Team:Slovenia/Project/Antigen-TLR fusion vaccine

Track 2: Antigen-TLR fusion vaccine

The most effective TLR agonists as vaccine adjuvants besides flagellin are CpG, that activates TLR9, poly(I:C) that activates TLR3 and MPLA that activates TLR4. Stimulation with those ligands triggers production of type I interferon, which is particularly suitable for vaccines, to avoid excessive inflammation. These TLR agonists are not proteins and can not be encoded into fusion combination vaccines as in case of TLR5 and flagellin.

Thus, in order to activate these Toll-like receptors, we had to choose a different approach that would not require the antigen to be fused to the TLR ligand and might perhaps bypass the use of TLR agonist altogether.

The second important consideration is that it has been shown that phagocytosed molecules are properly processed for antigen presentation only in vesicles that contain TLR agonists or where TLR activation occurs (Blander and Medzhitov, 2006; Blander 2007). Even activation of the cell by soluble TLR agonist may not trigger proteolytic processing of phagocytosed proteins unless activation occurs and unless those proteins are within the same vesicles as activated TLRs. This is the cellular mechanism to ensure that only pathogenic microbes that activate TLRs trigger immune response but not endogenous cells, such as apoptotic cells.

The goal is therefore to join activation of TLRs and antigens within the same cellular compartment. Our solution for this problem was to prepare a construct that would fuse antigen with constitutively activated TLRs, which therefore ensures the same cellular localization as well as activation of innate immunity and antigen processing and presentation for adaptive response.

We decided to combine the following elements into a single fusion protein:

- cytosolic TIR domain of selected TLRs, responsible for the activation of the intracellular signaling cascade (leading to type I infterferon production, therefore preferably TLR3, TLR9 and TLR4),

- a transmembrane domain (TM), responsible for the correct localization of the construct,

- a dimerization domain, causing constitutive dimerization of constructs and thus constitutive recruitment of downstream signaling adaptors,

- an antigen, allowing for presentation of selected epitopes by MHC II on cell surface and,

- several peptide tags, that would allow us to track the cellular localization and activity of the constructs.

Scheme of the composition of constitutively active dimeric antigen-TLR fusion constucts

Our idea was basically to short-circuit the TLR signaling and trigger activation from selected constitutively active TLRs. This could be accomplished by forcing the cytosolic signaling domain to dimerize by the added dimerization domain. Addition of a selected antigen at the N-terminal segment, which is at the core of our idea, provides a further advantage, namely the fact that the antigen is (in case of fusion with endosomal TLRs, such as TLR3 or TLR9) localized within the endosomal vesicles, which are maturing due to the activated TLR. The antigen is processed in the endosome and loaded to MHC class II molecules that activate CD4 T-lymphocytes.

By combining several constructs our invention allows activation of several TLRs simultaneously, which resembles cellular activation by microorganisms, a common property of some of the most effective vaccines. Additionally it has been shown that activation of TLRs that signal through different signaling pathways act synergistically and significantly increase cell activation.

We selected TLR4, TLR3 and TLR9 on the basis of previous research, including the project from last year's Ljubljana iGEM 2007 team. TLR3 and TLR9 seem to be the most appropriate receptors to support efficient vaccination, since their activation leads to the production of type I interferon, more desirable than inflammatory mediators, production of which is triggered by most other TLRs. This is the reason for the use of TLR9 and TLR3 agonists, CpG and poly(I:C) as adjuvants in some of the most modern vaccines.

TLR4 is able to use a set of different adaptor molecules and thus cause very strong cell activation. On the other hand, TLR3 uses a different signaling pathway, interacting mostly with the adaptor molecule TRIF instead of the usual MyD88 utilized by the other TLRs. The final effect of TLR3 signaling is, in contrast to TLR4’s mainly NFκB activation, increased production of interferon β (IFNβ) and can be measured by IFNβ inducible gene expression. TLR3, naturally involved in nucleic acid recognition, is also not expressed on the cell surface, but is rather recruited to endosomes, where it can bind to double stranded RNA released from lysed viruses.

Researchers in our laboratory (Panter et al., manuscript in preparation) have discovered that activation of TLR4 can be easily triggered by fusing the TM and TIR domains to almost any extracellular domain and that no additional dimerization domain is required. Therefore, in our constructs with TLR4 we simply fused the TLR4 TM and TIR domain with selected antigen without the addition of dimerization domains. Activation of TLR3 was not as simple, but it has been shown, that fusing the extracellular domain of CD4 to the TM and TIR domain of any TLR causes a large increase in its activity (Hasan et al. 2004).

Localization of TLRs and therefore probably also fusion proteins can be regulated by the selection of transmembrane segments, which we also planed to test on the example of TLR3, which is localized within the ER and endosomes, white the replacement of transmembrane domain by the transmembrane segment of CD4 was expected to increase the surface localization. Antigens used in this part of the project were the same H. pylori urease protein and the synthetic multiepitope protein that were used in Track 1.