Four biochemical tests for identification of probable enteroinvasive Escherichia coli strains

Enteroinvasive Escherichia coli (EIEC) share important features with Shigella spp., but EIEC strains are difficult to identify because their biochemical reactions are variable, and Sereny tests or other biological and molecular assays are expensive or hard to perform. The aim of this work was to detect probable enteroinvasive E. coli strains by using four biochemical tests, in children under 5 years of age with and without acute diarrhea. 330 strains of E. coli isolated from children with diarrhea, and 660 strains from children without diarrhea were studied. All strains were tested with the following tests: mucus , lysine and ornithine decarboxylase and motility. The strains which were negative to the four tests were tested by Sereny assay. Twelve strains (3.6%) isolated from children with diarrhea were negative to the tests proposed; eleven were lactose positive and only one was lactose negative. Three strains (0.5%) from children without diarrhea were negative to the tests proposed and were lactose positive. All the 15 strains (100%) were positive in Sereny assay. We recommend the use of these four biochemical tests for initial detection of EIEC strains, because their cost is very low and it is feasible carry out them in small diagnostic laboratories.

Characterisation of dermonecrotic toxin-producing strains of Pasteurella multocida subsp. multocida isolated from man and swine

Thirty-six isolates, from man or swine, of Pasteurella multocida subsp. multocida producing (n = 13) or not producing (n = 23) the dermonecrotic toxin (DNT) were studied by numerical analysis, capsular typing and ribotyping. Toxigenic strains were also characterised by restriction fragment length polymorphism (RFLP) of the toxA gene and pulsed-field gel electrophoresis (PFGE). Numerical analysis differentiated the Pasteurella species and subspecies, but did not discriminate between toxigenic and nontoxigenic strains. RFLP demonstrated that toxA was located in a conserved part of the chromosome of all toxigenic strains. Ribotyping provided evidence of a close association between DNT production and one of the six EcoRI ribotypes designated as E2. In contrast, PFGE provided evidence for significant DNA polymorphism amongst the toxigenic strains. Results of phenotypic and genotypic studies suggested that toxigenic strains do not form a clone within the subspecies multocida. No difference was found between toxigenic strains of porcine or human origin by biochemical characterisation, capsular serotyping or genomic typing methods.

Salmonella typhi VI antigen co-agglutination test for the rapid diagnosis of typhoid fever

A slide Co-agglutination test for the detection of Salmonella typhi Vi antigen in blood was evaluated for its efficiency in rapid diagnosis of Typhoid fever. The results were compared with conventional methods like Blood culture and Widal test. The test showed a sensitivity of 86.67% and specificity of 88.83% when compared with blood culture positivity or Widal titre above 160. This is a useful rapid diagnostic test for the early diagnosis of Typhoid fever.

Demonstration that Australian Pasteurella multocida isolates from sporadic outbreaks of porcine pneumonia are non-toxigenic (toxA-) and display heterogeneous DNA restriction endonuclease profiles compared with toxigenic isolates from herds with progressive atrophic rhinitis

Capsular types A and D of Pasteurella multocida cause economic losses in swine because of their association with progressive atrophic rhinitis (PAR) and enzootic pneumonia. There have been no studies comparing whole-cell DNA profiles of isolates associated with these two porcine respiratory diseases. Twenty-two isolates of P. multocida from diseased pigs in different geographic localities within Australia were characterised genotypically by restriction endonuclease analysis (REA) with the enzyme CfoI. Seven of 12 P. multocida isolates from nasal swabs from pigs in herds where PAR was either present or suspected displayed a capsular type D phenotype. These were shown to possess the toxA gene by polymerase chain reaction (PCR) and Southern hybridisation, and further substantiated by production of cytotoxin in vitro. The CfoI profile of one of these seven isolates, which was from the initial outbreak of PAR in Australia (in Western Australia, WA), was identical with profiles of all six other toxigenic isolates from sporadic episodes in New South Wales (NSW). The evidence suggests that the strain involved in the initial outbreak was responsible for the spread of PAR to the eastern states of Australia. Another 10 isolates, representing both capsular types A and D, were isolated exclusively from porcine lung lesions after sporadic outbreaks of enzootic pneumonia in NSW and WA. CfoI restriction endonuclease profiles of these isolates revealed considerable genomic heterogeneity. Furthermore, none of these possessed the toxA gene. This suggests that P. multocida strains with the toxA gene do not have a competitive survival advantage in the lower respiratory tract or that toxin production does not play a role in the pathology of pneumonic lesions, or both. REA with polyacrylamide gel electrophoresis and silver staining was found to be a practical and discriminatory tool for epidemiological tracing of P. multocida outbreaks associated with PAR or pneumonia in pigs.

