- Email: [email protected]
Noticeable differences in bacterial defence on tonsillar surfaces between bacteria-induced and virus-induced acute tonsillitis
International Journal of Pediatric Otorhinolaryngology (2003) 67, 1075 /1082
www.elsevier.com/locate/ijporl
Noticeable differences in bacterial defence on tonsillar surfaces between bacteria-induced and virus-induced acute tonsillitis ¨isa ¨nenb Lars-Eric Stenforsa,*, Helga-Marie Byea, Simo Ra a
Department of Otolaryngology, Institute of Clinical Medicine, University of Tromsø, N-9037 Tromsø, Norway b Clinical Laboratory, Central Hospital of Keski-Pohjanmaa, FIN-67200 Kokkola, Finland Received 30 January 2003; received in revised form 22 May 2003; accepted 1 June 2003
KEYWORDS Antibacterial activity; Tonsillar disease; Mucosal immunity; Streptococcus pyogenes ; Epstein /Barr virus
Summary Objective : Oral and pharyngeal cavities harbor a commensal bacterial flora which is kept in check by several innate and acquired agents. In this study, we focused on the proportions in which some antibacterial moderators (lysozyme, lactoferrin, IgG and S-IgA) coat the tonsillar surface bacteria in healthy individuals, in patients with acute tonsillitis (AT) culture-positive for Streptococcus pyogenes , and in patients with infectious mononucleosis (IM) caused by Epstein /Barr virus (EBV). Methods : Bacterial samples were collected for aerobic culturing and immunocytochemical evaluation from the tonsillar surfaces of eight healthy individuals (four males, four females; age range 16 /22 years), eight patients with current AT (two males, six females; age range 16 /29 years) and seven patients with IM (four males, three females; age range 15 /21 years). The immunocytochemical assay was based on gold-labeled antiserum to human lysozyme, lactoferrin, IgG and S-IgA followed by gold particle tracing in the transmission electron microscope. Results : During AT, a significant increase in lysozyme coating (P B/0.05) and lactoferrin coating (P B/ 0.0005) of the bacteria was noted, whereas the S-IgA coating was significantly reduced (P B/0.0005). During IM infection, a significant increase in lactoferrin coating was noted (P B/0.0005) whereas immunoglobulin coating was significantly reduced (IgG P B/0.025; S-IgA P B/0.0005) compared with healthy controls. During IM, all antibacterial moderators evaluated were significantly reduced compared with the situation during AT. Conclusions : Noticeable changes in the local innate and acquired bacterial defence system were observed during tonsillar infections, particularly during IM. – 2003 Elsevier Ireland Ltd. All rights reserved.
*Corresponding author. Tel.: /47-776-2-7390; fax: /47-776-4-4772. E-mail address: [email protected] (L.-E. Stenfors). 0165-5876/03/$ - see front matter – 2003 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/S0165-5876(03)00194-0
1076
1. Introduction It is evident that the oropharyngeal cavity possesses a high degree of immunity. Large wounds resulting from tonsillectomy normally heal rapidly despite contamination by the vast bacterial flora that proliferates in this region. This fortunate outcome is probably results from local immune defence, in which saliva, salivary components and flow [1], non-immunological and immunological antibacterial substances [2 /4], leukocytes [5] and desquamation of epithelial cells all play a central role. Numerous antibacterial agents (lysozyme, lactoferrin, peroxidase, antibacterial peptides, immunoglobulins, etc.) which are secreted into the saliva, inhibit the metabolism, adherence and viability of microorganisms in vitro, but their clinical role on the palatine tonsils and tonsillar fossae is largely unknown. It would seem, however, that they are important for the control of microbial overgrowth in the oropharynx, though just how selective they are against pathogens in situ is not known. The main bacterial pathogen causing roughly 30% of all acute tonsillitis (AT) cases is Streptococcus pyogenes [6,7]. By means of an immunocytochemical technique, using gold-labeled antiserum to S. pyogenes and several antimicrobial agents, we have succeeded in localizing these pathogens on tonsillar surfaces during AT [8,9] and their coating with IgG and secretory IgA (S-IgA) immunoglobulins [10] as well as lactoferrin [11]. When using the same immunocytochemical technique, it was found that coating of the tonsillar surface bacteria with IgG and S-IgA immunoglobulins was greatly suppressed during infectious mononucleosis (IM) [12,13], whereas bacterial coating with lactoferrin increased slightly compared with healthy controls [14]. During IM, caused mostly by Epstein/Barr virus (EBV), a characteristic sign is an immense bacterial colonization of the palatine tonsils, forming a membranous tonsillitis [15,16]. In both of these diseases, it is obvious that the overall bacterial defence on the mucosal membranes of the palatine tonsils is weakened in some way. The purpose of the present study was to compare the extent to which some antibacterial substances attached themselves to tonsillar surface bacteria during bacteria-induced and virus-induced infection of the palatine tonsils. Of the numerous antimicrobial factors present in sputum, we chose the two non-immunological substances having the greatest concentration, viz. lysozyme [1,17] and
L.-E. Stenfors et al.
lactoferrin [1,18]. Furthermore, bacterial coating with IgG and S-IgA was examined.
2. Material and methods 2.1. Controls and patients 2.1.1. Control group For control purposes we used eight healthy individuals with good oral health (four males, four females; age range 16 /22 years, median age 20) attending a normal health check-up. None of the control subjects had taken an antibiotic within 2 months of bacterial sampling from the palatine tonsils. 2.1.2. AT group Eight patients (two males, six females; age range 16 /29 years, median age 25) with classical symptoms of AT (sore throat, dysphagia, axillary body temperature /38 8C, local inflammatory changes in the palatine tonsils and yellowish exudate on the tonsils) comprised the AT group. None of the patients had taken an antibiotic within 2 months of bacterial sampling. All patients displayed a positive rapid test result for S. pyogenes (Quick Vue In-Line One-Step Strep A Test system, Quides, Rosny, France) and subsequent throat culture revealed growth of S. pyogenes . 2.1.3. IM group Seven patients (three males, four females; age range 15 /21 years, median age 17) with IM and concomitant membranous tonsillitis who were referred to our out-patient department comprised the IM group. The IM diagnosis was based on clinical and serological findings. All patients evidenced lymphadenopathy, lymphocytosis with /10% atypical McKinley cells, a positive heterophil antibody reaction (Monosticon Dri-Dot slide test, Organon Diagnostics, Durham, NC, USA) and a Paul /Bunnell test with titre /40. All patients showed positive IgG and/or IgM titres to capsid EBV [19]. Moreover, the IgG avidity was very low, hinting at a primary EBV infection of recent date [20]. None of the patients had a subnormal serum immunoglobulin level according to an immunoturbidimetric assay. The study was approved by the Medical Ethics Committee of our hospital. Informed consent was always obtained from the patients and/or patient’s parents before sampling.
Differences in bacterial defence between bacteria- and virus-induced tonsillitis
2.2. Aerobic culturing To obtain throat samples from the palatine tonsils, a cottonwool-tipped wooden swab, which had been dipped in physiological saline, was rotated against the mucosa. Samples were then smeared over blood-agar and chocolate-agar plates for standard aerobic culturing according to routine laboratory procedures. The plates were kept incubated in 5% CO2 at 37 8C for 48 h. Growth of S. pyogenes was identified by means of colony morphology, b-haemolysis, a rapid agglutination test of material obtained from bacterial colonies (Streptex A, Diagnostic Products Corporation, Los Angeles, CA, USA) and PYR test (Api 20 Strep, bioMe ´rieux, Marcy l’Etoile, France). Semiquantification of S. pyogenes was performed by calculating the colony-forming units (CFU) of S. pyogenes versus other aerobic colonies in equal sections of the agar plates.
