J.

During intraperiplasmic growth of Bdellovibrio bacteriovorus 109J, the substrate cell surface becomes more hydrophobic. This was shown (i) by comparing the sensitivity to hydrophobic antibiotics of wild-type and lipopolysaccharide mutant strains of Salmonella typhimurium to that of the bdellovibrio growing on these strains and (ii) by measuring the binding efficiency of these strains, Escherichia coli, and their derived bdelloplasts to octyl Sepharose. The kinetics of increase in surface hydrophobicity was similar to the kinetics of the conversion of the substrate cell peptidoglycan to a lysozyme-resistant form (M. Thomashow and S. Rittenberg, J. Bacteriol. 135:1008-1014, 1978), and hydrophobicity reached a maximum at about 60 min in a synchronous culture. The change in hydrophobicity was inhibited by chloramphenicol, suggesting that bdellovibrio protein synthesis was required. Control experiments revealed that the free-swimming bdellovibrio had a more hydrophobic surface than the deep rough mutants of S. typhimurium.

Effect of rat polymorphonuclear leukocyte granule components on the growth and survival of Pseudomonas aeruginosa and Salmonella typhimurium

Granule contents from rat polymorphonuclear neutrophils were prepared by extraction with 0.2 M acetate buffer (pH 4.0), dialyzed against phosphate-buffered saline (pH 7.0), and tested for bactericidal activity against selected target bacteria. Salmonella typhimurium LT-2 and a series of progressively rough lipopolysaccharide outer membrane mutants derived from it were used to monitor antimicrobial activity. Although an antimicrobial potential was present in rat granule contents for S. typhimurium, the growth of Pseudomonas aeruginosa PAO-1 in antimicrobial assay mixtures containing rat granule contents was substantially enhanced over control values. The growth enhancement property of the granule protein was heat resistant and promoted increased oxygen consumption by whole cells.

The role of the O and K antigens in determining the resistance of Klebsiella aerogenes to serum killing and phagocytosis.

The presence of both K and O antigens of Klebsiella aerogenes was found necessary to protect the organism from either complement-mediated serum killing or phagocytosis in the absence of specific antisera. Optimal phagocytic ingestion of K. aerogenes NCTC 5055 could be achieved in the presence of either anti-K or anti-O sera or to a much smaller extent in antisera raised against a rough unencapsulated mutant (M10B) derived from NCTC 5055. Anti-O sera failed to opsonize a clinical klebsiella isolate (DL1) possessing immunologically identical lipopolysaccharide, but did so when the amount of capsule was physically reduced. The serum sensitivity of the encapsulated strains was unaffected by the addition of specific antisera. Fresh serum was bacteriostatic for an unencapsulated smooth mutant (M10) derived from NCTC 5055. This bacteriostatic effect was reduced by heat-inactivation of the serum or by the addition of anti-O serum. M10 was rendered sensitive to the bactericidal action of serum in the presence of antisera raised against M10B or after chelation with MgEGTA to isolate alternative complement pathway activity. The rough unencapsulated mutant (M10B) was rapidly killed by fresh serum, an effect which could be delayed by chelation with MgEGTA. The serum sensitivity of M10B was unaffected by the presence of anti-M10B sera. Thus, the O antigen, unlike the K antigen, of these klebsiella strains is not antiphagocytic but it does confer some protection against the rapid bactericidal activity of serum complement.

Polycations sensitize enteric bacteria to antibiotics.

Polymyxin B nonapeptide, a polymyxin B derivative which lacks the fatty acyl part and the bactericidal activity of polymyxin, was shown to sensitize smooth encapsulated Escherichia coli (O18:K1) and smooth Salmonella typhimurium to hydrophobic antibiotics (novobiocin, fusidic acid, erythromycin, clindamycin, nafcillin, and cloxacillin). The polymyxin B nonapeptide-treated bacteria were as sensitive to these antibiotics as are deep rough mutants. A lysine polymer with 20 lysine residues (lysine 20) had a largely similar effect. Larger lysine polymers and the protamine salmine were bactericidal but, at sublethal concentrations, sensitized the strains to the antibiotics mentioned above, whereas lysine4, streptomycin, cytochrome c, lysozyme, and the polyamines cadaverine, spermidine, and spermine had neither bactericidal nor sensitizing activity.