Identification of two glycosylated components of Mycoplasma penetrans: a surface-exposed capsular polysaccharide and a glycolipid fraction

Among the wall-less mycoplasmas only a few species have been identified with a capsule at their cell surface. Mycoplasma penetrans is a recently identified mycoplasma with unique morphology, isolated from HIV-infected patients. Using transmission electron microscopy, it was found that M. penetrans is surrounded by capsular material 11 nm (strain GTU-54-6A1) to 30 nm (strain HF-2) thick, which can be stained with ruthenium red and labelled with cationized ferritin. The polysaccharide composition of this capsule was indicated by its staining with periodic acid-thiocarbohydrazide silver proteinate and the abolition of ruthenium red staining of the cell surface by neuraminidase treatment. In addition, proteinase K treatment of the M. penetrans cells resulted in removal of the capsule, suggesting that polypeptides may contribute in anchoring it to the membrane or in its stability. Two different types of glycosylated material were detected in mycoplasma extracts by SDS-PAGE and periodic acid-Schiff staining. The first component was a high-molecular-mass material, which was heat- and proteinase-K-labile and which probably constitutes the capsular polymer. The other component was a low-molecular-mass glycolipid fraction, which was proteinase-K-, heat- and EDTA-resistant. The identification of a capsule at the M. penetrans cell surface is of particular interest for a mycoplasma which has been shown to adhere to various host cells and to penetrate into their intracellular compartments. The capsule may have significance in the pathogenesis of disease associated with infection by this organism.

Re-evaluation of the serotypes of Serratia marcescens and separation into two schemes based on lipopolysaccharide (O) and capsular polysaccharide (K) antigens

Chemical and serological analysis has revealed that many of the 29 O serotype reference strains of Serratia marcescens contain both neutral and acidic polysaccharides which correspond to LPS O antigens and capsular K antigens, respectively. New O and K antigen typing schemes have therefore been devised, based on the known chemical structures of the surface polysaccharides of the organism. These schemes were designed to allow the specific detection of these antigens on unknown strains using ELISAs. O antigens were detected using whole cells cultured in broth then autoclaved to remove capsular material, while K antigens were detected using formolized whole cells which had been cultured on glycerol agar to enhance capsule production. After testing with the 29 reference strains as well as 423 distinct clinical strains, it was apparent that different aspects of chemical structure were associated with different degrees of serological reactivity and the typing schemes were modified further to accommodate this. In general, the O antigen repeating unit structures were chemically simple with di- or trisaccharide backbones. Serological specificity was often provided solely by the presence or absence of an O-acetyl substituent, or a change in the linkage between two sugar residues. Five of the O serotypes in the new scheme were represented by 12 of the 29 reference strains, while three reference strains lacked O antigens altogether, resulting in the elimination of 10 of the original O types. In contrast, the K antigen repeating unit structures were more complex and chemically diverse, having at least four sugar residues. Three K types were each seen in two reference strains while 12 of the 29 reference strains were acapsular. Thus, the resulting schemes contain 19 O types and 14 K types and allow the definitive serotype identification of S. marcescens.

Association of intrastrain phase variation in quantity of capsular polysaccharide and teichoic acid with the virulence of Streptococcus pneumoniae

The pneumococcus undergoes spontaneous phase variation between an opaque and a transparent colony form. In an animal model of systemic infection following intraperitoneal inoculation of mice, the opaque phenotype was significantly more virulent than the transparent for each of 3 strains examined. The opaque phenotype was associated with 1.2- to 5.6-fold greater amounts of capsular polysaccharide compared with the transparent using a sandwich ELISA. A similar technique comparing the amount of total teichoic acid showed that the transparent phenotype had 2.1- to 3.8-fold more immunodetectable teichoic acid. This difference was confirmed by comparing the incorporation of [3H]choline into teichoic acid. Cell fractionation revealed that variation in quantity of incorporated choline was due to differences in cell wall-associated teichoic acid. Results suggest that the pneumococcus phase varies between a virulent form with more capsular polysaccharide and less teichoic acid and an avirulent form with less capsular polysaccharide and more teichoic acid.

