Fifty years of dysentery research at the Pécs University.

Károly Rauss was appointed as head of the Department of Public Health of the Elisabeth University of Pécs in 1946, Professor Rauss's carrier had started working with Professor Hugó Preisz in Budapest. During his residency years he was already appointed to the Department of Bacteriology chaired by Lovrekovics at the National Institute of Public Health. In this institution--as in all organizations affiliated with the Rockefeller Institute--the state of art diagnostic work together with research focusing on problems derived from everyday medical and public health practice was considered as to be of primary importance. Stimulated by this scientific environment Rauss's interest turned towards enteric pathogens. In cooperation with Lovrekovics he developed a typhoid vaccine containing a trichloracetic acid extract of the pathogen adsorbed to aluminum hydroxide. This vaccine was introduced in 1938 when ca. 6-8000 enteric fever cases were registered in Hungary annually. The vaccination, supported by the public health work concerning carriers, eventually lead to the eradication of the disease in Hungary.

A model for testing drug susceptibility of Pseudomonas aeruginosa and Staphylococcus aureus grown in biofilms on medical devices.

UV-sterilized polyethylene rings infected with Pseudomonas aeruginosa or with Staphylococcus aureus were implanted into artificial wounds of mice. During 5 days incubation biofilm was formed on the plastic surfaces. To determine the Minimal Bactericidal Dose on sessile bacteria, rings with biofilm were removed and incubated in Luria broth containing serial dilutions of different antibiotics. Parallel, the sensitivities of planktonic phase organisms were also determined using cells grown in broth. The biofilm mode of growth strongly reduced the sensitivity of the strains against most of the antibiotics used, especially against polymyxin B. On the other hand, beta-lactam type antibiotics were equally effective against bacteria both in the sessile and planktonic phase of growth.

Biofilms produced by Pseudomonas aeruginosa and by Staphylococcus aureus on model medical devices.

Polyethylene, teflon, tygon, polypropylene, silicon rubber, and rubber tubes or rings contaminated with Pseudomonas aeruginosa or Staphylococcus aureus were implanted subcutaneously into mice. After 5 days the colony forming units developing on, and attaching to them were determined. The highest numbers of bacteria were observed on rubber and silicon rubber, polyethylene was next in order, while significantly lower values were obtained on teflon and on tygon and polypropylene. Rubber devices were better colonized after heat than after UV sterilization. The number of bacteria rose further, if the already used rubber implant was resterilized, recontaminated and reimplanted. The model seems suitable to test the development of bacterial biofilms on different materials pretreated in different manners.

Effectiveness of antibiotics on the autochthonous Escherichia coli of mice in the intestinal biofilm versus its planktonic phase.

The effectiveness of antibiotics was tested on the autochthonous Escherichia coli in biofilm mode of growth in large bowel pieces as well as on the predominant faecal E. coli isolated from the same SPF mice in planktonic phase of growth. Aminoglycoside antibiotics, chloramphenicol, oxytetracycline, erythromycin and lincomycin-clindamycin treatment had a very limited effect in the intestinal biofilm. Surprising ineffectiveness was found with polymyxins: polymyxin B showed a Minimal Bactericidal Dose of 0.78 microgram in planktonic phase, while it was 400 micrograms for E. coli incorporated in the biofilm matrix. In contrast to the above groups of antibiotics, the beta-lactam drugs were effective both in the biofilm and in the planktonic phase growth of E. coli and their derivatives with broad or broader spectrum exerted an increased biofilm activity. Polymyxin B showed no sign of penetration into the colonic mucus, but on the other side ampicillin concentrated about three-four times in the intestinal matrix.

[Effect of antibiotics on autochtonous intestinal E. coli population in SPF mice and on planktonic phase isolates]. / Antibiotikumok hatékonysága SPF egerek autochton Escherichia coli intestinális populációjára és planktonikus fázisú izolátumaira.

Representatives of different groups of antibiotics were tested on the autochtonous Escherichia coli of mice being in biofilm mode of growth, as well as on their isolates in planktonic phase. Aminoglycoside antibiotics, chloramphenicol, oxytetracycline, erythromycin, lincomycin-clindamycin, and polymyxins showed a very reduced effectivity on Escherichia coli embedded in the biofilm matrix. On the other hand, beta-lactam antibiotics were equally effective both for sessile and planktonic bacteria. Derivates with broader, or broad spectra showed an increased biofilm activity. Testing the possibilities of penetration or concentration on mouse colonic pieces exposed to 2 Minimal Bactericidal Doses--polymyxin B showed no sign of penetration into the colonic mucus, while 2.9% of streptomycin and 60% of the exposed quantity of carbenicillin was detected in the intestinal mucus. Ampicillin, however, has concentrated about three-four times in the intestinal matrix.

