This brings this journey (or should it be called climb) across the top concepts to the third principal hypothesis of cancer metabolism which is well supported by a considerable number of experimental results. Its main protagonist is Potter 198 189 who on frequent occasions has promulgated, described and defended it. It is known as "deletion hypothesis" and postulates, when expressed in simplified terms, that during the development towards the malignant state and after, the cell has lost certain essential constituents, the presence of which, under normal conditions, exerts a vital influence on the regulation of growth and cell division. While the Millers,181 at least originally, gave emphasis to the possible removal of proteins for instance during carcinogenesis by azo dyes in the liver, and Weiler 256,257,258 pro. duced in a most elegant manner immunological evidence for the disappearance of proteinous antigens in the liver of the rat and the kidney of the hamster (both animals having been exposed to different kinds of carcinogenic agents), other investigators put forward different ideas as to the true nature of the deleted constituents (see, e.g., Farber87). In verbal or printed contributions to this subject Haddow 116117 sided with those who favoured the idea of key protein deficiencies, essential for the control of growth, in the precancerous and cancer cell, although he advocated that the deletion hypothesis should not be accepted too uncritically. Burnett60 in his deliberation on the biology of cancer has given preference to controls dependent on cell surface properties, and has taken hormonal and immunological factors into consideration. With regard to the latter he agreed in principle with H. N. Green 108 who formulated the hypothesis that the protein or proteins lost in the cancer cells were those with specific immunological functionst Whatever the case may be for or against these predilections, Potter,198 if one summarises his opinions, stressed the deletion of enzyme forming, systems (see Fig. 4; EFS in his abbreviation). These probably consist of RNA-proteins which in many instances are themselves under the control of nuclear DNA-genes, and placed between the latter and the enzymes and small molecules oi endogenous or exogenous origin, such as substrates, products, precursors of coenzymes, and which by feedback mechanisms could influence the whole character of the cell. Allowing for this feedback device which in these days of electronic computers and brains appears to be very appropriate, his whole scheme based on practical examples of biochemistry, to be mentioned further on, becomes very convincing. This is the more so, because not only do some of the features of the Warburg-Quastel etc. phenomena fit in with this concept but it also permits explanation of the deficiency of coenzymes, as these non proteinous compounds owe their existence to the action of specific enzymes. The deletion hypothesis of course does not exclude changes by loss in the cell surface system nor does it specifically reject events concerning immunobodies, the formation and interaction of which have always been likened to enzymic happenings. Even the present day problem of viruses as cause of some cancers, and naturally that of the diversity of carcinogenic agents can be accommodated, at least in a provisional fashion. On the other side, it is a little difficult to amalgamate all this with the Greenstein postulate of a quasi uniform tumor enzyme pattern which does not allow for marked disappearances of certain biocatalytic systems nor for significant increases in others which, according to Potter,198 198 are only a consequence of losses somewhere else in the overall processes of the cell. This interdependence of biocatalytic systems is also brought out by the findings of Hinshelwood131 who studied adaptation phenomena in bacteria, governed by the exponential law of auto catalysis.

Fig. 3. Xanthine oxidase during carcinogenesis in mouse breast. (From Bergel et al.: Ciba Foundation Symposium." Courtesy of J. and A. Churchill, Ltd., London.).

While it is rather difficult to pinpoint enzyme forming systems (EFS), as long as one has only scanty knowledge about the special structure of RNA and RNA-proteins responsible for this process, there are a considerable number of examples in the literature concerning enzymes, coenzymes and even metal activators which will support the deletion hypothesis at least with regard to biocatalysts. However any enthusiasm over new and convincing additions to this list of enzymic changes in malignant conditions has to be tempered by a sound portion of scepticism. Do these changes, with regard to deficiences, go below threshold levels for the efficient and normal functioning of enzymes and are thus rate determining? Is the cancer cell in which such alterations are observed a true descendant of the corresponding healthy cell used as a standard? Or is it necessary to ignore quantitative differences and only pay attention to those cases in which certain enzyme activities have completely disappeared? In the following chapter a more detailed consideration will be given to these points, and also to practical results which mainly have a bearing on shifts, imbalances, deficiencies and excess in metabolic events. One cannot stress sufficiently the possibility that there may exist considerable variations in these respects between tumors and tumors, between solid and ascites tumors, reticuloses and leukemias, and between experimental (transplantable or spontaneous) and clinically recognisable ones.

Fig. 4. Potter's feedback mechanism. (From Potter: Univ. Michigan Med. Bull. Courtesy of the University of Michigan Medical Bulletin, Ann Arbor. Michigan).