Correlation between phagocytic activity and metabolic response of polymorphonuclear leukocytes toward different strains of Escherichia coli

Increased susceptibility to Escherichia coli infection in mice pretreated with Corynebacterium parvum

The contribution of activated macrophages to protection against Escherichia coli was studied in mice treated intravenously with Corynebacterium parvum 7 days before infection. C. parvum-treated mice showed increased phagocytic activity and enhanced resistance to Listeria infection. In contrast, these mice showed increased susceptibility to a subsequent challenge with E. coli that correlated closely with a reduction in the LD50 of lipopolysaccharide (LPS) in these mice. The peritoneal macrophages obtained from C. parvum-treated mice had a strong ability to phagocytize and kill E. coli in in vitro experiments. A rapid decline in the number of bacteria in the liver of C. parvum-treated mice was observed in the early period of infection. However, the number of bacteria in liver and spleen increased progressively to a lethal dose from 6 hr after infection. At this time, a significant increase in beta-glucuronidase, a lysosomal acid hydrolase, was found in the serum of these mice. In vitro experiments revealed that the peritoneal macrophages from C. parvum-treated mice were highly susceptible to the cytotoxic effect of LPS after 6 hr of incubation with LPS. It is suggested that the hypersensitivity of activated macrophages to the cytotoxic effect of endotoxin derived from E. coli may be partly responsible for the increased susceptibility of C. parvum-treated mice to E. coli infection.

Incidence of myeloperoxidase deficiency in an area of northern Italy: histochemical, biochemical and functional studies

Forty-five subjects with a complete deficiency of myeloperoxidase were identified in an area of the region Friuli-Venezia Giulia in north-eastern Italy using the Hemalog D system as the screening technique. Histochemical and biochemical tests performed on the leucocytes of some of these subjects confirmed the defects shown by the Hemalog D system. The defect was of genetic origin in seven subjects. The genetic origin could be suspected in another eight subjects since more than two affected members were present in a given family. Eosinophil peroxidase, which is present in MPO deficient subjects, interfered with the guaiacol assay of MPO, and in several cases masked the genetic transmission. An assay was developed using o-dianisidine as the electron donor which considerably reduced the interference by EPO. With this assay an autosomal recessive pattern of inheritance was found. The MPO deficient leucocytes had a higher respiratory burst than control cells and an impaired bactericidal activity, at early post-phagocytic periods, which became comparable to that of control cells at later stages. Particle ingestion by the MPO-deficient cells was normal.

A new combined assay of phagocytosis and intracellular killing of Escherichia coli by polymorphonuclear leukocytes

A new combined radiometric assay is described in which adherence, and phagocytosis and killing of Escherichia coli by human polymorphonuclear leucocytes (PMN) are simultaneously measured in the same sample. Pure monolayers of PMN in Petri dishes are allowed to ingest [14C]phenylalanine labelled E. coli and excess bacteria are removed by washing. A period of incubation allows intracellular killing to occur while polymyxin-B is added to half the dishes to kill extracellular bacteria. The remaining viable bacteria in all dishes are labelled with [3H]thymidine. The number of ingested bacteria and the percentage of intracellular organisms killed is determined from the 14C and 3H counts by a simple subtraction technique. By performing protein assays on representative monolayers, the number of PMN adhered in the monolayers and hence the mean bacterial uptake per PMN is estimated. The assay detected killing efficiencies reduced below the normal range, in monolayers treated with sodium azide, phenylbutazone, in polymorphonuclear leukocytes from patients with chronic granulomatous disease, and in immature neutrophils from the promyelocytic leukaemic cell line, HL60. The assay was adapted to measure phagocytosis and killing by cells in suspension.

Phagocytic and bactericidal activity of human neutrophils against two isolates of Group B streptococci Type Ic of differing pathogenicity

The phagocytic and bactericidal activities of normal adult human neutrophils against 2 strains of Group B streptococci Type Ic of differing pathogenicity were examined. Both isolates were phagocytosed by the neutrophils in the presence of normal and homologous immune serum. However, the highly pathogenicity streptococci were killed less readily in the presence of immune serum than were the streptococci of low pathogenicity in the presence of immune or normal serum. This difference in killing ability was not due to a defect in phagocytosis by the neutrophils, but to a defect in bactericidal activity. The highly pathogenic streptococci were not killed in the presence of normal serum, but were readily phagocytosed by the neutrophils, in which they accumulated and eventually caused their destruction. The streptococci of low pathogenicity, however, were killed equally as well in the absence of specific antibody as in its presence. It is suggested that an in vitro assessment of neutrophil function against streptococci of differing pathogenicity for mice may provide a useful method by which the pathogenicity of streptococci for man can be compared.

