differing from the Neisser-Wechsberg leucocidin, which does not affect human CLASSIFICATION OF 110 STRAINS OF STAPHYLOCOCCUS AUREUS

Transcription

1 CLASSIFICATION OF 110 STRAINS OF STAPHYLOCOCCUS AUREUS Lederle Laboratories, Inc., Pearl River, New York Received for publication April 24, 1942 A collection of strains of Staphylococcus aureus was studied with a view to finding out what relationship, if any, could be detected between the agglutinative grouping and the production of polysaccharide, alpha hemolysin, or leucocidin by the various strains. A number of serological groupings for staphylococci have been described in the literature, the most recent ones being those of Blair and Hallman (1936), Yonemura (1936), Cowan (1938), and Christie and Keough (1940). The classification put forward by Cowan has probably been the most widely adopted, and is the one employed in this study. Of the polysaccharide classifications, the "A" and "B" classification of Julianelle and Wieghard (1935) remains fundamental for pathogenic staphylococci. Thompson and Khorazo (1937) have found three additional groups containing mostly non-pathogenic staphylococci, and Cowan (1938) places pathogenic strains giving irregular polysaccharide reactions into a general class "C". The Julianelle classification is the one employed for the work described in this paper. A staphylococcal leucocidin which destroys rabbit leucocytes was described by Van der Velde (1894) and studied further by Neisser and Wechsberg (1901). More recently (1936) Wright has worked with this Neisser-Wechsberg leucocidin, and she suggests, as do also Valentine (1936) and Proom (1937), that it is identical with staphylococcal alpha-hemolysin. Panton and Valentine (1932), Valentine (1936), and Valentine and Butler (1939), however, have demonstrated that certain strains of staphylococci elaborate a leucocidin which is distinct from the Neisser-Wechsberg type. The Panton-Valentine leucocidin is active against human as well as rabbit leucocytes, thus differing from the Neisser-Wechsberg leucocidin, which does not affect human leucocytes. The two leucocidins also differ in their mode of attack upon the leucocytes. Both cause disintegration of the cell nucleus, but the appearance of the damaged particles is different. Panton-Valentine leucocidin in stronger concentration produces complete disintegration of the cytoplasm. Panton-Valentine leucocidin is the one referred to in this paper. Dr. John E. Blair of the Hospital for Joint Diseases, New York, supplied us with Cowan group I and group III strains, and Dr. Champ Lyons of Harvard Medical School, the group II strain. Dr. L. A. Julianelle of Washington University School of Medicine, Saint Louis, sent us his representative polysaccharide "A" and "B" strains. All the cultures included in the study were isolated from human infections. The origin of the strains is shown in table

2 590 METHODS AND PROCEDURES Agglutinating sera were prepared from the three Cowan group strains and the cultures under study were classified by slide agglutination. The sera were prepared by immunizing rabbits with organisms from 18-hour beef-infusion broth cultures. The growths were resuspended in physiological salt solution containing 0.5 per cent formalin. The group II serum as prepared, contained considerable group III agglutinin, which was absorbed out. No preliminary absorption was necessary with the antisera of the other groups. For antigens intended for the method of slide agglutination, 3-hour growths were used, as recommended by Cowan (1939). It was found that the employment of the trypsin-digest beef-heart broth described by Pauli and Coburn (1937) for streptococci, greatly reduced the incidence of granular growth or autoagglutination, and this broth was used to grow the antigens. The antigens were heated rapidly to boiling according to Cowan's method, centrifuged, and resuspended in trypsinized broth, with a density of. 100 billion organisms per milliliter. TABLE 1 Origin of strains Infection of skin Infections of bone Septicemia and bacteremia Miscellaneous internal infections... 7 Type of lesion not stated in history For the tests, 0.01 ml. of antigen and 0.03 ml. of serum (1-5 dilution) were mixed on pa glass plate and rocked 5 minutes. Readings were made with a hand lens against an illuminated dark background. Extracts for carbohydrate detection were prepared from 103 strains of the series by the method of Julianelle and Wieghard (1935). A uniform volume of broth, 600 ml., was employed in all cases, and the extract obtained was made up to a volume of 60 ml. The extracts were tested against sera produced in rabbits with the Julianelle "A" strain 13 and the "B" strain Mll. Forty-eight of the strains studied were tested for alpha-hemolysin production. Cultures were grown 48 hours in veal-infusion, semi-solid agar in an atmosphere of 80 per cent oxygen and 20 per cent carbon dioxide. The fluid obtained from the medium was Berkefeld-filtered and tested for toxin with a 1 per cent suspension of rabbit erythrocytes. Strains that had been carried in the laboratory for a number of years were not tested for hemolysin-production. Recently isolated strains were tested for leucocidin production. The strains were grown on veal-infusion agar, according to the method described by Valentine (1936), and tested for potency by his method, with human leucocytes. RESULTS When tested against the Cowan sera, the cultures were divided as shown in table 2.

