Search

A broad panel of agents including serum, interleukin-1, double-stranded RNA, and platelet-derived growth factor (PDGF) stimulate transcription of the "slow" immediate-early gene MCP-1. These disparate inducers act through a tight cluster of regulatory elements in the distal 5'-flanking sequences of the MCP-1 gene. We describe a 22-base element in this cluster which, in single copy, confers PDGF-inducibility to a tagged MCP-1 reporter gene. In mobility shift assays, the element binds a PDGF-activated form of NF-kappaB, and a 90-kDa protein (p90) which binds constitutively. Antibody supershift and UV cross-linking experiments indicate that the PDGF-activated NF-kappaB species is a Rel A homodimer. The DNA binding form of p90 is a nuclear-restricted serine/threonine phosphoprotein. Mutagenesis of the 22-base element shows that the NF-kappaB and p90 binding sites overlap, but binding of the two species is mutually independent. Both sites, however, are required for optimum PDGF induction of MCP-1. Therefore, p90 appears to be a coactivator with NF-kappaB in PDGF-mediated induction of MCP-1.

We have developed an assay that can detect relative changes in the amount of a non-cAMP inhibitor of maturation present in cumulus cells (Eppig et al., 1983, Dev. Biol., 100:39-49). Using this assay in which accelerated maturation of a group of treated cumulus cell-oocyte complexes relative to untreated complexes indicates a decrease in the amount of inhibitor, results of the experiments described here suggest a possible relationship between elevation of cAMP levels and subsequent decreased amounts of a non-cAMP inhibitor. Mouse oocytes obtained from cumulus cell-oocyte complexes treated with luteinizing hormone (LH) resumed meiosis prior to oocytes obtained from untreated complexes; the degree of acceleration of maturation was dependent on LH concentration. A similar result was obtained with follicle-stimulating hormone (FSH). Correlated with LH- or FSH-acceleration of maturation was an LH- or FSH-induced elevation of cumulus cell cAMP levels. Inhibiting LH-induced elevation of cumulus cell cAMP levels inhibited LH-induced acceleration of maturation. An initial incubation of complexes in medium containing dibutyryl cAMP (dbcAMP) also promoted acceleration of maturation. In contrast, maturation of denuded oocytes was not altered by treatment with either LH, FSH, or dbcAMP. Complexes initially incubated in dbcAMP-containing medium still demonstrated acceleration of maturation after a subsequent 2 h incubation in dbcAMP-free medium. Relative to untreated complexes, none of these treatments disrupted intercellular communication between cumulus cells and the oocyte. Elevating follicle cAMP levels with cholera toxin induced maturation of follicle-enclosed oocytes when cumulus cell-oocyte coupling was still fully maintained. These results are interpreted to indicate that gonadotropin-mediated acceleration of maturation is via a cAMP-dependent reduction in the level of a maturation inhibitor present in granulosa/cumulus cells.

Aminoacylation of tRNA occurs with a high degree of accuracy in many cases because misacylated molecules are effectively proofread on the aminoacyl-tRNA synthetase by preferential hydrolysis. This hydrolysis releases an ATP equivalent of energy. An explicit relationship between cost of proofreading and the resulting degree of accuracy is presented. Experimental data from Escherichia coli for isoleucyl-tRNA synthetase proofreading valyl-tRNAIle are examined by means of this relationship, and a conjecture concerning the natural selection of accuracy and proofreading costs is put forth and tested. We have found the energy cost of accurate proofreading to be high. The minimum error, derived in previous theoretical studies, is never actually reached. Instead, higher values, determined by the balance between energy wasted in the cell as a consequence of error and the energy cost of proofreading, appear to be selected. The total cost of proofreading all types of tRNA aminoacylations is estimated to be approximately 2% of the energy required to synthesize a bacterial cell.