2.3. Immunocytochemical assay Swabbed cell material was obtained from both palatine tonsils and loosened from the swab by rinsing it in physiological saline. No food, snacks, coffee or toothbrushing was allowed within 2 h prior to bacterial sampling. The cell mixture was given a light centrifugation to harvest the cells at the bottom of the tubes. Supernatant was carefully discharged and the precipitates carefully adjusted to 0.5 ml with physiological saline. To this cell mixture, 0.5 ml 8% formaldehyde buffered in 200 mM HEPES fixative was added to give a working concentration of 4% formaldehyde. After fixation, the mixture was centrifugated at 16 000 rpm for 1 / 2 min. The supernatant was pipetted away and the cells diluted in 100 ml 10% gelatin. The samples were centrifugated briefly (15 s) and kept on ice for 2 h. The cell pellet was excised, immersed in 2.3 M sucrose overnight, placed on aluminium specimen pins and frozen by immersion in liquid nitrogen. Ultrathin sections were made using a Reichert Ultracut S ultramicrotome with a FCS cryochamber (Leica, Deerfield, IL, USA). The sections (ca. 30/60 nm) were cut with diamond knives and retrieved from the knife surface with a drop of methyl cellulose mixed with sucrose. The specimens were then mounted on carbon-coated grids, which were placed in 1% cold fish-skin gelatin (FSG) for 15 min. Each grid was placed on a 5-ml droplet of the actual antiserum, which was diluted in accordance with earlier titration of the antiserum. After renewed washing in PBS, the grids were placed on protein A-gold diluted in 1% FSG (gold particle 5
1077
nm) and then washed with PBS. When doublestained, the grids were fixed with 1% glutaraldehyde (blocking step to prevent second A-gold probe from binding to the first antibody), followed by washing in PBS, after which the grids were placed in 0.12% glycine. After further washing in PBS the specimens were placed on droplets of the second antiserum. After washing in PBS, the grids were placed on protein A-gold (gold particle 10 nm) diluted in 1% FSG for 15 min. Further washing with distilled water was followed by drying in the presence of 3% uranyl acetate in 2% methyl cellulose. The sections were examined in a JEOL JEM 1010 transmission electron microscope operated at 80 kV. From each patient, three grids, either single or double stained for each substance, were evaluated. From each grid roughly 1000 bacteria were scrutinized by two of the authors (H.-M. B., L.-E. S.) independently without knowledge of the origin of the sample. The following antisera were used: rabbit antiserum to S. pyogenes (diluted in FSG 1:200; Difco Labs, Detroit, MI, USA); rabbit antiserum to human lysozyme (dilution in FSG 1:150; DAKO A/S, Glostrup, Denmark); rabbit antiserum to human lactoferrin (dilution in FSG 1:200; DAKO); rabbit antiserum to human IgG (dilution in FSG 1:100; Organon Teknika, Westchester, PA, USA); goat antiserum to human S-IgA (dilution in FSG 1:100; Organon Teknika Co). Prior to investigation, the antisera were titrated in order to minimize background attachment of the gold particles. As positive controls we used plain cultures of S. pyogenes and Staphylococcus aureus coated with lysozyme, lactoferrin, IgG or S-IgA. As negative controls we used plain cultures of Streptococcus mitis , Escherichia coli , Neisseria sicca , Streptococcus salivarius and Streptococcus equisimilis . Handling the specimens with 1% FSG and protein-A gold without antisera, or with antisera absorbed with excess antigen, proved negative. All samples were handled in exactly the same way. In the double-label experiments, care was taken to ensure that labeling followed the patterns in single-label experiments, and also that they were the same in crossover controls.
2.4. Statistical analysis Wilcoxon’s rank sum test was used to compare values within the groups and Mann /Whitney’s U test to compare values between groups. P -values below 0.05 were considered statistically significant.
1078
L.-E. Stenfors et al.
3. Results
infection culture-positive for S. pyogenes and IM caused by EBV, are summarized in Table 1 (median values and ranges) and Table 2 (differences with P values).
3.1. Aerobic culture 3.1.1. Healthy controls Aerobic culturing of bacterial samples obtained from tonsillar surfaces of the healthy control individuals evidenced growth of mainly a-haemolytic viridans Streptococci , Corynebacterium spp., Neisseria spp., and coagulase-negative Staphylococci . In four controls, upper respiratory tract pathogens such as Haemophilus influenzae (two controls), Streptococcus pneumoniae (one control) and S. aureus (one control) could be identified in small numbers. None of the controls showed growth of S. pyogenes or other b-haemolytic Streptococci . 3.1.2. AT patients All samples obtained from the tonsillar surfaces of the AT patients exhibited heavy growth of S. pyogenes . Semiquantitative evaluation of the aerobic bacterial colonies showed that 47% (median value; range 27/92%) of those cultured on agar plates consisted of S. pyogenes . The remainder were mainly a-haemolytic Streptococci , with small numbers of Corynebacterium spp., and Neisseria spp. 3.1.3. IM patients Aerobic culturing revealed massive growth of the bacterial flora normally harbored on the tonsillar surfaces, viz. Streptococcus viridans , Neisseria spp., Corynebacterium spp., and various spirochaetes. None of the patients in this group revealed growth of S. pyogenes or other bhaemolytic Streptococci.
3.2.1. Healthy controls The great majority of tonsillar surface bacteria were coated with IgG, S-IgA and lysozyme (Table 1); less than 10% were coated with lactoferrin. Differences in bacterial coating between these antimicrobial agents were significant (Table 2). 3.2.2. AT patients AT patients showed a remarkable increase in bacterial coating of S. pyogenes microorganisms with lactoferrin, vis-a ´-vis healthy controls (Tables 1 and 2). The increase in coating with lysozyme was not as prominent though significant (Table 2). No significant alteration in IgG-coating was noted (Fig. 1). In contrast, bacterial coating with S-IgA antibodies was considerably reduced (Fig. 2). 3.2.3. IM patients A highly significant increase in lactoferrin coating of the tonsillar bacteria was noted, whereas the lysozyme coating appeared about the same as in healthy controls (Tables 1 and 2 and Fig. 3). In contrast, bacterial coating with IgG and S-IgA antibodies was appreciably reduced. Compared with the situation during AT infection culture-positive for S. pyogenes , IM infection caused a dramatic reduction in both innate and acquired bacterial defence. The quantities of all antibacterial substances attached to the bacterial walls studied were significantly reduced (Table 2).
