A screening of streptococci freshly isolated from human and animal sources for binding of human IgG

Zbl. Bakt. Hyg. A 265, 420-429 (1987)

A Screening of Streptococci Freshly Isolated from Human and Animal Sources for Binding of Human IgG KARL-HERMANN SCHMIDT, OTTO KOHNEMUND, and WERNER KOHLER Akademie der Wissenschaften der DDR, Zentralinstitut fur Mikrobiologie und experirnentelle Therapie, 6900 Jena

With 4 Figures· Received June 27, 1986 . Accepted November 10, 1986

Abstract Human isolates of groups A, C, and G streptococci as well as animal isolates of group C were investigated with respect to their binding capacity for human IgG by using the direct fluorescence technique and the Mancini test. From each serological group of human isolates, more than 100 strains were tested. The results were evaluated statistically with respect to serological group, type and source of isolates. Between human isolates of groups A, C and G, no statistical differences concerning IgG binding were found. However, group C streptococci isolated from pigs showed a significantly higher number of strains with high IgG binding than the human isolates of groups A, C and G. Group A streptococci isolated from suppurating lesions showed an increased IgG uptake when compared with isolates from scarlet fever patients or patients with throat infections. However, strong IgG binding by group A steptococci seems not to be restricted to certain types. By using selected streptococcal strains, it was found that IgG absorption from a solution of purified IgG was much higher than from IgG solutions containing albumin or from diluted serum. The results are discussed in connection with the competition of different plasma proteins for binding sites on the streptococcal cell surface and with a possible influence of IgG receptors on the virulence of streptococci.

Zusammenfassung Menschliche Isolate von Gruppe A, C und G Streptokokken sowie auch tierische Isolate der Gruppe C wurden mittels direkter Fluoreszenztechnik und Mancini-Test hinsichtlich ihrer IgG-Bindungskapazitat untersucht. In jeder serologischen Gruppe wurden mehr als 100 Stamme untersucht und die Ergebnisse statistisch ausgewertet. Zwischen den Gruppen A, C und G ergaben sich keine Unterschiede hinsichtlich der IgG-Bindung. In jeder Gruppe wurden stark, maf~ig und schwach bindende Starnme gefunden, Dagegen zeigten die von Schweinen isolierten Gruppe C-Stamme eine signifikant erhohte IgG-Bindung. Gruppe A-Streptokokken, die vorrangig aus Wundinfektionen isoliert wurden, absorbierten ebenfalls erhoht IgG, verglichen mit Stammen, die von Scharlachpatienten oder Patienten mit Rachenrauminfektionen stammten.

IgG Binding of Streptococci

421

Eine erhohte IgG-Bindungeinze1ner Gruppe A Starnme scheint jedoch nicht an bestimmte Typen gebunden zu sein. Mit ausgewahlten Stammen wurde die IgG-Aufnahme aus einer Losung mit gereinigtem IgG, aus einer IgG-Lasung, enthaltend Serumalbumin, und aus verdiinntem Humanserum bestimmt. Die IgG-Absorption aus der reinen IgG-Lasung war wesentlich hoher als aus den andere Proteine enthaltenden Losungen. Die Ergebnisse werden diskutiert hinsichtlich moglicher, die Virulenz beeinflussender Funktionen solcher Rezeptoren und beziiglich der Konkurrenz verschiedener Plasmaproteine urn Bindungsstellen an der Bakterienzelle.

Introduction Recently, a considerable number of investigations dealt with bacterial receptors for human or animal serum proteins (3, 8, 9, 10, 11, 16,23). For example staphylococcal protein A as an IgG Fe-receptor has become an universal tool for use in the microbiological, immunological, and clinical practice (5). Receptors of other bacterial species, e.g. streptococci, were investigated with increasing interest, because such receptors are possibly virulence factors. Fibrinogen binding to group A streptococci has been reported to decrease their phagocytosis by human neutrophils (24,25) and albumin has also been described to influence the phagocytic process (24). Injection of IgG Fe-reactive protein isolated from a type 15 group A streptococcus increased the susceptibility of mice to virulent group A streptococci of type 12 (2). Investigations of streptococci of several serological groups for IgG binding receptors have been published by a number of groups (14, 15, 17,22) since Kronvall in 1973 (9) showed non-immune reactivity of streptococci of groups A, C and G with IgG. The interaction between streptococci and IgG has been found to take place between the Fe fragment of IgG and the streptococci (3a). In this study we determined quantitatively the binding of IgG by streptococci of human origin of groups A, C and G by using two methods: a) measuring of fluorescence of fluorescein-labelled IgG to the bacterial sediments and b) quantification of bound IgG absorbed from supernatants by using the Mancini technique. More than 110 strains of each group were tested by the first method. The results were evaluated statistically with respect to serological group, type and source of the isolates. Group C streptococci isolated from pig and cattle were also included.

