No abstract

A study was made of the origin, occurrence, and properties of natural antibodies to Gram-negative bacteria in the normal serum of several species. Antibody was measured by a procedure based on the bactericidal reaction carried out under conditions in which activity was a function of the amount of antibody contributed by the test serum. Antibody to seven genera of the family Enterobacteriaceae were demonstrated in normal human serum. The specificity of these antibodies was affirmed by absorption with homologous bacteria and by inhibition of bactericidal activity with purified homologous somatic antigen. Absorption with graded amounts of bacterial suspensions showed that a large excess of bacteria led to non-specific removal of antibodies. Analogous findings were also made with immune antibody. As determined by quantitative absorption tests no difference could be found in the avidity of natural and immune antibody. Natural antibody in the serum of various species differed considerably as regards their lability to heat, but in parallel tests immune antibody of each species was significantly more heat-stable. The time and appearance of the antibodies varied in young animals, of different species. Mice developed these antibodies at the earliest age, with guinea pigs, rats, and rabbits following in that order. Serum from germ-free rats and chickens had no demonstrable antibodies to E. coli or S. typhosa whereas these antibodies were present in the serum of litter mates reared under conventional conditions. On the other hand germ-free and conventional mice did not differ appreciably as regards the levels of these same antibodies.

The attention of immunologists has long been focused on acquired immunity produced by specific antigens. In the classical sense this type of immunity to infection, which generally is of a high order, depends upon the appearance of protective antibodies. The obvious importance of such artificial immunity, and the relative ease with which these antigens and antibodies could be studied, contributed to the comparative neglect of other components of bacterial cells which may also give rise to protective reactions. Nevertheless, it has been recognized that there is also a type of resistance which transcends the limits of antigenic specificity, and from time to time this kind of protection against infection was observed even though demonstrable antibodies could not be detected. In fact it was noted by a number of workers that the injection of various substances of bacterial and of other origin produced an increase in normal resistance. While the level of resistance thus achieved was lower than that which followed the injection of specific immunizing agents, it was sufficient to indicate that it might nonetheless contribute to immunity.These observations extend back to some of the early work in microbiology and were first brought together by Kolle and Prigge (1). The more recent literature has been assembled by Brandis (2) who also reported the rapid production of a protective effect against challenge with Salmonella derby by the injection into mice of a number of bacterial vaccines or unrelated products. He called this effect "proimmunity" and considered it to be a result of non-specific stimulation of the cellular defenses of the host. Rowley (3) showed that mice injected with cell walls of Esckerichla coil, Salmonella typhimurlum, or zyrnosan (4), within the first few hours were more susceptible than normal mice to challenge with E. coll.

Investigations of factors affecting the immunological state of the host constitute a voluminous literature, most of which has been reviewed in a comprehensive treatise by Perla and Marmorston (1). However, of the great array of substances investigated, only a few have proved to be of consequence in that they provided new and productive approaches to improvement in the antibody response of the host.The contributions of major importance in this respect are: (a) the observations of Ramon (2) and of Glenn}" et al. (3) on the augmentation of antibody response to antigens adsorbed on particulate carriers; (b) the studies of Freund (4--6) which established the enhancement and prolongation of antibody response to antigens administered in water-in-oil emulsions; and (c) the finding that certain bacterial vaccines, notably pertussis and typhoid, incorporated in combined prophylactics, exert a synergistic effect on the antigenieity of toxoids (7).Study of the factors responsible for the elevated antibody levels following combined immunization has been limited, and the mechanisms by which these vaccines augment antitoxin formation are unknown. The effect of bacterial components on antibody response to antigenic stimulus has received less attention. Staphylotoxin has been reported to raise the level of antibody developed in response to weakly antigenic substances such as ragweed pollen and crystalline beef lens, when administered combined with, or several hours prior to these antigens (8,9). This synergistic action of staphylotoxin has been used to advantage by Hecht et al. in producing specific antiskin antibodies in rabbits (10). In addition, the type A toxoid of Clostridium

Endotoxins of low lipid content prepared from S. enteritidis by the aqueous ether method have been further treated to remove bound lipid by non-hydrolytic procedures. Such endotoxins, containing as little as 2 per cent lipid A, were as potent in stimulating a variety of physiological responses as those prepared by the well known phenol-water or Boivin procedures which yield products containing as much as 30 per cent lipid A. To verify the difference in lipid content between the aqueous ether preparations and other types of endotoxins, three different methods of lipid analysis were employed: determination of chloroform-soluble material released by hydrolysis with hydrochloric acid (lipid A) or with acetic acid (lipid W), and estimation of total bound fatty acids. These methods were in accord in showing the magnitude of the difference. No more than one-half of the fatty acids present in endotoxin were associated with the fraction designated lipid A. Methods are described for the preparation of potent endotoxins with analytical values for nitrogen, phosphorus, hexosamine, carbohydrate, and fatty acid which do not differ appreciably from those of the classical, non-toxic, haptenic polysaccharides.