Relative opsonic and protective activities of antibodies against K1, O and lipid A antigens of Escherichia coli

The K1 Escherichia coli capsular antigen has been implicated as a virulence factor because of the frequency of isolation of K1 containing strains from certain invasive human infections. In the study of the interaction between K1 strains, normal human polymorphonuclear cells (PMNs) and fresh human serum, we have found varying susceptibility to phagocytosis and killing; thus, the in vitro opsonophagocytic and in vivo protective role of K1, somatic O and core glycolipid antibodies remain unclear. We have therefore examined strains of E. coli with defined susceptibility to phagocytosis by normal PMNs and sera and compared the effect of K1, somatic O and lipid A antibodies in opsonophagocytic tests and mouse protection experiments. K1 E. coli strains demonstrating relative resistance to phagocytosis and killing were effectively opsonized only with specific K1 capsular antisera. Similarly, K1 capsular antisera, but not anti-O or lipid A antisera, also provided protection in mice challenged with a LD100 of K1 E. coli that were "resistant" to phagocytosis. The ability of purified capsular antigens from Neisseria meningitidis group B and K1 "resistant" E. coli to inhibit the phagocytosis of a "sensitive" non-K1 and a K1 E. coli strain of "intermediate" susceptibility to opsonophagocytosis was also investigated. Purified K1 and group B capsular antigens were able to block specific capsular-antibody mediated opsonophagocytosis, yet these capsular antigens failed to inhibit the phagocytosis of non-K1 "sensitive" or K1 "intermediate" E. coli. These studies suggest that K1 antibodies are obligatory for the in vitro and in vivo opsonophagocytosis of "resistant" K1 E. coli and that the K1 antigen must remain in situ on the bacterial surface to exert an anti-phagocytic effect.

Role of Escherichia coli K capsular antigens during complement activation, C3 fixation, and opsonization

Escherichia coli strains with K capsular polysaccharides are relatively resistant to phagocytosis by polymorphonuclear leukocytes, in contrast to E. coli strains without K antigens. This inhibition of phagocytosis is related to an impaired recognition of the K+ strains by the phagocytes due to ineffective opsonization. All five strains without K antigens were readily phagocytized after opsonization in 5% normal serum, compared with no uptake of the K+ strains. Evidence is presented that the decreased opsonization of the K+ strains in normal serum is caused by a low rate of complement activation of the strains, with subsequent absence of C3b fixation or C3d fixation or both to the cell wall of the bacteria. After removal of the K+ antigens by heating of a K+ E. coli strain, the strain was able to activate complement, to bind C3b or C3d or both, and to become opsonized. Complement was then activated via the classical and alternative pathways, which was comparable to the complement consumption by K- E. coli.

Effect of antibodies and complement on the interaction between Escherichia coli 0111:B4 and polymorphonuclear leukocytes

The interaction of Escherichia coli 0111:B4 with polymorphonuclear leukocytes in the presence of specific antibodies and complement was studied. This strain, which is resistant to phagocytosis by polymorphonuclear leukocytes, may be ingested and killed by the phagocytes in the presence of both antibodies and fresh serum. The ineffectiveness of fresh serum to promote ingestion of E. coli 0111:B4 by the phagocytes in the absence of antibodies reflects the inability of this strain to activate the complement system through the alternative pathway. Investigation of the mechanisms of the bacterial killing by polymorphonuclear leukocytes showed that both antibodies and complement were required for the oxygen-independent bactericidal system, whereas they were not needed for the oxygen-dependent system.

Capsular K1 polysaccharide of Escherichia coli: relationship to virulence in newborn rats and resistance to phagocytosis

The virulence of Escherichia coli strains for newborn rats was related to opsonic requirements of the strains, sensitivity to the bactericidal activity of serum, and K1 capsular polysaccharide content. K1 E. coli strains were more virulent than non-K1 strains after intraperitoneal injection in newborn rats (P less than 0.05) and were more resistant to phagocytosis than non-K1 strains when the classical complement pathway was blocked with Mg-ethyleneglycoltetraacetic acid (P less than 0.0005). Sensitivity to the bactericidal activity of serum was similar among K1 and non-K1 E. coli strains. Two groups of K1 E. coli strains were defined on the basis of opsonic requirements. Group I strains were efficiently opsonized by the alternative complement pathway, while group II strains required the classical complement pathway for opsonization. Group I strains had less detectable K1 polysaccharide in the washed whole cell fraction than group II strains (10.3 versus 18.9 microgram of K1 polysaccharide per 10(10) colony-forming units) and were less virulent than group II strains (mortality, 44 versus 77%, P less than 0.05). The K1 capsular polysaccharide appears to play an important role in determining virulence in newborn rats and opsonic requirements of these strains, but does not contribute to the sensitivity of strains to the bactericidal activity of serum.

