- Email: [email protected]
Internalization of non-toxigenic Corynebacterium diphtheriae by cultured human respiratory epithelial cells
Microbial Pathogenesis 37 (2004) 111–118 www.elsevier.com/locate/micpath
Internalization of non-toxigenic Corynebacterium diphtheriae by cultured human respiratory epithelial cells Lucia Bertuccini, Lucilla Baldassarri, Christina von Hunolstein* Dipartimento di Malattie Infettive, Parassitarie ed Immunomediate Istituto Superiore di Sanita`, Viale Regina Elena 299, 00161 Roma, Italy Received 23 February 2004; accepted 10 June 2004
Abstract Although infection by Corynebacterium diphtheriae is a model of extracellular mucosal pathogenesis, and diphtheria is one of the most worried diseases, this microorganism can be associated also with invasive infections such as endocarditis, septic arthritis, and osteomyelitis. Invasive infections are usually caused by non-toxigenic C. diphtheriae strains. Over the last years severe pharyngitis/tonsillitis associated with the isolation of non-toxigenic C. diphtheriae have been described. Penicillin treatment failure of these infections could only partially be explained by penicillin tolerance of the causing strain. Thus, we examined the in vitro ability of non-toxigenic C. diphtheriae throat clinical isolates to adhere to, and enter human respiratory epithelial cells. Trasmission and scanning electron microscopy demonstrated intracellular C. diphtheriae in laryngeal (HEp-2 cells) and pharyngeal (Detroit D562 cells) tissue culture. Live intracellular bacteria were detectable up to 48 h post-infection. Using a variety of compound that act on eukariotic cell structures, the internalization of C. diphtheriae seems to occur via a zipper-like mechanism. It is likely that internalization of C. diphtheriae can be involved in throat colonization contributing to bacterial eradication failure and asymptomatic carriage. q 2004 Elsevier Ltd. All rights reserved. Keywords: Corynebacterium diphtheriae; Adherence; Internalization; Infection; HEp-2 cells; Detroit D562 cells
1. Introduction In the last decade it has become evident that Corynebacterium diphtheriae should not be considered only the etiologic agent of diphtheria. The introduction of mass immunization against diphtheria in the 1950s in most European countries, as well as in USA and Canada, resulted in a very low incidence of this disesase [1] and circulation of toxigenic strains declined in all countries with a good vaccination coverage [2,3]. In contrast, non-toxigenic strains of C. diphtheriae have been increasingly documented as a cause of invasive disease, which include endocarditis, bacteraemia, septic arthritis, osteomyelitis and splenic abscesses [4–9]. Nasopharyngeal carriage and disease caused by this organism has been documented * Corresponding author. Tel.: C39-649-902-659; fax: C39-649902-934. E-mail address: [email protected] (C. von Hunolstein). 0882-4010/$ - see front matter q 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.micpath.2004.06.002
amongst homeless, intravenous drug users and alcoholics [10,11]. Emergence of non-toxigenic C. diphtheriae has also been observed in Italy, England and Wales; in particular, clusters of cases of sore throat associated with the isolation of C. diphtheriae were observed during an enhanced surveillance study [3,12]. The upper respiratory tract is the primary site of colonization by this microorganism. Besides toxin production, little is known about other C. diphtheriae virulence factors. In a recent study [13], penicillin treatment failure of severe pharyngitis/tonsillitis could only partially be explained by penicillin tolerance of the respective isolated strains. Previous studies dealt with the adhesive properties of C. diphtheriae to human buccal epithelial cells and erythrocytes [14,15] and found that this feature was independent from toxin production. Systemic infections caused by C. diphtheriae suggest that the organism is not only able to adhere to host epithelial cells, but must be able to gain access to deeper tissues.
112
L. Bertuccini et al. / Microbial Pathogenesis 37 (2004) 111–118
In the present report, we examined the in vitro ability of non-toxigenic C. diphtheriae clinical isolates to adhere to and enter into human epithelial cells. Using a variety of compounds that act on eukaryotic cells, we investigated the dependence of these processes on microfilaments and microtubules and evaluated the route of entry of C. diphtheriae into cells. 2. Results 2.1. Interaction of non-toxigenic C. diphtheriae strains with epithelial cells To evaluate the ability of non-toxigenic strains of C. diphtheriae to adhere, enter and survive within epithelial cells, HEp-2 cells grown on glass coverslips were infected with three different strains (Table 1). Two strains (3319 and 4060) were non-toxigenic clinical isolates of patients affected by pharyngitis/tonsillitis; C. diphtheriae C7 (b197) is a lysogenic strain which produces a non-toxic diphtheria protein due to a mutation in the fragment A, resposible for enzymatic activity. After one hour of infection, bacteria showed a specific association with cell monolayers with a localized adherence pattern, when observed by light microscopy (Fig. 1a). After further 2 h of incubation with gentamicin, bacteria were visible inside cells vacuoles (Fig. 1b), the average number of viable bacteria, recovered on agar plates after cell lysis, being slightly different for the three strains (Table 1). Adhesion and internalization seemed independent from the capacity to produce diphtheria toxin, as evidenced by the adherence and internalization rates of the strain C7 (b197). Strain 3319 was chosen for further investigations to better understand the kinetic of entry of non-toxigenic C. diphtheriae into epithelial cells. The invasion efficiency of C. diphtheriae in HEp-2 cells was compared with that observed in human pharyngeal epithelial cells (Detroit 562, D562). As shown in Fig. 2, at time T0 viable bacteria associated with HEp-2 cells were significantly more than those associated with D562 cells. After 3 and 6 h of treatment with gentamicin, the average number of viable bacteria recovered in D562 cells continued to be
Fig. 1. After 1 h of infection of HEp-2 cells with non-toxigenic C. diphtheriae (strain 3319) a localized pattern of adherence was visible (a); after further 2 h of incubation with gentamicin, bacteria were visible inside HEp-2 cellular vacuoles (b). (1000!).
lower than that in HEp-2 cells, but after more than 24 h of incubation the survival of internalized bacteria into D562 was higher. At 48 h post-infection, bacteria were still recovered on agar plates after lysis of D562 cells (Fig. 2). Similar internalization rate was also observed for the other clinical isolates used in this study (data not shown).
Table 1 Adhesion to and internalization into HEp-2 cells of three strains of C. diphtheriae (2!105 cells/ml were infected with 2!107 CFU/ml (MOI 100) Strain
Viable bacteria (CFU/ml)a Adherentb
3319 4060 C7(b197) a
6
Internalizedc 6
5.6!10 G4.3!10 5.8!105G3.4!105 2.2!106G0.9!106
1.1!105G0.7!105 5.3!103G2.8!103 9.0!104G3.0!104
Values represent the means of three separate experiments. b Average number of viable bacteria recovered on agar plates after 1 h of infection. c Average number of viable bacteria recovered on agar plates after 1 h of infection and further 2 h of treatment with gentamicin.
Fig. 2. Viable (Log10 CFU/mlGSD) C. diphtheriae (strain 3319) recovered from HEp-2 (-*-) and D562 (-B-) cells after different time of infection. At 1 h post-infection (T0), monolayers were washed to remove unadsorbed bacteria and further incubated with fresh medium containing gentamicin (100 mg/ml) for 3 (T3), 6 (T6), 24 (T24), and 48 (T48) hours. Data are from three independent experiments.
