Brazilian J Med Biol Res (1991) 24: 399-406 RELATIONSHIP OF BIOTYPE AND SOURCE TO THE HEMAGGLUTINATION AND ADHESIVE PROPERTIES OF C. DIPHTHERIAE A.L, MATTOS-GUARALDI and L.C.D, FORMIGA* Departamento de Patologia e Laboratérios, Faculdade de Ciéncias Médicas, Universidade do Estado do Rio de Janeiro, 20550 Rio de Janeiro, RJ, Brasil “Departamento de Microbiologia Médica, Instituto de Microbiologia, Universidade Federal do Rio de Janeiro, 21944 Rio de Janeiro, RJ, Brasit 1. A. study was conducted on the hemagglutinating end adhesive properties of sucrose-fermenting, and non-sucrose-fermenting Corynebacterium diphtheriae strains. Sheep erythrocytes ‘wore agglutinated by 72% of toxigen C. diphtheriae strains and by 87% of nontoxigenic strains, by 67% of the strains isolated from throats and 94% of skin isolates, and by 65% of sucrose-fermenting organisms and 93% of non-sucrose-fermenting strains. 2. Neither presence nor absence of toxin production was correlated with the hemagglutinating property. 3, Different levels of adherence to glass surfaces occurred among C. diphtheriae strains. 4 Strains from skin lesions and of the non-sucrose-fermenting biotype showed strong hemagglutinating activity. In contrast, sucrose-fermenting strains exhibited less hemagglutinating activity. but most of them (715%) were predominantly effective adherers to glass surfaces. Key words: adherence, Corynebacteriuo diphtheriae, diphtheria, hemagglutination, Introduction The ability of different diphtheria bacillum strains to agglutinate sheep erythrocytes has been reported (Formiga, 1985; Kostyukova et al., 1985; Mattos-Guaraldi and Formiga, 1986). There is considerable variation in the hemagglutinating activity (HA) of the positive strains. Furthermore, some strains were unable to agglutinate intact sheep erythrocytes (Mattos-Guaraldi and Formiga, 1986). The possibility that the property of specific adherence is involved in determining the site of establishment of initial infection by C. diphtheriae led us to evaluate the IIA of different groups of strains of C. diphtheriae based on their toxin production and the source of isolates. In contrast to. most other countries, sucrose- fermenting strains are highly Research supported by CNPq and FINED, Part of a thesis submitted by A.L.M.G, to the Instituto de Microbiologia, UFRJ, in partial fulfillment of the requirements for the Master's degree. Correspondence: Dr. L.C.D. Formiga, Departamento de Microbiologia Médica, Instituto de Microbiologia, UPRJ, Caixa Postal 68040, 21944 Rio de Janeiro, RJ, Brasil,400 ALL. Mattos-Guaraldi and U.C.D. Formiga prevalent in Rio de Janeiro, Brazil. Until the present time, we have been unable to explain the reason for the prevalence of this biotype in this community (Formiga, 1985), which could possibly be related to the adhesive properties of the organism, Since we are not aware of reports which relate hemagglutination or other adhesive properties of C. diphtheriae to sucrose fermentation biotypes, in the present study wwe evaluated the HA properties of sucrose-fermenting and non-fermenting strains. We have also evaluated adherence to glass surface of the organisms and compared it with the HA of both biotypes. Material and Methods Eighty strains of C. diphtheriae isolated in Rio de Janeiro, Brazil were examined. The organisms were identified by the UV fluorescence test (screening test), by the toxi genicity test (radial immunodiffusion method), and on the basis of pyrazine carboxylamidase activity (PYZ) (Sulea at al,, 1980; Formiga et al., 1983; Formiga, 1985, 1986). Strains which showed autoagglutination were excluded from the study. In addition, two non-sucrose- fermenting strains, C. diphtheriae var. mitis CDC-E8392 and var. intermedius CDC-7920, provided by the Centers for Disease Control, Georgia, USA, were also evaluated. The HA for sheep erythrocytes was determined by the microtechnique (Mattos- Guaraldi and Tormiga, 1986). The microorganisms were grown at 37°C for 24 hur 48 hon nutrient agar medium containing 5% calf serum (Yanagawa and Honda, 1976). Bacterial suspensions were prepared in PBS (turbidity equivalent to 0.8 absorbance at 570 nm). ‘Twenty-five 141 of the 0.25% sheep erythrocyte suspension was added to each 25 jl of five serial two-fold dilutions of the microorganism suspension and incubated at 37°C for 3-4 ‘Twenty-eight strains (14 sucrose-fermenting and 14 non-sucrose- fermenting) were also tested for glass surface adherence. Except for the culture