3 CLASSIFICATION OF STAPHYLOCOCCUS AUREUS 591 Several of the negative strains showed a common antigen not contained in any of the Cowan groups, and these apparently constitute a supplementary group. The type strain of this group, no. 235 (Smith strain) of the collection, was received by us from Dr. Ren6 J. Dubos of the Rockefeller Institute and had been obtained from a human infection. The relationship between the agglutinative grouping and the polysaccharide reactions of the strains is shown in table 3. As indicated in table 3, extracts of 85 strains gave strongly positive reactions for Julianelle "A" polysaccharide, while 18 reacted weakly or not at all. None of the extracts gave "B" polysaccharide reactions. All strains produced coagulase except two members of the "ungrouped" negative agglutinative group. Another TABLE 2 Agglutinative groups Group I... Group II... Group III Group I or III... Negative (ungrouped)... TABLE 3 Agglutinative and polyeaccharide relationships 38 strains 10 strains strains 17 strains 9 strains COWAN AGGLUTINATIVE JULIANELL "A" POLYSACCHA3IDE JULIANELLE "B" POLYSACCEARIDE GROUP Positive Slight or negative Positive Negative I II III I or III group Ungrouped member of the negative group was the sole non-fermenter of mannitol. The three strains that gave these irregular reactions did not produce detectable "A" polysaccharide. Table 3 shows that of 10 group II strains, 4 or 40 per cent yielded a lessened amount of "A" polysaccharide, but of 93 strains exclusive of group II, 14 or only 15 per cent yielded a lessened amount of "A" polysaccharide. This is in agreement with the observation of Cowan (1939), that irregularities of polysaccharide production are most frequently associated with group II. All our recently isolated group II strains were leucocidin producers. It was considered useless to test old strains for leucocidin production. Most of the leucocidin-producing strains belonged to group II, and the majority of them appeared to produce less than the average amount of "A" polysaccharide. Of 4 members of group III that were weak in "A" polysaccharide, two were leucocidin

4 592 strains. Of the total number of cultures examined, 10 per cent belonged to group II, but of leucocidin-producing strains, 55 per cent belonged to group II. Of the total number of cultures examined, 17 per cent produced a lessened amount of "A" polysaccharide, but of leucocidin-producing strains, 66 per cent produced a lessened amount of "A" polysaccharide. It is hardly safe to generalize from so small a number of cultures, but so far as this series is concerned, it appears that leucocidin producers frequently fall into the Cowan II group, and that leucocidin-producing strains show a tendency to produce less Julianelle "A" polysaccharide than other strains. It is not the intention to suggest that these strains are entirely devoid of "A" polysaccharide. If a more concentrated extract were prepared, perhaps all of them would show a strongly positive reaction. However, when carbohydrates were prepared from all the strains studied, employing a uniform volume of broth and a uniform volume of final product, the leucocidin-positive strains referred to, TABLE 4 Slide-agglutination with Cowan and 208 Cowan grouping GROUP I SERUM GROUP sera SERUM GROUlP III SERUM 235 sum (Sil) (S80) (S33) S1l S S _ appeared to be weaker as "A" carbohydrate producers than the majority of strains in the series. No agglutinative group seemed to be associated with any particular type of infection, but on the contrary, strains obtained from different lesions were distributed indiscriminately. Alpha-hemolysin producers occurred among members of all the groups. Strain 235 produces a very weak alpha-hemolysin and no leucocidin, but is unusually pathogenic for mice. It gives an entirely negative reaction to sera of the three Cowan groups, but is strongly positive to its homologous serum. Table 4 shows the relationship between the Cowan strains Sll (group I), S80 (group II) and S33 (group III) and strains related to 235. It will be noted that certain strains show a strong affinity for the serum prepared with strain 235 and are non-reactive toward the Cowan sera. The Cowan group strains, on the other hand, show but slight reactivity toward the 235 serum. Agglutinin absorption tests were run using the Cowan and 235 sera and cultures, and the absorbed sera were tested against the strains shown in table 4.

