Kelsey C. Milner

PHYSICAL ASPECTS OF REVERSIBLE INACTIVATION OF ENDOTOXIN

Most early investigations of the chemistry of endotoxins were limited to gross analyses. I t soon became apparent that subtler means of degradation, such as enzymes, were necessary to determine structural configurations and intramolecular relationships within the endotoxin complex. Just such a subtle endotoxin-degrading system was suggested by the observation of Hegemann (1954) that human plasma factors inactivated the pyrogenic properties of endotoxin. The applicability of this reaction became particularly evident when further studies showed that smaller-sized antigenic units of endotoxin appeared to be produced during the plasma-endotoxin reaction (Rudbach and Johnson, 1961, 1962; Skarnes and Chedid, 1964). The kinetics of the endotoxin-altering reaction, measured by reduction in the pyrogenicity of endotoxin (Keene, Landy and Shear, 1961) and by reduction in ability of endotoxin to precipitate quantitatively with homologous antiserum (Stauch and Johnson, 1959), lent credence to the theory that a plasma enzyme was degrading the endotoxin polymer. This endotoxin-altering system was shown to be inhibited by divalent cations (Rosen et al., 1958) and by other plasma components (Yoshioka and Johnson, 1962). These and other observations (cf. Atkins, f960; Landy, 1960) supported the contention that a plasma enzyme was degrading the endotoxin polymer. However, proof of this must be the isolation of degraded products which have come from a previously intact endotoxin molecule. This evidence has not been presented to date. Furthermore, other evidence has indicated that endotoxin may not be cleaved by plasma hydrolases. Rudbach and Johnson (1964) demonstrated that, after alteration of endotoxin by plasma, activity could be restored by procedures not considered likely to reverse an enzymatic hydrolysis.

IMMUNOTHERAPY WITH NONVIABLE MICROBIAL COMPONENTS

Structural components of microorganisms have been studied for immunopotentiating effect with the aid of transplantable (line 10) tumors in syngeneic guinea pigs. Microbial components were associated with oil droplets, suspended in Tween-saline, and injected intralesionally. BCG cell walls, given in this way, produced regression and cure of 50-60% of established tumors, as did viable BCG. Lipid extraction markedly reduced the tumor-regressing potency of cell walls, but P3, a trehalose mycolate present in the extract, restored full activity to the cell wall residue. P3 alone was nonsensitizing and had no antitumor activity, but it enhanced the latter property of various other microbial products. For example, the cure rates produced by cell walls of M. tuberculosis, M. bovis, M. phlei, or M. smegmatis were enhanced from 20-60% to as much as 90% by addition of P3. P3 also conferred antitumor activity on products from unrelated microbes, such as cell walls of E. coli, and in combination with endotoxins from rough Re mutant salmonellae, it produced cure rates of up to 93%. These results suggest that P3 is essential to the immunopotentiating activity of mycobacteria and that it may be broadly applicable in immunotherapy of cancer with microbial agents.

Tumor regression with Q fever rickettsiae and a mycobacterial glycolipid

SummaryRegression of established tumors by Coxiella burneti, the rickettsial agent which causes Q fever, was studied using the transplantable line-10 tumor in strain 2 guinea pigs. Suspensions of formalin-killed, purified rickettsiae induced an average of 42% tumor regression after intratumor injection. The activity of C. burneti was enhanced by incorporation of the rickettsiae into oil droplet vaccines containing the mycobacterial glycolipid, P3. Such vaccines induced 64% tumor regression. The activity of C. burneti that was enhanced by P3 was found in phenol-water extracts of the rickettsiae. These extracts contained lipopolysaccharides which were less toxic than bacterial endotoxins, and they induced 63% tumor regression when combined with P3. These lipopolysaccharides may provide agents of high therapeutic activity but relatively low toxicity for cancer immunotherapy.

Further studies on the structural requirements of agents for immunotherapy of the guinea pig line-10 tumor

SummaryIn this laboratory and elsewhere, cord factor or some less complex analogue has been found to be an important constituent of agents for suppression and immunotherapy of cancer. In further attempts to delineate structural requirements, we have tested several such analogues, in combination with the optimum quantity (150 μg) of glycolipid from an Re mutant salmonella (Re glycolipid), for ability to produce regression of transplantable one-week-old line-10 tumors in guinea pigs. The synthetic diester, trehalose 6,6′-dipalmitate, has been reported to be a useful antibacterial prophylactic and a tumor-suppressive agent, but neither alone nor in combination with Re glycolipid was it effective in therapy of established line-10 tumors, even in doses up to 1500 μg. Trehalose myocolates (cord factor, P3) were also ineffective when given alone, but as little as 15 μg of P3, combined with Re glycolipid and oil droplets, produced a high rate of regression. Some analogues of higher molecular weight were effective but, within the limits of these experiments, only when the fatty acid residues contained side chains at the alpha carbon atom. It was striking that a naturally-occurring 6,6′-trehalose diester of a branched chain fatty acid(s) containing the carbon equivalent of two condensed palmitic acid residues was as effective as any of the higher-molecular weight compounds, including the mycolates. Thus, it appears that there may be requirements for a certain molecular size and/or for a particular molecular conformation. Only such a compound, in conjunction with Re glycolipid or other suitable immunogen, has been found to bring about complete cures, including regression of the primary tumor, elimination of metastases in the regional lymph node, and specific systemic immunity to rechallenge with the line-10 tumor.

