Sialic acid in the lipopolysaccharide of Haemophilus influenzae: strain distribution, influence on serum resistance and structural characterization.

A survey of Haemophilus influenzae strains indicated that around one-third of capsular strains and over two-thirds of non-typeable strains included sialic acid in their lipopolysaccharides (LPS). Mutation of the CMP-Neu5Ac synthetase gene (siaB) resulted in a sialylation-deficient phenotype. Isogenic pairs, wild type and siaB mutant of two non-typeable strains were used to demonstrate that sialic acid influences resistance to the killing effect of normal human serum but has little effect on attachment to, or invasion of, cultured human epithelial cells or neutrophils. We determine for the first time the site of attachment of sialic acid in the LPS of a non-typeable strain and report that a small proportion of glycoforms include two sialic acid residues in a disaccharide unit.

Using the yeast two-hybrid system to identify human epithelial cell proteins that bind gonococcal Opa proteins: intracellular gonococci bind pyruvate kinase via their Opa proteins and require host pyruvate for growth.

Neisseria gonorrhoeae opacity-associated (Opa) proteins are a family of outer membrane proteins involved in gonococcal adherence to and invasion of human cells. We wanted to identify additional roles for Opa in the infectious process and used the yeast two-hybrid system to identify human epithelial cell proteins that interact with Opa proteins. Although this system has been used successfully to identify many types of interacting proteins, it has not been used to screen a human cell cDNA library for binding partners of a prokaryotic outer membrane protein. Therefore, we were also interested in exploring the versatility of the yeast two-hybrid system in identifying bacteria-host interactions. Using OpaP from strain F62SF as bait, we screened a HeLa cell cDNA library for Opa-interacting proteins (OIPs). We identified five different OIPs, designated OIP1-OIP5, two of which are homologous to human proteins--thyroid hormone receptor interacting protein (TRIP6) and pyruvate kinase isoenzyme M2 (PK). In the studies presented here, we investigated the interaction between Opa proteins and PK in more depth. Opa-PK interactions were confirmed by in vitro and in vivo assays independent of the yeast two-hybrid system. Escherichia coli expressing six different Opa proteins from gonococcal strain FA1090 all bound more PK than Opa-negative E. coli in in vitro binding assays. Using anti-PK antibody and fluorescence microscopy, we showed that human epithelial cell PK co-localizes with intracellular Opa+ gonococci and E. coli expressing Opa proteins. Using a mutant of N. gonorrhoeae unable to grow on pyruvate or lactate, it appears that intracellular pyruvate is essential for gonococcal growth and survival. These results suggest a novel mechanism in bacterial pathogenesis, i.e. the requirement for direct molecular interaction with a host metabolic enzyme (PK) for the acquisition of an essential intracellular carbon source and growth substrate (pyruvate). These results demonstrate that the yeast two-hybrid system is a valuable tool for identifying biologically relevant interactions between bacteria and host proteins, providing valuable leads for further investigations into novel mechanisms of bacterial pathogenesis.

Regulation of gonococcal sialyltransferase, lipooligosaccharide, and serum resistance by glucose, pyruvate, and lactate.

Strain F62 of Neisseria gonorrhoeae gonococci (GC) is sensitive to normal human serum unless CMP-N-acetylneuraminic acid (CMP-NANA) is present. NANA is transferred primarily to a 4.5-kDa lipooligosaccharide (LOS) structure by a GC sialyltransferase (Stase). We investigated LOS and Stase expression and serum resistance in strain F62 grown in different carbon sources and growth conditions. Pyruvate-grown GC expressed 1.9- to 5.6-fold more Stase activity than did glucose-grown GC, whereas lactate-grown GC generally expressed intermediate Stase activities. Broth-grown GC expressed two- to fourfold more Stase activity than did plate-grown GC in all carbon sources. Pyruvate- or lactate-grown GC expressed significantly more of the sialylateable 4.5-kDa LOS species than did glucose-grown GC. Anaerobically, the 4.5-kDa LOS species was expressed in greater quantity than the 4.9-kDa N-acetyl galactosamine-terminating species in all carbon sources. Pyruvate-grown GC also incorporated up to threefold more radiolabelled CMP-NANA onto the 4.5-kDa LOS species than did glucose-grown GC. In serum resistance studies, pyruvate-grown GC were 6.5- to 16.1-fold more serum resistant than glucose-grown GC at limiting CMP-NANA concentrations (1.56 to 12.50 microg/ml). Taken together, these results indicate that gonococcal expression of Stase activity is up-regulated by growth in pyruvate or lactate, which correlates with enhanced expression of the sialylateable 4.5-kDa LOS and, for growth in pyruvate, correlates with enhanced sialylation of gonococcal LOS and greater serum resistance. In different in vivo niches, gonococcal LOS sialylation, serum resistance, and interaction with host cells can be highly regulated.