4. Discussion 3.2. Immunocytochemical assay The overall findings regarding bacterial coating on tonsillar surfaces with lysozyme, lactoferrin, IgG and S-IgA, respectively, among healthy control individuals, and in patients suffering from AT
It is well known that the oral and pharyngeal cavities harbor a commensal bacterial flora composed of both non-pathogenic and pathogenic species. In contrast, the inferior respiratory tract [21], the middle ear cavities [22] and paranasal
Table 1 Proportions (%) of tonsillar surface bacteria coated with lysozyme (LZ), lactoferrin (LF), immunoglobulins IgG and S-IgA, in healthy controls, in patients with streptococcal tonsillitis (AT) and in patients with IM Controls (n/8)
LZ LF IgG S-IgA
AT-patients (n /8)
IM-patients (n/7)
Median
Range
Median
Range
Median
Range
67 9 92 78
48 /85 6 /12 79 /98 68 /85
89 59 92 30
33 /94 42 /68 83 /97 12 /72
56 28 64 12
47 /82 19 /35 26 /90 5 /32
¡/LZ ¡/LF ¡/IgG ¡/S-IgA
B/0.025 /0.0005 B/0.005 /0.001
1079
B/0.05 B/0.0005 NS B/0.0005 B/0.005 B/0.0005 B/0.0005 B/0.025 B/0.0005 B/0.005
Fig. 1 Cryosection double-stained with antiserum to S. pyogenes (gold particles 10 nm) and antiserum to human IgG immunoglobulin (gold particles 5 nm). From a patient culture-positive for S. pyogenes . Original magnification /30 000. Bar 200 nm.
Fig. 2 Cryosection double-stained with antiserum to S. pyogenes (gold particles 5 nm) and with antiserum to human S-IgA (gold particles 10 nm). From a patient culture-positive to S. pyogenes . Original magnification /30 000. Bar 200 nm.
IgG/S-IgA /LZ /LF IgG vs. S-IgA IgG vs.LZ IgG vs. LF S-IgA vs. LZ S-IgA vs. LF LZ vs. LF
IgG/LZ /LF /S-IgA /LZ /LF IgG ¡/S-IgA
IgG /LZ /LF /S-IgA ¡/LZ /LF ¡/IgG ¡/S-IgA
NS B/0.0005 B/0.025 B/0.0005
P -values IM patients vs. AT patients P -values IM patients vs. healthy P -values AT patients vs. healthy P -values Healthy controls
Table 2 Differences (with P -values) in bacterial coating with lysozyme (LZ), lactoferrin (LF), IgG and S-IgA between healthy controls, AT patients and IM patients Wilcoxon’s rank sum and Mann /Whitney’s U -test were used to compare the values
Differences in bacterial defence between bacteria- and virus-induced tonsillitis
sinuses [23] are normally void of microorganisms. Thus, the normal oral bacterial flora must constantly be kept in check, otherwise there will be a complete overgrowth of bacterial species in the fauces. The saliva appears to play a key role in the transport and neutralization of bacterial species, as the sputum is continuously moving towards the stomach, where a low pH in combination with
1080
Fig. 3 Cryosection of sample obtained from tonsillar surfaces of patient with IM showing bacteria coated and uncoated with gold-labeled antiserum to human lactoferrin (gold particles 10 nm). Bar 200 nm. Original magnification /20 000.
effective enzymes can kill a large proportion of the bacterial invaders [1]. Of the innate antibacterial substances in sputum, lysozyme and lactoferrin are found in substantial concentrations in secretions bathing the oral cavity and pharynx. Many of the antimicrobial agents in saliva interact with each other and tend to vary, both together and in some cases independently [21]. Unstimulated whole saliva contains 10 /200 mg/ ml lysozyme [1,17]. The classical concept of the antimicrobial action of lysozyme is based on its muramidase activity. Successful hydrolysis by lysozyme results in lysis and release of intracellular cytoplasmic constituents. Furthermore, lysozyme is also known to aggregate oral bacteria, although this is probably of less importance than other aggregating factors in saliva. Lysozyme is highly active against many Gram-positive bacteria, but appears to be ineffective against Gram-negative bacteria unless potentiated by a co-factor such as lactoferrin or an antibody-complement complex [24]. Unstimulated whole saliva also contains 8.5 /24 mg/ml lactoferrin [1]. Lactoferrin has a bacteriostatic effect on a wide variety of microorganisms, including streptococci, staphylococci, candida and enteric microorganisms [25]. Furthermore, lactoferrin can also exert an irreversible bactericidal effect on a variety of microorganisms, including Streptococcus mutans , the main bacterial species responsible for dental caries [1]. Although lactoferrin is present in human saliva at concentrations
L.-E. Stenfors et al.
that ought to be bactericidal, most native samples from normal individuals tested do not kill S. mutans . Thus, in human saliva the bacteriostatic activity is probably lactoferrin’s major antimicrobial mechanism [1]. The chief immunoglobulin on mucosal membranes S-IgA, can agglutinate microbes and block interactions between microbial adhesins and their receptors on the apical surfaces of mucosal epithelial cells [2 /4]. The antibacterial effect of IgA can be potentiated by such nonspecific antibacterial substances as lysozyme and lactoferrin [21]. IgA antibodies against viruses may prevent their binding to cells, as well as their internalization and replication [4]. The IgG antibodies may enhance phagocytosis and killing of oral microorganisms through opsonization [2,5]. A large proportion of the IgG in sputum originates in the palatine tonsils, as was shown by a significant reduction of IgG antibodies, observed following tonsillectomy [26]. The present study showed that the great majority of tonsillar surface bacteria in healthy individuals were coated with IgG and S-IgA antibodies, but also with lysozyme (Table 1). In contrast, only about 10% of the bacteria were coated with lactoferrin, possibly due to the fact that most of the lactoferrin on the oropharyngeal surfaces derives from leukocytes, which are relatively sparse on the oral mucosa under healthy conditions. During an AT infection caused by S. pyogenes , the innate bacterial defence directed towards the pathogens was particularly activated (lysozyme-coating increase P B/0.05; lactoferrincoating increase P B/0.0005). In contrast, the acquired bacterial defence was highly suppressed by the S-IgA immunoglobulins (S-IgA-coating reduction vis-a ´-vis controls P B/0.0005). As no significant change in IgG-coating of the pathogens was noted, the latter should be sufficiently opsonized with IgG for phagocytosis to take place. However, lysozyme, lactoferrin- and IgG-coating of the S. pyogenes bacteria seemed to be insufficient to compensate for the lack of S-IgA-coating of the pathogens in order to prevent infection. During IM infection, in these cases caused by EBV, the antibacterial defence was particularly defective regarding the acquired defence system (IgGcoating reduced vis-a ´-vis controls P B/0.025; S-IgA reduction P B/0.0005). A significant activation of the innate defence system represented by lactoferrin was noted, however, (P B/0.0005). Most lactoferrin derives from the specific granulae of leukocytes, which naturally are attracted to the target region during an infection of the palatine tonsils [27]. The antibacterial effect of an increased lactoferrin-coating of the bacteria cannot,
Differences in bacterial defence between bacteria- and virus-induced tonsillitis
however, compensate the deficient immunoglobulin-coating of the bacteria. Thus, it seems that the binding of the bacteria to the mucus layer effected by S-IgA, opsonization of the bacteria by IgG and their subsequent removal from the tonsillar surfaces is central to the etiopathogenesis of bacterial colonization of mucosal membranes. When comparing AT culture-positive for S. pyogenes with IM, it became evident that EBV suppressed both of the innate and the acquired defence systems. Significant reductions in bacterial coating with lysozyme, lactoferrin, IgG and SIgA were noted (Table 2). These findings can explain our clinical experience that IM patients culture-positive for S. pyogenes exhibit the most severe symptoms from the pharyngeal region regarding tonsillar coverings, tonsillar enlargement and swelling, and risk of developing peritonsillar abscesses due to bacterial penetration into tonsillar tissue [28]. Although the human pharynx is a favorable environment for colonization by microorganisms, it is protected by a powerful defence mechanism. This mechanism can be explained as trapping of bacteria by mucus, local production of immunoglobulins and non-immunoglobulin mediators, mobilization of leukocytes, shedding of bacteria bound to exfoliated epithelial cells, and transport of mucus into the stomach. In particular, defective S-IgAcoating of pharyngeal microorganisms seems to be of the utmost importance in the etiopathogenesis of tonsillar infections. The extent to which other cationic antibacterial agents than lysozyme and lactoferrin */such as secretory leukoprotease inhibitors and defensins [27] */contributes to the overall bacterial defence of the palatine tonsils in vivo requires further evaluation.
