Early studies aimed at characterising tumour antigens were based on in vivo transplantation/rejection techniques with chemically-induced tumours in mice, which identified cell surface antigens unique to each tumour. Gradually evidence appeared supporting the concept of tumour antigens shared by tumours of the same histological type (in man) or induced by the same virus (in animals). As more sophisticated in vitro methods were developed for demonstrating tumour antigens, antigens shared by more than one type of tumour, and even present on normal tissues, were discovered. This review emphasises those tumour antigens which could be called ^common' tumour antigens and outlines the development of the concept of a ^common' tumour antigen related to myelin basic protein (MBP). This tumour antigen, cancer basic protein (CaBP), has been proposed as a component of malignant cells which sensitises the tumour host; as a consequence, host lymphocytes recognise the related protein MBP. Methods for detecting tumour antigens are reviewed with emphasis on the Erosette augmentation assay, which forms the basis for the experimental work.
E-rosette Augmentation (ERA)
ERA was chosen as an in vitro assay because it is a simple well defined technique for demonstrating cell-mediated immunity. It is an uncomplicated assay that employs a relatively short incubation time. Important characteristics of the assay were initially defined with haptenated proteins as antigens. The requirement for optimum concentration to observe antigenicity was established, and a bell-shaped dose-response curve demonstrated for all antigens tested.
A useful technique was developed for determining whether antigens were related and contained a common epitope. Optimum antigen reactivity in ERA could be suppressed by another antigen in excess concentration if it was related (analogous to the tailing off of reactivity observed with antigen concentrations above the optimum), whereas unrelated antigens were ineffective.
Myelin Basic Protein (MBP)
The association between the presence of a tumour and apparent sensitisation to MBP was confirmed using ERA assays with spleen cells (SO of mice bearing chemically induced tumours. The SC of normal mice did not recognise MBP. T lymphocytes and antigen-presenting cells were required to demonstrate ERA with MBP thus establishing the immunological nature of the association.
Several peptides of MBP were recognised by SC of mice bearing tumours. One peptide (1-19) was identified as being absent from the sensitising tumour antigen since the peptide was not reactive with SC of tumour-bearing mice but was reactive with MBP-sensitised SC. Another peptide (64-83) showed a different pattern of reactivity in optimum/excess experiments with tumour-bearing mice than with MBP-sensitised mice; the unique sequence of murine MBP in this region suggested that the sensitising tumour antigen resembled the latter.
Cancer Basic Protein
Acid extraction of tumour tissue (also the method for extraction of MBP from nervous tissue) with purification by freeze-drying and ion exchange chromatography, identified an antigen highly reactive in ERA. This appeared to be CaBP. Extracts from different tumour sources but not normal tissue, were recognised by SC of mice bearing various tumours, mice sensitised to MBP and mice sensitised to CaBP. These ERA reactions confirmed a close relationship between tumour antigens, MBP and CaBP.
MBP and CaBP are similar but not identical since an excess of either antigen did not suppress the optimum reactivity of the other in ERA optimum/excess reactions using SC of mice bearing tumours. That is, both proteins had unique and common epitopes. Some of these results were confirmed with delayed-type hypersensitivity tests.
A tumour antigen resembling MBP appears to exist in all tumours but not normal tissue, and that tumour antigen sensitises animals so that their lymphocytes can recognise CaBP or MBP.