Two companion papers in this series have characterized the interaction between Vibrio cholerae and the surfaces of eukaryotic cells. The present paper reports studies of the association between vibrios or Salmonella enteritidis and intact slices of intestinal tissue. A significant number of differences were noted in the characteristics of bacterial adhesion in these systems. The results are interpreted to indicate the presence of at least two receptors for vibrio adhesion on the mucosal surface of the rabbit small intestine. The receptor mediating the adhesion of salmonella appeared to be distinct from these. A primary role for bacterial motility in the process of adhesion of vibrios to mucosal surfaces could not be demonstrated in the assay systems studied. Rather, loss of motility in mutant vibrios appeared to be correlated with the simultaneous loss of adhesive factors (adhesins) from the bacterial surface. The inhibition of vibrio adhesion to slices of intestinal tissues by antibody to the heat-stable antigens of V. cholerae occurred in the absence of bacterial agglutination. Agglutination in this assay system appeared to be an artifact in that it could be observed only in experiments where extremely high concentrations of vibrios were used. We speculate that such high vibrio concentrations are not likely to be present in humans at the time of infection and that agglutination in the lumen of the intestine might therefore play only a minor role in prophylactic immunity against natural cholera and other enteric infections of humans.

Nonmotile vibrio mutants lacked the ability to adhere to rabbit intestinal brush border membranes and to agglutinate human group O erythrocytes, but motile revertant vibrios isolated from such strains expressed adhesiveness equivalent to that of the original parent. Two possible explanations for the relation between vibrio motility and adhesion in these assays systems are (i) that the rate of adhesion depends on the rate of chance contact brought about by motility, and (ii) that the flagellum either acts as a carrier for the bacterial adhesin or may itself be the adhesin. The present study indicates, however, that the lack of adhesion by nonmotile vibrios did not depend on motility as such and did not involve greater rates of elution. Increasing the rate of contact between nonmotile vibrio mutants and brush border membranes by compaction did not restore the adhesive properties of the defective strains. Accordingly, we speculate that the flagellum may function in some indirect way that allows the expression of the adhesive properties, such as by acting as a carrier for a specific vibrio adhesin. Adhesion to brush borders and agglutination of human group O erythrocytes was specifically inhibited by L-fucose and various glycosides of L-fucose and to a lesser extent by D-mannose. Vibrios adhered specifically to agarose beads that carried covalently linked L-fucose on their surfaces. The results suggest that L-fucose-containing structures of eukaryotic cell surfaces may function as receptors for the vibrio adhesin and may therefore be an important determinant of host susceptibility.

Adhesion of vibrios to the small intestine may occur (i) by association of the bacteria with secreted mucus gel or (ii) by adherence of the bacteria to the surface of epithelial cells. In the present study, vibrios readily adhered to isolated brush border membranes obtained from rabbit intestinal epithelial cells. Adhesion was temperature dependent and required the presence of divalent cations such as calcium. The agglutination of human O erythrocytes by Vibrio cholerae was observed also, and the hemagglutination test appeared to detect the same mechanism that was involved in the adhesion of vibrios to brush borders. When the bacteria were grown in broth they were adhesive and hemagglutinating, but vibrios grown on agar plates or suspended in buffer for 15 min at 37 C lacked these abilities, even though they retained undiminished motility. These two model systems differed, however, in that strontium promoted only adhesion to brush borders. The significance of this difference remains to be determined. Vibrios were observed to penetrate intestinal mucus gel and occasionally to become entrapped in it. However, there was no evidence that vibrios attached to mucus gel.

The function of indigenous lactobacilli in the control of other intestinal microbial species is not clear. Still more controversial is the effect of dietary bacterial supplements containing lactobacilli or other species. This situation is unlikely to change unless the mechanisms that control the colonization of ingested bacteria are better understood, and until more detailed information becomes available on the mechanisms by which certain populations of indigenous bacteria can affect the population sizes of other species. We used gnotobiotic mice and a continuous-flow culture system to study the interactions between Escherichia coli and (i) clostridia (in chloroform-treated cecal suspensions from conventional mice) and (ii) three strains of lactobacilli isolated from conventional mice. In gnotobiotic mice, the lactobacilli suppressed E. coli multiplication in the stomach and the small intestine, but had no demonstrable effect on E. coli multiplication in the large intestine. In contrast, clostridia were most effective in controlling E. coli multiplication in the large intestine. In the presence of both lactobacilli and clostridia, E. coli populations in the various regions of the gastrointestinal tract resembled those found in conventionalized control animals. The control of E. coli populations was not related to changes in pH or intestinal motility. In vitro stimulation of the above-described in vivo interactions required a two-stage continuous-flow culture in which the effluent from the first stage represented the influx to the second. The first stage was inoculated with lactobacilli, and the second stage was inoculated with either a pure culture of E. coli or E. coli and clostridia. In these instances, the E. coli populations in the second stage of the culture resembled in size those found in the large intestine of gnotobiotic mice harboring a similar flora. Although there are some current shortcomings of this in vitro model, we expect that a multistage continuous-flow culture can be developed to satisfactorily model the interactions among bacterial populations along the entire gastrointestinal tract.

