CuMV_TT-VLPs expression and production was performed as described in detail in [5].

Physical stability and integrity of CuMV_TT-VLPs were visualized by transmission electron microscopy using the Philips CM12 EM. For imaging, sample-grids were glow discharged and 5 μl of VLP solution was added for 30 s. The grids were then washed 3x with ddH₂O and negativly stained with 5 μl of 5% uranyl acetate for 30 s. Finally, excess uranyl acetate was removed by pipetting and the grids were air dried for 10 min. Images were taken with 84,000x and 110,000x magnification.

Wild type C57BL/6 mice were purchased from Harlan. RAG2 ^−/− mice on a C57BL/6 background were provided by Ochsenbein’ lab and were bred in our pathogen-free animal facility. All in vivo experiments used 8–12-week-old female. All animal procedures were performed in accordance with the Swiss Animals Act (455.109.1) (September 2008, 5th) of University of Bern.

WT C57BL/6 mice (8–12 weeks; Harlan) were anesthetized and prepared for imaging by shaving their right leg. Skin and adipose tissues were removed to expose the popliteal lymph node (LN) as described in detail in [18]. The anesthetized mice were then stabilized on a customized platform for imaging. The popliteal LN was located by bright field illumination imaging. A dose of 10 μg of CuMV_TT-VLPs vaccine was labelled with Alexa Fluor 488 (AF488) according to the manufacturer’s instructions (Thermo Fisher SCIENTIFIC) and injected subcutaneously (s.c.) into the mouse footpad to study the draining kinetics of CuMV_TT-VLPs. Fluorescent light illumination with a CCD Nikon camera was used for imaging.

CuMV_TT-VLPs were derivatized using 10-fold molar excess of DBCO cross-linker (Dibenzocyclooctyne-N-hydroxysuccinimidyl ester) (Sigma-Aldrich) in 2 mM EDTA and 20 mM NaP, pH 7.5 for 30 min at RT in a shaker at 400 rpm. Excess uncoupled DBCO was removed by diafiltration steps. Modified p33 peptide H-KAVYNFATMGGCK(N3)-NH2 was purchased from (Pepscan PRESTO) and reconstituted using DMSO. 10-fold molar excess of the modified peptide was then added to the derivatized CuMV_TT-VLPs. 5-fold molar excess of TCEP was added to liberate cysteine residues at the C-terminus on the VLPs. The coupling was performed for 1 h at RT in a shaker at 400 rpm. Excess peptide was removed using 100 kDa MWCO amicon centrifuge tubes (Sigma Aldrich). The efficiency of the coupling was tested by SDS-PAGE (Bio-RAD) and assessed by densitometric analysis of SDS-PAGE of CuMV_TT-VLP monomer bands compared to CuMV_TT-VLP monomer plus p33 after coupling.

10 μg of AF488 CuMV_TT-p33 nano-vaccine formulated or not formulated with 50 μl of 4% MCT adjuvant was injected in WT C57BL/6 mice footpads (8–12 weeks; Harlan) using isoflurane anesthesia. Popliteal LNs were collected 3 h, 24 h, 48 h, 96 h and 216 h after injection. Lymph nodes were fixed in PFA 2% for 3 h at RT. The LNs were transferred to CUBIC 1 medium [19] for 8 days at 37 °C and scanned at Leica SP8 with 10x lense, 1024 × 1024 resolution, tile scan was performed in LNs that were not fitting in the field of view of the 10x lense. Images were analyzed and segmented the individual CMV particles using Imaris Software v9.2.1 (Bitplane).

10 μg of AF488 CuMV_TT-p33 nano-vaccine formulated or not formulated with 50 μl of 4% MCT adjuvant was injected in WT C57BL/6 mice footpads (8–12 weeks; Harlan) using isoflurane anesthesia. Popliteal LNs were collected 3 h, 24 h, 48 h, 96 h and 216 h after injection and treated with collagenase D (Roch) in 10%FSC containing DMEM for 25 min at 37 °C. Cells were stained with live/dead dye (eBioscience) and analyzed for total number of FITC⁺ (CuMV_TT-p33). Naïve mice were used as a control.

Six groups of WT C57BL/6 mice (8–12 weeks old; Harlan) were vaccinated s.c. with a single dose of: 1st group 70 μg of CuMV_TT-VLPs, 2nd group 70 μg of CuMV_TT-Actin nano-vaccine, 3rd group 70 μg of CuMV_TT-p33 nano-vaccine_, 4th group 70 μg of CuMV_TT-p33 nano-vaccine admixed with 15 nmol of B type CpGs 5′′-TCC ATG ACG TTC CTG ATG CT-3′′) (20 mer) (Invivogen), 5th group 70 μg of CuMV_TT-p33 nano-vaccine formulated with 4% MCT adjuvant (40 mg/ ml) (Allergy Therapeutics Ltd. Worthing, UK) and the 6th group 70 μg of CuMV_TT-p33 nano-vaccine formulated with 100 μl of Alhydrogel adjuvant 2% (InvivoGen). Formulating CuMV_TT-p33 with MCT or Alum requires prolonged mixing of both components for 1 h at RT in shaker at 400 rpm to ensure adequate adsorption of the VLPs on MCT or Alum surface. Seven days later, spleens were collected and staining was performed using Fc-block, live/dead, anti-CD8 (eBioscience) and p33 (KAVYNFATM) tetramer designed using H-2Db allele and PE fluorochrome (TCMetrix).

Intra-cellular cytokine staining was performed on spleens and TILs of vaccinated WT C57BL/6 mice for measuring of IFN-γ and TNF-α cytokines as described in detail in [20].