Expression of sialyltransferase is not required for interaction of Neisseria gonorrhoeae with human epithelial cells and human neutrophils.

Sialyltransferase (Stase) in Neisseria gonorrhoeae organisms (gonococci [GC]) transfers sialic acid (N-acetylneuraminic acid [NANA]) from cytidine 5'-monophospho-N-acetylneuraminic acid (CMP-NANA) mainly to the terminal galactose (Gal) residue in the Gal beta-1,4 N-acetylglucosamine (Gal-GlcNAc)-R lipooligosaccharide (LOS) structure. Sialylated GC resist killing by normal human serum, sometimes show reduced invasion of epithelial cells, and have reduced adhesion to and stimulation of human neutrophils. We questioned whether Stase itself modulates the interactions of GC with human epithelial cells and neutrophils in the absence of exogenous CMP-NANA. To that end, we treated strain F62 with ethyl methanesulfonate and grew approximately 175,000 colonies on CMP-NANA plates, and screened them with monoclonal antibody 1B2-1B7 (MAb 1B2). MAb 1B2 is specific for Gal-GlcNAc and reacts only with asialylated GC. We isolated 13 MAb 1B2-reactive mutants, including five null mutants, that had Stase activities ranging from barely detectable to fivefold less than that of wild-type (WT) F62. The LOS phenotype of Stase null mutants was identical to that of WT F62, yet the mutants could not sialylate their LOS when grown with CMP-NANA. The Stase null phenotype was rescuable to Stase+ by transformation with chromosomal DNA from WT F62. Stase null mutants remained serum sensitive even when grown with CMP-NANA. One Stase null mutant, ST94A, adhered to and invaded the human cervical epithelial cell line ME-180 at levels indistinguishable from that of WT F62 in the absence of CMP-NANA. In human neutrophil studies, ST94A stimulated the oxidative burst in and adhered to human neutrophils at levels similar to those of WT F62. ST94A and WT F62 were also phagocytically killed by neutrophils at similar levels. These results indicate that expression of Stase activity is not required for interaction of GC with human cells.

Interaction of pathogenic Neisseria with host defenses. What happens in vivo?

N. gonorrhoeae initiates infection by adhering to and invading columnar epithelial cells. Over time these activities often induce inflammation, with the influx of neutrophils and serum into the urethral lumen, cervical os, conjunctiva, and the like. At least some of these infected niches contain CMP-NANA (cytidine monophospho-N-acetyl neuraminic acid, also called CMP-sialic), contain sialylated gonococci, and are relatively or strictly anaerobic due to neutrophil and gonococcal metabolism and to the site of disease, that is, the peritoneal cavity. Gonococci thus encounter environmental conditions, reagents, and substrates in the human body that are not normally present in vitro. Knapp and Clark were the first to successfully grow gonococci anaerobically in an easily reproducible system, allowing researchers to begin to investigate in vitro the effects of anaerobiosis on gonococcal virulence traits. As a result of a series of elegant and in depth studies, Smith and Parsons and their colleagues showed that growth in CMP-NANA confers on the gonococcus a high degree of phenotypic (readily reversible) serum resistance and that CMP-NANA is available in vivo at sites of gonococcal infection and disease; gonococci become covalently coated with sialic acid and they become serum resistant (reviewed in refs. 8-10). Given that gonococci growing in the absence of oxygen or in the presence of CMP-NANA probably more closely resemble gonococci growing inside the human host, we studied several possible virulence traits of gonococci cultivated under these conditions. We first observed that anaerobic growth (in the absence of CMP-NANA) increases gonococcal resistance to killing by low (but not high) concentrations of normal human serum. We also asked whether anaerobic growth affected gonococcal association with host cells. Contrary to the effects on serum killing, anaerobic growth (in the absence of CMP-NANA) does not appear to affect the ability of gonococci (expressing certain adhesive outer membrane proteins called Opa proteins) to bind to and enter human epithelial cell lines or to bind to or resist killing by human neutrophils. The results from studies investigating the modulatory role of CMP-NANA were more striking. Growth in CMP-NANA dramatically inhibits the adherence of Opa+ gonococci to human neutrophils. It does not, however, appear to significantly decrease their sensitivity to phagocytic killing or to in vitro killing by lysosomal contents (aqueous extracts of human neutrophil granules).(ABSTRACT TRUNCATED AT 400 WORDS)