Regression of line-10 hepatocellular carcinomas following treatment with water-soluble, microbial extracts combined with trehalose or arabinose mycolates

SummaryIntratumor injections of the aqueous phase of phenol-water extracts of Re mutant Salmonella typhimurium (Re glycolipid) in combination with trehalose dimycolate at dose levels of 150 to 15 μg were consistently and highly effective (65–93%) in producing regression of line-10 tumors in strain-2 guinea pigs. We observed that the rate of regression was more rapid than that seen after treatment with cell walls from Mycobacterium bovis strain Bacillus of Calmette and Guèrin (BCG). Arabinose mycolate could be substituted for trehalose dimycolate in the Re glycolipid-mycolate mixture without appreciably compromising antitumor activity, providing that the level of arabinose mycolate was not reduced below 15 μg. In addition to the Re glycolipid preparation, similarly prepared aqueous extracts from Mycobacterium bovis strain BCG and strain AN5 in combination with trehalose dimycolate also possessed tumor-regressive activity. The activity of these last extracts was reduced when the arabinose mycolate was substituted for the trehalose dimycolate. The aqueous extract of a rickettsia, Coxiella burnetii, in combination with either trehalose dimycolate or arabinose mycolate was also active (50 and 80% tumor regression rates, respectively). Intracutaneous administration of Re glycolipid or aqueous extracts from BCG in combination with trehalose or arabinose mycolates did not produce life-threatening, clinical signs of toxicity in young mice. If additional toxicity studies demonstrate that adverse side effects can be satisfactorily controlled, these watersoluble extracts may prove beneficial in the treatment of spontaneous tumors of humans and other animals.

STRUCTURE AND BIOLOGICAL FUNCTIONS OF MYCOBACTERIA

The relationship between cell structure and function is obviously too large a subject to be covered in a short presentation. However, certain fundamental concepts can be considered by limiting ourselves to only one function of mycobacterial cells, that of their ability to stimulate an immune response in a suitable animal host. To investigate the relationship between acquired immunity in the host and the part of the whole bacterial cell which incites this immunity, it is necessary to accomplish two things: first, devise a suitable test system for immunity and, second, isolate the specific component of the mycobacterial cell which incites the immunity. To begin with, we must define what we mean by an immune response and, if possible, describe i t in a quantitative fashion. In our laboratory, we place greatest reliance upon a test system which involves challenge of mice by aerosol with small numbers of virulent tubercle bacilli and modification of the effects of such challenge by prior i.v. injection of experimental vaccines.13 Assessment of the response to vaccination includes: 1. The count of grossly visible lung lesions four weeks after challenge. 2. Enumeration of the bacilli in the target organ, the lung, at prescribed inter-

Preparation and properties of an aluminum citrate-endotoxin complex from Salmonella enteritidis.

Summary Treatment of conventional aqueous ether extracted endotoxins from Salmonella enteritidis with aqueous phenol, followed by high speed centrifugation, yielded material having high biological potency with a nitrogen content of 0.4–0.5% and a fatty acid content of 3–4%. Upon reacting this purified endotoxin with lithium aluminum hydride and citric acid solution, an aluminum citrate-endotoxin complex was recovered which had the desirable features of high biological potency and a rapid rate of solution in water. Non-phenolized endotoxins extracted from S. enteritidis with aqueous ether or trichloroacetic acid also formed aluminum citrate complexes under the same conditions. These were biologically potent but somewhat more difficult to dissolve. After prolonged treatment with lithium aluminum hydride, the aluminum citrate complexes still contained 1–2% esterified fatty acids. Lipid fractions isolated from acid hydrolysates of the complexes were biologically inactive.

REACTIVATION OF PAPAIN-TREATED ENDOTOXIN.

Summary The pyrogenicity of endotoxin which had been diminished by incubation with papain was restored by treatment of the inactive complex with pronase and ethanol, indicating that a hydrolytic cleavage of the endotoxin molecule had not occurred. In addition, it was shown that the reduction of the second fever peak in papain-inactivated endotoxin could be explained on the basis of a normal dose-response phenomenon. These findings suggest that there is no disagreement between the inactivation of a purified endotoxin by papain and the observed correlation of the biological activity of endotoxin with the polysaccharide portion of the molecule.

Specificity of resistance to tuberculosis and to salmonellosis stimulated in mice by oil-treated cell walls.

Summary Coating with oil, which was essential to render cell walls of BOG protective to mice against challenge with tubercle bacilli by aerosol, has been found not to affect the specificity of reactions conditioned by cell walls in this and other systems. When the interval between vaccination and challenge was short (24 hours), mice were protected against heavy intravenous challenge with M. tuberculosis principally by nonspecific factors; after a 30-day interval, protection appeared to depend upon a mixture of specific and nonspecific effects.

Specific and nonspecific stimulation of resistance in mice against infection with mycobacterium tuberculosis H37Rv

Intravenous vaccination of mice with oil-treated cell walls of BCG orM. tuberculosis, H37Rv protected against aerosol challenge with virulent tubercle bacilli, whereas protoplasm of these strains or cell walls in aqueous suspension failed to elicit resistance.The aerosol challenge system has been shown to be the only test system for resistance against tuberculosis with a remarkable specificity, since cell walls of unrelated species likeS. typhimurium orL. monocytogenes were non-protective in this test, in contrast to the conventional “massive intravenous challenge test” where specific and non-specific factors can become effective.