5 CLASSIFICATION OF STAPHYLOCOCCUS AUREUS Each serum was absorbed with a suspension containing 200 billion organisms, per ml. The dilutions of the absorbed sera were made to one tenth of the maximum titer which could be obtained with homologous organisms. TABLE 5 Agglutinin ab8orption8 of group sera ANTIGEN Sll S80 S CONTROL, UNABSORBED SERUM Absorption of group I serum Sil 1:5 1:5 1:5 1: :5 1:500 S80 1:25 S33 1:5 1: Absorption of group II serum S80 1:5 1: :5 1:25 1:5 1:50 Sil 1: :5 1:50 S33 1: Absorption of group III serum S33 1:50 S1l 1:5 1: :5 1:50 SW 1:5 1: Absorption of 235 serum 235 1:25 1:25 1:25 1: :100 1: :5 1:50 1:5 1: :50 1:50 Sl. S80 S Results are shown in table 5. It appears from table 5 that strains 235 and 207 are nearly identical in antigenic consitution and but slightly related to the Cowan groups, while strain 114

6 594 contains some group II antigenic substance. However, strain S80 (group II) is only partially able to absorb agglutinins for 114 from the sera. Strain 198 reacts as a group I or group III strain that contains some 235 antigenic substance. Strain 198 is the Gamma-hemolysin strain of Llewellyn Smith (1938) and, like 235, is unusually pathogenic for mice. S33, when used in anounts of 200 billion organisms per milliliter, absorbs out nearly all agglutinins from the sera, so far as is shown by slide-agglutinative tests. Cowan (1939) warns of this tendency of S33 to become non-specific, and advises frequent passages through rabbits to maintain its specificity. Mice were employed instead of rabbits for this study. The S33 culture was passed through mice frequently and the cultures for use were inoculated from the blood, or in some cases the blood cultures were stored in vacuo until used. In spite of this treatment S33 showed considerable nonspecific reactivity. SUMMARY The agglutinative and polysaccharide-forming qualities of a collection of Staphylococcus aureus strains were studied. Most of the strains belonged to the Cowan groups I and III and produced Julianelle "A" polysaccharide. Leucocidin-positive strains belonged mainly to group II. Most leucocidin-positive strains showed a tendency to produce less "A" polysaccharide than other strains. A supplementary agglutinative group is described in addition to the three Cowan groups. REFERENCES BLAIR, J. E., Aim HALLMAN, F. A Serologic grouping of Staphylococci. J. Bact., 31, 81. CHRISTE, R., xd KEOUGH, E. V Physiological and serological characteristics of staphylococci of human origin. J. Path. Bact., 51, CowAN, S. T The classification of staphylococci by precipitation and biological reactions. J. Path. Bact., 40, CowAN, S. T Classification of staphylococci by slide agglutinations. J. Path. Bact., 48, JULIANELLE, L. A., AND WIEGHARD, C. W The immunological specificity of staphylococci. J. Exptl. Med., 62, NEISSER, M., AND WECHSBERG, F tjber das Staphylotoxin. Z. Hyg. Infektionskrankh., 36, PANTON, P. N., AND VALENTINE, F. C Staphylococcal Toxin. Lancet, 222, PAuLI, R. H., AND COBtJN, A. F Studies on the serological typing of streptococcus hemolyticus. J. Exptl. Med., 65, PROOM, H The interrelationships of staphylococcal leucocidins. J. Path. Bact., 44, SMIT, M. L., AND PRICE, S. A Staphylococcus gamma hemolysin. J. Path. Bact., 47, THOMPSON, R., AND KHORAZO, D Correlated antigenic and biochemical properties of staphylococci. J. Bact., 34, VALEiNTNE, F. C Further observations on the role of the toxin in staphylococcal infection. Lancet, 230,

7 ICLASSIFICATION OP STAPHYLOCOCCUS AUREUS 595 VA ENE, F. C. O., ANi BUTrER, E. C. B Specific immmunity in acute staphylococal osteomyelitis. Lancet, 236, VAN DER VELDE, H Etude sur le mecanisme de la virulence du staphylocoque pyogene. La Cellule, 10, 401. W=IGHT, J Staphylococcal leucocidin (Neisser-Wechsberg type) and antileucocidin. Lancet, 230, YoNmEuRA, N tuber die immunisatorische Einteilung von Staphylococcus pyogenes. Z. Immunit&ts., 89,