L. Bertuccini et al. / Microbial Pathogenesis 37 (2004) 111–118 Table 2 Adhesion and internalization of C. diphtherae strain 3319. D562 cells (2!105 cells/ml) were infected with different bacterial concentrations, ranging from 2!106 to 2!109 CFU/ml, corresponding to MOI 10–10,000 Multiplicity of infection (MOI)
Viable bacteria (CFU/ml)a Adherentb
Internalizedc
10 100 1000 10,000
9.6!103G8.5!103 3.2!104G2.4!104 3.0!105G1.1!105 8.2!105G5.4!105
41.0G33.0 4.7!102G4.1!102 1.7!103G1.0!103 1.1!103G1.0!103
a
Values represent the means of three separate experiments. Average number of viable bacteria recovered on agar plates after 1 h of infection. c Average number of viable bacteria recovered on agar plates after 1 h of infection and further 2 h of treatment with gentamicin (100 mg/ml). b
Furthermore, it was investigated how interaction between C. diphtheriae and D562 cells could be modified by changing the bacteria/cell ratio (multiplicity of infection, MOI). When it was varied from 10:1 to 10,000:1,
113
the number of bacteria associated to D562 cells increased continuously up to a plateau between MOI 1000 and 10,000, without damage to the monolayers (Table 2). After 2 h of treatment with gentamicin, the level of intracellular bacteria had the same trend at different MOIs; from MOI 100 to MOI 1000 a slight increase of internalized C. diphtheriae was observed while at higher MOIs internalization remained costant (Table 2). 2.2. Ultrastructural analysis of the internalization process of C. diphtheriae Monolayers were infected with C. diphtheriae for 1, 2 and 3 h and processed for scanning electron microscopy to analyze the first step of interaction between bacteria and host cells. After 1 h of incubation, most of epithelial cells were colonized by clumped bacteria which showed a localized pattern of adherence (Fig. 3a and b), confirming
Fig. 3. Interaction of C. diphtheriae (strain 3319) with D562 cells viewed by scanning (a)–(e) and trasmission electron microscopy (f) at 1 h (a)–(c), 3 h (d) and (e) and 48 h (f) of post-infection. (a) Small bacterial clumps adhere to the epithelial cell surface. (b) Cellular microvilli appear in close contact with the polar side of adherent rods. (c) Bacterial contact to cellular microvilli induced membrane protrusion towards the entire length of the rod. (d) Cellular protrusions engulf a bacterial cell with a zipper-like mechanism of phagocytosis. (e) A peculiar ring-like structure, originating from the host cell membrane internalizes a single rod. (f) Forty-eight hours post-infection, numerous C. diphtheriae bacilli are still present inside cells in tight fitting vacuoles, without sign of digestion.
114
L. Bertuccini et al. / Microbial Pathogenesis 37 (2004) 111–118
the light microscopy observations. Adhesion was mediated by cellular microvilli which seemed to contact bacterial cells firstly in the polar region and then extend to the entire length of the rod (Fig. 3c), establishing a close contact with the bacterial cell wall. The extention of host cell protrusions, guided by cytoskeletal rearrangements, induced a sequential engagement of cell membrane which engulfed the bacterium and formed a tight-fitting vacuole (Fig. 3d). In the 3 h samples these events were more frequently observed. Fig. 3e shows a very peculiar ring-like structure, with a ruffle-like appearance although strictly adherent to the bacterial surface, which directly originates from the host cell surface to internalize a single rod. Inside cells, C. diphtheriae was present in tight-fitting vacuoles (Fig. 3f). At 24 and 48 h post-infection, internalized bacteria were still found without evident signs of degradation (Fig. 3f). Fluorescence actin staining with tetramethyl rhodamine isothiocyanate-conjugated phalloidin, which binds
specifically to actin filaments, showed a partial condensation of actin filaments around bacteria that were just internalized by cells (not shown). 2.3. Effects of inhibitors on the internalization process Specific inhibitors of cell functions were used to examine the mechanism of internalization induced in epithelial cells by C. diphtheriae. Cytocalasin D, which blocks F-actin polymerization, showed a dose-dependent inhibitory effect on the internalization of bacteria, indicating the involvement of actin microfilaments in the bacterial uptake. The decrease of viable bacteria was more evident at the highest dose utilized that determined a reduction of internalization of 87% (Fig. 4a). In contrast, microtubules did not seem to be involved in C. diphtheriae entry into D562 cells, as colchicine had no significant effects on the average
Fig. 4. Effect of different compounds on C. diphtheriae (strain 3319) invasion process into D562 cells. Internalization rates were strongly affected by treatment with Cytocalasin D (a), while colchicine (b) had no effect. Staurosporin (c), but not wortmannin (d), enhanced bacterial adhesion (-B-) and entry (->-) into cells. Monolayers were pretreated with each chemical for 30 min at indicated concentrations. Only cytocalasin D and colchicine were maintained troughout the 2 h internalization period. Viable bacteria were determined as Log10 CFU/mlGSD, recovered on agar plates after cells lysis. Data are from three independent experiments.
L. Bertuccini et al. / Microbial Pathogenesis 37 (2004) 111–118
number of viable bacteria after gentamicin protection assay (Fig. 4b). Monolayers were pre-treated with monodansylcadaverine (200 mM), a primary amine that blocks receptor recycling to cellular membrane, to evaluate the involvement of membrane receptors in the internalization process of C. diphtheriae. Cells treatment affected bacterial uptake, reducing the percentage of viable bacteria to 18% of those recovered in non-treated cells (data not shown). To investigate whether the binding to membrane involved a signal cascade mediated by protein kinases (PKs), cellular monolayers were treated with specific inhibitors of these enzymes. Exposure of cells to increasing doses of staurosporine, a potent inhibitor of different protein kinase classes including protein kinase C (PKC), some phosphotyrosinekinases (TPKs) and cAMP-dependent PK, produced a significant enhancement of the bacterial adhesion and internalization into cells (23-fold). This effect was more relevant at concentrations in the nanomolar range (Fig. 4c). Moreover, treatment of D562 cells with increasing doses of wortmannin, a compound which blocks phosphatidilinositol-trisphosphate kinase (PI3-K), did not affect the level of both adherence and internalization of C. diphtheriae (Fig. 4d). As well, no effect was observed after treatment with orthovanadate, an inhibitor of cell phosphatases (not shown).
