Potential pathogenic role of aggregative-adhering Corynebacterium diphtheriae of different clonal groups in endocarditis

1Departamento de Microbiologia, Imunologia e Parasitologia, Faculdade de Ciências Médicas, 2Instituto de Biologia Roberto Alcântara Gomes, Universidade do Estado do Rio de Janeiro, Rio de Janeiro, RJ, Brasil 3Instituto de Microbiologia Professor Paulo de Góes, Universidade Federal do Rio de Janeiro, Rio de Janeiro, RJ, Brasil 4Escola de Medicina e Cirurgia, Universidade Federal do Estado do Rio de Janeiro, Rio de Janeiro, RJ, Brasil


Introduction
Invasive diseases add new aspects to the infectious processes caused by Corynebacterium diphtheriae, and identify this species as an emerging pathogen.In the early years of the new millennium, infective endocarditis still proves to be difficult to diagnose and is associated with a high death rate (21-41%) worldwide (1,2).Infective endocarditis due to C. diphtheriae is perhaps more common than supposed and on occasion may be an aggressive fatal disease (3).Few reports have dealt with the microbiological aspects of community-acquired endocarditis caused by C. diphtheriae (4)(5)(6).Entry of C. diphtheriae by invasive processes was described as being caused by percutaneous trauma, skin (7,8) and throat colonization (9).Unlike classical diphtheria, invasive disease caused by C. diphtheriae affects both vaccinated and non-vaccinated per- sons, and is mostly induced by nontoxigenic isolates (10).Diphtheria epidemics are likely to spread with a clonal character (11).Likewise, systemic diseases caused by C. diphtheriae have been related to invasive clones (7,12,13).
In addition to host factors, bacterial virulence determinants may contribute to the outcome of invasive infections.Microbial adhesive properties may contribute to the spread and outcome of invasive processes (14).Despite the destructive feature of invasive infections, the pathogenicity aspects that favor the invasive ability of certain C. diphtheriae strains remain unclear (10,15).The ability to survive within cultured epithelial cells was observed for C. diphtheriae strains isolated from throat and blood.The blood isolate showed a higher percentage of adherence and internalization compared with throat isolates (5).Further study showed that all sucrose fermenting and non-fermenting strains isolated from throat and skin lesions exhibited localized and diffuse adherence patterns, respectively (16).However, the binding properties of C. diphtheriae strains related to invasive disease have not yet been investigated.Moreover, identifying organisms belonging to invasive clones, including locations where diphtheria is rarely encountered, may provide valuable information and allow timely preventive measures.
In addition to describing clinical and microbiological characteristics that illustrate the destructive features of endocarditis due to C. diphtheriae, this study was undertaken to assess the patterns of adherence to HEp-2 cells of endocarditis-associated C. diphtheriae strains representative of five different clones.

Bacterial strains and growth conditions
Five C. diphtheriae strains isolated from blood samples of patients with endocarditis were used to illustrate specific biological aspects of invasive disease and to evaluate phenotypic methods that could be used to discriminate potentially invasive strains (Table 1).Microorganisms were recovered from clinical specimens routinely submitted for culturing to the Bacteriology Laboratory of Hospital Universitário Pedro Ernesto, HUPE (LABAC) and the Laboratory of Diphtheria and Non-diphtheria Corynebacterial Infections at Faculdade de Ciências Médicas, Universidade do Estado do Rio de Janeiro (UERJ), Rio de Janeiro, RJ, Brazil.
For phenotypic and genotypic analysis, the microorganisms were grown in trypticase soy broth (Difco Lab., USA) or agar medium for 48 h at 37°C.Two toxigenic strains (CDC-E8392 from Centers for Disease Control, USA, and 241 from Rio de Janeiro, Brazil, isolated from the respiratory tract of patients with diphtheria) were used for comparison.

Biochemical identification and toxigenicity test
Identification, biotype, and toxigenicity determination were performed by conventional microbiological methods as described elsewhere and by the API Coryne System (bioMérieux, France).The toxigenicity test was performed by the Elek assay (6,15,17,18).