Material and Methods

Bacteria Human clinical isolates of Streptococcus pyogenes (group A; 127 isolates), S. equisimilis (group C; 113 isolates), and group G streptococci (114 isolates) as well as 51 strains of S. equisimilis isolated from pig and 8 from cattle were investigated. Staphylococcus aureus strains Cowan I and Wood 46 were included as reference strains for IgG binding studies.

Grouping and typing of streptococci The ~-hemolytic streptococci were grouped ad modum Fuller (4) using formamide extracts. Additionally, group C strains were confirmed by agglutination of trypsinized streptococci with the lectin of Helix pomatia (7). Group A streptococci were typed by T agglutination (6). Briefly, streptococci were incubated overnight at 30°C in Pope broth, sedimented, washed with saline, and trypsinized at 37°C for 2 h and at 52°C for 30 min 27 Zbl. Bakt. Hyg. A 265/3-4

422

K.-H. Schmidt, O. Kiihnemund, and W. Kohler

with trypsin emulsion ad modum Cole and Onslow. The supernatant was discarded and the streptococcal suspension tested with anti-T antisera. Whereever possible, the type determination was confirmed by M precipitation. Arrangement of types in Fig. 3 follows the T agglutination patterns. Strains showing agglutination pattern 121B3264were most probably type 22.

Cultivation of bacteria

3rc.

Strains were cultivated for 18 h in 3 ml of Todd-Hewitt broth at Then bacteria were sedimented by centrifugation and successively washed with 0.015 molll phosphate buffer pH 7.5 containing 0.25 molll NaCl, followed by 0.02 molll citric acid buffer, pH 3.0 and finally with the first buffer as described previously (22). As reference strains, Staphylococcus aureus Cowan I and Wood 46 representing strong and poor IgG binding, respectively, were grown in parallel to each test series. For studies with the Mancini technique, selected strains were grown in 80 ml ToddHewitt broth and washed as described above.

Direct fluorescence test for determination of IgG binding The procedure has been described in detail previously (22). Each bacterial sediment was mixed with 0.15 ml of a 1 mg/ml IgG-FITC1 solution and incubated for 30 min at room temperature. Then 5 ml phosphate buffer (see above) containing 0.1 % Tween 20 were added and after mixing, the suspensions were centrifuged at 4000 g. The supernatants were carefully removed. The fluorescence of the bacterial sediments, which were transferred to glass slides, was determined at the contact layer of the adhered drops on the inverted slide. A "Zetopan" microscope equipped with a fluorescence unit and a photometer type 600 b from Reichert, Vienna (Austria), was used. The equipment was calibrated daily with a uranyl glass standard. The flourescence intensity of the strains to be screened was compared with the data, obtained for the reference strains Staphylococcus aureus Cowan I (the average fluorescence intensity of bound IgG-FITC of 13 different culture sediments of strain Cowan I was defined to represent 100% binding) and Wood 46 (5% IgG binding) giving a quantitative information of streptococcal IgG binding(22).

Mathematical calculations The significance of differences of IgG binding between the bacterial strains was determined by the Wilcoxon test.

Determination of IgG binding capacity of selected strains Absorbed IgG was quantified on selected strains by using the Mancini technique (3). The washed bacterial sediments of 80 ml culture were incubated for 1 h with 3 ml of a 1: 32 dilution of human serum (dilution buffer: 0.05 molll phosphate, pH 8.0) or with 3 ml of human IgG (0.5 mg/ml) dissolved in phosphate buffer. In some experiments,S mg/ml bovine serum albumin were added to the IgG solution. After centrifugation,S ul of the supernatants were applied to 4 mm wells punched in 1.5% agarose gels containing 4% antihuman-IgG serum (from rabbits). The IgG concentration was evaluated after 24h. To enable the calculation of IgG absorption by a specific number of bacteria, single cocci were counted in a Thoma chamber before incubation with IgG.