3. Discussion and conclusions Non-toxigenic C. diphtheriae strains have been increasingly isolated from different types of infection, from cutaneous lesions and pharyngitis to bacteremia and endocarditis [3,5,7,12,16,17]. Several studies have underlined that immunization against diphtheria toxin does not protect from the insurgence of non-toxigenic C. diphtheriae diseases, supporting the hypothesis that diphtheria bacilli may possess other factors enhancing their virulence [5,7,11]. C. diphtheriae is generally considered a non-invasive pathogen that causes localized infection of the higher respiratory tract, with extensive peripheral tissue damages due to the diffusion of the diphtheria toxin [18]. In this study we show that non-toxigenic C. diphtheriae strains can be internalized by non-professional phagocytes. Two different cell lines were used, a larynx- (HEp-2) and a pharyngealderived (D562), as in vivo both larynx and pharynx are colonized by C. diphtheriae. Infection of HEp-2 monolayers with non-toxigenic strains indicated that C. diphtheriae is able to adhere to the epithelial cell surface and, by light and electron microscopy, that bacilli are present in intracellular vacuoles. Bacterial cells internalization seemed to be a strain-dependent process as the entry efficiency was slighty different for each strain examined. Hirata et al. have previously described the ability of toxigenic C. diphtheriae strains to enter HEp-2 cells with
115
different levels of uptake [19]. Our results showed that also non-toxigenic C. diphtheriae isolates were able to penetrate respiratory epithelial cells of both pharyngeal and laryngeal origin and that internalization does not seem to be influenced by toxin production. C. diphtheriae showed different behaviours with the two cell lines used. Adhesion and internalization rates in HEp-2 cells were significantly higher than in D562 cells; however, intracellular survival was more efficient in the latter and intact bacterial cells were still recovered after 48 h of infection. Thus, both cellular and bacterial determinants appear to be involved in C. diphtheriae internalization by epithelial cells. Although the number of internalized bacteria was lower than that observed for invasive pathogens sensu stricto (i.e. Shigella, enteroinvasive Escherichia coli, etc.), the possibility to hide into epithelial cells even at low numbers and to survive for at least two days, would provide C. diphtheriae with the opportunity to escape from the host’s humoral immunity and/or antibiotic therapy and possibly to spread to distant sites. Results obtained by changing bacteria/cell ratio have shown that C. diphtheriae invasion of D562 cells is a saturable process, reaching maximal number of internalized bacteria at a MOI between 100 and 1000. This may reflect a limited number of host cell receptors or entry sites [20], a limitation of some other host biochemical requisite for entry, or host cell modifications occuring during internalization, preventing penetration of further C. diphtheriae cells. Supporting the hyphothesis of a receptor-mediated entry, scanning and thin-section electron microscopy images are consistent with a zipper-like mechanism of phagocytosis with an initial binding of the bacteria to the host cell receptors leading to the formation of membrane protrusions and subsequent extention of cell membrane around the bacteria, until complete engulfment [21]. The intimate contact of cellular membrane protrusions with the bacterial cell wall was clearly visible in electron micrographs and, even after 48 h of infection, vacuole membrane remained strictly adherent to the C. diphtheriae cell wall, that did not show any sign of degradation. Other bacterial pathogens, such as Yersinia and Listeria, are internalized by a similar mechanism by non-professional phagocytes [22,23]; furthermore, Helicobacter pylori, considered a non-invasive pathogen, was found entering gastric mucosa cells through a zipper-like mechanism [24]. The inhibitory effect induced by treatment of D562 monolayers with monodansylcadaverine confirmed that C. diphtheriae internalization process needs cellular membrane receptors, as this compound blocks receptor recycling to plasma membrane [25]. Also, the dose dependent inhibition effect of cytocalasin D on bacterial entry and the partial condensation of actin observed around the rods, support the idea that bacterial binding to specific sites of epithelial cell surface may induce the intracellular signal transduction responsible for actin microfilaments
116
L. Bertuccini et al. / Microbial Pathogenesis 37 (2004) 111–118
rearrangement and pseudopods movement [26]. Staurosporine, an inhibitor of different classes of PKs, upregulated C. diphtheriae internalization, suggesting that activation of PKC or TPKs or cAMP-depending ATPases negatively controlled C. diphtheriae uptake; this effect was specific, since other inhibitors such as wortmannin, that blocks the PI3K activity [27], and orthovanadate, an inhibitor of cell phosphatases [24], had no effect. The increased internalization of C. diphtheriae is in contrast to other reports where PKs inhibition led to reduced bacterial entry [26,27]. Genistein, a specific inhibitor of TPKs, was shown to block C. diphtheriae entry into HEp-2 cells [19]. Thus, it is likely that PKC is involved in this particular mechanism of enhanced internalization. Our hypothesis is that C. diphtheriae binding to cell membrane induces a reorganization of the actin cytoskeleton which in turn affects the dynamic of receptors involved in internalization of C. diphtheriae, limiting the number of internalized bacteria; alteration of this feed-back inhibition process would eliminate such limit. This hypotesis is supported by the trend of adhesion that parallels that of internalization, as seen by pretreatment with staurosporine. On the basis of the results of this study, non-toxigenic C. diphtheriae should not be considered exclusively an extracellular microorganism. Its behaviour resemble that of Streptococcus pyogenes also found to be internalized by non-phagocytic cells [28,29]. The internalization of C. diphtheriae by cultured laryngeal and pharyngeal cells may resembles the events occurring in human host with recurrent pharyngitis/tonsillitis after seemingly adequate antibiotic treatment [3,12], or in asymptomatic carriage.
4. Materials and methods 4.1. Bacterial strains and growth conditions Adherence and internalization of C. diphtheriae was studied using non-toxigenic clinical strains isolated from patients affected by pharyngis/tonsillitis: 3319, 4060, 95354, 95325, 95492, and 96161. A lysogenic strain of C. diphtheriae, C7 (b-197), was kindly provided by Dr R. Rappuoli. Clinical isolates were identified as C. diphtheriae using API Coryne (bioMe´rieux). Strains were identified as non-toxigenic by the Elek test, PCR and Southern blotting [3]. Gentamicin, used in the cell assays, was bactericidal for all strains at concentration of 100 mg/ ml for 2 h. For cell assays, bacteria were grown overnight at 37 8C in Todd–Hewitt broth (THB, Oxoid)) with agitation, resuspended in tissue-culture medium (see below) and adjusted spectrophotometrically at 600 nm at a concentration of 107 CFU/ml as checked by plating on THB agar of serial dilutions of the culture and CFU counts.
4.2. Cell cultures HEp-2 cells (human larynx carcinoma, ATCC, Manassas, VA) were cultured in Minimal Essential Medium (Eagle’s) (MEM, Life Technologies) supplemented with 0.1 mM non-essential aminoacids, 2 mM glutamine, penicillin 100 UI/ml, streptomycin 100 mg/ml, and 10% heat inactivated fetal calf serum (FCS, Life Technologies) in a humidified atmosphere of 5% CO2 and 95% air at 37 8C. Cells were passaged at a ratio of 1:5 twice each week. Detroit 562 cells (D562) (human pharynx carcinoma, ATCC CCL-138) were cultured in MEM supplemented with 0.1 mM non-essential amino acids, 2 mM glutamine, 1 mM sodium pyruvate, and 10% heat inactivated FCS in a CO2 incubator. Cells were passaged at a ratio of 1:3 twice each week. 4.3. Adherence and internalization assays Assays were performed as previously described [30] with slight modifications. Semi-confluent cells, grown in 24-well plates, were infected with stationary phase bacteria at different MOIs (bacteria-to-cell ratio 100:1) for 1 h at 37 8C in a humidified atmosphere of 5% CO2 and 95% air. At 1 h post-infection (T0), monolayers were washed three times with phosphate-buffered salt solution (PBS), pH 7.4, to remove unadsorbed bacteria, and further incubated with fresh medium containing gentamicin 100 mg/ml for 3 (T3), 6 (T6), 24 (T24), and 48 (T48) hours. At different time points cells were lysed with triton X-100 (0.02% in H2O) for 5 min, serially diluted in PBS and 10 ml of each dilution plated in triplicates on THB agar. As each condition was performed in duplicate, the mean numberGSD of CFU recovered for the two wells was determined. For light microscopy examination, monolayers were grown on glass coverslips in 24 wells/ plate, fixed at the different time points with methanol, and stained by MayGrunwald-Giemsa [31]. All experiments were repeated at least three times. 4.4. Chemicals Cytochalasin D and colchicine (Sigma-Aldrich, Milan, Italy), diluted in phosphate buffered saline (PBS), (pH 7.4), were prepared as stock solutions and added to the cell culture medium at different concentrations. To analyze cell structures involved in internalization of corynebacteria, inhibitors were added to the incubation medium 30 min before infection and were present throughout the assay. Monodansylcadaverine, wortmannin and staurosporin (all from Sigma-Aldrich, Milan, Italy) were stocked in DMSO and diluted at different concentrations for monolayers pretreatment. After 30 min of incubation with each inhibitor, cells were washed with fresh medium and infected with corynebacteria for internalization assay.
L. Bertuccini et al. / Microbial Pathogenesis 37 (2004) 111–118
All inhibitors were tested for possible effects on cell and bacterial viability by trypan blue exclusion test and CFU counting of treated and untreated samples, respectively. 4.5. Fluorescence microscopy To monitore the depolymerization of microfilaments (MF) and microtubules (MT), cells grown on glass coverslips were fixed with paraformaldeyde for 15 min and permeabilized with Triton X-100 (0.5% in PBSC1% BSA) for 5 min. MF were subsequently treated with diluted (1:200) tetramethyl rhodamine isothiocyanate-labelled phalloidin (Sigma-Aldrich) for 30 min. MT were visualized by monoclonal mouse anti a- and b-tubulin antibodies (1:50) (Sigma-Aldrich), incubated for 30 min at 37 8C. After extensive washing with PBS, cells were further incubated for 30 min with TRITC-labelled anti-mouse antibodies (Sigma-Aldrich). Slides were mounted in PBS-glycerine and examined by fluorescence microscopy. 4.6. Scanning electron microscopy (SEM) Infected monolayers were first fixed with 2.5% glutaraldehyde in 0.1 M cacodylate-buffered (Merck, Darmstadt, Germany) for 20 min at room temperature and post-fixed with 1% OsO4 for an additional hour. Samples were then dehydrated through a graded series of ethanol, critical point dried and gold sputtered, and examined by a Cambridge SE360 scanning electron microscope. 4.7. Transmission electron microscopy (TEM) Infected monolayers were processed at different times according to Perry et al. [32]. Cells were fixed with 2.5% glutaraldehyde, 2% paraformaldehyde and 2 mM CaCl2 in 0.1 M sodium cacodylate buffer (pH 7,4) overnight at 4 8C. Monolayers were washed in cacodylate buffer and postfixed with 1% OsO4 in 0.1 M sodium cacodylate buffer for 1 h at room temperature, treated with 1% tannic acid in 0.05 M cacodylate buffer for 30 min and rinsed in 1% sodium solphate in 0.05 cacodylate for 10 min. Fixed specimens were washed, dehydrated through a graded series of ethanol solutions (30–100% ethanol) and embedded in Agar 100 (Agar Scientific Ltd, UK). Ultrathin sections obtained by a MT-2B Ultramicrotome (LKB—Pharmacia) were stained with uranyl acetate and lead citrate and examined by an EM 208 Philips electron microscope.