HEp-2 adherence assays
Adherence patterns to HEp-2 of C. diphtheriae blood isolates were assayed by methods previously described (16).Adherence assays were performed within semi-confluent HEp-2 cells grown on circular coverslips in DMEM (Sigma, USA) supplemented with 5% fetal calf serum (Gibco-BRL, USA), 50 µg/mL gentamicin, 2.5 µg/mL amphotericin B and 0.5% L-glutamine at 37°C in a 5% CO 2 atmosphere.Microorganisms were washed twice in 10 mM phosphate-buffered saline (PBS, pH 7.2) and resuspended in DMEM to a concentration of 10 7 colony-forming units per mL, and used in adherence assays (3-h incubation).Giemsa-stained coverslips were examined by bright field microscopy.

SDS-PAGE analysis
Protein profiles of crude extracts of bacterial strains were obtained after bacterial cell lysis by treatment with 5 mg/mL lysozyme (Sigma) in PBS for 2 h at 37°C.Thereafter, protein samples were submitted to treatment with sample buffer [0.5 mM Tris-HCl, pH 6.8, 4% sodium dodecyl sulfate (SDS), 20% glycerol, 10% ß-mercaptoethanol, and 0.001% bromophenol blue] and heating at 100°C for 15 min.Total crude cell lysates were cleared by centrifugation at 14,000 g for 5 min, and submitted to electrophoresis on 10% polyacrylamide gels in the presence of SDS (SDS-PAGE).The protein profiles were observed after staining with Coomassie brilliant blue R250 (Sigma) (19).

RAPD analysis
RAPD was performed using the Ready-to-go RAPD Kit (Pharmacia Biotech, USA), according to manufacturer instructions on a Minicycler PTC-100 Programmable Thermal Controller (MJ Research, USA) (20).Amplified products were electrophoresed on a 2% agarose gel for 3 h at 120 V. Gels were stained with ethidium bromide and the R. Hirata Jr. et al.