Acknowledgements Financial support for this study was provided by the Medical Society of Finland.
References [1] J. Tenovuo, Nonimmunoglobulin defence factors in human saliva, in: J. Tenovuo (Ed.), Human Saliva: Clinical Chemistry and Microbiology, CRC Press, Boca Raton, FL, 1989, pp. 55 /91. [2] P. Brandtzaeg, Immunocompetent cells of the upper airway: functions in normal and diseased mucosa, Eur. Arch. Otorhinolaryngol. 252 (Suppl. 1) (1995) 8 /21. [3] M.E. Lamm, Interaction of antigens and antibodies at mucosal surfaces, Annu. Rev. Microbiol. 51 (1997) 311 / 340.
1081
[4] H. Marcotte, M.C. Lavoie, Oral microbial ecology and the role of salivary immunoglobulin A, Microbiol. Mol. Biol. Rev. 62 (1998) 71 /109. [5] M. Ivarsson, C. Lundberg, Phagocytosis in the nasopharyngeal secretion by cells from the adenoid, Acta Otolaryngol. 121 (2001) 517 /522. [6] A. Stjernquist-Desatnik, K. Prellner, P. Christensen, Clinical and laboratory findings in patients with acute tonsillitis, Acta Otolaryngol. (Stockholm) 104 (1987) 351 /359. [7] J. Ylikoski, J. Karjalainen, Acute tonsillitis in young men: etiological agents and their differentiation, Scand. J. Infect. Dis. 21 (1989) 169 /174. [8] L.-E. Stenfors, F. Fredriksen, S. Ra ¨isa ¨nen, R. Myklebust, Identification of Streptococcus pyogenes on tonsillar epithelium during infection, Acta Otolaryngol. 529 (Suppl) (1997) 212 /214. [9] M. Lilja, J. Silvola, S. Ra ¨isa ¨nen, L.-E. Stenfors, Where are the receptors for Streptococcus pyogenes located on the tonsillar surface epithelium, Int. J. Pediatr. Otorhinolaryngol. 50 (1999) 37 /43. [10] M. Lilja, J. Silvola, H.-M. Bye, S. Ra ¨isa ¨nen, L.-E. Stenfors, S-IgA and IgG-coated Streptococcus pyogenes on the tonsillar surfaces during acute tonsillitis, Acta Otolaryngol. 119 (1999) 718 /723. [11] L.-E. Stenfors, H.-M. Bye, L.H. Vorland, Remarkable attachment of lactoferrin to Streptococcus pyogenes during acute pharyngotonsillitis, Acta Otolaryngol. 121 (2001) 637 /642. [12] L.-E. Stenfors, S. Ra ¨isa ¨nen, Immunoglobulin-coated bacteria on the tonsillar surface during infectious mononucleosis, J. Laryngol. Otol. 110 (1996) 339 /342. [13] L.-E. Stenfors, H.-M. Bye, S. Ra ¨isa ¨nen, Bacterial coating with immunoglobulins on the palatine tonsils during infectious mononucleosis: immunocytochemical study with gold markers, J. Laryngol. Otol. 115 (2001) 101 /105. [14] L.-E. Stenfors, H.-M. Bye, S. Ra ¨isa ¨nen, Immunocytochemical localization of lysozyme and lactoferrin attached to surface bacteria of the palatine tonsils during infectious mononucleosis, J. Laryngol. Otol. 116 (2002) 264 /268. [15] L.-E. Stenfors, S. Ra ¨isa ¨nen, The membranous tonsillitis during infectious mononucleosis is nevertheless of bacterial origin, Int. J. Pediatr. Otorhinolaryngol. 26 (1993) 149 / 155. [16] I. Brook, F. De Leyva, Immune response to Fusobacterium nucleatum and Prevotella intermedia in patients with infectious mononucleosis, J. Med. Microbiol. 44 (1996) 131 /134. [17] J. Tenovuo, O.-P. Lehtonen, A.S. Aaltonen, P. Vilja, P. Tuohimaa, Antimicrobial factors in whole saliva of human infants, Infect. Immun. 51 (1986) 49 /53. [18] C. DiPaolo, I.D. Mandel, Lactoferrin concentration in human parotid saliva as measured by an enzyme linked immuno-sorbent assay (ELISA), J. Dent. Res. 59 (1980) 1463 /1465. [19] M. Votava, V. Bartosova, A. Krchnakova, K. Crhova, L. Kubinova, Diagnostic importance of heterophile antibodies and immunoglobulins IgA, IgE, IgM and low-avidity IgG against Epstein /Barr virus capsid antigen in children, Acta Virol. 40 (2) (1996) 99 /101. [20] J.J. Gray, Avidity of EBV VCA-specific IgG antibodies: distinction between recent primary infection, past infection and reactivation, J. Virol. Methods 52 (1 /2) (1995) 95 /104. [21] P.K. Singh, B.F. Tack, P.B. McCray, M.J. Welsh, Synergistic and additive killing by antimicrobial factors found in human airway surface liquid, Am. J. Physiol. Lung Cell Physiol. 279 (5) (2000) 799 /805.
1082 [22] L.-E. Stenfors, Non-specific and specific immunity to bacterial invasion of the middle ear cavity, Int. J. Pediatr. Otorhinolaryngol. 49 (Suppl. 1) (1999) 223 /226. [23] J. Sobin, S. Engquist, C.E. Nord, Bacteriology of the maxillary sinus in healthy volunteers, Scand. J. Infect. Dis. 24 (1992) 633 /635. [24] R.T. Ellison, T.J. Giehl, Killing of gram-negative bacteria by lactoferrin and lysozyme, J. Clin. Invest. 88 (4) (1991) 1080 /1091. [25] L.H. Vorland, Lactoferrin: a multifunctional glycoprotein, APMIS 117 (1999) 971 /981.
