New knowledge in the chemistry of antimetabolites.

Exotoxin from B. thuringiensis was purified by chromatography and its structure was determined by means of chemical and physical methods. The formula of the substance as given above indicates that the exotoxin possesses between ribose and glucose an unusual ether bond which up to now has not been found in Nature. The study of the mechanism of action of the exotoxin shows that it competes with ATP in the RNA polymerase reaction. The anomalous structure of the exotoxin has been confirmed by the synthesis of its substantial moiety, i.e. of glucosyladenosine. In the Institute of Organic Chemistry and Biochemistry, Czechoslovak Academy of Sciences, Prague, we have been interested for a long time in antimetabolites of nucleic acids as potential cancerostatics and virostatics. Most attention has been paid to synthetic analogues of bases and nucleosides. Of these analogues, 6-azauridine is now being used clinically. Recently it was suggested on the basis of chemical speculations another aza-analogue of pyrimidine nucleosides, namely 5-azacytidine, might be of value and it was therefore prepared synthetically in this Institute'. We never expected that 5azacytidine might occur in Nature. Surprisingly enough, this substance was identified as the cancerostatically active antibiotic and naturally occurring antimetabolite of Streptoverticillium ladacanus2. This finding drew our attention to substances which interfere with the biosynthesis of nucleic acids, and to investigations on the exotoxin of Bacillus thuringiensis which is known to exhibit insecticidal activity. It has been found in the biochemical laboratories of our Institute that this exotoxin is a specific inhibitor of DNA-dependent ribonucleic acid polymerase and that it competes specifically with adenosine triphosphate3. The structure of exotoxin was established in this Institute4. In this lecture, I should like to present the complex data relating to the proposal of the unusual structure of exotoxin as well as to the synthesis of a portion of its molecule. Nothing was known about the chemical structure of exotoxin until recently. Its heat stability indicated a low molecular weight. The first not quite pure preparations of exotoxin were obtained by de Barjac and Dedonder5 of the Pasteur Institute in Paris. These workers analysed it chemically and found it to 253 F. ORM contain adenine, ribose and phosphorus in equimolar ratios. This and the acid character of the exotoxin suggested a nucleotide of anomalous composition. In Czechoslovakia, research on B. thuringiensis was for a long time in the hands of entomologists, in particular Dr Va6kovâ who developed from it an insecticidal preparation. In cooperation with her we set out in 1966 to isolate and identify the exotoxin of B. thuringiensis. The credit here goes mainly to Dr ebesta (a biochemist) and to Dr Farka (an organic chemist). For producing the exotoxin on a large scale we used a variety of B. thuringiensis, the so-called B. gelechiae, which produces no endotoxin. We know now that the exotoxin of this variety is identical with the substance used by the French workers. In the first part of the work we developed a procedure for the preparation of about 10 g amounts of exotoxin. We grew Bacillus gelechiae in a simple synthetic medium and, after filtration, adsorbed the exotoxin on charcoal. After desorption of the exotoxin with ethanol, and extraction of ballast substances into phenol, the principal part of the purification procedure was chromatography on Dowex 1. Gradient elution with ammonium formate yielded a uniform fraction absorbing in the u.v. region (Figure 1). After sublimation of ammonium formate, thç purified toxin appeared as a yellowish amorphous powder. The product was fully active biologically as was shown by astandardtest on the caterpillars of Galleria mellonella. The compound was readily soluble in water and displayed a spectrum characteristic of adenine nucleotides. The i.r. spectrum showed a particularly characteristic absorption band 1691 cm', corresponding to a carboxyl. From the content of nitrogen (8.7 per cent) and phosphorus (4.2 per cent) we concluded that the molecular weight of the compound would be about 750. The electrophoretic behaviour of the exotoxin showed that we were dealing with an acidic compound. Its titration curve was determined (Figure 2), from which it followed that the exotoxin molecule contains four dissociable protons; three of these with a pK of 3.8 apparently belong to two carboxyl groups and to 254 100 1.0 n Figure 1. Chromatography of exotoxin on Dowex 1 x 2 (formate) NEW KNOWLEDGE IN THE CHEMISTRY OF ANTIMETABOLITES the more acid hydrogen atom of an ester-linked phosphoric acid, the fourth with a pK of 6.6 to the less dissociable hydrogen atom of the phosphate residue. Naturally we tried first to split off the ester-linked phosphoric acid of the exotoxin which was readily achieved by applying calf intestinal alkaline phosphatase. The product was again purified by gradient elution on a column of Dowex 1. The fraction absorbing u.v. light was freeze-dried and a white amorphous, biologically completely inactive compound was obtained. The i.r. spectrum showed a principal band belonging to a carboxyl group. From the content of nitrogen, assuming that all nitrogen was present in the adenine moiety (10.4 per cent), we concluded that the dephosphorylated exotoxin has a molecular weight of about 650. In agreement with this assumption the titration curve of dephosphorylated exotoxin corresponds to the presence of only two carboxyl groups.