Gene therapy of cancer is an active field of research that was even more hope-raising at the late nineties. At the sight of the first results with the agonistic anti-CD137 monoclonal antibodies, it became a reasonable alternative to endow tumor cells with surface expression of the only characterized ligand for this surface costimulatory molecule (Goodwin et al., 1993). In order to test this idea a panel of retrovirus encoding the CD137 ligand cDNA were generated (Melero et al., 1998a). Transfected cells were able to bind a CD137 chimeric protein. Stable trans-fectants expressing high levels of CD137 ligand were selected in various tumors that included P815 mastocytoma and AG104A (Melero et al., 1998a). Transfected P815 cells were rejected after a transient growth in 92% of the cases raising systemic immunity against subsequent challenge. By contrast Ag104A CD137L+ transfectants progressed as lethal tumors regardless of expressing CD137 ligand (Melero et al., 1998a). This finding was reminiscent of the fact that this tumor cell line was able to graft as a lethal tumor even if transfected to express high levels of B7-1 (CD80) on its surface (Chen et. al., 1994). Interestingly, double transfectants co-expressing both CD137 ligand and CD80 were rejected in 60% of cases due to a stronger CTL response. This finding had important implications that suggested that the CD80-CD28 pathway and the CD137-CD137 ligand pathways of costimulation could be cross-talking (Melero et al., 1998a). Cross-talk was possible at several levels that importantly included the effect of CD28 liga-tion as enhancing the expression of CD137 on the plasma membrane of primed T lymphocytes (Melero et al., 1998a). Previous studies had nonetheless reported that CD137 and CD28 could operate independently (Chu et al., 1997). However, in vivo blockade of the CD28 pathway eliminated the stronger induction of CTLs displayed by the CD137L+ transfectants (Melero etal., 1998a). It is of note that in these models, rejection was dependent on CD8+ lymphocytes, but not on CD4+ T cells.
From the beginning it became apparent that attempts of treatment with these CD137L+ transfectants for concomitant untransfected tumors were not successful, raising questions on the viability of this treatment option. It was therefore clear that the therapeutic effect of the agonistic anti-CD137 antibodies was more potent. Reasons are still unclear, although two explanations have been raised: higher affinity of the antibodies acting as artificial ligands; and the ability of the antibodies to chase antigen activated T cells everywhere in the organism without need for cell to cell contact to deliver costimulation (Murillo et al., 2003). The duration of the costimulatory interaction conceivably longer with mAbs could also be critical.
One year later, these results were confirmed using CD137 ligand stable transfectants in the A20 lymphoma cell line (Guinn et al., 1999). It was found that CD137 ligand, CD80 and CD86 contributed to the generation of antitumor CTLs as deduced from the in vivo effect of blocking antibodies or chimeric proteins. These findings again indicated a clear role for the coexpression of the CD28 ligands (CD80 and CD86) along with CD137 ligand, a concept that was reinforced with subsequent studies (Guinn et al., 1999). Interestingly, some tumors that escaped rejection showed loss of CD137 ligand (Guinn et al., 1999), thus providing evolutionary evidence for selection against the expression of this costimulatory molecule. In the case of the A20 lymphoma, tumor cells expressed MHC class II, but the in vivo role of CD4 T cells was not explored. Nonetheless, indirect experiments suggested that the expression of CD137 ligand was not very efficient at costimulating the primary MLR reaction that is mainly mediated by alloreactive naive CD4+ T cells (Guinn et al., 1999).
Surprisingly a latter report showed that human carcinoma cell lines and primary carcinomas expressed low but detectable levels of functional CD137 ligand (Salih et al., 2000). In subsequent studies by the same group it was found that CD137 ligand could become a soluble molecule with increased concentrations in patients with hematologic malignancies (Salih et al., 2001). Understanding the role of this soluble functional CD137 ligand, that is also found increased in chronic inflammation, might shed light to understand the reasons behind ectopic CD137 ligand expression in tumors. The still unclear reasons behind this observation cast further doubts on whether transfection of tumor cells with the natural ligand of CD137 holds any hope for clinical translation.
Was this article helpful?