Further evidence has been obtained which supports the view that alpha toxin from Staphylococcus aureus "Wood 46" is a protease which requires activation by erythrocyte membrane proteases. Rabbit erythrocyte antiprotease prepared in mice inhibited degradation of rabbit ghosts by the toxin. Supernatant fluid of toxin-treated ghosts incubated with EDTA and then passed through a column of Sephadex G-75 yielded a fraction which was hemolytic and proteolytic. These activities were both neutralized by alpha antitoxin prepared in rabbits, but not by control sera.It was also observed that alpha antitoxin inhibited the proteolytic activity of rabbit ghosts not exposed to toxin, in contrast with control sera. Inhibitory ability was removed from the antitoxin by adsorption with a heavy suspension of ghosts, and this treatment destroyed the antitoxin's capacity to neutralize hemolytic activity of the alpha toxin.

The mode of action of highly punned beta hemolysin derived from the R-1 strain of Staphylococcus aureus has been investigated. Sphingomyelin was absent from lipid extracts of sheep erythrocytes treated with beta hemolysin when compared to normal cells. A correlation was also established between the sphingomyelin content of other erythrocyte species and their sensitivity to beta hemolysin. Further investigations revealed that sphingomyelin is hydrolyzed to yield phosphorylcholine and N-acyl sphingosine. Thus, the mode of action of the beta hemolysin is like that of phospholipase C. Various phosphate compounds other than sphingomyelin, including RNA, glycerophosphate, phenyl-phosphate, phosphatidylethanolamine, and phosphatidylcholine were tested as substrates, but virtually no hydrolysis was observed. In contrast with the results of other workers, R-1 beta hemolysin did not release detectable amounts of carbohydrate from staphylococcal cell walls.

Analysis of glycose and fatty acid content of lipopolysaccharide extracted from 38 strains of Neisseria gonorrhoeae indicated that glycoses common to colonial types 1 to 5 were glucose, mannose, and galactose. N-acetylneuraminic acid, 2keto-3-deoxyoctulosonic acid (KDO), glucosamine, and galactosamine were also invariably present. Virulent colonial types 1 and 2 contained no rhamnose, in contrast to avirulent types 3 to 5 and several strains ofthe nonpathogenic species N. sicca and N. lactamica. Fucose, characteristic of these nonpathogenic species, was not present in the gonococci. Variation in the concentration of individual glycoses in different strains was also noted. Mannose-KDO, galactose-KDO, and glucose-KDO ratios of virulent gonococci exceeded those of avirulent organisms, except that the correlation for glucose was not quite so striking. This relationship was not found in N. sicca and N. lactamica strains. Fatty acid analyses of lipid A from gonococci showed that 10-, 12-, 14-, 16-, and 18-carbon acids, as well as 3-hydroxytetradecanoic acid, were present, but differences in concentration between colonial types, although evident in some cases, appeared less significant than glycose content.

An improved procedure for the purification of the delta toxin of Staphylococcus aureus strain E-delta has been devised which relies upon precipitation at pH 4.0 and further treatment with ammonium sulphate. A final step consists of passage twice through a column of DEAE-cellulose. Toxin obtained by this method appeared to be homogeneous on the basis of immunodiffusion and electrophoresis studies. However, two peaks with sedimentation coefficients of 2.8 S and 9.8 S were obtained when toxin was examined in the ultracentrifuge. Proline was identified as the N-terminal amino acid. No other N-terminal amino acids were detected in the purified toxin.

The alpha, beta, and delta hemolysins of Staphylococcus aureus are implicated in the pathogenicity of this microorganism for man and animals. However, only the mode of action of the beta hemolysin is known with certainty. This enzyme hydrolyzes sphingomyelin, a common constituent of cellular membranes, resulting in the liberation of water-soluble phosphorylcholine and waterinsoluble K-acyl sphingosine ( 1,2). In the present experiments, we have shown that two highly purified preparations of the delta hemolysin liberate organic phosphorus from phospholipid extracts of mammalian erythrocytes. Of several phospholipids investigated as substrates, phosphatidylinositol was the most susceptible to degradation by the delta hemolysin.Materials and Methods. Crude delta hemolysin from the Newman and E-delta strains of S. aureusl was prepared in stainless steel trays of Dolman-Wilson agar (3) overlain with sterile cellophane and covered with aluminum foil. The cellophane surface was inoculated with 2 ml of a saline suspension of a 24-hour agar slope culture of the organisms, which was spread with a glass rod. After 24-hours incubation at 37OC in a sealed plexiglass tank in an atmosphere of 25% C 0 2 in air, 0.01 A1 phosphate buffered saline at pH 7.0 was added to the trays. The growth was taken up in the buffer, pooled, and centrifuged at 11,OOOg for 30 min. The supernatant fluid containing the hemolysin was stored at -20°C.Crude delta hemolysin prepared in this manner was dialyzed against 0.05 M acetate buffer a t pH 4.0 for 48 hours. The resulting ~~ * Supported by the Medical Research Council.precipitate which contains the active material was collected by centrifugation and dissolved in 0.05 M tris (hydroxymethyl) aminomethane (Tris) buffer a t pH 9.0 to effect solution. Further dialysis of the preparation was then carried out against 0.1 M phosphate buffer a t pH 7.0 for 48 hours. I t was centrifuged and the supernatant fluid was added to hydroxylapatite (4) in the amount of 10 m1/0.2 gm of dry weight of adsorbent. Elution of activity was accomplished after 1-hour standing a t 4°C by addition of 2 M NaCl in buffer to the adsorbent, followed by centrifugat ion.Further purification of the delta hemolysin was achieved on a column of diethylaminoethyl (DEAE) cellulose equilibrated with 0.02 M phosphate buffer a t p H 7.0. The hemolysin was eluted from the column against a linear gradient of NaCl (W.5 M ) in phosphate buffer prepared in a Varigrad.2 About 40 mg of protein was applied to the column and 10 ml fractions were collected, the active material appearing in the first of several peaks eluted.A 30-fold increase in specific activity was achieved by this method when crude hemolysin was compared to that eluted from the DEAE column. One strong line of precipitation was observed in Ouchterlony plates when purified hemolysin was incubated with a rabbit-produced antiserum to crude material. However, Jackson and Little ( 5 ) and Gladstone and Yoshida (6) have observed that protein components of normal serum strongly inhibit hem...