It has been reported (1) that a petroleum ether-soluble material can be extracted from virulent tubercle bacilli. Since it was obtained only from "cordforming" organisms (2) it was called "cord factor." Mter being subjected to extraction with petroleum ether, the bacteria were still able to grow normally in suitable culture media. This fact suggested that the cord factor was located at the surface of the bacterial cell and that its removal did not disturb any essential structures of the cell. On the other hand, the importance of the bacterial surface in the establishment of an infection was emphasized by the fact that the isolated cord factor exerts some of the effects of intact virulent organisms, while the behavior of virulent bacteria after removal of the cord factor was in some respects more characteristic of avirulent than of virulent organisms. It is noteworthy that young cultures yielded considerably more cord factor than older ones.These findings were consistent with the previous work of Middlebrook, Dubos, and Pierce (2) calling attention to the cord formation by virulent tubercle bacilli and correlating this phenomenon with properties of the bacterial surface concerned in the virulence of tubercle bacilli. Further evidence for differences in the surface structure of virulent and non-virulent mycobacteria was brought out in another series of experiments (3). Virulent organisms were shown to be considerably less permeable to the dye methylene blue than noncord-forming avirulent bacteria. Mter extraction with petroleum ether, however, they became equally permeable. Apparently the petroleum ether extracts a surface material which is responsible for the low permeability of virulent strains of tubercle bacilli to even small molecules like methylene blue. Moreover, when cord-forming cultures were allowed to age in vitro, the ceils became increasingly permeable to methylene blue; the greater the virulence of a strain, the longer its organisms remained relatively impermeable to the dye.This observation parallels the small yield of cord factor obtained from older cultures, on one hand, and the fact that avirulent bacteria which did not contain any measurable amount of cord factor were completely permeable to methylene blue, on the other.Our earlier observations had already suggested that the cord factor was in some way concerned in the virulence of tubercle bacilli. If this was true, young cultures possessing more cord factor, should also be more virulent for susceptible animals than older cultures of the same strain.The experiments reported in the present paper seem to bear out this working hypothesis.
Materials and MethodsBacteria.--Three strains of tubercle bacilli were used in these experiments: The human strain H37Rv, obtained from the collection of Dr. R..L Dubos; a human strain Jamaica No. 22, received from Dr. J. Freund; and a bovine strain Vall~e, originally obtained from the Pasteur Institute in Paris and propagated on egg media for at least 20 years at the Institute for Hygiene of the Universtiy of Ba...