Acknowledgements This work was partially supported by the Istituto Superiore di Sanita` grant N8 1024/RI (C.v.H.). We gratefully acknowledge Dr. A. Efstratiou (Health Protection Agency, London, UK) for providing us the strains 95354, 95325, 95492 and 96161.
117
References [1] Galazka A. The changing epidemiology of diphtheria in the vaccine era. J Infect Dis 2000;181(Suppl 1):S2–S9. [2] Saragea A, Maximescu P, Meiter E. Corynebacterium diphtheriae. In: Bergan T, Norris JR, editors. Microbiological methods used in clinical and epidemiological investigations—methods in microbiology, vol. 13. London, New York, Toronto, Sydney, San Francisco: Academic Press; 1979, p. 62–176. [3] von Hunolstein C, Alfarone G, Scopetti F, et al. Molecular epidemiology and characteristics of Corynebacterium diphtheriae and Corynebacterium ulcerans strains isolated in Italy during the 1990s. J Med Microbiol 2003;52:181–8. [4] Alexander D. Splenic abscess caused by Corynebacterium diphtheriae. Clin Pediatr 1984;23:591–2. [5] Funke G, Altwegg M, Frommel L, von Graevenizt A A. Emergence of related non-toxigenic Corynebacterium diphtheriae biotype mitis strains in Western Europe. Emer Infect Dis 1999;5: 477–80. [6] Poilane I, Fawaz F, Nathanson M, et al. Corynebacterium diphtheriae osteomyelitis in an immunocompetent child: a case report. Eur J Pediatr 1995;154:381–3. [7] Patey O, Bimet F, Riegel P, et al. Clinical and molecular study of Corynebacterium diphtheriae systemic infections in France. J Clin Microbiol 1997;35:441–5. [8] Belko J, Wessel D, Malley R. Endocarditis caused by Corynebacterium diphtheriae: case report and review of the literature. Pediatr Infect Dis J 2000;19:159–63. [9] Mattos-Guaraldi AL, Formiga LC, Camello TC, Pereira GA, Hirata Jr R, Halpern M. Corynebacterium diphtheriae threats in cancer patients. Rev Argent Microbiol 2001;33:96–100. [10] Gubler J, Huber-Schneiner C, Gruner E, Altwegg M. An outbreak of non-toxigenic Corynebacterium diphtheriae infection: single bacterial clone causing invasive infection among Swiss drug users. Clin Infect Dis 1998;27:1295–8. [11] Wilson APR. The return of Corynebacterium diphtheriae: the rise of non-toxigenic strains. J Hosp Infect 1995;30:306–12. [12] Reacher M, Ramsay M, White J, et al. Non-toxigenic Corynebacterium diphtheriae: an emerging pathogen in England and Wales? Emer Infect Dis 2000;6:640–5. [13] von Hunolstein C, Scopetti F, Efstratiou A, Engler K. Penicillin tolerance amongst non-toxigenic Corynebacterium diphtheriae isolated from cases of pharyngitis. J Antimicrob Chemother 2002;50: 125–8. [14] Colombo AVC, Hirata Jr R, Rocha de Souza CM, et al. Corynebacterium diphtheriae surface proteins as adhesins to human erythrocytes. FEMS Microbiol Lett 2001;197:235–9. [15] Mattos-Guaraldi AL, Formiga LCD, Pereira GA. Cell surface components and adhesion in Corynebacterium diphtheriae. Microbes Infect 2000;2:1507–12. [16] Hogg GG, Strachan JE, Huayi L, Beaton SA, Robinson PM, Taylor K. Non-toxigenic Corynebacterium diphtheriae biovar gravis: evidence for an invasive clone in a south-eastern Australian community. Med J Australia 1996;164:72–5. [17] Mattos-Guaraldi AL, Formiga LCD. Bacteriological properties of a sucrose-fermenting Corynebacterium diphtheriae strain isolated from a cases of endocarditis. Curr Microbiol 1998; 37:156–8. [18] Mac Gregor RR. In: Mandell GL, Bennett JR, Dolin R, editors. Corynebacterium diphtheriae. Mandell, Douglas and Bennett’s principles and practice of infectious diseases. New York: Churchill Livingstone; 1995, p. 1865–72. [19] Hirata RJr, Napolea˜o F, Monteiro-Leal LH, et al. Intracellular viability of toxigenic Corynebacterium diphtheriae strains in HEp-2 cells. FEMS Microbiol Lett 2002;215:115–9.
118
L. Bertuccini et al. / Microbial Pathogenesis 37 (2004) 111–118
[20] Lan Hu, Kopecko DJ. Campylobacter jejuni 81–176 associates with microtubules and dynein during invasion of human intestinal cells. Infect Immun 1999;67:4171–82. [21] Swanson JA, Baer SC. Phagocytosis by zipper and triggers. Trends Cell Biol 1995;5:89–93. [22] Iseberg RR. Discrimination between intracellular uptake and surface adhesion of bacterial pathogens. Science 1991;252:934–8. [23] Mengaud J, Ohayon H, Gounon P, Me´ge RM, Cossart P. E-cadherin is the receptor for internalin, a surface protein required for entry of Listeria monocytogenes into epithelial cells. Cell 1996; 84:923–32. [24] Kwok T, Backert S, Schwarz H, Berger J, Meyer T. Specific entry of Helicobacter pylori into cultured gastric epithelia cells via a zipperlike mechanism. Infect Immun 2002;70:2108–20. [25] Wileman T, Harding C, Stahl P. Receptor mediated endocytosis. Biochem J 1985;232:1–14. [26] Rosenshine I, Duronio V, Finlay B. Tyrosine protein kinase inhibitors block invasin-promoted bacterial uptake by epithelial cells. Infect Immun 1992;60:2211–7.
[27] Benjamin P, Federman M, Wanke CA. Characterization of an invasive phenotype associated with enteroaggregative Escherichia coli. Infect Immun 1995;63:3342–417. [28] Osterlund A, Popa R, Nikkila T, Scheynius T, Engstrand L. Intracellular reservoir of Streptococcus pyogenes in vivo: a possible explanation for recurrent pharyngotosillitis. The Laryngoscope 1997; 107:640–7. [29] Marouni MJ, Sela S. Fate of Streptoccocus pyogenes and epithelial cells following internalization. J Med Microbiol 2004;53:1–7. [30] Bertuccini L, Ammendolia MG, Superti F, Baldassarri L. Invasion of HeLa cells by Enterococcus faecalis clinical isolates. Med Microbiol Immunol 2002;191:25–31. [31] Baldassarri L, Caprioli A, Donelli G. Adherence to and penetration of cultured cells by an invasive strain of Escherichia coli: an ultrastructural study. Microbiologica 1987;10:317–23. [32] Perry MM, Gilbert AB. Yolk transport in the ovarian follicle of the hen (Gallus domesticus): lypoprotein-like particle at the perifery of the oocyte in the rapid growth phase. J Cell Sci 1979;39:257–72.