Anaerobic growth and cytidine 5'-monophospho-N-acetylneuraminic acid act synergistically to induce high-level serum resistance in Neisseria gonorrhoeae.

In vivo, gonococci encounter a myriad of conditions not present in vitro. At some stages of infection and disease, gonococci may grow anaerobically, probably by using sodium nitrite as a terminal electron acceptor. Also, gonococci sialylate their lipooligosaccharide (LOS) in vivo, by using low concentrations of cytidine 5'-monophospho-N-acetylneuraminic acid (CMP-NANA) present in host tissue. This sialylation is responsible for the acquired resistance of gonococci to both normal and immune human serum. Given that gonococci grown in the absence of oxygen or in the presence of CMP-NANA probably more closely resemble gonococci grown inside a human host, we studied the serum resistance of gonococci cultivated under these conditions. In the absence of CMP-NANA, anaerobically grown (anaerobic) gonococci were somewhat less sensitive to serum killing than were aerobically grown (aerobic) gonococci. However, anaerobic gonococci grown with 6 micrograms of CMP-NANA per ml exhibited almost complete serum resistance, while aerobic gonococci required 16-fold-higher CMP-NANA concentrations to achieve significant serum resistance. Anaerobic gonococci incubated in CMP-NANA converted to serum resistance two to three times faster than did similarly treated aerobic gonococci and incorporated up to six times as much sialic acid into their LOS. Gonococci can express several different LOS molecules. Anaerobic gonococci expressed the LOS molecule that acts as an acceptor for sialic acid from CMP-NANA in greater quantity than aerobic gonococci did. Finally, Triton X-100 extracts of anaerobic gonococci contained about four times more sialyltransferase activity than did extracts of aerobic gonococci. Sialyltransferase activity in these extracts was not inhibited by oxygen or enhanced by anaerobiosis. These data indicate that anaerobic conditions lead to altered LOS biosynthesis and to induction of sialyltransferase activity in gonococci. In vivo, where decreased oxygen levels and relevant concentrations of CMP-NANA are found, gonococci could readily become resistant to killing by normal and immune human serum.

Escherichia coli expressing a Neisseria gonorrhoeae opacity-associated outer membrane protein invade human cervical and endometrial epithelial cell lines.