1

Human IgG labelled with fluorescein isothiocyanate.

IgG Binding of Streptococci

423

Results

Screening of streptococci of groups A, C, and G for binding of IgG-FITC Among the three streptococcal groups tested, strains with strong (> 40%), moderate (20 to 40%) or poor « 20%) binding of IgG were found (Fig. 1, Table 1). No statistical differences in binding capacity were found to exist between the human isolates of groups A, C, and G. However, among the group C streptococci isolated from pigs, a significantly higher number of strongly IgG-binding strains was found (p < 0.01). Among groups C and G some strains were found to bind more than 70% IgG. The eight group C strains isolated from cattle did not bind noteworthy amounts of IgG (Table 1).

IgG binding of human isolates of different sources Streptococcal isolates from patients suffering from scarlet fever, throat infections or other infections (suppurating lesions of skin, wounds etc.) were evaluated statistically with respect to their IgG binding (Fig. 2). No remarkable differences were found to exist between the isolates of groups C and G from different infections. However, an elevated number of strongly IgG-binding group A streptococcal strains was found in the group "other infections" representing mainly suppurating lesions (p < 0.05). The streptococci of group C and G isolated from scarlet fever patients can be assumed to have been secondary isolates.

:::!( 0 (lJ

o c (lJ o

UI

(lJ L-

a

~

'6-

100 80 ~

60 40

(lJ

.~ ....,

0

aJ L-

20

~

..

..... ::: .:::: :::.......

11::::':":::_'

.. :.. !!!:..

r

~~:

t"

...

f:···

:

human

group A

I

.. ..

!.

iL-

l!: '

!!::. n::-:sa pi-

It.

t

Wood 46

pig

group C

group G

Fig. 1. Bindingof fluoresceine-labelled IgG to human isolates of streptococci of groups A, C and G and to porcine isolates of group C. Staphylococcus aureus strains Cowan I and Wood 46 were used as reference strains.

424

K.-H. Schmidt, O. Kiihnemund, and W. Kohler

Table 1. Percentage of streptococci of groups A, C and G binding IgG strongly (> 40% ), moderately (20-40% ), and poorl y « 20%) IgG bindung

group A n' %2

strong moderate poor

18 63 46

total I

2

14.2 49.6 36.2

127 100

human group C % n 10 47 56

group G n %

8.8 41.6 49.6

15 54 45

113 100

13.1 47.4 39.5

114 100

porcine group C n %

bovine group C n %

24 16 11

0 1 7

47 31.4 21.6

51

0 12.5 87.5

8 100

100

Number of strains Percentage of the absolute number of strains tested

IgG binding of different serological types of group A streptococci The isolates of group A strepto cocci represented 20 different serological types. In Fig. 3, the number of strains which showed a higher IgG binding than the median value of 23 % established for group A is given for each type (hatched and black bars). Among the isolates of types 3, 13, 6, 12, and 48 , no strongly IgG-binding strains were found. Statistical calculations to compare IgG binding of different type s could not give

_100 ~ 0

.,0 .,0c ., II)

80

60

.....

L.-

-., a

:J

-

40

> 20

~ G

"

_: ..-. ...-

L.-

a

..· · · ·

...i,'... ~~ . .... .....- .. i::.. ........ :

'"

b group A

.....

:

...

c

.

a

b group C

::::-

c

.... a

b

..

g;~.

m_ .

...... ..... ::~\.

c

group G

Fig. 2. IgG binding of streptoco ccal strains of groups A, C and G of human origin isolated from patients suffering from a) scarlet fever b) throat infection c) suppur ating lesions

IgG Binding of Streptococci

425

No. of

strains16

14 12 10

8 6 4

2 L..ILJ--J;:L~-JII!ILII!LJ--~~....t;:iU;;~~:=:S-J?~~~...L.Ji!L.Ii!:L..Jnolli!t:""

t ypes

: typed 12/B3264

Fig. 3. IgG binding of different types of group A streptococcal strains

o Straisn binding IgG to below 23% (Median value of group A)