Internalization of non-toxigenic Corynebacterium diphtheriae by cultured human respiratory epithelial cells Lucia Bertuccini, Lucilla Baldassarri, Christina von Hunolstein* Dipartimento di Malattie Infettive, Parassitarie ed Immunomediate Istituto Superiore di Sanita`, Viale Regina Elena 299, 00161 Roma, Italy Received 23 February 2004; accepted 10 June 2004
Abstract Although infection by Corynebacterium diphtheriae is a model of extracellular mucosal pathogenesis, and diphtheria is one of the most worried diseases, this microorganism can be associated also with invasive infections such as endocarditis, septic arthritis, and osteomyelitis. Invasive infections are usually caused by non-toxigenic C. diphtheriae strains. Over the last years severe pharyngitis/tonsillitis associated with the isolation of non-toxigenic C. diphtheriae have been described. Penicillin treatment failure of these infections could only partially be explained by penicillin tolerance of the causing strain. Thus, we examined the in vitro ability of non-toxigenic C. diphtheriae throat clinical isolates to adhere to, and enter human respiratory epithelial cells. Trasmission and scanning electron microscopy demonstrated intracellular C. diphtheriae in laryngeal (HEp-2 cells) and pharyngeal (Detroit D562 cells) tissue culture. Live intracellular bacteria were detectable up to 48 h post-infection. Using a variety of compound that act on eukariotic cell structures, the internalization of C. diphtheriae seems to occur via a zipper-like mechanism. It is likely that internalization of C. diphtheriae can be involved in throat colonization contributing to bacterial eradication failure and asymptomatic carriage. q 2004 Elsevier Ltd. All rights reserved. Keywords: Corynebacterium diphtheriae; Adherence; Internalization; Infection; HEp-2 cells; Detroit D562 cells
1. Introduction In the last decade it has become evident that Corynebacterium diphtheriae should not be considered only the etiologic agent of diphtheria. The introduction of mass immunization against diphtheria in the 1950s in most European countries, as well as in USA and Canada, resulted in a very low incidence of this disesase [1] and circulation of toxigenic strains declined in all countries with a good vaccination coverage [2,3]. In contrast, non-toxigenic strains of C. diphtheriae have been increasingly documented as a cause of invasive disease, which include endocarditis, bacteraemia, septic arthritis, osteomyelitis and splenic abscesses [4–9]. Nasopharyngeal carriage and disease caused by this organism has been documented * Corresponding author. Tel.: C39-649-902-659; fax: C39-649902-934. E-mail address: [email protected] (C. von Hunolstein). 0882-4010/$ - see front matter q 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.micpath.2004.06.002
amongst homeless, intravenous drug users and alcoholics [10,11]. Emergence of non-toxigenic C. diphtheriae has also been observed in Italy, England and Wales; in particular, clusters of cases of sore throat associated with the isolation of C. diphtheriae were observed during an enhanced surveillance study [3,12]. The upper respiratory tract is the primary site of colonization by this microorganism. Besides toxin production, little is known about other C. diphtheriae virulence factors. In a recent study [13], penicillin treatment failure of severe pharyngitis/tonsillitis could only partially be explained by penicillin tolerance of the respective isolated strains. Previous studies dealt with the adhesive properties of C. diphtheriae to human buccal epithelial cells and erythrocytes [14,15] and found that this feature was independent from toxin production. Systemic infections caused by C. diphtheriae suggest that the organism is not only able to adhere to host epithelial cells, but must be able to gain access to deeper tissues.
112
L. Bertuccini et al. / Microbial Pathogenesis 37 (2004) 111–118
In the present report, we examined the in vitro ability of non-toxigenic C. diphtheriae clinical isolates to adhere to and enter into human epithelial cells. Using a variety of compounds that act on eukaryotic cells, we investigated the dependence of these processes on microfilaments and microtubules and evaluated the route of entry of C. diphtheriae into cells. 2. Results 2.1. Interaction of non-toxigenic C. diphtheriae strains with epithelial cells To evaluate the ability of non-toxigenic strains of C. diphtheriae to adhere, enter and survive within epithelial cells, HEp-2 cells grown on glass coverslips were infected with three different strains (Table 1). Two strains (3319 and 4060) were non-toxigenic clinical isolates of patients affected by pharyngitis/tonsillitis; C. diphtheriae C7 (b197) is a lysogenic strain which produces a non-toxic diphtheria protein due to a mutation in the fragment A, resposible for enzymatic activity. After one hour of infection, bacteria showed a specific association with cell monolayers with a localized adherence pattern, when observed by light microscopy (Fig. 1a). After further 2 h of incubation with gentamicin, bacteria were visible inside cells vacuoles (Fig. 1b), the average number of viable bacteria, recovered on agar plates after cell lysis, being slightly different for the three strains (Table 1). Adhesion and internalization seemed independent from the capacity to produce diphtheria toxin, as evidenced by the adherence and internalization rates of the strain C7 (b197). Strain 3319 was chosen for further investigations to better understand the kinetic of entry of non-toxigenic C. diphtheriae into epithelial cells. The invasion efficiency of C. diphtheriae in HEp-2 cells was compared with that observed in human pharyngeal epithelial cells (Detroit 562, D562). As shown in Fig. 2, at time T0 viable bacteria associated with HEp-2 cells were significantly more than those associated with D562 cells. After 3 and 6 h of treatment with gentamicin, the average number of viable bacteria recovered in D562 cells continued to be
Fig. 1. After 1 h of infection of HEp-2 cells with non-toxigenic C. diphtheriae (strain 3319) a localized pattern of adherence was visible (a); after further 2 h of incubation with gentamicin, bacteria were visible inside HEp-2 cellular vacuoles (b). (1000!).
lower than that in HEp-2 cells, but after more than 24 h of incubation the survival of internalized bacteria into D562 was higher. At 48 h post-infection, bacteria were still recovered on agar plates after lysis of D562 cells (Fig. 2). Similar internalization rate was also observed for the other clinical isolates used in this study (data not shown).
Table 1 Adhesion to and internalization into HEp-2 cells of three strains of C. diphtheriae (2!105 cells/ml were infected with 2!107 CFU/ml (MOI 100) Strain
Viable bacteria (CFU/ml)a Adherentb
3319 4060 C7(b197) a
6
Internalizedc 6
5.6!10 G4.3!10 5.8!105G3.4!105 2.2!106G0.9!106
1.1!105G0.7!105 5.3!103G2.8!103 9.0!104G3.0!104
Values represent the means of three separate experiments. b Average number of viable bacteria recovered on agar plates after 1 h of infection. c Average number of viable bacteria recovered on agar plates after 1 h of infection and further 2 h of treatment with gentamicin.
Fig. 2. Viable (Log10 CFU/mlGSD) C. diphtheriae (strain 3319) recovered from HEp-2 (-*-) and D562 (-B-) cells after different time of infection. At 1 h post-infection (T0), monolayers were washed to remove unadsorbed bacteria and further incubated with fresh medium containing gentamicin (100 mg/ml) for 3 (T3), 6 (T6), 24 (T24), and 48 (T48) hours. Data are from three independent experiments.
