To be effective, a classical vaccine should supply two types of immune signals: an antigen (Ag) and a suitable accessory signal. In fact, recognizing the Agby itself is not sufficient to protect the host; antigen recognition has to elicit a biological response suited to the nature of the pathogen or the tumour cells. We have shown that a soluble LAG-3 molecule (a high affinity MHC class II ligand, see
http://www.ncbi.nlm.nih.gov/prow/guide/1656481751_g.htm) can deliver in vitro a powerful accessory signal to human DCs. In vivo, LAG-3 expressed following tumour cell transfection or injection of a murine LAG-3Ig recombinant molecule in cancer vaccines has been shown to induce specific CTLs and tumour regression in syngeneic tumour models such as breast or renal adenocarcinoma, melanoma and sarcoma.
LAG-3 co-localizes with the TCR and the two other MHC ligands (i.e. CD4 and CD8) on the surface of activated T cells. LAG-3Ig (a fusion molecule linking the 4 extra cellular domains of LAG-3 and the Fc fraction of an Ig) was identified as a molecule involved in mediating maturation signals for DCs.
LAG-3, like CD40L, also contributes to DC maturation and increases signal 2 deliveries. It therefore increases the immunogenicity of Ag-loaded mature DCs reaching the lymph node, leading to efficient Ag presentation and priming of naïve T and B cells. In addition to increased signal 2 delivery, inflammation induction is often associated with increased immunogenicity and inflammatory cytokines, such as IL-1 and IL-12, have been shown to provide a third signal for activation of naïve CD4+ and CD8+ T cells. Our results indicate that LAG-3, like LPS, clearly induced inflammatory cytokines, such as IL-8 and MIP-1α/CCL3, and signal 3 may thus be involved in the elicitation of Th1 type response in mice, when using a soluble mLAG-3Ig molecule as an adjuvant. In addition to signals 2 and 3, the induction of an immune response also critically depends on Ag (signal 1) reaching lymphoid organs. Immune responsiveness that is increased or initiated by adjuvants may simply be a result of enhanced translocation of vaccine antigen from the peripheral site of injection towards the draining local lymph node. Along this line, LAG-3, like CD40L (but not LPS), induces MDC and TARC production (chemokines that are required for DC migration to lymph nodes) that may participate to the adjuvant effect observed in mice by contributing to the delivery of an appropriate TCR signal.
Finally, we have characterized the molecules involved in MHC class II signalling following LAG-3 binding on human DC. These include PLCgamma2, p70syk and a substrate for PI3K (Akt).
Overall, our data show that MHC class II signalling into DCs induced after their engagement with LAG-3 permits an increased conditioning of DCs and, therefore, a better activity of vaccines. Indeed, we found that LAG-3 is highly effective when given as adjuvant into the same site of the proteinaceous vaccine. Thus, new powerful generation of therapeutic cancer vaccines could consist of recombinant protein Ags and recombinant LAG-3Ig. These results have to be confirmed in humans following the production of GMP batches of LAG-3Ig by Immutep S.A. (http://www.immutep.com/), a start-up company developing the LAG-3 technology. This information can be exploited, to optimize the schedules of immunization for anti-cancer vaccines. End users will be hospitals and cancer centres.