The role of heparin in guinea pig gram negative bacterial sepsis

The ability of antibody directed against shared antigenic determinants of gram negative organisms to protect against a challenge of diverse gram negative bacterial species remains controversial in the experimental setting. Attention has focused, however, on the use as immunogens of rough mutants of Escherichia coli and Salmonella minnesota, which express a portion of core lipopolysaccharide (LPS) extensively on their cell surface. Core LPS is a structure present on the outer membrane of most, if not all, gram negative bacteria. In this study rabbits were immunized with E. coli J5, a rough mutant of E. coli, to produce anti- E. coli J5 rabbit antiserum (anti-J5 RS). Anti-J5 RS was found to cross react extensively by enzyme-linked immunosorbent assay with various gram negative bacterial whole cell or LPS antigens, compared to normal rabbit serum (NRS). Anti-J5 RS ± heparin was also compared to NRS ± heparin pretreatment in a guinea pig model of sepsis utilizing E. coli 0111:B4 as the challenge organism. Anti-J5 RS ± heparin augmented systemic bacterial clearance compared to NRS ± heparin, but only the combination of anti-J5 RS and heparin enhanced survival 48 hr after bacterial challenge. It was concluded that pretreatment with anti-J5 RS was a necessary, but not sufficient condition for enhanced survival, and that the addition of heparin to anti-J5 RS pretreatment might diminish the otherwise lethal consequences of complement activation and disseminated intravascular coagulation in this model system.

Quantitation of metal cations bound to membranes and extracted lipopolysaccharide of Escherichia coli.

Inductively coupled plasma emission spectroscopy was used to quantitate the metal cations bound to outer and cytoplasmic membranes and to extracted lipopolysaccharide from several Escherichia coli K12 strains. The outer membrane was found to be enriched in both calcium and magnesium relative to the cytoplasmic membrane. Both membranes contained significant levels of iron, aluminum, and zinc. The multivalent cation content of the lipopolysaccharide resembled that of the intact outer membrane. Lipopolysaccharide extracted from wild-type k12 strains contained higher levels of Mg than Ca regardless of the growth medium, but the medium used for growth did affect the relative amounts of bound Mg as well as the levels of the minor cations iron, aluminum, and zinc. In contrast, lipopolysaccharide isolated from a deep rough mutant strain, D21f2, contained more Ca than Mg. Electrodialysis of lipopolysaccharide from wild-type k12 strains removed 1 mol of Mg per mol of lipopolysaccharide but did not significantly affect the level of other bound metal ions. Dialysis of lipopolysaccharide against sodium (ethylenedinitrilo)tetraacetate removed most of the Mg and Ca, resulting in a sodium salt. The equimolar replacement of divalent cations with sodium in the sodium salt resulted in a net loss of counterion change. The sodium salt was dialyzed against either tris(hydroxymethyl)aminomethane hydrochloride, CaCl2, MgCl2, or TbCl3, and the resulting lipopolysaccharide salts were analyzed for their ionic composition. It was shown that tris(hydroxymethyl)aminomethane and Ca can replace some but not all of the Na bound to the sodium salt, but all of the other multivalent cations tested replaced Na, resulting in uniform lipopolysaccharide salts. Lipopolysaccharide isolated from the deep rough mutant strain D21f2 was also converted into a sodium salt. Relative to the wild-type lipopolysaccharide, Na was able to neutralize the anionic charge to a greater extent in the mutant lipopolysaccharide. Our results suggest that the loss of specific groups in the core region of the lipopolysaccharide from the mutant strain results in a more open structure that allows the binding of larger cations and of more monovalent cations.

Effect of lipopolysaccharide mutations on the pathogenesis of experimental Salmonella gastroenteritis.