Extracellular enzyme activities potentially involved in the pathogenicity of Mycobacterium tuberculosis

To evaluate the potential contribution of extracellular enzymes to the pathogenicity of mycobacteria, the presence of selected enzyme activities was investigated in the culture filtrates of the obligate human pathogen Mycobacterium tuberculosis, M. bovis BCG, the opportunistic pathogens M. kansasii and M. fortuitum, and the non-pathogenic species M. phlei and M. smegmatis. For M. tuberculosis and M. bovis, 22 enzyme activities were detected in the culture filtrates and/or cell surfaces, of which eight were absent from the culture fluids of non-pathogens: alanine dehydrogenase, glutamine synthetase, nicotinamidase, isonicotinamidase, superoxide dismutase, catalase, peroxidase and alcohol dehydrogenase. These activities, which correspond to secreted enzymes, formed a significant part (up to 92%) of the total enzyme activities of the bacteria and were absent from the culture fluids and the cell surfaces of the non-pathogenic species M. smegmatis and M. phlei. The extracellular location of superoxide dismutase and glutamine synthetase seemed to be restricted to the obligate pathogens examined. The difference in the enzyme profiles was not attributable to the growth rates of the two groups of bacteria. The presence of the eight enzyme activities in the outermost compartments of obligate pathogens and their absence in those of non-pathogens provides further evidence that these enzymes may be involved in the pathogenicity of mycobacteria. In addition, the eight enzyme activities were demonstrated in the cell extract of M. smegmatis. Stepwise erosion of the cell surface of M. smegmatis to expose internal capsular constituents showed that the various enzyme activities, with the possible exception of superoxide dismutase, were located more deeply in the cell envelope of this bacterium. This suggests that the molecular architecture of the mycobacterial envelopes may play an important role in the pathogenicity of these organisms.

Characterization of Shigella dysenteriae type 1 capsular polysaccharide by immunochemical methods

We have isolated the capsular polysaccharide from the strain of Shigella dysenteriae type 1 8337. The product was purified by ultracentrifugation, treated with enzymes (proteinase K, DNA-RNAase) and analyzed by immunochemical methods. Polyclonal antibodies were obtained from rabbits immunized by whole cell antigens prepared from Shigella by ultrasonic treatment and by purified capsular polysaccharide. Crossed immunoelectrophoresis, PAGE and Western blot analysis showed that this product containing mainly the polysaccharide component also contained glycoprotein and lipopolysaccharide. Double diffusion in agarose gel confirmed that the capsular preparation contained at least three antigens reacting with rabbit polyclonal antiserum.

The capsule and S-layer: two independent and yet compatible macromolecular structures in Bacillus anthracis

Bacillus anthracis, the etiological agent of anthrax, is a gram-positive spore-forming bacterium. Fully virulent bacilli are toxinogenic and capsulated. Two abundant surface proteins, including the major antigen, are components of the B. anthracis surface layer (S-layer). The B. anthracis paracrystalline S-layer has previously only been found in noncapsulated vegetative cells. Here we report that the S-layer proteins are also synthesized under conditions where the poly-gamma-D-glutamic acid capsule is present. Structural and immunological analyses show that the capsule is exterior to and completely covers the S-layer proteins. Nevertheless, analysis of single and double S-layer protein mutants shows that the presence of these proteins is not required for normal capsulation of the bacilli. Similarly, the S-layer proteins assemble as a two-dimensional crystal, even in the presence of the capsule. Thus, both structures are compatible, and yet neither is required for the correct formation of the other.

In vivo presence of capsular polysaccharide in coagulase-negative staphylococci of ovine origin

Three strains of coagulase-negative staphylococci identified as Staphilococcus warnery, Staphylococcus simulans and Staphylococcus haemolyticus were examine for evidence of capsular polysaccharide, in vitro by negative staining with India ink, and in vivo by transmission electron microscopy. In vitro, unstained materials surrounded clusters or single cocci. In vivo, capsula materials were surrounding phagocytosed bacteria cells. These capsular materials were either closely or loosely bound to bacterial cells, or projecting like "pili" from the surface of the cell wall. The evidence suggests that capsular polysaccharide is a common characteristic of coagulase-negative staphylococci causing prolonged intramammary infection of sheep.