L.-E. Stenfors et al. [26] V. Kirstila, J. Tenovuo, O. Ruuskanen, J. Suonpaa, O. Meurman, P. Vilja, Longitudinal analysis of human salivary immunoglobulins, nonimmune antimicrobial agents, and microflora after tonsillectomy, Clin. Immunol. Immunopathol. 80 (2) (1996) 110 /115. [27] T. Ganz, Antimicrobial polypeptides in host defence of the respiratory tract, J. Clin. Invest. 109 (6) (2002) 693 /697. [28] L.-E. Stenfors, H.-M. Bye, S. Ra ¨isa ¨nen, R. Myklebust, Bacterial penetration into tonsillar surface epithelium during infectious mononucleosis, J. Laryngol. Otol. 114 (2000) 848 /852.
www.elsevier.com/locate/ijporl
Noticeable differences in bacterial defence on tonsillar surfaces between bacteria-induced and virus-induced acute tonsillitis ¨isa ¨nenb Lars-Eric Stenforsa,*, Helga-Marie Byea, Simo Ra a
Department of Otolaryngology, Institute of Clinical Medicine, University of Tromsø, N-9037 Tromsø, Norway b Clinical Laboratory, Central Hospital of Keski-Pohjanmaa, FIN-67200 Kokkola, Finland Received 30 January 2003; received in revised form 22 May 2003; accepted 1 June 2003
KEYWORDS Antibacterial activity; Tonsillar disease; Mucosal immunity; Streptococcus pyogenes ; Epstein /Barr virus
Summary Objective : Oral and pharyngeal cavities harbor a commensal bacterial flora which is kept in check by several innate and acquired agents. In this study, we focused on the proportions in which some antibacterial moderators (lysozyme, lactoferrin, IgG and S-IgA) coat the tonsillar surface bacteria in healthy individuals, in patients with acute tonsillitis (AT) culture-positive for Streptococcus pyogenes , and in patients with infectious mononucleosis (IM) caused by Epstein /Barr virus (EBV). Methods : Bacterial samples were collected for aerobic culturing and immunocytochemical evaluation from the tonsillar surfaces of eight healthy individuals (four males, four females; age range 16 /22 years), eight patients with current AT (two males, six females; age range 16 /29 years) and seven patients with IM (four males, three females; age range 15 /21 years). The immunocytochemical assay was based on gold-labeled antiserum to human lysozyme, lactoferrin, IgG and S-IgA followed by gold particle tracing in the transmission electron microscope. Results : During AT, a significant increase in lysozyme coating (P B/0.05) and lactoferrin coating (P B/ 0.0005) of the bacteria was noted, whereas the S-IgA coating was significantly reduced (P B/0.0005). During IM infection, a significant increase in lactoferrin coating was noted (P B/0.0005) whereas immunoglobulin coating was significantly reduced (IgG P B/0.025; S-IgA P B/0.0005) compared with healthy controls. During IM, all antibacterial moderators evaluated were significantly reduced compared with the situation during AT. Conclusions : Noticeable changes in the local innate and acquired bacterial defence system were observed during tonsillar infections, particularly during IM. – 2003 Elsevier Ireland Ltd. All rights reserved.
*Corresponding author. Tel.: /47-776-2-7390; fax: /47-776-4-4772. E-mail address: [email protected] (L.-E. Stenfors). 0165-5876/03/$ - see front matter – 2003 Elsevier Ireland Ltd. All rights reserved. doi:10.1016/S0165-5876(03)00194-0
1076
1. Introduction It is evident that the oropharyngeal cavity possesses a high degree of immunity. Large wounds resulting from tonsillectomy normally heal rapidly despite contamination by the vast bacterial flora that proliferates in this region. This fortunate outcome is probably results from local immune defence, in which saliva, salivary components and flow [1], non-immunological and immunological antibacterial substances [2 /4], leukocytes [5] and desquamation of epithelial cells all play a central role. Numerous antibacterial agents (lysozyme, lactoferrin, peroxidase, antibacterial peptides, immunoglobulins, etc.) which are secreted into the saliva, inhibit the metabolism, adherence and viability of microorganisms in vitro, but their clinical role on the palatine tonsils and tonsillar fossae is largely unknown. It would seem, however, that they are important for the control of microbial overgrowth in the oropharynx, though just how selective they are against pathogens in situ is not known. The main bacterial pathogen causing roughly 30% of all acute tonsillitis (AT) cases is Streptococcus pyogenes [6,7]. By means of an immunocytochemical technique, using gold-labeled antiserum to S. pyogenes and several antimicrobial agents, we have succeeded in localizing these pathogens on tonsillar surfaces during AT [8,9] and their coating with IgG and secretory IgA (S-IgA) immunoglobulins [10] as well as lactoferrin [11]. When using the same immunocytochemical technique, it was found that coating of the tonsillar surface bacteria with IgG and S-IgA immunoglobulins was greatly suppressed during infectious mononucleosis (IM) [12,13], whereas bacterial coating with lactoferrin increased slightly compared with healthy controls [14]. During IM, caused mostly by Epstein/Barr virus (EBV), a characteristic sign is an immense bacterial colonization of the palatine tonsils, forming a membranous tonsillitis [15,16]. In both of these diseases, it is obvious that the overall bacterial defence on the mucosal membranes of the palatine tonsils is weakened in some way. The purpose of the present study was to compare the extent to which some antibacterial substances attached themselves to tonsillar surface bacteria during bacteria-induced and virus-induced infection of the palatine tonsils. Of the numerous antimicrobial factors present in sputum, we chose the two non-immunological substances having the greatest concentration, viz. lysozyme [1,17] and
L.-E. Stenfors et al.
lactoferrin [1,18]. Furthermore, bacterial coating with IgG and S-IgA was examined.
2. Material and methods 2.1. Controls and patients 2.1.1. Control group For control purposes we used eight healthy individuals with good oral health (four males, four females; age range 16 /22 years, median age 20) attending a normal health check-up. None of the control subjects had taken an antibiotic within 2 months of bacterial sampling from the palatine tonsils. 2.1.2. AT group Eight patients (two males, six females; age range 16 /29 years, median age 25) with classical symptoms of AT (sore throat, dysphagia, axillary body temperature /38 8C, local inflammatory changes in the palatine tonsils and yellowish exudate on the tonsils) comprised the AT group. None of the patients had taken an antibiotic within 2 months of bacterial sampling. All patients displayed a positive rapid test result for S. pyogenes (Quick Vue In-Line One-Step Strep A Test system, Quides, Rosny, France) and subsequent throat culture revealed growth of S. pyogenes . 2.1.3. IM group Seven patients (three males, four females; age range 15 /21 years, median age 17) with IM and concomitant membranous tonsillitis who were referred to our out-patient department comprised the IM group. The IM diagnosis was based on clinical and serological findings. All patients evidenced lymphadenopathy, lymphocytosis with /10% atypical McKinley cells, a positive heterophil antibody reaction (Monosticon Dri-Dot slide test, Organon Diagnostics, Durham, NC, USA) and a Paul /Bunnell test with titre /40. All patients showed positive IgG and/or IgM titres to capsid EBV [19]. Moreover, the IgG avidity was very low, hinting at a primary EBV infection of recent date [20]. None of the patients had a subnormal serum immunoglobulin level according to an immunoturbidimetric assay. The study was approved by the Medical Ethics Committee of our hospital. Informed consent was always obtained from the patients and/or patient’s parents before sampling.
Differences in bacterial defence between bacteria- and virus-induced tonsillitis
2.2. Aerobic culturing To obtain throat samples from the palatine tonsils, a cottonwool-tipped wooden swab, which had been dipped in physiological saline, was rotated against the mucosa. Samples were then smeared over blood-agar and chocolate-agar plates for standard aerobic culturing according to routine laboratory procedures. The plates were kept incubated in 5% CO2 at 37 8C for 48 h. Growth of S. pyogenes was identified by means of colony morphology, b-haemolysis, a rapid agglutination test of material obtained from bacterial colonies (Streptex A, Diagnostic Products Corporation, Los Angeles, CA, USA) and PYR test (Api 20 Strep, bioMe ´rieux, Marcy l’Etoile, France). Semiquantification of S. pyogenes was performed by calculating the colony-forming units (CFU) of S. pyogenes versus other aerobic colonies in equal sections of the agar plates.