medium and the incubation period, the adherence assay employed is based on the assay described by ‘Murchison et al. (1981) for S. mutans. Glass tubes (13 x 100 mm) containing 4 ml of TSB medium (Trypticase Soy Broth, Baltimore Biological Laboratory, Inc.) were inoculated with one C. diphtheriae strain each and incubated stationarily for 48 h at 37°C. Each culture was gently shaken by hand for 5's to remove those cells which grew in close contact with the glass surface but did not actually adhere. The culture fluid was then added and the tubes were incubated for 48 b. Thi procedure was repeated twice. After pouring off the culture fluid for the third time the tubes were scored on a scale of I to IV for adherence: I (strong), confluent coat of cells on sides of tube and localized adherence on glass surface where culture medium is in contact with air; 11 (intermediate), confluent coat of cells on sides of tube; II (weak), localized adherence ‘on glass surface where culture medium is in contact with air; IV (negative), no visible adherence. A suctose-fermenting strain and a non-fermenting-strain were also cultured in TSB medium supplemented with 5% sucrose for measurement of adherence to a glass surface.Hemagglutination and adhesive properties of C. diphtberise 401 Data were analyzed statistically by determining the difference between the parameters of two binominal distributions (Remington and Schork, 1970). Results The HA results obtained for organisms isolated from two sites of infection are given in Table 1, for toxigenic and nontoxigenic strains in Table 2 and for sucrose fermentation biotypes in Table 3. Sheep erythrocytes were agglutinated by 72% of the toxigenic strains and by 87% of the nontoxigenic strains, by 67% of the strains isolated from throats and 94% of the strains from skin lesions, and by 65% of sucrose-fermenting and 93% of non-sucrose- {fermenting strains. HA was significantly higher in throat than in ski isolates and in non-sucrose- Table 1 - Hemagglutination of C, diphtheriae with sheep erythrocytes in relation 10 the origin of strains from throats or skin sources determined as the differences between the parameters of two binomial distributions. ‘The 95 percent confidence interval for p-p; -0.45 -0.1. Site of infection Hemagglutinating Throat Skiniesion activity — No.of % ‘Statistical ~—sNowof_ = % Statistical strains analysis strains analysis Positive 3167.39 O.674(p]) «32S 0.941(p2) Negative 15 3261 0.326 2 5.88 0.059 Subtotals 46 100.00 1.000 3499.99 0,999 ‘Table 2 - Hemagglutination of C. diphtheriae with sheep erythrocytes in relation to toxin production determined as the difference between the parameters of two binomial distributions. The 95 percent confidence interval for p]-p2 is (+0.3; +0.02). Toxigenic Nontoxigenic Hemagglutinating activity No.of % ~— Statistical ~=—=«Nowof == % += ‘Statistical strains analysis strains analysis Positive 31 7210 0.724(p1) 34 87.18 0.872(p2) Negative 12 27.90 0.279 s 12.82 0.128 Subtotals 43 100.00 1,000 39 100.00 1.000402 A.L, Mattos-Guaraldi and L.C.D. Formiga Table 3 - Hemagglutinating activity with sheep erythrocytes of C. diphtheriae secrose fermentation biotypes determined as the difference between the parameters of two binomial distributions. ‘The 95 percent confidence interval for p1-p2 is (-0.45 -0.1). ‘Non-fermenting strai Fermenting-strains Hemagglutinating activity No. % Statistical No. % Statistical analysis analysis Negative 3 Tl 0.071 i4 35.0 0.350 Positive 39 929 0.92%) 26 65.0 0.650(p1) Subtotals 42 100.0 1.000 40 100.0 1.000 fermenting than in sucrose-fermenting biotypes. The presence or absence of toxin produc- tion did not correlate with HA activity. Correlation of HA with biotypes (gravis, mitis and intermedius) could not be satisfactorily studied because of the unequal numbers of different biotypes and also because the gravis strains showed autoagglutination, The results of adherence to glass tests for 14 sucrose-fermenting strains and 14 non- fermenting strains are given in Table 4, Most of the strains (86%) were able to adhere to fa glass surface independently of their sucrose-fermenting activity. Non-adhering strains were also observed (14%). The adherence levels of C. diphtheriae strains are presented in Table 5. Differences occurred in adherence levels between sucrose-fermenting and non- fermenting biotypes. We observed that, although sucrose-fermenting strains exhibited less HA than the non-sucrose-fermenting strains, most of them