L. Bertuccini et al. / Microbial Pathogenesis 37 (2004) 111–118 Table 2 Adhesion and internalization of C. diphtherae strain 3319. D562 cells (2!105 cells/ml) were infected with different bacterial concentrations, ranging from 2!106 to 2!109 CFU/ml, corresponding to MOI 10–10,000 Multiplicity of infection (MOI)
Viable bacteria (CFU/ml)a Adherentb
Internalizedc
10 100 1000 10,000
9.6!103G8.5!103 3.2!104G2.4!104 3.0!105G1.1!105 8.2!105G5.4!105
41.0G33.0 4.7!102G4.1!102 1.7!103G1.0!103 1.1!103G1.0!103
a
Values represent the means of three separate experiments. Average number of viable bacteria recovered on agar plates after 1 h of infection. c Average number of viable bacteria recovered on agar plates after 1 h of infection and further 2 h of treatment with gentamicin (100 mg/ml). b
Furthermore, it was investigated how interaction between C. diphtheriae and D562 cells could be modified by changing the bacteria/cell ratio (multiplicity of infection, MOI). When it was varied from 10:1 to 10,000:1,
113
the number of bacteria associated to D562 cells increased continuously up to a plateau between MOI 1000 and 10,000, without damage to the monolayers (Table 2). After 2 h of treatment with gentamicin, the level of intracellular bacteria had the same trend at different MOIs; from MOI 100 to MOI 1000 a slight increase of internalized C. diphtheriae was observed while at higher MOIs internalization remained costant (Table 2). 2.2. Ultrastructural analysis of the internalization process of C. diphtheriae Monolayers were infected with C. diphtheriae for 1, 2 and 3 h and processed for scanning electron microscopy to analyze the first step of interaction between bacteria and host cells. After 1 h of incubation, most of epithelial cells were colonized by clumped bacteria which showed a localized pattern of adherence (Fig. 3a and b), confirming
Fig. 3. Interaction of C. diphtheriae (strain 3319) with D562 cells viewed by scanning (a)–(e) and trasmission electron microscopy (f) at 1 h (a)–(c), 3 h (d) and (e) and 48 h (f) of post-infection. (a) Small bacterial clumps adhere to the epithelial cell surface. (b) Cellular microvilli appear in close contact with the polar side of adherent rods. (c) Bacterial contact to cellular microvilli induced membrane protrusion towards the entire length of the rod. (d) Cellular protrusions engulf a bacterial cell with a zipper-like mechanism of phagocytosis. (e) A peculiar ring-like structure, originating from the host cell membrane internalizes a single rod. (f) Forty-eight hours post-infection, numerous C. diphtheriae bacilli are still present inside cells in tight fitting vacuoles, without sign of digestion.
114
L. Bertuccini et al. / Microbial Pathogenesis 37 (2004) 111–118
the light microscopy observations. Adhesion was mediated by cellular microvilli which seemed to contact bacterial cells firstly in the polar region and then extend to the entire length of the rod (Fig. 3c), establishing a close contact with the bacterial cell wall. The extention of host cell protrusions, guided by cytoskeletal rearrangements, induced a sequential engagement of cell membrane which engulfed the bacterium and formed a tight-fitting vacuole (Fig. 3d). In the 3 h samples these events were more frequently observed. Fig. 3e shows a very peculiar ring-like structure, with a ruffle-like appearance although strictly adherent to the bacterial surface, which directly originates from the host cell surface to internalize a single rod. Inside cells, C. diphtheriae was present in tight-fitting vacuoles (Fig. 3f). At 24 and 48 h post-infection, internalized bacteria were still found without evident signs of degradation (Fig. 3f). Fluorescence actin staining with tetramethyl rhodamine isothiocyanate-conjugated phalloidin, which binds
specifically to actin filaments, showed a partial condensation of actin filaments around bacteria that were just internalized by cells (not shown). 2.3. Effects of inhibitors on the internalization process Specific inhibitors of cell functions were used to examine the mechanism of internalization induced in epithelial cells by C. diphtheriae. Cytocalasin D, which blocks F-actin polymerization, showed a dose-dependent inhibitory effect on the internalization of bacteria, indicating the involvement of actin microfilaments in the bacterial uptake. The decrease of viable bacteria was more evident at the highest dose utilized that determined a reduction of internalization of 87% (Fig. 4a). In contrast, microtubules did not seem to be involved in C. diphtheriae entry into D562 cells, as colchicine had no significant effects on the average
Fig. 4. Effect of different compounds on C. diphtheriae (strain 3319) invasion process into D562 cells. Internalization rates were strongly affected by treatment with Cytocalasin D (a), while colchicine (b) had no effect. Staurosporin (c), but not wortmannin (d), enhanced bacterial adhesion (-B-) and entry (->-) into cells. Monolayers were pretreated with each chemical for 30 min at indicated concentrations. Only cytocalasin D and colchicine were maintained troughout the 2 h internalization period. Viable bacteria were determined as Log10 CFU/mlGSD, recovered on agar plates after cells lysis. Data are from three independent experiments.
L. Bertuccini et al. / Microbial Pathogenesis 37 (2004) 111–118
number of viable bacteria after gentamicin protection assay (Fig. 4b). Monolayers were pre-treated with monodansylcadaverine (200 mM), a primary amine that blocks receptor recycling to cellular membrane, to evaluate the involvement of membrane receptors in the internalization process of C. diphtheriae. Cells treatment affected bacterial uptake, reducing the percentage of viable bacteria to 18% of those recovered in non-treated cells (data not shown). To investigate whether the binding to membrane involved a signal cascade mediated by protein kinases (PKs), cellular monolayers were treated with specific inhibitors of these enzymes. Exposure of cells to increasing doses of staurosporine, a potent inhibitor of different protein kinase classes including protein kinase C (PKC), some phosphotyrosinekinases (TPKs) and cAMP-dependent PK, produced a significant enhancement of the bacterial adhesion and internalization into cells (23-fold). This effect was more relevant at concentrations in the nanomolar range (Fig. 4c). Moreover, treatment of D562 cells with increasing doses of wortmannin, a compound which blocks phosphatidilinositol-trisphosphate kinase (PI3-K), did not affect the level of both adherence and internalization of C. diphtheriae (Fig. 4d). As well, no effect was observed after treatment with orthovanadate, an inhibitor of cell phosphatases (not shown).
3. Discussion and conclusions Non-toxigenic C. diphtheriae strains have been increasingly isolated from different types of infection, from cutaneous lesions and pharyngitis to bacteremia and endocarditis [3,5,7,12,16,17]. Several studies have underlined that immunization against diphtheria toxin does not protect from the insurgence of non-toxigenic C. diphtheriae diseases, supporting the hypothesis that diphtheria bacilli may possess other factors enhancing their virulence [5,7,11]. C. diphtheriae is generally considered a non-invasive pathogen that causes localized infection of the higher respiratory tract, with extensive peripheral tissue damages due to the diffusion of the diphtheria toxin [18]. In this study we show that non-toxigenic C. diphtheriae strains can be internalized by non-professional phagocytes. Two different cell lines were used, a larynx- (HEp-2) and a pharyngealderived (D562), as in vivo both larynx and pharynx are colonized by C. diphtheriae. Infection of HEp-2 monolayers with non-toxigenic strains indicated that C. diphtheriae is able to adhere to the epithelial cell surface and, by light and electron microscopy, that bacilli are present in intracellular vacuoles. Bacterial cells internalization seemed to be a strain-dependent process as the entry efficiency was slighty different for each strain examined. Hirata et al. have previously described the ability of toxigenic C. diphtheriae strains to enter HEp-2 cells with
115