Strains binding IgG to between 23-40% • Strains binding IgG to more than 40%

~

0.5 A

0.4

Q3

Q2

0.1

I

51 2

3

I 4

Jl 5

6

7

~

8

n 9

10

Fig. 4. Quantitation of IgG absorbed by selected strains (Mancini technique) A = mg IgG absorbed by 1010 bacteria from a solution of purified IgG 0 or from diluted human serum •. Streptococcal strains: Group A; 1. No. 27338 (64%h 2. No. 27684 (5%) Group C; 3. No. 26429 (82%) 4. No. 27623 (6%) Group G; 5. No. 24094 (73%) 6. No. 24114 (2%) (human isolates) Group C (porcine isolates) 7. No. 12628 (100%) 8.No. 11862 (11%) Staphylococcal strains: 9. Cowan I (100%) 10. Wood 46 (5%) 2

Percent IgG binding as determined by the fluorescence technique

426

K.-H. Schmidt, O. Kiihnemund, and W. Kohler

a clear evidence because the number of strains within most single types was not sufficient. However, it can be concluded that group A strains binding large amounts of IgG did not belong to certain types.

Quantification of IgG-binding capacity of selected strains On the basis of the screening results shown in Fig. 1, we selected strains from each group showing high and low fluorescence. The amount of IgG absorbed by standardized numbers of bacteria from these strains was determined by the Mancini technique. As shown in Fig. 4, the amount of IgG absorbed by streptococci as well as by staphylococci from a solution of purified IgG was higher than that absorbed from diluted serum. The addition of bovine serum albumin(5 mg/ml) to the IgG solution reduced the IgG uptake by streptococci (results not illustrated). Strongly IgG-binding streptococci bound the same amount (group A) or more (group C and G) IgG/10 10 cells than Staphylococcus aureus Cowan I. For such strains, a binding capacity of between 100 ug and 450 flgIgG/l0 10 cells was found. The IgG absorption from IgG solutions of those strains which showed a week IgG binding in the fluorescence test was also reduced when the Mancini technique was applied.

Discussion The binding of IgG to streptococci has been described to take place between the Fc fragment of IgG and the streptococci (3a, 14, 15, 17). In this investigation the quantitative IgG binding of 413 steptococcal strains from different sources was determined by the direct fluorescence technique. The number of strains was sufficient to enable a statistical evaluation. No statistical differences concerning IgG binding were found between human isolates of streptococci of groups A, C and G. In all three groups strains with poor, moderate or strong IgG binding were found. However, group C streptococci isolated from pigs showed a significantly higher number of strains with high IgG binding (p < 0.01) than the human isolates of groups A, C and G. Only 7.6% of all tested strains showed the same weak IgG binding capacity « 10%) as the protein-A-negative Staphylococcus aureus strain Wood 46. Myhre and Kronvall (15) reported that human isolates of group C and G streptococci showed a more pronounced IgG binding than group A streptococci. We found that only some strains of human isolates of groups C and G showed an outstanding IgG uptake (> 70%), however, the main part of strains did not differ statistically from group A (Fig. 1). Our results support data published by Lebrun et al. (14) who found, by using an indirect qualitative immunofluorescence technique, about 30 to 50% out of 162 streptococcal strains of groups A, C and G isolated from humans to be negative in IgG binding. However, the number of strains of groups C and G used in their investigation was to small to obtain reliable data. Our results obtained with a sufficient number of strains of each group confirm the percentage of poor IgG binding strains reported by those authors (Table 1).