2.3. Immunocytochemical assay Swabbed cell material was obtained from both palatine tonsils and loosened from the swab by rinsing it in physiological saline. No food, snacks, coffee or toothbrushing was allowed within 2 h prior to bacterial sampling. The cell mixture was given a light centrifugation to harvest the cells at the bottom of the tubes. Supernatant was carefully discharged and the precipitates carefully adjusted to 0.5 ml with physiological saline. To this cell mixture, 0.5 ml 8% formaldehyde buffered in 200 mM HEPES fixative was added to give a working concentration of 4% formaldehyde. After fixation, the mixture was centrifugated at 16 000 rpm for 1 / 2 min. The supernatant was pipetted away and the cells diluted in 100 ml 10% gelatin. The samples were centrifugated briefly (15 s) and kept on ice for 2 h. The cell pellet was excised, immersed in 2.3 M sucrose overnight, placed on aluminium specimen pins and frozen by immersion in liquid nitrogen. Ultrathin sections were made using a Reichert Ultracut S ultramicrotome with a FCS cryochamber (Leica, Deerfield, IL, USA). The sections (ca. 30/60 nm) were cut with diamond knives and retrieved from the knife surface with a drop of methyl cellulose mixed with sucrose. The specimens were then mounted on carbon-coated grids, which were placed in 1% cold fish-skin gelatin (FSG) for 15 min. Each grid was placed on a 5-ml droplet of the actual antiserum, which was diluted in accordance with earlier titration of the antiserum. After renewed washing in PBS, the grids were placed on protein A-gold diluted in 1% FSG (gold particle 5
1077
nm) and then washed with PBS. When doublestained, the grids were fixed with 1% glutaraldehyde (blocking step to prevent second A-gold probe from binding to the first antibody), followed by washing in PBS, after which the grids were placed in 0.12% glycine. After further washing in PBS the specimens were placed on droplets of the second antiserum. After washing in PBS, the grids were placed on protein A-gold (gold particle 10 nm) diluted in 1% FSG for 15 min. Further washing with distilled water was followed by drying in the presence of 3% uranyl acetate in 2% methyl cellulose. The sections were examined in a JEOL JEM 1010 transmission electron microscope operated at 80 kV. From each patient, three grids, either single or double stained for each substance, were evaluated. From each grid roughly 1000 bacteria were scrutinized by two of the authors (H.-M. B., L.-E. S.) independently without knowledge of the origin of the sample. The following antisera were used: rabbit antiserum to S. pyogenes (diluted in FSG 1:200; Difco Labs, Detroit, MI, USA); rabbit antiserum to human lysozyme (dilution in FSG 1:150; DAKO A/S, Glostrup, Denmark); rabbit antiserum to human lactoferrin (dilution in FSG 1:200; DAKO); rabbit antiserum to human IgG (dilution in FSG 1:100; Organon Teknika, Westchester, PA, USA); goat antiserum to human S-IgA (dilution in FSG 1:100; Organon Teknika Co). Prior to investigation, the antisera were titrated in order to minimize background attachment of the gold particles. As positive controls we used plain cultures of S. pyogenes and Staphylococcus aureus coated with lysozyme, lactoferrin, IgG or S-IgA. As negative controls we used plain cultures of Streptococcus mitis , Escherichia coli , Neisseria sicca , Streptococcus salivarius and Streptococcus equisimilis . Handling the specimens with 1% FSG and protein-A gold without antisera, or with antisera absorbed with excess antigen, proved negative. All samples were handled in exactly the same way. In the double-label experiments, care was taken to ensure that labeling followed the patterns in single-label experiments, and also that they were the same in crossover controls.
2.4. Statistical analysis Wilcoxon’s rank sum test was used to compare values within the groups and Mann /Whitney’s U test to compare values between groups. P -values below 0.05 were considered statistically significant.
1078
L.-E. Stenfors et al.
3. Results
infection culture-positive for S. pyogenes and IM caused by EBV, are summarized in Table 1 (median values and ranges) and Table 2 (differences with P values).
3.1. Aerobic culture 3.1.1. Healthy controls Aerobic culturing of bacterial samples obtained from tonsillar surfaces of the healthy control individuals evidenced growth of mainly a-haemolytic viridans Streptococci , Corynebacterium spp., Neisseria spp., and coagulase-negative Staphylococci . In four controls, upper respiratory tract pathogens such as Haemophilus influenzae (two controls), Streptococcus pneumoniae (one control) and S. aureus (one control) could be identified in small numbers. None of the controls showed growth of S. pyogenes or other b-haemolytic Streptococci . 3.1.2. AT patients All samples obtained from the tonsillar surfaces of the AT patients exhibited heavy growth of S. pyogenes . Semiquantitative evaluation of the aerobic bacterial colonies showed that 47% (median value; range 27/92%) of those cultured on agar plates consisted of S. pyogenes . The remainder were mainly a-haemolytic Streptococci , with small numbers of Corynebacterium spp., and Neisseria spp. 3.1.3. IM patients Aerobic culturing revealed massive growth of the bacterial flora normally harbored on the tonsillar surfaces, viz. Streptococcus viridans , Neisseria spp., Corynebacterium spp., and various spirochaetes. None of the patients in this group revealed growth of S. pyogenes or other bhaemolytic Streptococci.
3.2.1. Healthy controls The great majority of tonsillar surface bacteria were coated with IgG, S-IgA and lysozyme (Table 1); less than 10% were coated with lactoferrin. Differences in bacterial coating between these antimicrobial agents were significant (Table 2). 3.2.2. AT patients AT patients showed a remarkable increase in bacterial coating of S. pyogenes microorganisms with lactoferrin, vis-a ´-vis healthy controls (Tables 1 and 2). The increase in coating with lysozyme was not as prominent though significant (Table 2). No significant alteration in IgG-coating was noted (Fig. 1). In contrast, bacterial coating with S-IgA antibodies was considerably reduced (Fig. 2). 3.2.3. IM patients A highly significant increase in lactoferrin coating of the tonsillar bacteria was noted, whereas the lysozyme coating appeared about the same as in healthy controls (Tables 1 and 2 and Fig. 3). In contrast, bacterial coating with IgG and S-IgA antibodies was appreciably reduced. Compared with the situation during AT infection culture-positive for S. pyogenes , IM infection caused a dramatic reduction in both innate and acquired bacterial defence. The quantities of all antibacterial substances attached to the bacterial walls studied were significantly reduced (Table 2).
