PATHOGENICITY AND TOXIN PRODUCTION 803 



from ground-up bacilli. It was lethal on intravenous injection into guinea-pigs, 

 and produced areas of necrosis on intradermal injection in the rabbit. Evans and 

 Maitland's ^^reparations also contained the agglutinogen ; this proved to be distinct 

 from the toxin, since antisera to the extract were protective against experimental 

 infection, agglutinated bacillary suspensions, but had no antitoxic activity as 

 judged by their ability to modify skin necrosis induced by the toxin. The toxin 

 was easily destroyed by formalin, was unstable at 37° C, and was rapidly destroyed 

 at bb° C. A similar toxin with very similar properties was obtained from H. bronchi- 

 septicus (Evans and Maitland 1939) and from H. parapertussis (Brueckner and 

 Evans 1939). The toxin, presumably owing to its marked instability, did not 

 at first appear to be antigenic. Evans (1940, 1942) later found that formolized 

 toxin was antigenic, and that antitoxin prepared against it neutralized the toxins 

 of H. pertussis, H. parapertussis and H. bronchisepticus. It is of practical interest 

 that he was unable to induce antitoxin formation in the rabbit by the injection 

 of whole bacilli, but found that toxic extracts of the bacilli were antitoxinogenic 

 (see also Katsampes, Brooks and Bradford 1942). These observations have been 

 confirmed and extended (Flosdorf, Bondi and Dozois 1941, Ehrich, Bondi, Mudd 

 and Flosdorf 1942, Eldering 1941, 1942). Flosdorf and his colleagues distinguish 

 in H. pertussis a feebly antigenic thermostable toxin in addition to the thermo- 

 labile toxin of Evans and Maitland. Toxin is produced by all phases of H. pertussis, 

 though most abundantly by Phase I (see also Wood 1940, Koberts and Ospeck 

 1942). Eldering obtained toxic fractious from H. pertussis, H. parapertussis and 

 H. bronchisepticus, and demonstrated varying degrees of cross-protection against 

 infection by living bacteria, in animals actively immunized by the fractions. It 

 is probable that her results reflect the antigenic similarity of both toxins and 

 agglutinogens present in the extracts. 



The mode of action of the toxin is at present obscure. Ehrich and his colleagues 

 (1942) describe generahzed degenerative changes in the viscera of intoxicated rabbits, 

 particularly in the lymphoid tissue. Given intratracheally, the heat-labile toxin produces 

 in the lungs of rabbits a severe cedematous reaction, followed by a characteristic accumu- 

 lation of macrophages in the alveoli, of lymphocytes round the blood vessels, and severe 

 necrosis in scattered areas — a histological picture not unhke that found in the lung in 

 whooping cough ; antitoxin protected rabbits against this eflFect (Sprunt and Martin 

 1943). Though the precise role of toxin and antitoxin in infection with H. pertussis 

 is not yet fully understood, the protective action of antitoxin in experimental infection 

 appears to be hmited to neutraUzing the toxin contained in the infecting dose, and thus 

 reducing the likehhood of the organisms establishing a foothold in the tissues. Thus, 

 Anderson and North (1943) protected mice against an intraperitoneal injection of H. per- 

 tussis with antitoxin, but not with antibacterial sera. SystemicaUy administered, anti- 

 toxin had no effect in animals infected by the nasal route, but antibacterial serum was 

 effective. In their hands toxin had no aggressive action (see Chapter 48) in nasal infections. 

 It should be noted, however, that North and his colleagues (1939) found that in intra- 

 nasaUy infected mice the protective effect of human sera from pertussis convalescents 

 could not be accounted for simply in terms of agglutinin or toxin-neutraUzing power, 

 though the effect as a whole appeared to be antibacterial. The labile toxm, however, 

 may have immunizing properties, for in a later study (1941) of active immunization they 

 found that whole bacilh treated with the minimum of heat (58° C. for 8 min.) or with 

 phenohc preservatives, induced a higher degree of active immunity than baciUi preserved 

 with formahn or heated to 60° C. for one hour. Evans (1944) confirmed the inability 

 of intravenous antitoxin to protect mice against intranasal infections, but found that 

 antitoxin mixed with the bacteria before instillation into the nose lowered their infectivity, 