Members of the opacity-associated (Opa) outer membrane protein family of Neisseria gonorrhoeae have been proposed to mediate adherence to and invasion of cultured human epithelial cells. We transformed Escherichia coli with a plasmid containing a gonococcal opa gene fused in-frame to the leader sequence of the beta-lactamase gene as described by Palmer et al. [Palmer, L., Brooks, G. F. & Falkow, S. (1989) Mol. Microbiol. 3, 663-671]. These transformed E. coli [E. coli (opa)] expressed the heat-modifiable opa gene product (the Opa protein) in their outer membrane and adhered to and invaded ME-180 human endocervical epithelial cells. In a 2-h adherence assay, an average of 26.7 E. coli (opa) adhered per ME-180 cell, whereas the control E. coli carrying only the expression vector (pKT279) did not adhere at all (less than 0.15 bacterium per cell). We investigated the ability of the adherent E. coli (opa) to invade ME-180 epithelial cells by using a gentamicin selection assay. We recovered up to 1 x 10(6) gentamicin-resistant bacteria per monolayer when ME-180 cells were infected with E. coli (opa) compared to less than 10 bacteria when the epithelial cells were infected with the same number of control E. coli (pKT279). The kinetics and level of invasion by E. coli (opa) were similar to invasion by Opa+ N. gonorrhoeae. Maximum invasion occurred 4 h after infection with 4 x 10(7) bacteria. Transmission electron microscopy studies confirmed that E. coli (opa) invaded ME-180 cells. In comparative studies, the number of E. coli (opa) that invaded HEC-1-B human endometrial epithelial cells was about an order of magnitude less than the number that invaded ME-180 cells, and E. coli (opa) did not invade Chang human conjunctival epithelial cells at all. The observations that early (less than 4 h) invasion by E. coli (opa) was dramatically inhibited, in a dose-responsive manner, by the actin-disrupting reagent cytochalasin D but later invasion (8-24 h) was not suggest that invasion mediated by Opa proteins may occur by two mechanisms, only one of which is dependent upon microfilament function. Transmission electron microscopy also revealed that infected epithelial cells had a dramatically increased amount of cytoplasmic fibrillar material surrounding the nucleus. The function and genesis of this material remain unclear. These studies indicate that at least one gonococcal Opa protein is an invasin.

Anaerobic growth of gonococci does not alter their Opa-mediated interactions with human neutrophils.

Gonococci grown anaerobically (anaerobic gonococci) in the presence of nitrite induce the expression of at least three novel outer membrane proteins (PANs 1 to 3). Although PAN 1 is expressed by gonococci during gonorrhea, the function of the PAN proteins remains unknown. In the absence of serum, gonococci possessing opacity-associated (Opa, formerly PII) outer membrane proteins adhere to, stimulate, and are phagocytically killed by human neutrophils. Gonococci lacking Opa proteins demonstrate none of these activities. We investigated whether the PAN proteins, or any other characteristics of anaerobic gonococci, altered the ability of nonpiliated, Opa+ or Opa- gonococci to adhere to, stimulate, or be phagocytically killed by neutrophils. The expression of Opa4 by strain F62, as determined by its relative mobility on sodium dodecyl sulfate-polyacrylamide gels, appeared to be unaltered by anaerobic growth, as seen previously (V. L. Clark, L. A. Campbell, D. A. Palermo, T. M. Evans, and K. W. Klimpel, Infect Immun. 55:1359-1364, 1987). Anaerobic and aerobic Opa+ gonococci adhered to and stimulated neutrophils to the same extent. Similarly, anaerobic and aerobic Opa- gonococci adhered to and stimulated neutrophils equally poorly. Finally, anaerobic and aerobic Opa+ gonococci were equally sensitive to phagocytic killing by neutrophils, while anaerobic and aerobic Opa- gonococci were equally resistant to killing. Thus, the role of Opa proteins in mediating the interactions of gonococci with human neutrophils appears unaltered by anaerobic growth, and the PAN proteins, or other cryptic properties of anaerobic gonococci, do not seem to modulate or mediate these phenomena.

Growth of Neisseria gonorrhoeae in CMP-N-acetylneuraminic acid inhibits nonopsonic (opacity-associated outer membrane protein-mediated) interactions with human neutrophils.