4. Discussion 3.2. Immunocytochemical assay The overall findings regarding bacterial coating on tonsillar surfaces with lysozyme, lactoferrin, IgG and S-IgA, respectively, among healthy control individuals, and in patients suffering from AT
It is well known that the oral and pharyngeal cavities harbor a commensal bacterial flora composed of both non-pathogenic and pathogenic species. In contrast, the inferior respiratory tract [21], the middle ear cavities [22] and paranasal
Table 1 Proportions (%) of tonsillar surface bacteria coated with lysozyme (LZ), lactoferrin (LF), immunoglobulins IgG and S-IgA, in healthy controls, in patients with streptococcal tonsillitis (AT) and in patients with IM Controls (n/8)
LZ LF IgG S-IgA
AT-patients (n /8)
IM-patients (n/7)
Median
Range
Median
Range
Median
Range
67 9 92 78
48 /85 6 /12 79 /98 68 /85
89 59 92 30
33 /94 42 /68 83 /97 12 /72
56 28 64 12
47 /82 19 /35 26 /90 5 /32
¡/LZ ¡/LF ¡/IgG ¡/S-IgA
B/0.025 /0.0005 B/0.005 /0.001
1079
B/0.05 B/0.0005 NS B/0.0005 B/0.005 B/0.0005 B/0.0005 B/0.025 B/0.0005 B/0.005
Fig. 1 Cryosection double-stained with antiserum to S. pyogenes (gold particles 10 nm) and antiserum to human IgG immunoglobulin (gold particles 5 nm). From a patient culture-positive for S. pyogenes . Original magnification /30 000. Bar 200 nm.
Fig. 2 Cryosection double-stained with antiserum to S. pyogenes (gold particles 5 nm) and with antiserum to human S-IgA (gold particles 10 nm). From a patient culture-positive to S. pyogenes . Original magnification /30 000. Bar 200 nm.
IgG/S-IgA /LZ /LF IgG vs. S-IgA IgG vs.LZ IgG vs. LF S-IgA vs. LZ S-IgA vs. LF LZ vs. LF
IgG/LZ /LF /S-IgA /LZ /LF IgG ¡/S-IgA
IgG /LZ /LF /S-IgA ¡/LZ /LF ¡/IgG ¡/S-IgA
NS B/0.0005 B/0.025 B/0.0005
P -values IM patients vs. AT patients P -values IM patients vs. healthy P -values AT patients vs. healthy P -values Healthy controls
Table 2 Differences (with P -values) in bacterial coating with lysozyme (LZ), lactoferrin (LF), IgG and S-IgA between healthy controls, AT patients and IM patients Wilcoxon’s rank sum and Mann /Whitney’s U -test were used to compare the values
Differences in bacterial defence between bacteria- and virus-induced tonsillitis
sinuses [23] are normally void of microorganisms. Thus, the normal oral bacterial flora must constantly be kept in check, otherwise there will be a complete overgrowth of bacterial species in the fauces. The saliva appears to play a key role in the transport and neutralization of bacterial species, as the sputum is continuously moving towards the stomach, where a low pH in combination with
1080
Fig. 3 Cryosection of sample obtained from tonsillar surfaces of patient with IM showing bacteria coated and uncoated with gold-labeled antiserum to human lactoferrin (gold particles 10 nm). Bar 200 nm. Original magnification /20 000.
effective enzymes can kill a large proportion of the bacterial invaders [1]. Of the innate antibacterial substances in sputum, lysozyme and lactoferrin are found in substantial concentrations in secretions bathing the oral cavity and pharynx. Many of the antimicrobial agents in saliva interact with each other and tend to vary, both together and in some cases independently [21]. Unstimulated whole saliva contains 10 /200 mg/ ml lysozyme [1,17]. The classical concept of the antimicrobial action of lysozyme is based on its muramidase activity. Successful hydrolysis by lysozyme results in lysis and release of intracellular cytoplasmic constituents. Furthermore, lysozyme is also known to aggregate oral bacteria, although this is probably of less importance than other aggregating factors in saliva. Lysozyme is highly active against many Gram-positive bacteria, but appears to be ineffective against Gram-negative bacteria unless potentiated by a co-factor such as lactoferrin or an antibody-complement complex [24]. Unstimulated whole saliva also contains 8.5 /24 mg/ml lactoferrin [1]. Lactoferrin has a bacteriostatic effect on a wide variety of microorganisms, including streptococci, staphylococci, candida and enteric microorganisms [25]. Furthermore, lactoferrin can also exert an irreversible bactericidal effect on a variety of microorganisms, including Streptococcus mutans , the main bacterial species responsible for dental caries [1]. Although lactoferrin is present in human saliva at concentrations
L.-E. Stenfors et al.
that ought to be bactericidal, most native samples from normal individuals tested do not kill S. mutans . Thus, in human saliva the bacteriostatic activity is probably lactoferrin’s major antimicrobial mechanism [1]. The chief immunoglobulin on mucosal membranes S-IgA, can agglutinate microbes and block interactions between microbial adhesins and their receptors on the apical surfaces of mucosal epithelial cells [2 /4]. The antibacterial effect of IgA can be potentiated by such nonspecific antibacterial substances as lysozyme and lactoferrin [21]. IgA antibodies against viruses may prevent their binding to cells, as well as their internalization and replication [4]. The IgG antibodies may enhance phagocytosis and killing of oral microorganisms through opsonization [2,5]. A large proportion of the IgG in sputum originates in the palatine tonsils, as was shown by a significant reduction of IgG antibodies, observed following tonsillectomy [26]. The present study showed that the great majority of tonsillar surface bacteria in healthy individuals were coated with IgG and S-IgA antibodies, but also with lysozyme (Table 1). In contrast, only about 10% of the bacteria were coated with lactoferrin, possibly due to the fact that most of the lactoferrin on the oropharyngeal surfaces derives from leukocytes, which are relatively sparse on the oral mucosa under healthy conditions. During an AT infection caused by S. pyogenes , the innate bacterial defence directed towards the pathogens was particularly activated (lysozyme-coating increase P B/0.05; lactoferrincoating increase P B/0.0005). In contrast, the acquired bacterial defence was highly suppressed by the S-IgA immunoglobulins (S-IgA-coating reduction vis-a ´-vis controls P B/0.0005). As no significant change in IgG-coating of the pathogens was noted, the latter should be sufficiently opsonized with IgG for phagocytosis to take place. However, lysozyme, lactoferrin- and IgG-coating of the S. pyogenes bacteria seemed to be insufficient to compensate for the lack of S-IgA-coating of the pathogens in order to prevent infection. During IM infection, in these cases caused by EBV, the antibacterial defence was particularly defective regarding the acquired defence system (IgGcoating reduced vis-a ´-vis controls P B/0.025; S-IgA reduction P B/0.0005). A significant activation of the innate defence system represented by lactoferrin was noted, however, (P B/0.0005). Most lactoferrin derives from the specific granulae of leukocytes, which naturally are attracted to the target region during an infection of the palatine tonsils [27]. The antibacterial effect of an increased lactoferrin-coating of the bacteria cannot,
Differences in bacterial defence between bacteria- and virus-induced tonsillitis
however, compensate the deficient immunoglobulin-coating of the bacteria. Thus, it seems that the binding of the bacteria to the mucus layer effected by S-IgA, opsonization of the bacteria by IgG and their subsequent removal from the tonsillar surfaces is central to the etiopathogenesis of bacterial colonization of mucosal membranes. When comparing AT culture-positive for S. pyogenes with IM, it became evident that EBV suppressed both of the innate and the acquired defence systems. Significant reductions in bacterial coating with lysozyme, lactoferrin, IgG and SIgA were noted (Table 2). These findings can explain our clinical experience that IM patients culture-positive for S. pyogenes exhibit the most severe symptoms from the pharyngeal region regarding tonsillar coverings, tonsillar enlargement and swelling, and risk of developing peritonsillar abscesses due to bacterial penetration into tonsillar tissue [28]. Although the human pharynx is a favorable environment for colonization by microorganisms, it is protected by a powerful defence mechanism. This mechanism can be explained as trapping of bacteria by mucus, local production of immunoglobulins and non-immunoglobulin mediators, mobilization of leukocytes, shedding of bacteria bound to exfoliated epithelial cells, and transport of mucus into the stomach. In particular, defective S-IgAcoating of pharyngeal microorganisms seems to be of the utmost importance in the etiopathogenesis of tonsillar infections. The extent to which other cationic antibacterial agents than lysozyme and lactoferrin */such as secretory leukoprotease inhibitors and defensins [27] */contributes to the overall bacterial defence of the palatine tonsils in vivo requires further evaluation.