Shigella-type pathomechanism in the "mouse model".

A strain of Escherichia coli K-12 carrying the 140-Megadalton virulence plasmid of the enteroinvasive E. coli--J53(pSPl)--showed high virulence in the "mouse model", in chick embryos, but not in the Serény test. It expresses the outer membrane proteins thermoregulatedly, encoded also by the virulence plasmid. In orally infected streptomycin-pretreated mice this strain infects only the large bowel, shows adherence to the epithelial surface, but in its first step preferentially to the mucus excreted by the goblet cells. Epithelial penetration and intracellular multiplication occurs with a characteristic localization of bacteria in the depth of crypts. Consequence of the infection is degeneration of the epithelial surface, its denudation.

Resistance of Escherichia coli to some antibiotics and biocides in the intestinal biofilm of mice.

An artificial monoflora of Escherichia coli in mice, as well as their autochtonous E. coli, exhibited enhanced resistance to streptomycin, chloramphenicol, sodium hypochlorite and silver nitrate. The level of resistance of the monoflora, which was 10-32 times higher than the in vitro determined Minimal Bactericidal Dose (MBCD), reached its maximum on the 7th-9th day after implantation. This latency is a requirement for the stabilization of the monoflora. Formaldehyde and carbenicillin were equally effective in the planktonic and in the biofilm mode of growth. In the case of carbenicillin the pieces of mouse colon contained about 60% of the dose used for exposure, in contrast to the 3% rate of streptomycin, showing the excellent penetrating ability of carbenicillin into the intestinal biofilm.

Binding of enteric bacteria to hog gastric mucin.

The binding features of enteric bacteria were studied using a model mucin of hog gastric origin. The time requirement of binding is short, it is temperature-independent, but dose-dependent. The binding effectiveness of Escherichia coli, Shigella sonnei and Shigella flexneri, as well as Salmonella minnesota had a narrow range: 1.5-9 germs pro pg of mucin. The bacterial ligand of the binding is certainly not a polysaccharide as proved by the uniform binding of the R-mutant series of S. sonnei and S. minnesota. On the basis of inhibition tests by an outer membrane protein fraction, the ligand may be a common outer membrane protein of the enteric bacteria. The outer membrane proteins encoded by the Shigella-EIEC invasivity plasmids do not take part in this binding. The inhibition by killed bacteria or by their culture supernatants of mucin binding of heterologous species may suggest a non-species specific common ligand, too. Similarly to the mucin utilization, the binding ability also seems to be a general phenomenon among the enteric bacteria.

[Bacterial biofilms and medical practice]. / Bakteriális biofilmek és az orvosi gyakorlat.

A suboptimal environment with limited supply of nutrients leads to marked changes in the microbial features. Similar conditions exist also on some tissue surfaces of the human body and on biomedical devices implanted. The basic feature of this environmental adaptation is the adherence of microbes to some surfaces and production of abundant exopolysaccharides. Microbes embedded in this hydrated, mostly anionic polysaccharide matrix called biofilm. The biofilm acts not only by trapping the nutrient or oxygen molecules, but protects the microbes against phagocytic cells, antibodies, biocids, and antibiotics, too. This protected form is responsible for the special nature of normal flora, as well as for some clinical and therapeutic characters of several diseases. Microbes liberated from the biofilm, being in the so called planctonic phase. In cases where the biofilms have developed on biomedical devices such planctonic microbes may cause persistent or relapsing infections. These planctonic microbes can be eliminated by the host responses or by antibiotic treatments, while those in the biofilm cannot, thus serving as a source of further infections.

Epidemiology of the enteroinvasive Escherichia coli. Observations in Hungary.

The Hungarian experiences on the epidemiology of enteroinvasive Escherichia coli (EIEC) concerning the dominant serogroup 0124 are summarized. One of the basic differences between Shigella and EIEC infections may be attributed to the higher environmental resistance of EIEC, therefore first of all water-borne outbreaks may be frequent. The other essential difference is in te age incidence: the infection rate of infants is low, the rise of incidence rate begins at the 3rd year and its peak is in the school-children age. EIEC, like to other nosological units of E. coli is not a zoonosis. Symptomless carriership is frequent and may be long lasting with the excretion of the virulent agent for over 1 year.

Biofilm model experiments in vitro with Escherichia coli and Shigella strains.

Escherichia coli and Shigella were cultured together in a glucose-free minimal medium completed only by hog gastric mucin. In mixed cultures inoculated simultaneously one member of the pair more or less overgrew the other. This organism remained predominant when superinfected with its strain pair after 1, 3, 5 or 7 days incubation. Other signs of the quasi biofilm character of such cultures are the higher level of free polysaccharides and enhanced viscosity in the medium and higher streptomycin resistance of the culture. Around the bacteria a capsule-like material can morphologically be demonstrated.