Results and Discussion
Clinical and microbiological features related to C. diphtheriae strains isolated from patients with native-valve endocarditis are presented in Table 1.Four strains (HC01, HC02, HC03, HC05) were identified as C. diphtheriae subsp mitis and one (HC04) as C. diphtheriae subsp gravis.All strains were toxigenic by the Elek assay.
We have found the incidence of infective endocarditis in the literature to be generally higher in males than in females (2:1), and the average age group affected is in the fifth decade (1).In contrast, the highest number of C. diphtheriae endocarditis in our community was found in children less than 18 years of age rather than in adults over 50 years of age (4:1), and in females more than in males (4:1).No cases were reported among children less than 5 years of age.The source of bacteremia was not precisely determined for all 5 cases.However, predisposing conditions to an increased risk of acquiring endocarditis (22), such as congenital heart disease and major dental treatment, were observed for patients HC03 and HC05, respectively.Patient HC02 experienced an episode of non-exudative pharyngitis 2 weeks before the onset of infection.Nevertheless, skin lesions were not seen in any patient.The preferred sites of attack of C. diphtheriae endocarditis isolates were mitral valves (80%).Complications observed in all five episodes of endocarditis included septicemia (N = 5), intracardiac abscess (N = 1), mycotic aneurysm (N = 1), peripheral (N = 2) and CNS emboli (N = 1), microaneurysm with brain hemorrhage (N = 2), arthritis (N = 2), and acute renal failure (N = 3).Surgical treatment was performed for valve replacement for patients HC02 and HC05.Patients HC03, HC04, and HC05 were submitted to antibiotic and serum therapy.The mortality rate (40%) due to infective endocarditis caused by C. diphtheriae was higher than that previously reported in the literature (27%) (23).
C. diphtheriae endocarditis isolates exhibited an aggregative-like adherence pattern, characterized by clumps of bacteria with a "stacked-brick" appearance that were attached to the surfaces of cultured epithelial cells and exposed areas of the glass slide among epithelial Table 1.Table 1.Table 1.Table 1 phenotype expressed by the C. diphtheriae blood isolates, distinct from those exhibited by strains isolated from throat and skin lesions.More information is necessary to correlate the aggregative adhesive properties with the pathogenicity of C. diphtheriae strains as observed for enteroaggregative Escherichia coli (25).Phenotypic and molecular typing methods have been used with varying success to distinguish between C. diphtheriae strains (26,27).Despite the fact that SDS-PAGE profiles are not considered to provide consistent results for the evaluation of clonality among bacterial strains, heterogeneity among protein profiles of C. diphtheriae has been used for the analysis of epidemic strains (28,29).In the present study, SDS-PAGE demonstrated C. diphtheriae endocarditis isolates characterized by a particular protein profile, as exemplified by the HC01 strain in Figure 2.However, the endocarditis isolates showed a protein profile distinct from those presented by C. diphtheriae 241 and CDC-E8392 strains isolated from patients with diphtheria.Thus, in addition to the aggregative pattern to HEp-2 cells, Figure 1. Figure 1. Figure 1. Figure 1.cells.Figure 1 illustrates the aggregative pattern of the endocarditis isolates, independent of C. diphtheriae subspecies.Adhesion to HEp-2 cells has provided a useful model for the study of toxigenic (5,16) and non-toxigenic C. diphtheriae strains (4).Similar to some enteric pathogens (24), differing degrees of attachment to HEp-2 monolayers with predominance of localized and diffuse adherence patterns have been reported for C. diphtheriae strains isolated from throat and skin lesions, respectively (16).In the present study, we have identified a third adherence Figure 3. Figure 3. Figure 3. Figure 3. the relationship among C. diphtheriae endocarditis isolates was verified by the expression of identical SDS-PAGE protein profiles.RAPD-PCR can be reliably and reproducibly used for rapidly screening C. diphtheriae strains of the predominant epidemic clonal group (30).C. diphtheriae endocarditis isolates showed different amplification patterns by RAPD-PCR for all 6 sets of primers used, indicating diversity among C. diphtheriae invasive clones.The use of RAPD-PCR was successful for detecting minor differences in closely related toxigenic C. diphtheriae endocarditis strains.The primer 4 of RAPD-PCR kit was more discriminating for the C. diphtheriae strains evaluated in this study (Figure 3).
Our data demonstrate the destructive feature of infective endocarditis due to C. diphtheriae and demonstrate the circulation of different C. diphtheriae clones exhibiting invasive properties.The data also suggest the pathogenic role of aggregative-adhering C. diphtheriae in invasive disease.
Additional studies are necessary to characterize the full range of virulence factors of the aggregative-adhering C. diphtheriae and the molecular pathophysiology of the invasive illness caused by these strains.It has been demonstrated that E. coli strains, which present aggregative adherence, are able to form biofilms (31).This property enables aggregative-adhering E. coli strains to produce more prolonged diseases.The possibility that C. diphtheriae can also form biofilms (15) and the potential of this property to aggravate the clinical outcome of invasive disease merits investigation.

Figure 1 .
Figure 1.Figure1.Figure1.Figure1.Figure1.Light micrographs illustrating different adherence patterns to HEp-2 cells of Corynebacterium diphtheriae strains.A and B, Aggregative adherence pattern characterized by clumps of bacteria with a "stacked-brick" appearance expressed by strains isolated from blood of patients with endocarditis.C, Localized and D, diffuse adherence patterns expressed by control strains of sucrose-fermenting (241 strain) and non-sucrose fermenting (CDC-E8392 strain) biotypes isolated from patients with classical respiratory diphtheria, respectively.Magnification: 1000X.

Table 1 .
. Clinical aspects of patients with endocarditis and microbiological properties of Corynebacterium diphtheriae blood isolates.
a Absence of skin lesions.b Nose and throat cultures were not performed.Toxin production evaluated by the Elek test.c Microbiological description of bacterial strain published by Mattos-Guaraldi et al. (15).d Pharyngitis without pseudomembrane formation two weeks before the onset of infection.e Patient submitted to prior dental surgery.*DTP, diphtheria, tetanus and pertussis.www.bjournal.com.br