Team:Slovenia/Results/Antigen-TLR fusion vaccine

For the Antigen-TLR fusion vaccine, we created several constructs with different TIR and transmembrane (TM) domains. Among these, the 5 construct shown schematically below have shown best results.

Five Antigen-TLR fusion constructs that have shown best results in various in vitro assays. CMV = CMV promotor, SS = CD4 signaling peptide, HA = HA tag, multi = synthetic multiepitope. The dimerization domain in the top four constructs is the ecto domain of CD4 (CD4e, yellow) where only the ecto domain of CD4 was used and CD4sh (brown) where the ecto and the TM domain of CD4 was used. In the construct with the TM domain of CD4 only the TIR domain of TLR3 was used (red), whereas in constructs without the CD4 TM domain the TM and TIR domains of TLR (red, pink) were used. The short names for these constructs, used in the graphs and the text below are composed only of the antigen, dimerization domain and TM and TIR domain names.

To test whether our antigen-TLR fusion vaccine is functional, we had to check for:

- correct cellular localization

By using different TM domains we aimed to achieve different localization of expressed proteins. Fusions expressed on the plasmalema would present the antigen on the surface of the cell, where it could be recognized by B-cells. Fusions expressed on the endosomal membrane would allow for antigen processing inside the endosome and the following presentation on MHC II.

- activation of immune system

Active TLRs stimulate production and release of costimulatory molecules that activate the immune response and processing of antigens. By using TIR domains of different TLRs and combining them in the same vaccine, we could activate both the MyD dependant and MyD independent signaling pathway, thus reiforcing immune response activation by many times.

All proteins expressed from DNA constructs were first tested for their cellular localization. It is known that the transmembrane region of proteins is usually responsible for correct localization of transmembrane proteins. Since CD4 is found in cell membrane, we expected that our protein-constructs with the CD4 transmembrane region will also be placed in cell membrane. On the other hand, Toll-like receptors can be divided into two groups - endosomal (TLR3, TLR7, TLR8, TLR9) and plasmalemic

TLRs (TLR4, TLR5). Therefore, we expected protein-constructs with transmembrane region of TLR3 and TLR4 to be localized within cellular vesicles and on the cell surface, respectively.

Proteins with CD4 transmembrane region are mostly localized in the cell membrane, while those with TLR3 transmembrane region are found within the cell. HEK293T cells were transfected with 2 µg of indicated constructs and after 48 h surface or intracellularly stained as indicated in Methods. Stained cells were then analyzed with flow cytometry. Results are shown in the left or right figure above.

Localization of protein-constructs was also confirmed by confocal microscopy:

We found CD4e-TMTIR3 localized intracellularly.

CD4sh-TIR3 was mainly localized in the cell membrane.

multi-TMTIR4 was also mainly located in cell membrane - bringing the multiepitope to the cell surface, and to a minor extent in the ER

There was no unspecific staining in control cells, imaged under same conditions.

Next, all DNA constructs that comprise the cytosolic signaling domain (TIR) of the Toll-like receptors TLR9, TLR3 or TLR4 and an agent of dimerization (for example CD4 extracellular domain), have been tested for constitutive signaling activity with luciferase reporter assay. The wild type TLR9 and TLR4 proteins, when stimulated by corresponding TLR ligands, both homodimerize and trigger a signaling cascade, eventually leading to activation of the transcription factor NFκBand thus transcription of any genes under control of NFκB inducible promoters. For our assays, plasmids harboring the firefly luciferase reporter gene under an NFκB promoter was cotransfected with the DNA constructs comprising TIR4 and TIR9 to evaluate their activity.

TLR3 on the other hand triggers a different signaling pathway, which leads to transcription of genes under control of IFNβ inducible promoters. Activity of DNA constructs comprising TIR3 was thus evaluated by cotransfection with plasmids harboring the firefly luciferase gene under control of the IFNβ promoter. Since all of our DNA constructs were based on constitutive dimerization of expressed chimeric proteins, cells were left unstimulated after transfection.

Transmembrane and intracellular domain of TLR4 with H. pylori multiepitope activates innate immunity. HEK293 cells were transfected with 40 ng or 80 ng of plasmid CMV-SS-HA-multi-(TMTIR)TLR4-Histop and luciferase reporter plasmids. 40 h after transfection cells were lysed and luciferase activity was measured.

Different combinations of CD4 and TLR3 can trigger inflammatory signaling cascade. HEK293 cells were transfected with 50 ng or 100 ng of various plasmid constructs and luciferase reporter plasmids. 30 h after transfection cells were lysed and luciferase activity was measured.

It was observed that constructs with TLR3 TIR domain required the extracellular domain of CD4 for dimerization, while construct with TLR4 TIR domain dimerized spontaneously when antigen was added on its N-terminus. To eliminate any possibility of untransfected cells reacting to TLR ligands or showing spontaneous NFκB or IFNβ activity, we also stimulated and measured the luciferase expression of cells cotransfected with the backbone vector containing only constitutive promoter with no inserts.

To sum it up, we have successfully shown correct localization and activity of the selected constructs. With a vaccine designed in this way, both MyD dependent and independent signaling pathways can be activated to stimulate the immune response against pathogen antigens presented on the cell surface (in case of TLR4 constructs) and on MHC II molecules after processing in endosomes.