different levels of uptake [19]. Our results showed that also non-toxigenic C. diphtheriae isolates were able to penetrate respiratory epithelial cells of both pharyngeal and laryngeal origin and that internalization does not seem to be influenced by toxin production. C. diphtheriae showed different behaviours with the two cell lines used. Adhesion and internalization rates in HEp-2 cells were significantly higher than in D562 cells; however, intracellular survival was more efficient in the latter and intact bacterial cells were still recovered after 48 h of infection. Thus, both cellular and bacterial determinants appear to be involved in C. diphtheriae internalization by epithelial cells. Although the number of internalized bacteria was lower than that observed for invasive pathogens sensu stricto (i.e. Shigella, enteroinvasive Escherichia coli, etc.), the possibility to hide into epithelial cells even at low numbers and to survive for at least two days, would provide C. diphtheriae with the opportunity to escape from the host’s humoral immunity and/or antibiotic therapy and possibly to spread to distant sites. Results obtained by changing bacteria/cell ratio have shown that C. diphtheriae invasion of D562 cells is a saturable process, reaching maximal number of internalized bacteria at a MOI between 100 and 1000. This may reflect a limited number of host cell receptors or entry sites [20], a limitation of some other host biochemical requisite for entry, or host cell modifications occuring during internalization, preventing penetration of further C. diphtheriae cells. Supporting the hyphothesis of a receptor-mediated entry, scanning and thin-section electron microscopy images are consistent with a zipper-like mechanism of phagocytosis with an initial binding of the bacteria to the host cell receptors leading to the formation of membrane protrusions and subsequent extention of cell membrane around the bacteria, until complete engulfment [21]. The intimate contact of cellular membrane protrusions with the bacterial cell wall was clearly visible in electron micrographs and, even after 48 h of infection, vacuole membrane remained strictly adherent to the C. diphtheriae cell wall, that did not show any sign of degradation. Other bacterial pathogens, such as Yersinia and Listeria, are internalized by a similar mechanism by non-professional phagocytes [22,23]; furthermore, Helicobacter pylori, considered a non-invasive pathogen, was found entering gastric mucosa cells through a zipper-like mechanism [24]. The inhibitory effect induced by treatment of D562 monolayers with monodansylcadaverine confirmed that C. diphtheriae internalization process needs cellular membrane receptors, as this compound blocks receptor recycling to plasma membrane [25]. Also, the dose dependent inhibition effect of cytocalasin D on bacterial entry and the partial condensation of actin observed around the rods, support the idea that bacterial binding to specific sites of epithelial cell surface may induce the intracellular signal transduction responsible for actin microfilaments
116
L. Bertuccini et al. / Microbial Pathogenesis 37 (2004) 111–118
rearrangement and pseudopods movement [26]. Staurosporine, an inhibitor of different classes of PKs, upregulated C. diphtheriae internalization, suggesting that activation of PKC or TPKs or cAMP-depending ATPases negatively controlled C. diphtheriae uptake; this effect was specific, since other inhibitors such as wortmannin, that blocks the PI3K activity [27], and orthovanadate, an inhibitor of cell phosphatases [24], had no effect. The increased internalization of C. diphtheriae is in contrast to other reports where PKs inhibition led to reduced bacterial entry [26,27]. Genistein, a specific inhibitor of TPKs, was shown to block C. diphtheriae entry into HEp-2 cells [19]. Thus, it is likely that PKC is involved in this particular mechanism of enhanced internalization. Our hypothesis is that C. diphtheriae binding to cell membrane induces a reorganization of the actin cytoskeleton which in turn affects the dynamic of receptors involved in internalization of C. diphtheriae, limiting the number of internalized bacteria; alteration of this feed-back inhibition process would eliminate such limit. This hypotesis is supported by the trend of adhesion that parallels that of internalization, as seen by pretreatment with staurosporine. On the basis of the results of this study, non-toxigenic C. diphtheriae should not be considered exclusively an extracellular microorganism. Its behaviour resemble that of Streptococcus pyogenes also found to be internalized by non-phagocytic cells [28,29]. The internalization of C. diphtheriae by cultured laryngeal and pharyngeal cells may resembles the events occurring in human host with recurrent pharyngitis/tonsillitis after seemingly adequate antibiotic treatment [3,12], or in asymptomatic carriage.
4. Materials and methods 4.1. Bacterial strains and growth conditions Adherence and internalization of C. diphtheriae was studied using non-toxigenic clinical strains isolated from patients affected by pharyngis/tonsillitis: 3319, 4060, 95354, 95325, 95492, and 96161. A lysogenic strain of C. diphtheriae, C7 (b-197), was kindly provided by Dr R. Rappuoli. Clinical isolates were identified as C. diphtheriae using API Coryne (bioMe´rieux). Strains were identified as non-toxigenic by the Elek test, PCR and Southern blotting [3]. Gentamicin, used in the cell assays, was bactericidal for all strains at concentration of 100 mg/ ml for 2 h. For cell assays, bacteria were grown overnight at 37 8C in Todd–Hewitt broth (THB, Oxoid)) with agitation, resuspended in tissue-culture medium (see below) and adjusted spectrophotometrically at 600 nm at a concentration of 107 CFU/ml as checked by plating on THB agar of serial dilutions of the culture and CFU counts.
4.2. Cell cultures HEp-2 cells (human larynx carcinoma, ATCC, Manassas, VA) were cultured in Minimal Essential Medium (Eagle’s) (MEM, Life Technologies) supplemented with 0.1 mM non-essential aminoacids, 2 mM glutamine, penicillin 100 UI/ml, streptomycin 100 mg/ml, and 10% heat inactivated fetal calf serum (FCS, Life Technologies) in a humidified atmosphere of 5% CO2 and 95% air at 37 8C. Cells were passaged at a ratio of 1:5 twice each week. Detroit 562 cells (D562) (human pharynx carcinoma, ATCC CCL-138) were cultured in MEM supplemented with 0.1 mM non-essential amino acids, 2 mM glutamine, 1 mM sodium pyruvate, and 10% heat inactivated FCS in a CO2 incubator. Cells were passaged at a ratio of 1:3 twice each week. 4.3. Adherence and internalization assays Assays were performed as previously described [30] with slight modifications. Semi-confluent cells, grown in 24-well plates, were infected with stationary phase bacteria at different MOIs (bacteria-to-cell ratio 100:1) for 1 h at 37 8C in a humidified atmosphere of 5% CO2 and 95% air. At 1 h post-infection (T0), monolayers were washed three times with phosphate-buffered salt solution (PBS), pH 7.4, to remove unadsorbed bacteria, and further incubated with fresh medium containing gentamicin 100 mg/ml for 3 (T3), 6 (T6), 24 (T24), and 48 (T48) hours. At different time points cells were lysed with triton X-100 (0.02% in H2O) for 5 min, serially diluted in PBS and 10 ml of each dilution plated in triplicates on THB agar. As each condition was performed in duplicate, the mean numberGSD of CFU recovered for the two wells was determined. For light microscopy examination, monolayers were grown on glass coverslips in 24 wells/ plate, fixed at the different time points with methanol, and stained by MayGrunwald-Giemsa [31]. All experiments were repeated at least three times. 4.4. Chemicals Cytochalasin D and colchicine (Sigma-Aldrich, Milan, Italy), diluted in phosphate buffered saline (PBS), (pH 7.4), were prepared as stock solutions and added to the cell culture medium at different concentrations. To analyze cell structures involved in internalization of corynebacteria, inhibitors were added to the incubation medium 30 min before infection and were present throughout the assay. Monodansylcadaverine, wortmannin and staurosporin (all from Sigma-Aldrich, Milan, Italy) were stocked in DMSO and diluted at different concentrations for monolayers pretreatment. After 30 min of incubation with each inhibitor, cells were washed with fresh medium and infected with corynebacteria for internalization assay.
L. Bertuccini et al. / Microbial Pathogenesis 37 (2004) 111–118
All inhibitors were tested for possible effects on cell and bacterial viability by trypan blue exclusion test and CFU counting of treated and untreated samples, respectively. 4.5. Fluorescence microscopy To monitore the depolymerization of microfilaments (MF) and microtubules (MT), cells grown on glass coverslips were fixed with paraformaldeyde for 15 min and permeabilized with Triton X-100 (0.5% in PBSC1% BSA) for 5 min. MF were subsequently treated with diluted (1:200) tetramethyl rhodamine isothiocyanate-labelled phalloidin (Sigma-Aldrich) for 30 min. MT were visualized by monoclonal mouse anti a- and b-tubulin antibodies (1:50) (Sigma-Aldrich), incubated for 30 min at 37 8C. After extensive washing with PBS, cells were further incubated for 30 min with TRITC-labelled anti-mouse antibodies (Sigma-Aldrich). Slides were mounted in PBS-glycerine and examined by fluorescence microscopy. 4.6. Scanning electron microscopy (SEM) Infected monolayers were first fixed with 2.5% glutaraldehyde in 0.1 M cacodylate-buffered (Merck, Darmstadt, Germany) for 20 min at room temperature and post-fixed with 1% OsO4 for an additional hour. Samples were then dehydrated through a graded series of ethanol, critical point dried and gold sputtered, and examined by a Cambridge SE360 scanning electron microscope. 4.7. Transmission electron microscopy (TEM) Infected monolayers were processed at different times according to Perry et al. [32]. Cells were fixed with 2.5% glutaraldehyde, 2% paraformaldehyde and 2 mM CaCl2 in 0.1 M sodium cacodylate buffer (pH 7,4) overnight at 4 8C. Monolayers were washed in cacodylate buffer and postfixed with 1% OsO4 in 0.1 M sodium cacodylate buffer for 1 h at room temperature, treated with 1% tannic acid in 0.05 M cacodylate buffer for 30 min and rinsed in 1% sodium solphate in 0.05 cacodylate for 10 min. Fixed specimens were washed, dehydrated through a graded series of ethanol solutions (30–100% ethanol) and embedded in Agar 100 (Agar Scientific Ltd, UK). Ultrathin sections obtained by a MT-2B Ultramicrotome (LKB—Pharmacia) were stained with uranyl acetate and lead citrate and examined by an EM 208 Philips electron microscope.