proved to be good (I, 57%) and moderate (II, 14%) adherers to glass surface, whereas strains of the non-sucrose-fermenting, biotype which exhibited stronger HA were predominantly poor adherers (II1, 64%) to the lass surface. Cultivation on media con- taining sucrose (5%) did not affect the original adherence features of the Table 4 - Adherence to glass surface in relation to sucrose fer- ‘mentation by C. diphtheriae. two biotypes. ‘The method of Murchison et al. (1981) was used. Tests with . scores of I, Il or III were taken as positive and IV was taken Discussion to be negative (see Methods). The ability of bacteria to Fermenting-strains Non- fermenting strains attach. to animal cells is usually as- Adherence sessed by hemagglutination. In this to glass No. % No. % study, most of the C. diphtheriae strains isolated from throats (67%) Positive Bo 87h 2 857) and from skin lesions (94%) were Neeive FM able to agglutinate intact sheep6 om Ht ot UH GUL UO Oe HOCH Zs Troy ry ye (ee or ee cece ees ’ Oe eae neces z oe ee ee i eae ee ee eee eee ees Leet ee Oe es ° % °N % ON % CN % “ON % “ON % "ON % “ON % "ON % “ON % “ON Taogns AL m1 1 I Taioigns AL ut 1 1 sion surezis 3unuowz2}-UON suyens Supwouss04 SuneunnyS3ew9} s]oda{ aoepns sse13 01 aousLOUPY soameBau ‘1 yea TI] ‘aveypouoru ‘y] Huong ‘J “soroompAse dats « suas souaipydip “> Stanuoutia/-uow pun Susuousiaf-asosons fo sjenay aaDfins ssop® 0} aouaseupy pun siaty Seamunns3ou2} - S18 403404 AL, Mattos-Guaraldi and L.C.D. Formiga erythrocytes (Table 1), suggesting a common surface component for the two groups of organisms. Fimbriae may be related to the HA property of the organism since these structures were observed in a hemagglutinating C. diphtheriae strain (Yanagawa and Honda, 1976). We cannot exchide the possibility of involvement of other different surface antigens in the C. diphtheriae adherence mechanism eventually not detected by sheep erythrocyte receptors. Quantitative variation of these components may also occur since we observed non-hemag- slutinating strains isolated from both sites of infection and principally because statistical analysis (Tables 1, 2 and 3) demonstrated significantly higher HA in bacteria isolated from skin than in bacteria isolated from throats. Tt seems that the factors involved in the HA and adhesive properties of the organisms are neither directly dependent on tox gene expression (Table 2) nor related to Porphyrin production (Formiga, 1986), since the latter property is observed in all C. diphtheriae strains together with other metabolic activities. In contrast to most other countries, a high incidence of sucrose-fermenting C. diphtheriae strains occurs in Rio de Janciro, Brazil (Formiga, 1985), Although most of the sucrose-fermenting and non-fermenting strains agglutinated weakly (1, 2 and 4 titers) some strains presented stronger HA (8 and 16 titers). Independently of their sucrose-fermenting property, diphtheria bacilli presented 20% non-agglutinating strains when sheep erythrocytes were used. However, when considered on the basis of their sucrose-fermenting property, sucrose-fermenting biotypes showed a higher percentage of negative strains (35%) than non-sucrose-fermenting biotypes (7%) (Table 3). All C. diphtheriae isolates were from the same geographical region, ie., Rio de Janeiro, Brazil. Most of the throat isolates were sucrose-fermenting strains while strains isolated from skin were mostly of the non-fermenting biotype, as previously observed by Formiga (1985) and Formiga ct al. (1986). Strains from skin lesions and strains of the non-fermenting biotype showed predominantly strong HA. Our data suggest an association between sucrose-fermenting biotype and the site of infection of C. diphtheriae strains. The selective pressure that occasionally occurs in C. diphtheriae, leading some strains to colonize the throat rather than the skin, may possibly be related to sucrose fermenting activity. Some strains may acquire the property of adherence to the throat simultaneously with acquiring the property of sucrose fermentation, possibly by plasmid jon, as is the case for C. renale (Deacock et al., 1983). A variety of bacteria have been reported to adhere to solid surfaces. Bacterial surface appendages such as fimbriae have been implicated in the adhesion of certain bacteria to both mammalian tissue and solid surfaces (Watt and Ward, 1977; Brinton Jr., 1985). In the present communication, evidence is provided for the attachment of C. diphtheriae