Host-clonal interactions in the generation of proviral gene deletion variants.

A nonproducer clone (clone A1) (from a retrovirus-infected guinea pig fibrosarcoma) has been described that under conditions of in vivo immunologic selection forms variants that lack the proviral gene. One trivial explanation for the apparent loss of the provirus from clone A1 is that clone A1 did not originate from a single cell. For evaluation of this possibility, subclones were derived from clone A1 and tested for tumor recurrence in nonimmune and virus-immune animals. Each of four subclones contained the A1 provirus and exhibited specific viral interference; tumor recurrences formed from each of these four subclones lacked the clone A1 provirus. Possible, when heterogeneous populations of retrovirus-infected cells are injected into nonimmune animals, some clones will elicit immunologic responses to retroviral antigens and subject other clones in the population to immunologic selective pressures. For testing this concept, clone A1 was injected in admixture with a producer clone (clone A4) into nonimmune Sewall Wright strain 2 guinea pigs. Tumors formed in nonimmune guinea pigs inoculated with clone A1 in admixture with clone A4 were shown to lack a detectable clone A1 provirus. The results supported the concept that a somatic mutational event (deletion of the proviral gene) occurs during growth of clone A1. When heterogeneous populations of retrovirus-infected cells are injected into animals, host-clonal interactions may occur leading to outgrowth of proviral gene deletion variants. These results supported the notion that interactions between tumor clones and the host can change the dominant clonal type of the tumor and provide a genetic basis for this change.