IgG Binding of Streptococci

427

A connection between IgG binding and the source of the isolates was only found for group A streptococci of human origin. Group A streptococci isolated mainly from suppurating lesions showed a significantly increased IgG binding if compared with isolates from scarlet fever patients or patients with throat infections. Comparable results also described by Lebrun et al. (14) suggest a contribution ofFc receptors to the virulence of streptococci causing such infections. Burova et al. (2) reported an increased susceptibility of mice to virulent group A type 12 streptococci when the animals had been injected with purified heterologous IgG Fe receptor of type 15 a few hours before infection. The production of homologous and heterologous anti-IgG antibodies observed after immunisation of rabbits with washed Fe-receptor-carrying streptococci (13, 19,20) may also contribute to the pathogenic processes during streptococcal infections. In further investigations, the influence of such phenomena on the phagocytic processes should be investigated. No relationship between serological type and the presence of IgG receptors on group A steptococci could be established. However, most strains of types 3, 13, 6, 12, and 48 were found to bind IgG weakly. Similar observations with type 12 strains were described recently (1,2). Mouse passages were found to increase the Fe-receptor content of group A streptococci (1, 18). Strong IgG binding of group A streptococci was not restricted to certain types as shown in Fig. 3. The quantitation of IgG binding of selected strains by the Mancini technique showed, that highly IgG-binding streptococci absorbed the same amount and partly even more IgG per 10 10 cells than did the Staphylococcus aureus strain Cowan 1. Interestingly, all strains absorbed a higher amount of IgG when the bacteria has been incubated with a solution of purified IgG instead of diluted human serum. A similar decreasing effect was found by addition of albumin to the IgG solution. We assume that a competition between different serum components for binding sites and their mutual steric hindrance on the steptococcal cell surface is responsible for decreased IgG binding. Several plasma components like albumin (16), haptoglobin (8), fibrinogen (11, 12,21), fibronectin (23), and ~2-microglobulin (10) were described to interact besides IgG with the streptococcal cell surface or with isolated cell wall components. The differences between IgG uptake from a solution of purified IgG and from diluted serum vary in a wide range. This could be influenced (besides competition) by differences in the affinity of IgG to corresponding receptors or by unspecific absorption (weak interaction) of exceeding IgG. In the fluorescence assay, unspecifically bound IgG was removed by washing with buffer containing Tween 20. In our test systems, we incubated bacteria with an excess of IgG (about 500 [1g110 10 cells). Other investigators used about a tenth of radiolabelled IgG in their test system (3, 15). Because some strains can bind more than 200 [1g IgG/I0 Io cells, the applied IgG concentration of 2-5[1g per 10 9 cells (15) should not be sufficient to occupy all binding sites on the cell surface. In such cases, differences between high IgG binding strains are no longer detectable.

Acknowledgements. The authors thank Mrs. B. Schafer for her excellent technical assistance.

428

K.-H. Schmidt, O. Kiihnemund, and W. Kohler

References

1. Buroua, 1. A., P.Christensen, R. Grubb, A. Johnsson, G. Samuelsson, C. Schalen, and M.-1. Svensson: Changes in virulence, M protein and IgG Fe receptor activity in a type 12 group A streptococcal strain during mouse passages. Acta path. microbioI. scand. Sect. B 88 (1980) 199-205 2. Buroua, 1. A., P. Christensen, A. Grubb, R. Grubb, A. Johnsson, C. Schalen, and 1. Truedsson: Streptococcal IgG Fe-receptor as a virulence factor . pp.205-206. In: Basic Concepts of Streptococci and Streptococcal Diseases. Proc. of the VIIIth International Symposium of Streptococci and Streptococcal Diseases, Lund, 1981, eds. S. E. Holm and P. Christensen. Reedbooks, Churtsey (1982) 3. Christensen, P., S. E. Holm, and]. Henrichsen: Deplation of IgG in rabbit sera by absorption with some group A streptococci. Int. Arch. Allergy appl. Immunol. 53 (1977) 133-136 3a. Christensen, P., B. G. Johansson, and G. Kronuall: Interaction of steptococci with the Fe fragment of IgG. Acta path . microbiol. scand. Sect. C 84 (1976) 73-76 4. Fuller, A. T.: The formamide method for the extraction of polysaccharides from haemolytic streptococci. Brit. j expoPath. 19 (1938) 13D-139 5. Goding,]. W.: The use of staphylococcal Protein A as an immunological reagent. J. Immunol. Merh, 20 (1978) 214-253 6. Kohler, W.: Die Typisierung harnolysierender Streptokokken der Gruppe A. I. Mitt. : Methoden der Typisierung und die Herstellung agglutinierender Sereno Z. Immun. Forsch. 120 (1960) 253-278 7. Kohler, W., O. Prokop, and O . Kiihnemund: Routine identification of group C streptococci by means of an agglutinin (protectin) from the albumen gland of the edible snail, Helix pomatia. J. Med. Microb iol. 6 (1973) 127-130 8. Kohler, W. and O. Prokop: Relationship between haptoglobin and Streptococcus pyogenes T4 antigens. Nature 271 (1978) 373 9. Kronuall, G.: A surface component in group A, C and G streptococci with non immune reactivity for immunoglobulin G. ]. Immunol. 111 (1973) 1401-1406 10. Kronuall, G., E. B. Myhre, L. Bjorck, and I. Berggdrd: Binding of aggregated human ~2­ microglobulin to surface protein structure in group A, C and G streptococci. Infect. Immun. 22 (1978) 136-142 11. Kronvall, G., C. Schonbeck, and E. Myhre: Fibrinogen binding structures in ~-hemoly­ tic streptococci group A, C and G. Acta path. microbiol. scand. Sect. B 87 (1979) 303-310 12. Kuhnemund, 0.,]. Havlicek, H. Knoll.]. Sjoquist, and W. Kohler: Interaction of group A streptococcal M protein with fibrinogen. Acta path. microbiol. scand. Sect. B 93 (1985) 201-209 13. Lebrun, 1., 1. Gangreot-Keros, and]. Pillet: Induction in rabbits of antibodies to human IgG by pathogenic streptococci. Role of Fe (0) receptors, pp. 207-208. In: Basic Concepts of Streptococci and Streptococcal Diseases. Proc. of the VIIIth International Symposium of Streptococci and Streptococcal Diseases, Lund 1981, eds. S. E. Holm and P. Christensen. Reedbooks, Churtsey (1982) 14. Lebrun, 1.,]. Pillot, 1. Grangeot-Keros, and M. T. Rannou: Detection of human Fe (y) receptors on streptococci by indirect immunofluorescence staining : a survey of streptococci freshly isolated from patients . ]. Clin. Microbiol. 16 (1982) 20D-201 15. Myhre, E. B. and G. Kronuall: Heterogenicity of nonimmune immunoglobulin Fe reactivity among gram-positive cocci: Description of three major types of receptors for human immunoglobulin G. Infect. Immun. 17 (1977) 475-482 16. Myhre, E. B. and G. Kronuall: Demonstration of specific binding sites for human serum albumin in group C and G streptococci. Infect. Immun. 27 (1980) 6-14 17. Reis, K. J., E. M. Ayoub, and M. D. P. Boyle: Detection of receptor s for the Fe region of IgG on streptococci. J. Immunol. Meth. 59 (1983) 83-94