Influence of O and K antigens on the surface properties of Escherichia coli in relation to phagocytosis

Strains of Escherichia coli with different O and K antigens were investigated with respect to physicochemical surface characteristics and liability to phagocytosis. Using two-phase partitioning analysis for the surface characterization, three main groups of strains emerged: Group I (O1:K1, O2:K1, O3:K2ab) showing both smooth hydrophilic O antigens and negatively-charged K antigen which rendered the strains maximally resistant to phagocytosis. Group II (O55:K59, O111:K58) showed no acidic K antigen but only smooth hydrophilic O antigen properties. However, these strains were as resistant to phagocytosis as the strains in group I. A third group (O14:K7, O24:K +) contained strains with rough, hydrophobic O antigen and negatively-charged K antigen. When the K antigen was removed by heat treatment these strains became more sensitive to phagocytosis. Certain other strains (O28:K-, O56:K + and O118:K-) did not fit into the three groups. These experiments show that the physicochemical surface effects and biological significance of the K antigen must be evaluated in relation to the properties conveyed by the corresponding O antigens.

Surface characteristics of Escherichia coli strains in relation to development of bacteraemia

Three groups of E. coli bacteria isolated from blood, urine and faeces were analysed with respect to physicochemical surface properties in aqueous polymer two-phase systems. Eighty-eight per cent of the bacteria isolated from blood cultures showed affinity for the dextran-rich bottom phase, whereas 60% and 30% of the bacteria isolated from urine and faeces, respectively, collected to the same phase. Further two-phase analysis indicated that the bacteraemia strains exposed more negative surface than the other groups of bacteria, and that this charge was reduced after heat treatment (70 degrees C, 40 min). These results thus indicate that the strains causing bacteraemia possess similar surface properties, and may have been selected from the more heterogeneous group of bacteria found in urine and faeces. It is further conceivable to propose that bacteria found in the blood expose more heatsensitive, negatively-charged K antigen than the other groups of bacteria.

Cellular elements in the resistance to candida infection in mice. I. Contribution of T lymphocytes and phagocytes at various stages of infection

Live organisms (cfu) of Candida albicans per organ were counted 1 hr and 1 to 20 days after an intravenous inoculation into various groups of mice which had distinct levels of immunologic or non-immunologic defense mechanisms. a) The number of cfu in the liver decreased progressively in normal mice, but those in the kidney maintained a constant level during the observation period. b) The number of cfu in the liver decreased progressively also in nude mice. In their kidneys, however, cfu increased progressively at a late stage of infection. c) In lethally irradiated AKR of nude mice in which phagocyte functions were severely depressed, the number of cfu increased progressively in both liver and kidney from the initial stage of infection. d) In immunized AKR mice, growth of C. albicans was suppressed at late stages of infection. Such protective immunity could be transferred partly with immune lymphoid cells but not with hyperimmune serum in the experimental system employed. In protection against candida infection, non-immune phagocytosis and T cell-mediated immunity appear to be required at the early and late stages of infection, respectively.

Correlation between the susceptibility of E. coli to phagocytosis and their ability to invade HeLa cells in vitro

The susceptibility of several strains of E. coli to phagocytic killing by polymorphonuclear lencocytes and the ability of the same strains to invade HeLa cells were studied. It was found that only the strains resistant to killing by leucocytes were able to penetrate and multiply within HeLa cells.

Interaction of E. coli strains with human serum: lack of relationship to K1 antigen

Twenty-eight strains of E. coli isolated from infants were compared with respect to opsonic requirements, sensitivity to serum, and ability to activate serum chemotactic factors. Six of the strains were isolated from stools of healthy newborn infants; 22 were isolated from the cerebrospinal fluid or blood of infants with meningitis and/or septicemia. Eighteen of the strains had K1 polysaccharide antigen. Fourteen of the strains (seven with K1 antigen) activated complement via the alternative pathway and all of these strains were well opsonized in 4% pooled human serum. A higher concentration of serum was necessary to opsonize 12 of the 14 strains that did not activate the alternative pathway. A wide variation was also found in opsonic requirements of E. coli strains isolated from healthy and sick infants. There was no relationship of the K1 antigen to opsonic requirements, to capacity to activate complement via the alternative pathway, to generation of chemotactic factors, or to sensitivity to serum cidal activity. Therefore, the association of E. coli with K1 antigen and neonatal meningitis did not appear to be related to these bacteria-serum interactions.

Rat polyvinyl sponge model for the study of infections: initial investigations

Polyvinyl sponges were implanted subcutaneously on both sides of young female rats. One sponge was infected with 10(8) of either Escherichia coli K-12 F-, Staphylococcus aureus ATCC 25923, or Pseudomonas aeruginosa CDC 7725. P. aeruginosa remained at the inoculum level and S. aureus declined by 1 log, whereas E. coli was reduced 1,000-fold. Only P. aeruginosa was recovered from the blood in 36% of the animals in 24 h and in 20% of the rats in 48 h. The nutrient potential of rat inflammatory fluid was compared to nutrient broth by growth of each bacterium in untreated and heat-inactivated sponge fluids and Trypticase soy broth.