Various Sephadex G-15 fractions of pepsin-digested mucosal extract inhibited the in vitro association of cholera vibrios with mucosal slices. Inhibitory activity paralleled the taxin activity of the fractions for these bacteria. This supports the theory that inhibition of mucosal association by pepsin-digested mucosal scrapings was due to the blocking of taxin receptors on the bacterial surface. Nonchemotactic mutants were significantly less efficient than the chemotactic parent or revertant strains in associating with mucosal slices in vitro. Control experiments in which filter paper disks replaced the mucosal slices showed a comparable extent of association of chemotactic and nonchemotactic vibrios with this material. Histological studies indicated that vibrios associated predominantly with the mucus gel of the intestinal slices rather than with the mucosal epithelium or the serosal surface. Intestinal slices attracted chemotactic vibrios even after prolonged washing, suggesting continuous production of the taxin by the tissue. Inert polystyrene particles 1.1 micrometers in diameter penetrated the mucus gel of intestinal slices very poorly, but nevertheless could be detected in low numbers in the deep intervillous spaces within 15 min. In contrast, chemotactic vibrios reached the deep intervillous spaces in significantly higher numbers, whereas motile, non-chemotactic vibrios reacted like the inert particles. It is concluded that mucus gel represents a partial barrier to the penetration of particles of bacterial size and that this barrier can be invaded efficiently by motile bacteria, but only when their locomotion is guided by chemotactic stimuli.

Contrary to earlier findings with all other in vivo and in vitro models of cholera studied, nonchemotactic vibrio mutants showed a relatively greater fitness in 5-day-old infant mice as compared with chemotactic parent or chemotactic revertant strains. This trend was manifest in the relatively greater number of nonchemotactic mutants recovered from the upper small intestine at 4 and 18 h after intragastric infection. The same trend was also revealed in the significantly greater virulence (in terms of time to death) of nonchemotactic mutants as compared with the chemotactic parent or revertant strains. Histological studies in infant mice of the penetration of chemotactic and nonchemotactic vibrios into the mucus gel of the small intestine yielded the same findings as in all other models studied, i.e., significantly greater penetration by chemotactic vibrios. There was no correlation between the relative fitness of nonchemotactic vibrios in the small intestine of infant mice and the rate of recovery of viable nonchemotactic vibrios from that site. In contrast, excellent correlation was found between the relative fitness of nonchemotactic vibrios and a decrease in the recovery of viable cells of the chemotactic strain from the small intestine. This indicates that the relatively greater fitness of the nonchemotactic vibrios in infant mice was only apparent and that the observed phenomenon was actually due to an antibacterial mechanism which prevented the accumulation of the chemotactic strains in the small intestine rather than to any stimulating effect on the nonchemotactic mutant itself. To study the in vivo fate of the inoculum in infant mice, vibrios were labeled with either 32P, 35S, or [3H]thymidine. Specific activity determinations of the 32P label were compatible with the assumption of an accelerated rate of death of the chemotactic parent strain in the small intestine. Results with the other isotopes, however, were significantly different. Indeed, the amount of radioactivity retained in the small intestine after feeding labeled bacteria correlated more closely with the isotope used than with the strain of vibrio under study. Consequently, considerable doubt must be cast on the general validity of this not uncommon technique for determining the in vivo location and the death or survival of radioactively labeled bacteria.

A motile, chemotactic, Ogawa strain of Vibrio cholerae was attracted by all 20 L-amino acids tested, in contrast to Escherichia coli AW 405, which did not react to several of these. The maximum number of vibrios entering a capillary was much lower when the capillary contained carbohydrates rather than amino acids, but the minimum effective concentrations of the carbohydrates and amino acids tested were of the same order of magnitude. L-Fucose, a sugar known to inhibit the adhesion of this vibrio strain to brush border membranes, had no attraction (taxin activity) for it. A pepsin digest of rabbit mucosal scrapings or tryptone attracted vibrios as strongly as the most active amino acids. Several nonchemotactic and one nonmotile mutant were selected from the parent vibrio. The nonchemotactic mutants were indistinguishable from the parent in their ability to attach in vitro to isolated intestinal brush border membranes, whereas the nonmotile mutant had lost this ability. Parent and nonchemotactic mutants had equal growth rates in stirred and still continuous flow cultures that were maintained in an anaerobic environment.