IgG Binding of Streptococci

429

18. Reis, K.,]., M. Yarnall, E. M. Ayoub, and M. D. P. Boyle: Effect of mouse passage on Fc receptor expression by group A streptococci. Scand. J. Immunol. 20 (1984) 433-439

19. Schalen, c, 1. A. Buroua, P. Christensen, R. Grubb, G. Samuelsson, and M. 1. Svensson: IgG F(abh of rabbit anti-M sera but not the unfractionated sera are bactericidal for some group A streptococci with IgG Fe receptor activity. Acta path. microbiol. scand. Sect. C 89 (1981) 247-252 20. Schaien.C; 1. A. Burova, P. Christensen, R. Grubb, M. 1. Svensson, and A. A. Totolian: Induction and specificity of anti-IgG following immunization with streptococci. pp. 89-90. In: Recent Advances in Streptococci and Streptooccal Diseases. Proceedings of the IXth International Symposium on Streptococci and Streptococcal Diseases, Tokyo, 1984, eds. Y. Kimura, S. Kotani, and Y. Shiokaioa. Reedbooks, Churtsey (1985) 21. Schmidt, K.-H. and W. Kohler: Interaction of streptococcal cell wall components with fibrinogen. I. Adsorption of fibrinogen by immobilized T-proteins of Streptococcus pyogenes. Immunobiol. 158 (1981) 330-337 22. Schmidt, K. H. and R. Stracke: Semi-quantitative determination of IgG-binding structures on bacteria by direct fluorescence technique. Zbl. Bakt. Hyg. Orig. A 262 (1986) 448-454 23. Simpson, W. A. and E. H. Beachey: Adherene of group A streptococci to fibronectin on oral epithelial cells. Infect. Immun. 39 (1983) 275-279 24. Wagner, B., K. H. Schmidt, M. Wagner, and W. Kohler: Albumin bound to the surface M protein-positive streptococci increased their phagocytosis by human polymorphonuclear leukocytes in the absence of complement and bactericidal antibodies. Zbl. Bakt. Hyg. Orig. A 261 (1986) 432-446 25. Whitnack, E. and E. H. Beachey: Antiopsonic activity of fibrinogen bound to M protein on the surface of group A streptococci. J. Clin. Invest. 69 (1982) 1042-1045 Dr. Karl-Hermann Schmidt, Akademie der Wissenschaften der DDR, Zentralinstitut fur Mikrobiologie und experimentelle Therapie, Beutenbergstr. 11, DDR-6900 Jena