Acknowledgements This work was partially supported by the Istituto Superiore di Sanita` grant N8 1024/RI (C.v.H.). We gratefully acknowledge Dr. A. Efstratiou (Health Protection Agency, London, UK) for providing us the strains 95354, 95325, 95492 and 96161.
117
References [1] Galazka A. The changing epidemiology of diphtheria in the vaccine era. J Infect Dis 2000;181(Suppl 1):S2–S9. [2] Saragea A, Maximescu P, Meiter E. Corynebacterium diphtheriae. In: Bergan T, Norris JR, editors. Microbiological methods used in clinical and epidemiological investigations—methods in microbiology, vol. 13. London, New York, Toronto, Sydney, San Francisco: Academic Press; 1979, p. 62–176. [3] von Hunolstein C, Alfarone G, Scopetti F, et al. Molecular epidemiology and characteristics of Corynebacterium diphtheriae and Corynebacterium ulcerans strains isolated in Italy during the 1990s. J Med Microbiol 2003;52:181–8. [4] Alexander D. Splenic abscess caused by Corynebacterium diphtheriae. Clin Pediatr 1984;23:591–2. [5] Funke G, Altwegg M, Frommel L, von Graevenizt A A. Emergence of related non-toxigenic Corynebacterium diphtheriae biotype mitis strains in Western Europe. Emer Infect Dis 1999;5: 477–80. [6] Poilane I, Fawaz F, Nathanson M, et al. Corynebacterium diphtheriae osteomyelitis in an immunocompetent child: a case report. Eur J Pediatr 1995;154:381–3. [7] Patey O, Bimet F, Riegel P, et al. Clinical and molecular study of Corynebacterium diphtheriae systemic infections in France. J Clin Microbiol 1997;35:441–5. [8] Belko J, Wessel D, Malley R. Endocarditis caused by Corynebacterium diphtheriae: case report and review of the literature. Pediatr Infect Dis J 2000;19:159–63. [9] Mattos-Guaraldi AL, Formiga LC, Camello TC, Pereira GA, Hirata Jr R, Halpern M. Corynebacterium diphtheriae threats in cancer patients. Rev Argent Microbiol 2001;33:96–100. [10] Gubler J, Huber-Schneiner C, Gruner E, Altwegg M. An outbreak of non-toxigenic Corynebacterium diphtheriae infection: single bacterial clone causing invasive infection among Swiss drug users. Clin Infect Dis 1998;27:1295–8. [11] Wilson APR. The return of Corynebacterium diphtheriae: the rise of non-toxigenic strains. J Hosp Infect 1995;30:306–12. [12] Reacher M, Ramsay M, White J, et al. Non-toxigenic Corynebacterium diphtheriae: an emerging pathogen in England and Wales? Emer Infect Dis 2000;6:640–5. [13] von Hunolstein C, Scopetti F, Efstratiou A, Engler K. Penicillin tolerance amongst non-toxigenic Corynebacterium diphtheriae isolated from cases of pharyngitis. J Antimicrob Chemother 2002;50: 125–8. [14] Colombo AVC, Hirata Jr R, Rocha de Souza CM, et al. Corynebacterium diphtheriae surface proteins as adhesins to human erythrocytes. FEMS Microbiol Lett 2001;197:235–9. [15] Mattos-Guaraldi AL, Formiga LCD, Pereira GA. Cell surface components and adhesion in Corynebacterium diphtheriae. Microbes Infect 2000;2:1507–12. [16] Hogg GG, Strachan JE, Huayi L, Beaton SA, Robinson PM, Taylor K. Non-toxigenic Corynebacterium diphtheriae biovar gravis: evidence for an invasive clone in a south-eastern Australian community. Med J Australia 1996;164:72–5. [17] Mattos-Guaraldi AL, Formiga LCD. Bacteriological properties of a sucrose-fermenting Corynebacterium diphtheriae strain isolated from a cases of endocarditis. Curr Microbiol 1998; 37:156–8. [18] Mac Gregor RR. In: Mandell GL, Bennett JR, Dolin R, editors. Corynebacterium diphtheriae. Mandell, Douglas and Bennett’s principles and practice of infectious diseases. New York: Churchill Livingstone; 1995, p. 1865–72. [19] Hirata RJr, Napolea˜o F, Monteiro-Leal LH, et al. Intracellular viability of toxigenic Corynebacterium diphtheriae strains in HEp-2 cells. FEMS Microbiol Lett 2002;215:115–9.
118
L. Bertuccini et al. / Microbial Pathogenesis 37 (2004) 111–118
[20] Lan Hu, Kopecko DJ. Campylobacter jejuni 81–176 associates with microtubules and dynein during invasion of human intestinal cells. Infect Immun 1999;67:4171–82. [21] Swanson JA, Baer SC. Phagocytosis by zipper and triggers. Trends Cell Biol 1995;5:89–93. [22] Iseberg RR. Discrimination between intracellular uptake and surface adhesion of bacterial pathogens. Science 1991;252:934–8. [23] Mengaud J, Ohayon H, Gounon P, Me´ge RM, Cossart P. E-cadherin is the receptor for internalin, a surface protein required for entry of Listeria monocytogenes into epithelial cells. Cell 1996; 84:923–32. [24] Kwok T, Backert S, Schwarz H, Berger J, Meyer T. Specific entry of Helicobacter pylori into cultured gastric epithelia cells via a zipperlike mechanism. Infect Immun 2002;70:2108–20. [25] Wileman T, Harding C, Stahl P. Receptor mediated endocytosis. Biochem J 1985;232:1–14. [26] Rosenshine I, Duronio V, Finlay B. Tyrosine protein kinase inhibitors block invasin-promoted bacterial uptake by epithelial cells. Infect Immun 1992;60:2211–7.
[27] Benjamin P, Federman M, Wanke CA. Characterization of an invasive phenotype associated with enteroaggregative Escherichia coli. Infect Immun 1995;63:3342–417. [28] Osterlund A, Popa R, Nikkila T, Scheynius T, Engstrand L. Intracellular reservoir of Streptococcus pyogenes in vivo: a possible explanation for recurrent pharyngotosillitis. The Laryngoscope 1997; 107:640–7. [29] Marouni MJ, Sela S. Fate of Streptoccocus pyogenes and epithelial cells following internalization. J Med Microbiol 2004;53:1–7. [30] Bertuccini L, Ammendolia MG, Superti F, Baldassarri L. Invasion of HeLa cells by Enterococcus faecalis clinical isolates. Med Microbiol Immunol 2002;191:25–31. [31] Baldassarri L, Caprioli A, Donelli G. Adherence to and penetration of cultured cells by an invasive strain of Escherichia coli: an ultrastructural study. Microbiologica 1987;10:317–23. [32] Perry MM, Gilbert AB. Yolk transport in the ovarian follicle of the hen (Gallus domesticus): lypoprotein-like particle at the perifery of the oocyte in the rapid growth phase. J Cell Sci 1979;39:257–72.