Molecular epidemiology of a large outbreak of multiresistant Klebsiella pneumoniae

An outbreak of multiresistant Klebsiella pneumoniae has continued in the Grampian Region of Scotland since 1992. The organism, which generally produces an extended-spectrum beta-lactamase (ESBL), has spread to several hospitals and nursing homes. DNA from 80 possible outbreak isolates was digested with the restriction endonucleases XbaI and SpeI, and the patterns obtained by pulsed-field gel electrophoresis were compared. Restriction patterns of 79 of the isolates were found to be highly similar with both restriction enzymes, whereas one isolate was unrelated. The outbreak isolates were divided into six subtypes with SpeI and 16 subtypes with XbaI. These subtypes were independent of antibiotic susceptibility pattern, date of isolation and ward of origin, but the XbaI subtype did correlate with the SpeI subtype. It was concluded that the Klebsiella isolates of this outbreak were clonally related.

Direct and indirect methods for molar-mass analysis of fragments of the capsular polysaccharide of Haemophilus influenzae type b

Two methods are described for direct molar-mass measurement of low-molar-mass fragments obtained by oxidative cleavage of the capsular polysaccharide of Haemophilus influenzae type b. Absolute molar masses were determined by size-exclusion chromatography (SEC) with detection by multiangle laser-light-scattering photometry (MALLS) and differential refractometry (RI). The end-group structure of the polysaccharide fragments allowed the direct measurement of average chain length by quantitative 1H NMR, from which molar masses were derived. Variation between the molar masses obtained by the two methods ranged from 5 to 7%. When molar masses estimated by indirect methods were compared to SEC-MALLS/RI data, significant deviations were observed. Analysis by SEC with secondary calibration with dextran standards gave molar masses that exceeded the SEC-MALLS/RI data by as much as 2.5-fold. Molar masses estimated by a combination of colorimetric assays varied from the SEC-MALLS/RI data by as much as 50%. These results demonstrated the applicability and superior accuracy of the direct methods of molar-mass determination of the polysaccharide fragments.

Recombinant V antigen protects mice against pneumonic and bubonic plague caused by F1-capsule-positive and -negative strains of Yersinia pestis

The purified recombinant V antigen from Yersinia pestis, expressed in Escherichia coli and adsorbed to aluminum hydroxide, an adjuvant approved for human use, was used to immunize outbred Hsd:ND4 mice subcutaneously. Immunization protected mice from lethal bubonic and pneumonic plague caused by CO92, a wild-type F1+ strain, or by the isogenic F1- strain C12. This work demonstrates that a subunit plague vaccine formulated for human use provides significant protection against bubonic plague caused by an F1- strain (C12) or against substantial aerosol challenges from either F1+ (CO92) or F1-(C12) Y. pestis.

The certification of a reference material for the evaluation of the ISO method for the detection of Salmonella

A reference material (RM) containing Salmonella typhimurium was certified as CRM 507 by the Standards, Measurements and Testing Programme of the European Commission. The material consists of a gelatin capsule filled with artificially contaminated milk powder. The material is certified for the evaluation of presence/absence methods based on the ISO 6579 procedure for the detection of Salmonella. In the certification study 11 laboratories determined the presence/absence of Salmonella from each of 50 capsules. They also determined the mean number of colony-forming particles (cfp) and the homogeneity of the batch of RM according to an enumeration procedure. Certified values were calculated for both procedures separately. Based on the presence/absence procedure a fraction of capsules in which no Salmonella could be detected of 2.7% (one-sided 95% confidence upper limit 4.4%) was certified, for the enumeration procedure this fraction was 0.61% (one-sided 95% confidence upper limit 1.6%). The certified mean number of Salmonella cfp in one capsule is 5.9 (two-sided 95% confidence interval 5.3-6.4). Data on the preparation, identification, stability (at storage and higher temperatures) and homogeneity of the material are presented.