Lipopolysaccharide mutants of Salmonella typhimurium provoked diminished amounts of fluid in rabbit ileal loops as compared with the response to the wild type. The responses elicited by these mutants ranged from 0 to 60% of that caused by the parent strain. Two completely rough mutants and one leaky rough mutant were chosen for further study. Purified lipopolysaccharide from the parent and the mutant strains failed to stimulate fluid exsorption in ileal loop experiments. Histological studies revealed that the three lipopolysaccharide mutants were less invasive than wild type and were less able to generate an inflammatory reaction in the rabbit ileum. A Salmonella enterotoxin was present in culture filtrates from one rough mutant and the wild type; however, the rough mutant appeared to produce less toxin. Enterotoxic activity was absent in culture filtrates from the two other rough mutants. These results suggest that reductions in both invasiveness and the ability to produce Salmonella enterotoxin decreased the ability of these mutants to provoke fluid exsorption. Also, the results indicate that lipopolysaccharide mutations can have a profound effect on the enteropathogenic properties of S. typhimurium.

Bactericidal activity of granule contents from rat polymorphonuclear leukocytes.

Granule contents from rat polymorphonuclear neutrophils were prepared by extraction with 0.2 M acetate (pH 4), dialyzed against phosphate-buffered saline (pH 7), and tested for bactericidal activity. Bactericidal assays consisted of mixing rat granule extract with 1 x 10(3) to 3 x 10(3) bacterial cells per ml at 37 degrees C for 1 h in a medium suited for bacterial growth. The granule extract demonstrated a distinctive dose-dependent bactericidal activity against outer membrane lipopolysaccharide mutants of Salmonella typhimurium LT-2, independent of added hydrogen peroxide or other active oxygen derivatives. The rough bacterial mutants showed an ordered increase in sensitivity to the rat lysosomal extracts inversely related to the length of their lipopolysaccharide carbohydrate side chains. Fractionation of the rat polymorphonuclear neutrophil granule extract with Sephadex G-100 column chromatography revealed an elution profile containing three major areas (peaks) of protein. Polyacrylamide gel electrophoresis and examination of enzymatic activity showed that these peaks contained myeloperoxidase (peak A), neutral protease (peak B), and lysozyme (peak C) activities. Also observed in peak C were cationic protein species whose cathodal electrophoretic migration was faster than that for lysozyme. Only peak C exhibited a bactericidal activity against the rough mutants of S. typhimurium LT-2 similar to that obtained for the unfractionated granule extract, with susceptibility of the bacterial mutants increasing with a progressive loss of carbohydrate residues in the lipopolysaccharide of the cell wall. The bactericidal activity of the peak C protein fraction was dose dependent. Boiling the unfractionated granule extract or peak C for 30 min had little affect on their antimicrobial activity when reacted against a deep-rough lipopolysaccharide mutant. However, trypsin pretreatment of these fractions significantly reduced their antimicrobial activity for the same mutant chemotype.

An intermediate step in translocation of lipopolysaccharide to the outer membrane of Salmonella typhimurium.

Evidence for transient localization of newly synthesized lipopolysaccharide at the periplasmic face of the inner membrane has been obtained by immunoelectron microscopic techniques. Salmonella typhimurium galE mutants in which O-antigen synthesis is dependent on addition of exogenous galactose were employed, and the distribution and fate of pulse-synthesized O antigen was examined by indirect ferritin labeling with anti-O-antigen IgG of spheroplasts prepared by treatment with lysozyme/EDTA. O-reactive lipopolysaccharide appeared rapidly at the exposed periplasmic face of the inner membrane after addition of galactose and was rapidly depleted upon termination of the pulse. Control experiments showed that secondary redistribution of lipopolysaccharide from outer membrane did not occur under the conditions employed for spheroplast formation and immunolabeling, and the pulse-chase kinetics were consistent with those expected for an intermediate in translocation of lipopolysaccharide to the outer membrane. In addition, undecaprenol-linked O antigen was detectable at the periplasmic face of the inner membrane within 30 sec after addition of galactose to a galE deep rough double mutant, and it accumulated stably in that location. The mutation in synthesis of the lipopolysaccharide core in the deep rough strain prevents transfer of O-antigen chains from undecaprenol phosphate to lipopolysaccharide. The result suggests that attachment of O antigen to lipopolysaccharide occurs on the extracytoplasmic side of the inner membrane and supports the conclusion that lipopolysaccharide is translocated to the outer membrane from the periplasmic, rather than the cytoplasmic, face of the inner membrane.