Gonococci possessing certain opacity-associated (Opa) outer membrane proteins adhere to and are phagocytosed by human neutrophils in the absence of serum. Recently, it has been shown that serum-sensitive strains of Neisseria gonorrhoeae possessing the appropriate lipooligosaccharide phenotype become serum resistant when grown in the presence of CMP-N-acetylneuraminic acid (CMP-NANA) because of sialylation of their lipooligosaccharide. We investigated whether such sialylation affects nonopsonic (antibody- and complement-independent) interactions of gonococci with human neutrophils in vitro. We grew Opa+ gonococci in the presence of up to 50 micrograms of CMP-NANA per ml, incubated them with neutrophils in vitro, and measured their abilities to adhere to neutrophils, stimulate neutrophil luminol-dependent chemiluminescence (LDCL), and be phagocytically killed by neutrophils. Growth in CMP-NANA dramatically inhibited (in a dose-dependent manner) the ability of Opa+ gonococci to adhere to neutrophils and stimulate neutrophil LDCL. Growth of Opa+ gonococci in 50 micrograms of CMP-NANA per ml appeared to delay, but did not inhibit, their killing by neutrophils. Sialidase treatment of sialylated Opa+ gonococci, i.e., gonococci grown with CMP-NANA, totally restored their abilities to adhere to neutrophils and stimulate neutrophil LDCL. Opa- gonococci grown in the presence of 50 micrograms of CMP-NANA per ml and opsonized with fresh human serum bound to neutrophils only about 30% less efficiently than did Opa- gonococci grown without CMP-NANA and opsonized. The results of our studies show that sialylated Opa+ gonococci have dramatically reduced nonopsonic interactions with neutrophils. Some gonococcal strains may resist killing by human neutrophils in vivo by such a mechanism.

Interactions of Neisseria gonorrhoeae with human neutrophils: studies with purified PII (Opa) outer membrane proteins and synthetic Opa peptides.

We investigated the role of gonococcal outer membrane protein PII (also called Opa protein) in nonopsonic adherence to human neutrophils. Gonococcal outer membranes, purified Opa in detergent (Opa), purified Opa in liposomes (Opa+ lips), and peptides composing the second hypervariable (HV2) region of OpaB (strain FA1090) in liposomes (pepHV2 lips) were tested for their abilities to inhibit subsequent gonococcal adherence to human neutrophils. Outer membranes from gonococci possessing adherent Opa, liposomes containing adherent Opa, purified adherent Opa, and two of three liposome preparations (pepHV2 lips) containing peptides from the HV2 region of an adherent Opa inhibited subsequent adherence to neutrophils of homologous Opa+ gonococci. On the other hand, outer membranes from Opa- gonococci, outer membranes containing a nonadherent Opa (OpaA from strain FA1090), purified OpaA, and OpaA lips had little or no inhibitory effect. Outer membranes containing adherent Opas, purified adherent Opas, and liposomes containing such Opas all bound to neutrophils, whereas preparations containing OpaA or no Opa protein did not. The results indicate that (i) Opa proteins can bind to neutrophils in a partially purified or purified form and (ii) the HV2 region of Opa appears to at least partially mediate Opa's biological role.

Stimulation of human neutrophil oxidative metabolism by nonopsonized Neisseria gonorrhoeae.

Nonopsonized gonococci possessing opacity-associated (Opa; previously PII) outer membrane proteins stimulate neutrophils to undergo a vigorous oxidative response when measured by luminol-dependent chemiluminescence (LDCL). In these studies, we characterized the mechanism of this stimulation. No gonococci that we tested induced measurable release of neutrophil superoxide anion (O2-) or hydrogen peroxide (H2O2) as measured by reduction of cytochrome c or the oxidation of scopoletin, respectively. Neutrophils pretreated with gonococci and then exposed to phorbol myristate acetate, the chemotactic peptide formylmethionylleucylphenylalanine, or opsonized zymosan released levels of neutrophil O2- and H2O2 comparable to controls, indicating that gonococci were not preventing or inhibiting neutrophil O2- or H2O2 release. To ascertain a possible explanation for these seemingly contradictory observations (i.e., induction of LDCL, but no release of O2- or H2O2), we further characterized the ability of Opa+ gonococci to stimulate LDCL. By using 1 mM azide and 4 U of horseradish peroxidase to monitor extracellular LDCL selectively and 2,000 U of catalase to monitor intracellular LDCL selectively, we determined that greater than 80% of total gonococcus-induced neutrophil LDCL occurred intracellularly. In addition, neutrophils stimulated with Opa+ gonococci showed a marked increase in O2 uptake and hexose monophosphate shunt activity. We conclude that Neisseria gonorrhoeae induces neutrophil oxidative metabolism without causing release of detectable amounts of reactive oxygen intermediates into the surrounding milieu. The gonococcus apparently directs oxidase assembly and activity to the phagolysosomal membrane. This could be a mechanism by which extracellular gonococci persist for extended periods in vivo in the presence of high concentrations of neutrophils.