A model for ascending unobstructed urinary tract infection was developed in mice to study the pathogenesis of urinary tract infection induced by Escherichia coli associated with urinary tract infection in humans. Specifically, the model was designed to monitor the initial stages of the infectious process, e.g., bacterial adhesion. Mice were selected since the specificity and intensity of bacterial attachment of pyelonephritogenic E. coli strains to human and mouse uroepithelial cells were similar. Female mice were infected by urethral catheterization and installation of bacteria in the urinary bladder. To maximize clearance of unattached bacteria, no obstructive manipulations were performed. After sacrifice, the persistence of bacteria in kidneys and bladder was determined by viable counts on homogenized tissues. The experimental infection was standardized by using one pyelonephritis (HU734) and one normal fecal (414) E. coli isolate. With both strains all of the bladders became infected, but E. coli 414 was eliminated more rapidly than HU734. The percentage of positive kidney cultures increased with the bacterial inoculum concentration and volume. An inoculum of 0.05 ml containing 10(10) bacteria per ml was selected, giving the highest percentage of positive kidney cultures without detectable bacterial spread to the blood stream. The variation in the percentage of positive kidney cultures possibly depended on the degree of vesicoureteric reflux in the individual animals. Both in the kidneys and in the urinary bladders, strain HU734 yielded higher numbers of bacteria at 24 h and persisted longer than did strain 414. Several E. coli pyelonephritis isolates with properties associated with virulence in the human urinary tract consistently were recovered from mouse kidneys and bladders in higher numbers than E. coli strains of human fecal origin lacking those properties. The role of bacterial adhesion per se is the topic of the accompanying paper.

The affinity of uropathogenic Escherichia colito kidneys and bladders of experimentally infected mice was shown to be determined in part by the adhesive properties of the infecting bacteria. Mice were infected with various pairwise combinations of two homogeneic sets of bacteria: (i) mutants derived from a human pyelonephritis E. coliisolate which were selected to express either or both adhesins specific for globoseries glycolipid receptors or for “mannosides”; and (ii) transformants of a normal fecal isolate which harbored recombinant plasmids encoding the genes for one or the other adhesin or which harbored only the vector plasmid. The relative efficiency of survival of the strains to be compared was evaluated in each animal by plating on selective media of samples of homogenized kidneys and bladders taken 24 h after intravesical inoculation. The presence of adhesins specific for globoseries glycolipid receptors, which mediate the in vitro mannose-resistant attachment to human and mouse uroepithelial cells, enhanced bacterial recovery from both kidneys and bladders of infected animals. The addition to the infecting strain of adhesins binding mannoside residues further improved bacterial recovery from the bladder, but not from the kidney. The mutants and transformants with adhesins binding only mannosides were recovered in higher numbers from the bladder than those expressing adhesins specific for the globoseries glycolipids only. There was apparent selection in vivo decreasing expression of mannoside binding adhesins in the kidneys, but not in the bladders, of animals infected with the mutant expressing both types of adhesins. Regardless of adhesive properties, the mutants of the pyelonephritis isolate were recovered in significantly higher numbers than the fecal isolate with adhesins encoded on recombinant plasmids. We conclude that the adhesive properties in part determine the localization and retention of bacteria in the mouse urinary tract. However, the addition of adhesins to a commensal E. colistrain was not sufficient to confer colonization capacity comparable to that of a pyelonephritis strain.

An anaerobic continuous-flow (CF) culture method has been developed which reproduces a number of bacterial interactions that occur in the large intestine of mice. These were determined in the following ways. (i) Bacterial counts in smears stained with 37 specific fluorescent antisera showed that the numeric balance between 37 strict anaerobes isolated from conventional mice was maintained in CF culture of conventional mouse flora in the same manner as in conventional mice. (ii) Mixed populations of various complexity of bacteria isolated from conventional mice were able to suppress Escherichia coli populations to similar levels in gnotobiotic mice and in CF cultures. (iii) Contents of CF cultures when fed to germfree mice were found to redress the germfree abnormalities studied, namely, cecal size and size of the E. coli population. Furthermore, dense layers of bacterial growth formed on the wall of CF cultures of mouse cecal flora, in a manner analogous to the colonization of mouse large intestinal mucosa. In the absence of such bacterial layers, the culture no longer exhibited these interactions. Because of the complexity and diversity of the interactions studied it is highly probable that at least the major underlying ecological control mechanisms operating in the culture model resemble those of the mouse intestine. We speculate that the somewhat surprising similarity between the ecology of the mouse large intestine and that of a CF culture in a glass vessel is due to the fact that both are dominated by thick layers of complex bacterial flora, the composition of which is controlled by their metabolic activities and by their relative ability to adhere to each other.