Novel polysaccharide capsular serotypes in Porphyromonas gingivalis

Recently van Winkelhoff et al. (1) described 3 novel serotypes in virulent Porphyromonas gingivalis strains, which were based on different polysaccharide antigens. These antigens probably represent capsular structures and have been designated K1, K2 and K3. In the present study we report on 3 novel capsular serotypes, which are represented by P. gingivalis strains ATCC 49417, HG 1690 and HG 1691. The strains, isolated from patients with periodontitis, showed obvious encapsulation in wet India ink preparations. Thermostable antigens could be detected in the supernatant fractions of autoclaved cells. These antigens appeared to be negatively charged, sensitive to periodate degradation, and resistant to proteinase K treatment. On the basis of these characteristics we conclude that the antigens are probably extra-cellular polysaccharides representing a bacterial capsular structure. These K-antigens did not cross-react with K1, K2 or K3 immune-sera of P. gingivalis, with the exception of the K2 antiserum, which partially recognized K5- and K6-antigens. In contrast, K5 and K6 antisera did not react with the K2-antigen. After absorbtion of the K2 antiserum with cells of strains HG 1690 (K5) and HG 1691 (K6) cross-reactivity was no longer present. We propose these novel serotypes to be designated: K4 (ATCC 49417), K5 (HG 1960) and K6 (HG 1691).

Effect of the surface composition of motile Escherichia coli and motile Salmonella species on the direction of galvanotaxis

We have reported that motile Escherichia coli K-12 placed in an electric field swims toward the anode but that motile Salmonella typhimurium strains swim toward the cathode, a phenomenon called galvanotaxis (J. Adler and W. Shi, Cold Spring Harbor Symp. Quant. Biol. 53:23-25, 1988). In the present study, we isolated mutants with an altered direction of galvanotaxis. By further analyses of these mutants and by examination of E. coli and Salmonella strains with altered cell surface structure, we have now established a correlation between the direction of galvanotaxis and the surface structure of the cell: motile rough bacteria (that is, those without O polysaccharide; for example, E. coli K-12 and S. typhimurium mutants of classes galE and rfa) swam toward the anode, whereas motile smooth bacteria (that is, those with O polysaccharide; for example, wild-type S. typhimurium LT2) swam toward the cathode. However, smooth bacteria with acidic polysaccharide capsules (K1 in E. coli and Vi in Salmonella typhi) swam toward the anode. Measurements of passive electrophoretic mobility of strains representative of each set were made. We propose that the different directions of galvanotaxis of rough (or capsulate) bacteria and of smooth bacteria are explicable if the negative electrophoretic mobility of flagellar filaments is less than that of rough bodies but greater than that of smooth bodies.

Bacterial capsules: a simple method for demonstration under the light microscope

It is sometimes desirable to demonstrate bacterial capsules during the routine examination of clinical isolates. Apart from the Indian ink method, methods of demonstrating bacterial capsules are not only tedious but are often non-reproducible. A combined positive-negative capsule staining procedure which is simple, rapid and reproducible is described.

[Comparison of detection methods for the Staphylococcus aureus capsule]

Eleven methods for capsule detection of Staphylococcus aureus were compared. The most suitable of them were transmission electron microscopy, determination of the presence of clumping factor, determination of colonial morphology in serum-soft agar, estimation of cell volume and staining with safranine. The determination of clumping factor is a fast and effective method for presumptive diagnosis of capsulated strains, but need to be confirmed by another method. The cell volume estimation is useful for determination of capsule production in liquid cultures, while staining with safranine is suitable for genetic studies of capsule production. The other methods analyzed in this work (Indian ink staining, use of anticapsular antisera, determination of virulence for mice, lisostaphin susceptibility, resistance to phages and resistance to phagocytosis) were laborious, too slow, or need components and/or equipment not available in all laboratories. In addition, two methods of induction of capsule production were assayed, one in vitro by several passages in broth with 10% bovine serum and the other in vitro by intraperitoneal inoculation in mice. Both methods induced capsule production.

[Virulence factors of coagulase-negative Staphylococcus of human and bovine origin]

Virulence factors such as hydrophobicity of cell surface, capsule formation and slime production were investigated in 393 CNS strains, 190 isolated from bovine mastitis and 203 from human infections; in addition a selected number of CNS species were experimentally infused into the teat cistern of lactating rabbits for pathogenicity tests. Most human (95%) and bovine (82.6%) strains showed cell hydrophobicity properties, however only 11.3% and 7.9% were capsulated strains, respectively. Slime production was detected in 10.5% of bovine strains but only in 0.9% of human strains. Using the lactating rabbit model it was shown that S. chromogenes, S. epidermidis, S. saprophyticus and S. haemolyticus were the most pathogenic species.