Feeding by mucin and intestinal growth of some enteric bacterial pathogens.

Enteroinvasive Escherichia coli, Salmonella typhi-murium, Shigella sonnei, Shigella flexneri, as well as E. coli K-12 show dose dependent growth in minimal medium completed with purified hog gastric "Granular Mucin". This ability is based on alpha-galactosidase production: defective, melibiose (and galactose) non-fermenting K-12 mutant were unable to utilize mucin. The viability of the parent K-12 strain in the cecal content of mice is significantly higher than that of its Mel- mutant phenotype. In mixed infections of mice the parent strain was the only one to be able to establish a monoflora against its Mel- or Gal- mutants. Among other mechanisms, the growing ability in the intestinal mucous layer may be an additional virulence factor when the enteric pathogens are exposed to a competitive antagonism of the normal flora.

Protective value of the plasmid-coded outer membrane protein of enteroinvasive Escherichia coli.

O-antigenically not related enteroinvasive Escherichia coli strains and rabbit sera prepared with them were used to study the role of plasmid-coded outer membrane proteins in protective immunity. Active immunization experiments were performed using a mouse model based on a long-lasting symptomless carriership after elimination of the bowel flora by streptomycin. Preliminary histological studies showed adhesion, penetration, intraepithelial multiplication, and epithelial desquamation after infection. In active immunization experiments only massive oral doses evoked protective immunity. Seroconversion against the plasmid-coded antigens was not observed in mice. Passive immunization was carried out in chick embryos with unabsorbed sera. A high level of protectivity was reached by serogroup-specific sera and a very low but significant protection was yielded by antibodies against the plasmid-coded protein antigens.

Urinary tract infectivity or R strains of Escherichia coli carrying various virulence factors.

The virulence factors of Escherichia coli supposed to act in urinary tract infections were studied on R strains in a suckling mouse model. The production of alpha-(diffusible-) haemolysin or the possession of antigen K1 enhanced the virulence significantly, while the type 1 (common) fimbriae failed to do so. An isogenic motile and non-motile pair of E. coli did not show any difference in infectivity in the model. The adhesins, the diffusible haemolysin, and the acidic polysaccharide K antigens (K1) are definitely additive virulence factors in the model. This is in good agreement with the experience of clinical bacteriology.

Immunological approach to the pathogenetical role of Escherichia coli adhesive factor "119" in a suckling mouse urinary tract model.

In the course of urinary tract infections, suckling mice with maternal anti-pilus ("119") immunity showed a massive protection against a 119+ strain of Escherichia coli. Animals could be protected against urinary tract infection by giving pilus antibody or pilus vaccine shortly after the infection. Results showed the importance of adhesive pili in initiating the urinary tract infection by E. coli.

Suckling mouse model of urinary tract infections caused by Escherichia coli.

The role in the vesico-renal pathogenicity of Escherichia coli of mannose resistant fimbriae provisionally designated "119" has been studied in a suckling mouse model. Two to 3 days old mice were infected by route of the urinary bladder and the occurrence and character of the developing infection were investigated. (i) The fimbriated strain caused infection and inflammation of the bladder in almost 100%. In about 60% of the cases it lasted till the end of observations, the 35-37th day following inoculation. The washed, vortexed urinary bladder taken up in 5ml of saline contained a high number (10(6)-10(9)/ml) of bacteria, the urine was often cloudy, containing numerous leucocytes and epithelial cells packed with bacteria. The kidneys were soon involved and by the 23th ad 37th day gross changes were seen in about 25%. In 10%, a classical contracted kidney developed. (ii) The nonfimbriated derivate also caused acute or chronic urinary tract infection, but with a lower frequency and the number of bacteria was also lower by 3-4 log 10 exponents. The model seems to be promising for the study of bacterial infections of the urinary tract and perhaps also for elucidation of its immunological and therapeutic aspects.

Estimation of Vibrio cholerae and Escherichia coli heat-labile enterotoxin by enzyme-linked immunosorbent assay (ELISA).

By Enzyme-Linked Immunosorbent Assay (ELISA) purified cholera and Escherichia coli enterotoxins can be detected as sensitively as by CHO cells. There is a linear relationship between toxin concentrations and extinction values. In plates sensitized with toxins, antitoxins can be titrated with high sensitivity. ELISA cross-titration experiments demonstrate the existing antigenic relationship between cholera toxin and heat labile E. coli enterotoxin. Plates sensitized with either anti-E, coli-IgG or anti-cholera-IgG are suitable for detecting both cholera toxin, and E. coli LT. ELISA seems to be a simple, sensitive and economic method for quantitation of enterotoxins and toxin-specific antibodies.