At the beginning of the twentieth century, Ehrlich was the first to propose that antibodies did exist prior to the introduction of antigen or immunogen and that antigenic stimulation only promoted amplification of molecules of the appropriate specificity. This proposal was quite pertinent, because it contained several key ideas, proven correct some 70 years later. One is the selective process of a preexisting repertoire, the other is that immunoglobulins (Igs) may exist as surface or secreted molecules. The only aspect that was not verified is that Ehrlich imagined a cell to express many different antibodies. After a long period during which hapten analysis indicated the exquisite precision of antibody-antigen interactions, it seemed impossible that antibody structures might preexist to recognize laboratory artifacts, like the haptens being studied. This resulted in the template theory of antibody formation, which required the presence of antigen prior to the synthesis of the corresponding antibodies. Revival of the selection theory was proposed in 1955 by Jerne, but it took another 4 years before the final form of the "clonal selection theory" was put forward by Burnet. The essence of the theory was that antibody specificities were preformed, but in a clonally distributed manner, so that one cell makes only one antibody of one given specificity. This was crucial to provide a mechanism accounting for acquisition of discrimination between self and nonself, after the observation on induction of immunological tolerance by the group of Medawar in 1953,when it was demonstrated that tolerance was an acquired phenomenon, resulting from a contact between the immature immune system and the self antigens during gestation or the perinatal period. Burnet most simply explained tolerance by a deletion of those clones that could interact with self components, leading to the concept of "forbidden clones." Once the basic anti-nonself repertoire established, an antigen was seen as merely selecting and amplifying the cell population of clones expressing the corresponding specific antibodies.
Many investigators worked to define models that would prove the clonal theory, and the literature is amply documented on this point, with many very elegant approaches, such as those based on limiting dilutions, either in vitro or in vivo after transfer in irradiated mice. The ultimate proof came from molecular analysis, first at the protein level on myeloma proteins and then when the mechanism of gene rearrangement of Ig genes was elucidated, under the control of allelic exclusion. Some exceptions to the clonal expression of Ig molecules have been periodically reported. Some were not exceptions, as in the case of lymphocytes coexpressing multiple isotypes that, in fact, share identical Vh and VL regions, and therefore having the same specificity. Upon reactivation of the recombinase genes upon immunization, for example, a transient coexpression of different V genes may occur, leading to secondary gene rearrangements. But this rarely results in long-term production of antibodies of different specificities. The clonal theory is widely accepted and considered proven, and the fantastic use of monoclonal antibodies is (for most biologists) daily living proof of its pertinence.