to glass surfaces (Table 4), Diphtheria bacilli adhered to glass regardless of the sucrose content of the culture medium, In contrast to S. mutans (Staat et al., 1980: Stinson et al., 1981), there was no production of extracellular material from sucrose. Different levels of adherence to glass occurred among C. diphtheriae strains (Table 5), Sucrose- fermenting organisms proved to be predominantly good and moderate adherers in contrast to nonHemagglutination and adhesive properties of C. diphtheriae 405 fermenting strains, which were mostly poor adherers. However, there were also non- adhering strains of both biotypes. When we related hemagglutination to adherence to a glass surface by C. diphtheriae we observed that most strains exhibited both adhesive properties. Moreover, a strain that did not exhibit either adhesive property was also detected. These observations reveal the possible multiplicity of interactions with the bacterial surface. C. diphtheriae adheres to a glass surface as Klebsiella and S. marcescens sp do. However, C. diphtheriae agglutinates sheep erythrocytes in contrast to what was reported for fimbrize of the latter two organisms, which do not agglutinate any species of erythro- cytes (Ottow, 1975). Although sucrose-fermenting strains were better adherers to a glass surface, the non-fermenting strains that adhered weakly to glass showed stronger HA. In summary, we conclude that the various C. diphtheriae strains studied differ in their adhesive properties. References Brinton Jr CC (1985). The structure, function, synthesis and control of bacterial pili and a molecular model for DNA and RNA transport in Gram negative bacteria. Transactions of the New York Academy of Sciences, 27: 1003-1054, Deacock SJ, Steward KA & Carne HR (1983). The role of adherence in determining the site of infection by Corynebacterivon diphtheriae. Journal of Hygiene, 90: 415-424, Formiga LCD (1985), Difteria. Enfoque microbiol6gico epidemiol6gico. Importincia das lesées cutfineas © ‘novas possiblidades de diagn6stico. Anais Brasileiros de Dermatologia, 60: 337-338, Formiga LCD (1986). Diagnéstico microbiolégico da diftera, Revista Brasileira de Patologia Cllnica, 22: 52-58, 90-93, 122-130. Formiga LCD, Camello TCF, Guaraldi ALM, Rangel LBA & Assis ACFCB (1983), Teste de fMuorescéncia € pesquisa da atividade pirazina-cerboxilamidase (PYZ) na identificagéo do bacilo Aiftérico atoxinogtnico. Revista de Microbiologia, 14: 172-173, Formiga LCD, Assis ACB, Rangel LBA, Camello TCF, Suassuna I, Martins Netto E & Marsden PD (1986). Isolamento de Corynebacterium diphtheriae de dlceres culdinco-mucosas por Leishmania, Revista Brasileira de Patologia Cltnica, 22: 202-204, Kostyukova NN & Pereversey NA (1985). Adhesion in Corynebacteriwn diphtheriae. Zhurnal Milrobiologi, Epidemiologi i bnmunobiologit SSSR, 11:30+33. Mattos-Guaraldi AL. & Formiga LCD (1986). Agglutination of sheep erythrocytes by Corynebacterion diphtheriae, Brasitan Journal of Medical and Biological Research, 19: 15-77. Murchison H, Larrimore $ & Curtis LIT R (1981). Isolation and characterization of Streptococcus mutans ‘mutants defective in adherence and agaregation. Infection and Inamunity, 34: 1044-1055. Ottow JCG (1975). Ecology, physiology and genetics of fimbrise and pili. Annual Review of Microbiology, 29: 79-108. Remington RD & Schork MA (1970). The difference between the parameters of two binominal distributions. In: Staristes with Application t0 the Biological and Health Sciences. Prentice-Hall, Princeton, New Jersey, 185-187. Staat R-H, Langley SD & Doyle RJ (1980). Streptococcus matans adherence; presumptive evidenee for protein-mediated attachment followed by glucan-dependent cellular accumulation. Infection and Inamaaity, 7: 675-681. Stinson MW, Tinks DC & Merrick JM (1981). Adherence of Streptococcus mutans and Streptococcus sanguis to salivary components bound to glass. Jnféction and nonumity, 32: 583-591 ulea IT, Pollice MC & Barksdale L (1980). Pyrazine-carboxylamidase activity in Corynebacteriun. Imvernational Journal of Systematic Bacteriology, 30: 446-472,406 ALL, Mattos-Guaraldi and L.C.D. Formiga Watt PJ & Ward ME (1977), The interaction of gonococei with human epithelial cells. In: Roberts RB (Editos), The Gonococeus. John Wiley and Sons, New York. Yanagawa R & Honda E (1976). Presence of pili of human and animal parasites and pathogens of the genus Corynebacterium. Infection and Immunity, 13: 1293-1295. Received June 11, 1990 Accepted April 10, 1991