Immune and non-immune control of neoplastic development: contrasting effects of host and tumor evolution.

Abstract

Tumor development is a play with many different dramatis personae. The main character is still the Cell. It is a Protean figure than can appear in many different forms and play very different roles, most of which can be classified as stages in the microevolutionary process known as tumor progression. The initiation of tumor development and most stages in tumor progression are most likely to involve changes at the genetic (DNA) level. This will clearly include both point mutations, chromosomal rearrangements, and changes in gene dosage by trisomy, for example. Infrequently, tumor induction and, perhaps more frequently, progression could also be due to epigenetic, i.e., essentially regulatory changes. On its road of progression, the preneoplastic cell encounters many predators, including various effectors of the immune system. Inter alia, it will depend on host genetics, on age and all of the many physiological modulators of the immune system, on just what predators are encountered, and in what frequencies. Their ability to kill shall, in turn, depend on their own recognition and killing equipment, their interactions, and the vulnerability of the cell. In our opinion, it would be important to replace current “all or none discussions” about the relative importance or unimportance of immune surveillance mechanisms in the protection against tumor development by a more sophisticated analysis that takes the biologic history of each system into consideration. It is clearly more difficult to build up reliable immune defenses against neoplastic cells in a vertebrate organism than to mount responses to outside parasites, including viruses and bacteria. In view of the close antigenic similarities between normal and tumor cells, this is not surprising. The fact that defenses can be built up nevertheless, and can even reach a state of complete watertight protection is, on the other hand, well illustrated by the polyoma system. The opposite, suppressive responses that can prevent the rejection of what would otherwise be highly antigenic tumor cells does also exist, however, as illustrated by the UV-induced carcinoma system of Margaret Kripke. Many, perhaps most tumors are probably in a limbo, ranging from no recognition to some but often quite haphazard responses. On the cellular side, the non-random chromosomal changes associated with tumors arising in a given cell type are highly suggestive of genetic systems that control differentiation and/or responsiveness to a specific type of growth control and are localized at defined regions of certain chromosomes. The wide involvement of chromosome 14 in human lymphoma is suggestive of genetic elements, concerned with the control of normal lymphocyte responsiveness. A variety of 14 q+ markers has been identified in different lymphomas, but the reciprocal 8;14 translocation appears to be specific, so far, for BL and B-cell derived ALL. Both represent the neoplastic proliferation of the same cell type. Could it be that the distal part of chromosome 14 is a “supergene” region, involved with a variety of lymphocyte subtypes and their responses, perhaps in analogy with the broad involvement of the MHC region with a whole variety of more or less loosely related immune and other membrane functions? In a similar way, the reciprocal 9;22 translocation represented by the Philadelphia chromosome may be postulated to reflect genetic regions concerned with the normal response of the granulopoietic series to differentiation-inducing stimuli. In experimental myeloid leukemia, the work of Sachs and his group has clearly shown the role of specific chromosomal changes in determining the ability of myeloid cells to differentiate. As an alternative explanation, one may consider the possibility that non-random chromosomal changes act by increasing the resistance of transformed cells to immune effectors. While I would consider this as the less likely explanation, it cannot be ruled out, particularly in view of the fact that diploid LCL, which are not tumorigenic in the subcutaneous space of nude mice (as contrasted to the chromosomally changed BL lines), can nevertheless.