Acknowledgements Financial support for this study was provided by the Medical Society of Finland.
References [1] J. Tenovuo, Nonimmunoglobulin defence factors in human saliva, in: J. Tenovuo (Ed.), Human Saliva: Clinical Chemistry and Microbiology, CRC Press, Boca Raton, FL, 1989, pp. 55 /91. [2] P. Brandtzaeg, Immunocompetent cells of the upper airway: functions in normal and diseased mucosa, Eur. Arch. Otorhinolaryngol. 252 (Suppl. 1) (1995) 8 /21. [3] M.E. Lamm, Interaction of antigens and antibodies at mucosal surfaces, Annu. Rev. Microbiol. 51 (1997) 311 / 340.
1081
[4] H. Marcotte, M.C. Lavoie, Oral microbial ecology and the role of salivary immunoglobulin A, Microbiol. Mol. Biol. Rev. 62 (1998) 71 /109. [5] M. Ivarsson, C. Lundberg, Phagocytosis in the nasopharyngeal secretion by cells from the adenoid, Acta Otolaryngol. 121 (2001) 517 /522. [6] A. Stjernquist-Desatnik, K. Prellner, P. Christensen, Clinical and laboratory findings in patients with acute tonsillitis, Acta Otolaryngol. (Stockholm) 104 (1987) 351 /359. [7] J. Ylikoski, J. Karjalainen, Acute tonsillitis in young men: etiological agents and their differentiation, Scand. J. Infect. Dis. 21 (1989) 169 /174. [8] L.-E. Stenfors, F. Fredriksen, S. Ra ¨isa ¨nen, R. Myklebust, Identification of Streptococcus pyogenes on tonsillar epithelium during infection, Acta Otolaryngol. 529 (Suppl) (1997) 212 /214. [9] M. Lilja, J. Silvola, S. Ra ¨isa ¨nen, L.-E. Stenfors, Where are the receptors for Streptococcus pyogenes located on the tonsillar surface epithelium, Int. J. Pediatr. Otorhinolaryngol. 50 (1999) 37 /43. [10] M. Lilja, J. Silvola, H.-M. Bye, S. Ra ¨isa ¨nen, L.-E. Stenfors, S-IgA and IgG-coated Streptococcus pyogenes on the tonsillar surfaces during acute tonsillitis, Acta Otolaryngol. 119 (1999) 718 /723. [11] L.-E. Stenfors, H.-M. Bye, L.H. Vorland, Remarkable attachment of lactoferrin to Streptococcus pyogenes during acute pharyngotonsillitis, Acta Otolaryngol. 121 (2001) 637 /642. [12] L.-E. Stenfors, S. Ra ¨isa ¨nen, Immunoglobulin-coated bacteria on the tonsillar surface during infectious mononucleosis, J. Laryngol. Otol. 110 (1996) 339 /342. [13] L.-E. Stenfors, H.-M. Bye, S. Ra ¨isa ¨nen, Bacterial coating with immunoglobulins on the palatine tonsils during infectious mononucleosis: immunocytochemical study with gold markers, J. Laryngol. Otol. 115 (2001) 101 /105. [14] L.-E. Stenfors, H.-M. Bye, S. Ra ¨isa ¨nen, Immunocytochemical localization of lysozyme and lactoferrin attached to surface bacteria of the palatine tonsils during infectious mononucleosis, J. Laryngol. Otol. 116 (2002) 264 /268. [15] L.-E. Stenfors, S. Ra ¨isa ¨nen, The membranous tonsillitis during infectious mononucleosis is nevertheless of bacterial origin, Int. J. Pediatr. Otorhinolaryngol. 26 (1993) 149 / 155. [16] I. Brook, F. De Leyva, Immune response to Fusobacterium nucleatum and Prevotella intermedia in patients with infectious mononucleosis, J. Med. Microbiol. 44 (1996) 131 /134. [17] J. Tenovuo, O.-P. Lehtonen, A.S. Aaltonen, P. Vilja, P. Tuohimaa, Antimicrobial factors in whole saliva of human infants, Infect. Immun. 51 (1986) 49 /53. [18] C. DiPaolo, I.D. Mandel, Lactoferrin concentration in human parotid saliva as measured by an enzyme linked immuno-sorbent assay (ELISA), J. Dent. Res. 59 (1980) 1463 /1465. [19] M. Votava, V. Bartosova, A. Krchnakova, K. Crhova, L. Kubinova, Diagnostic importance of heterophile antibodies and immunoglobulins IgA, IgE, IgM and low-avidity IgG against Epstein /Barr virus capsid antigen in children, Acta Virol. 40 (2) (1996) 99 /101. [20] J.J. Gray, Avidity of EBV VCA-specific IgG antibodies: distinction between recent primary infection, past infection and reactivation, J. Virol. Methods 52 (1 /2) (1995) 95 /104. [21] P.K. Singh, B.F. Tack, P.B. McCray, M.J. Welsh, Synergistic and additive killing by antimicrobial factors found in human airway surface liquid, Am. J. Physiol. Lung Cell Physiol. 279 (5) (2000) 799 /805.
1082 [22] L.-E. Stenfors, Non-specific and specific immunity to bacterial invasion of the middle ear cavity, Int. J. Pediatr. Otorhinolaryngol. 49 (Suppl. 1) (1999) 223 /226. [23] J. Sobin, S. Engquist, C.E. Nord, Bacteriology of the maxillary sinus in healthy volunteers, Scand. J. Infect. Dis. 24 (1992) 633 /635. [24] R.T. Ellison, T.J. Giehl, Killing of gram-negative bacteria by lactoferrin and lysozyme, J. Clin. Invest. 88 (4) (1991) 1080 /1091. [25] L.H. Vorland, Lactoferrin: a multifunctional glycoprotein, APMIS 117 (1999) 971 /981.
L.-E. Stenfors et al. [26] V. Kirstila, J. Tenovuo, O. Ruuskanen, J. Suonpaa, O. Meurman, P. Vilja, Longitudinal analysis of human salivary immunoglobulins, nonimmune antimicrobial agents, and microflora after tonsillectomy, Clin. Immunol. Immunopathol. 80 (2) (1996) 110 /115. [27] T. Ganz, Antimicrobial polypeptides in host defence of the respiratory tract, J. Clin. Invest. 109 (6) (2002) 693 /697. [28] L.-E. Stenfors, H.-M. Bye, S. Ra ¨isa ¨nen, R. Myklebust, Bacterial penetration into tonsillar surface epithelium during infectious mononucleosis, J. Laryngol. Otol. 114 (2000) 848 /852.