We studied the interactions between the entire cecal flora of hamsters and the pathogens Clostridium difficile and Escherichia coli in gnotobiotic mice and in a continuous-flow (CF) culture system in which the growth medium consisted of an extract of fecal pellets from germfree mice. CF cultures and germfree mice were colonized first with C. difficile and E. coli and then with the cecal flora of hamsters. Both in vivo and in vitro hamster flora markedly suppressed the potential pathogens. Contents of CF cultures inoculated with hamster flora were introduced into gnotobiotic mice previously colonized with C. difficile and E. coli. These mice were compared with mice given homogenates of hamster ceca. In both groups, the C. difficile population decreased by a factor of more than 10(6) and the E. coli population decreased by a factor of 10(4) to 10(5). CF culture contents also reduced the size of the dilated germfree mouse cecum to normal. When veal infusion broth was used as a medium, contents of CF cultures colonized with hamster flora failed to eliminate C. difficile from mice. Thus, the extract of fecal pellets appeared to contain a substance important for sustained colonization by important components of the cecal flora. We also studied the ability of collections of isolates to suppress the potential pathogens in both gnotobiotic mice and CF cultures. A total of 150 isolates obtained from predominant hamster flora at the ecologic climax stage (C flora) suppressed C. difficile and E. coli to 10 and 1 to 3%, respectively, of the population sizes attained in monoassociated mice. A total of 67 isolates obtained during ecologic succession combined with a C flora consisting of 100 isolates suppressed the potential pathogens to 0.3 and 0.03% of their original levels, respectively. Similar degrees of suppression occurred in CF cultures, further indicating that anaerobic CF cultures are promising models for investigation of the microbial ecology of C. difficile.

This paper describes studies of the mechanisms by which intestinal antibody may be delivered to the lumen of the intestine. The results indicate that in all animals studied (conventional mice, conventionalized mice, and germfree mice), antibody may appear in the intestine as a result of local synthesis or may be derived from serum antibody. Serum antibody was degraded in the intestine of all types of mice studied, but degradation in germfree mice was so rapid that no serum antibody was detectable in the intestine. The presence, and probably the quality, of intestinal flora was therefore shown to affect the host by a new mechanism: the sparing of intestinal antibody. High antibody titers in the intestine of orally vaccinated germfree mice were due to secretory IgA antibodies that appeared to be more resistant to intestinal degradation than the IgG and IgM fractions derived from the serum.Purified intestinal IgA antibody was shown to protect in vivo isolated intestinal loops of mice against experimental cholera. The implications of these findings for the problems of prophylactic immunity to bacterial enteric pathogens and of immunological control of normal enteric bacterial flora were disscussed.

Intestinal antibody (coproantibody) significantly reduced the adsorption of heat-killed Vibrio choleraeto the mucosa of in vivo isolated ileal loops of adult rabbits. This suggests a direct effect of coproantibody on adsorption, which appears to function in addition to the antibacterial mechanism described earlier. When antivibrio serum was administered passively into intestinal loops, it showed a predilection for the intestinal mucosa. In vivo adsorption of vibrios appeared to parallel their viability, i.e., vibrios killed by heat or in the presence of neomycin adsorbed significantly less than live vibrios. In contrast, in vivo adsorption was only slightly affected in the presence of bacteriostatic concentrations of tetracycline. Adsorption of Salmonella senftenbergand V. choleraeto slices of rabbit ileum in Krebs-Ringer solution appeared to involve different mechanisms, in that the former was strongly removed from the intestinal tissues by sodium lauryl sulfate, whereas vibrios were not affected by this agent. Triton X-100 increased the adsorption of vibrios, whereas rabbit bile and changes in pH had no effect.