The effect of plasmid acquisition on potential virulence attributes of Pasteurella multocida

Several plasmids that were isolated from complement-resistant Pasteurella multocida or Escherichia coli were evaluated for phenotypic markers. Plasmid p2267, isolated from a tetracycline-resistant, complement-resistant fowl cholera field isolate of P. multocida (PM2267), was used to transform a K-12 E. coli (C600); this resulted in increased complement resistance, which was eliminated by curing. Either of two plasmids (p1870 or p70-1, isolated from P. multocida and E. coli, respectively) conferred an increase in complement resistance and invasiveness to turkey epithelial cells when expressed in the Clemson University (CU) vaccine strain of P. multocida. Additionally, the CU strain containing p1870 was more virulent in turkey challenge, and the plasmid appeared amplified in vivo. No detectable differences in major outer-membrane proteins, capsule, or carbohydrate fermentation were found to be associated with the acquisition of these plasmids.

Capsular types and antibiotic susceptibility of pneumococci isolated from patients in Belgium with serious infections, 1980-1993

During the 13-year period from 1 November 1980 to 31 January 1993, we received and serotyped a total of 5,619 clinically significant strains of Streptococcus pneumoniae isolated in more than 75 laboratories in Belgium (4,079 [72.6%] were from blood or pleural fluid, 462 [8.2%] were from cerebrospinal fluid, 691 [12.3%] were from middle ear aspirates, and 387 [6.8%] were from various other body fluids). The isolates belonged to 64 of the 84 currently recognized serotypes. Among the 4,722 isolates tested for susceptibility since 1983, 22% were resistant to at least one antimicrobial agent. Resistance to penicillin has slowly increased since 1985 but remained stable at a level of 2%-4% between 1986 and 1993. Of the 119 isolates with reduced penicillin susceptibility, only 23 were fully resistant (MIC, > or = 2 micrograms/mL) and none of these proved to be resistant to cephalosporins. Resistance to erythromycin increased significantly from 5.2% in 1986 to 21.5% in 1993. Resistance to penicillin and erythromycin was also more frequently recognized in a smaller number of capsular types of S. pneumoniae.

Antibiotic resistance, plasmid content, phage type, and capsule type of Staphylococcus aureus isolates at a children's hospital

One hundred clinical isolates of Staphylococcus aureus were obtained from Al-Solimaniah Children's Hospital, Riyadh, Saudi Arabia. The isolates were screened for their antibiotic susceptibility, plasmid profiles, phage types, and capsular polysaccharide types. A total of 29 antibiotic resistance patterns were obtained. Nine plasmids were detected and 18 plasmid profiles were obtained. Thirteen isolates were non-phage-typeable and 14 different phage-typing-patterns were obtained from the other 87 isolates. Thirty eight isolates were capsular polysaccharide type 8 and 56 were capsular polysaccharide type 5; the other 6 isolates were non typeable. According to the above typing criteria, the isolates were divided into 13 groups. Plasmid profiles were found to be superior to phage typing, whereas, phage typing was superior to the antibiogram as a technique for determining similarities and differences among S. aureus hospital isolates.

Antimicrobial resistance in Haemophilus influenzae isolated from blood, cerebrospinal fluid, middle ear fluid and throat samples of children. A nationwide study in Finland in 1988-1990

A nation-wide survey of the prevalence of antimicrobial resistance in Haemophilus influenzae was conducted on isolates collected in 1988-90 from middle ear fluid (MEF), blood, or cerebrospinal fluid (CSF) in infected children or throat samples of healthy children. Altogether 885 strains were examined regarding capsular type b, beta-lactamase production and the minimal inhibitory concentration (MIC) of ampicillin, cefaclor, erythromycin, tetracycline, chloramphenicol, trimethoprim and trimethoprim-sulfamethoxazole for these strains was determined by the agar dilution method. 99% (578/585) of MEF isolates, 93% (112/121) of throat isolates, but only 6% (10/179) of blood/CSF isolates were not of type b (Hib). The rate of beta-lactamase production was 11.4% among Hib strains, 8.0% among non-type b MEF isolates, and 4.5% among non-type b throat isolates. No increase in the prevalence of beta-lactamase production in H. influenzae has taken place in Finland since the early 1980s. Resistance to ampicillin among strains that lacked beta-lactamase activity was rare (0.2%). Of the non-type b MEF and throat isolates, 5.9% and 2.7%, respectively, were resistant to trimethoprim and 3.6% and 2.7%, respectively, to trimethoprim-sulfamethoxazole. Resistance to other drugs was rare (< 2%) in all isolate groups.