PATHOGENICITY 647 



found that by direct agglutination the swarming strains could be divided into 

 3 main groups, and that by absorption these main groups could be divided into a 

 number of sub-groups. The non-swarming strains differed from the swarming 

 strains in their absence of an H antigen, and differed among themselves in the type 

 of their antigen. The relation of the vulgaris to the X strains is still somewhat 

 doubtful. The antigens of the X strains differ from those of the vulgaris strains, 

 but there appears to be a certain group relationship between the H antigens (Yacob 

 1932). Among the antigens of the X strains there are at present three fairly 

 well-defined groups, represented by the strains OX 2, OX 19, and OX K (see Chapter 

 83). White (1933) has brought evidence to show that the antigen of OX 19 

 contains two receptors, one of which is alkali-labile and is mainly responsible for 

 agglutination of this organism by an antiserum prepared against it, the other of 

 which is alkali-stable, and is responsible for the reaction of the bacillus with the 

 sera of patients suffering from typhus fever — -the Weil-Felix reaction. Meisel and 

 Mikulaszek (1933) and Castaneda (1934, 1935) have reported the extraction of 

 soluble specific polysaccharide substances from Proteus X strains. Castaneda's 

 results agree closely with those obtained by White. They show that the alkali- 

 stable polysaccharide, referred to as X, is common to both Proteus Z^ 19 and 

 Rickettsia prowazeki (see Chapter 39), while the alkali-labile polysaccharide, referred 

 to as P, is specific to Proteus X 19. 



Our own observations on a limited number of strains suggest that there is a wide 

 variety of H and antigens in Proteus vulgaris, the H antigens being distributed 

 to some extent independently of the antigens. 



With regard to Proteus morgani, Rauss states that the H antigen tends to be 

 group-specific, and the antigen type-specific. In an examination of 48 strains, 

 7 types of H antigen were differentiated, but as many as 17 types of antigen. 

 A relationship was found between the H receptor of one group of morgani and a 

 strain of Proteus vulgaris. 



Pathogenicity. — Proteus bacilli are frequently found in, and appear to be respon- 

 sible for, a number of inflammatory and suppurative conditions in man. They 

 are a very common cause of cystitis and may be isolated in pure culture from the 

 urine of infected patients. They are not uncommonly found in abscesses, either 

 alone or in combination with other organisms. Metchnikoff and his co-workers (see 

 Chapter 71) found them almost constantly in the faeces of infants with summer 

 diarrhoea. They are frequently present, usually as secondary invaders, in wounds 

 and burns, where they probably favour the development of the pathogenic anae- 

 robes. And they have been isolated from a variety of conditions, such as volvulus, 

 peritonitis, croupous pneumonia, acute gastro-enteritis of the food-poisoning type 

 (Wichels and Earner 1925, Plahn 1937, Cooper et al. 1941), empyema, gangrene of 

 the lung, and septicaemia. Jensen (1913) considers them responsible for one form 

 of epidemic calf dysentery ; and Wyss (1898) has encountered them in an epidemic 

 disease of fish in Lake Ziirich. They are also responsible for the black rot of eggs. 



An organism described as P. hi/drophilus, has been held responsible for the Red leg 

 disease of frogs, but as this organism is monotrichate, does not swarm on agar, and does 

 not decompose urea, it is very doubtful whether it should be included in the Proteus group 

 (see Kulp and Borden 1942). 



The relation of Proteus Z 19 to typhus fever is discussed in Chapter 83. 



Their pathogenicity to laboratory animals is variable ; virulent strains on 



introduction into the tissues are able to proliferate and invade the blood stream. 