The role of lipopolysaccharide structure in the recipient cell during plasmid-mediated bacterial conjugation

The role of the lipopolysaccharide (LPS) structure in the recipient cell during bacterial conjugation was studied using a series of well-defined LPS mutations in Salmonella minnesota. The plasmids F lac ( IncFI) and R1 drd19 ( IncFII) transferred at a high frequency to the smooth S218 parent strain and the rough LPS mutants. However, R64 drd11 ( IncIα) transferred poorly to the LPS mutants compared with transfer to the smooth LPS parent strain. The decrease in R64 drd11 transfer frequency correlated with the extent of the defect in LPS structure, suggesting that intact LPS on the recipient is a major requirement for R64 drd11 mating. Transfer of the P-group plasmid, RK2, was not affected by changes in LPS structure. These results show that plasmids use different cell surface structures during conjugation, and that LPS is particularly important for R64 drd11 transfer.

Influence of the culture conditions of Brucella abortus on its antitumor properties.

A heat-inactivated rough mutant of B. abortus increased resistance to a transplantable lymphoma when cultured on solid agar (ag) but not when cultured on a fluid medium (fl). The ag organisms were smaller, more toxic and induced more interferon 2 h after injection than fl. They also contained a phospholipid (cardiolipin) absent from fl preparations.

Differential determination of the 3-Deoxy-D-mannooctulosonic acid residues in lipopolysaccharides of Salmonella minnesota rough mutants.

Different applications of the thiobarbituric acid assay were used to determine C4/5‐unsubstituted, heptosyl‐substituted and total 3‐deoxy‐d‐mannooctulosonic acid (dOclA) in lipopolysaccharides of Salmonella minnesota rough mutants. C4/5‐unsubstituted dOclA is determined without prior hydrolysis of the lipopolysaccharides by periodate oxidation near neutral pH at 4°C using the methyl ketoside of dOclA as a standard. C4/5‐unsubstituted dOclA, and dOclA which is substituted by other dOclA residues, is quantified after hydrolysis in 0.1 M acetate buffer pH 4.4 at 100°C/1 h. Total dOclA, including heptosyl‐substituted dOclA, is measured after hydrolysis in 1 M HCl at 100°C for 1–4 h. For S. minnesota R595, 760–790 nmol/mg lipopolysaccharide are determined independently of the hydrolysis conditions employed. In R7, one half, and in R345, two thirds of the total dOclA content are detected after mild acid hydrolysis. Hydrolysis in 1 M HCl at 100°C for 1 h (R7) and 4 h (R345) are required to liberate the heptosyl‐substituted dOclA residues. For the three lipopolysaccharides tested, determination of C4/5‐unsubstituted dOclA revealed that about one third of the total dOclA is reactive in this application of the thiobarbituric acid assay. By strong acid hydrolysis, a 3‐deoxy‐2‐ketoaldonic acid is identified in the lipopolysaccharide of Vibrio cholerae. Upon mild and strong acid hydrolysis, a 3‐deoxy‐2‐ketoaldonic acid is identified in the lipopolysaccharide of Vibrio cholerae. Upon mild and strong acid hydrolysis, dOclA undergoes degradation which is recognized by (a) decreased reactivity in the thiobarbituric acid assay (b) changed ultraviolet spectra and (c) development of compounds which are reactive in the thiobarbituric acid assay after reduction with sodium borohydride. The rate of degradation depends on the strength of the acid but is the same for reference and lipopolysaccharide‐liberated dOclA under a given hydrolysis condition.

Isolation of a 3-deoxy-D-mannooctulosonic acid disaccharide from Salmonella minnesota rough-form lipopolysaccharides.

Lipopolysaccharides of Salmonella minnesota rough mutants were treated with 20 mM acetate buffer pH 4.4 at 70 degrees C/3 h. After dialysis of the hydrolysates about one third of the total 3-deoxy-D-mannooctulosonic acid (dOclA) content but no neutral sugars were found in the dialysate. By high-voltage paper electrophoresis, a compound with the mobility of 1.2 relative to dOclA could be isolated from the dialysate. It was identified as a dOclA disaccharide by hydrolysis without or after reduction with sodium borohydride and by analysis with the thiobarbituric acid assay under different conditions. The ketosidic linkage in the disaccharide is assumed to be 2.4 or 2.5.