1 × 10⁶ cells of B16F10p33 melanoma cell line (From Ochsenbein lab) was injected into the flank of RAG2 ^−/− C57BL/6 mice (From Ochsenbein lab). Twelve days later the growing tumours were collected and processed into ~2mm² fragments for transplantation into the flank of WT C57BL/6 mice (8–12 weeks old; Harlan) under full anesthesia. The transplanted WT C57BL/6 mice were treated 3 times over 14 days (mice in the control group reached the humane end-point at day 14). In the first tumor experiment mice were vaccinated s.c. as follows: 1st group 70 μg of CuMV_TT-VLPs, 2nd group 70 μg of CuMV_TT-p33 nano-vaccine and 3rd group 70 μg of CuMV_TT-p33 nano-vaccine formulated with 4% MCT adjuvant (40 mg/ ml) (Allergy Therapeutics Ltd. Worthing, UK). In the second tumour experiment mice were vaccinated s.c. as follows: 1st group 70 μg of CuMV_TT-VLPs, 2nd group 70 μg of CuMV_TT-p33 nano-vaccine admixed with 15 nmol of B type CpGs 5′′-TCC ATG ACG TTC CTG ATG CT-3′′) (20 mer) (Invivogen), 3rd group 70 μg of CuMV_TT-p33 nano-vaccine formulated with 4% MCT adjuvant (40 mg/ ml) (Allergy Therapeutics Ltd. Worthing, UK) and 4th group 70 μg of CuMV_TT-p33 nano-vaccine formulated with 100 μl of Alhydrogel adjuvant 2% (InvivoGen). Tumour growth was followed daily and measured using calipers. Tumours were collected and measured on day 14. TILs were isolated by treating the tumours with collagenase D (Roch) in 10%FSC containing DMEM for 25 min at 37 °C. Cells were passed through a cell strainer of 100 μm (Corning) and TILs were separated using Ficoll (Sigma-Aldrich). TILs were stained with Fc block, live/dead, anti-CD8 (eBioscience) and p33 tetramers (TCMetrix).

Tumour growth curves were compared by calculating the area-under curve (AUC) and analyzed by One-Way ANOVA (Turkey’s Multiple Comparison Test). Other data has been analyzed and presented using Unpaired Student’s t test. GraphPad Prism7 or 8 software was used for the analysis.

In a first step, we produced the engineered CuMV_TT-VLPs and confirmed their morphology, integrity and nano-size by electron microscopy (Fig. 1a). Next, we studied the draining-kinetics of CuMV_TT-VLPs utilizing stereomiscroscopic imaging. To this end, CuMV_TT-VLPs were labelled with the fluorescent dye AF488 and injected s.c. in the footpad of WT C57BL/6 mice. Our results show that the labelled nanoparticles accumulate in the popliteal LN in less than 1 min, demonstrating fast and efficient draining of free 30 nm CuMV_TT-VLPs (Fig. 1b). To study the efficacy of CuMV_TT-VLPs as a cytotoxic T cell based nano-vaccine, we have used the H-2D^b restricted p33 peptide derived from LCMV as a model antigen. The model peptide was coupled to CuMV_TT-VLPs using our developed method based on “biorthogonal Cu-free click chemistry” as illustrated in Fig. 1c. The efficiency of the coupling was tested using SDS-PAGE. The additional bands above the VLP monomer show efficient coupling of p33 peptide to CuMV_TT-VLP monomers (Fig. 1d).Fig1Fig. 1

CuMV_TT-VLPs demonstrate fast kinetics and constitute an efficient vaccine platform for displaying target peptides/epitopes. a Electron microscopy imaging of CuMV_TT-VLPs, (3.5 mg/ml) adsorped on carbon grids and negatively stained with uranyl acetate solution, scale bar 200 nm, CuMV_TT-VLPs sized ∼30 nm. b Stereomicroscopy images of mice popliteal LN following s.c. injection of AF488 CuMV_TT-VLPs in mice footpad. [1] bright field of the popliteal LN (identified by the arrowhead) [2] fluorescent image prior to injection of AF488 CuMV_TT-VLPs, [3–5] 1 min, 5 min and 10 min post injection of AF488 CuMV_TT-VLPs taken with the appropriate fluorescent filters. c A sketch illustrating the coupling of p33 epitope to CuMV_TT-VLPs using Cu-free click chemistry (Dibenzocyclooctyne-N-hydroxysuccinimidyl ester (DBCO) cross-linker). DBCO cross-linker reacts with Lys residues on CuMV_TT-VLPs and incorporates a cyclooctyne moiety. The formed dibenzocyclooctyne will then react with azide-labelled p33 peptide forming a stable triazole linkage without Cu catalyst. d SDS-PAGE of CuMV_TT-p33 nano-vaccine using DBCO “Dibenzocyclooctyne-N-hydroxysuccinimidyl ester” cross-linker; arrows indicate CuMV_TT-VLP monomers and dimers (formed by DBCO cross-linking of 2 monomers) with coupled p33 peptide

Microcrystalline tyrosine (MCT) is considered to be a depot-forming adjuvant facilitating the slow but prolonged release of antigens. Formulating the nano-vaccine CuMV_TT-p33 with the micron-sized MCT adjuvant may therefore enhance the slow release of the nanoparticles displaying the target epitope and extend their exposure to the immune system. To test that, we have first formulated the AF488 CuMV_TT-p33 nano-vaccine with MCT in vitro to visualize the binding of the nanoparticles to the micron-sized MCT adjuvant by confocal microscopy. The results showed that the labelled nanoparticles bind and decorate the surface of the micron-sized crystals (Fig. 2a and S1). To further study this hypothesis in vivo, we injected the AF488 CuMV_TT-p33 nano-vaccine (alone or formulated with MCT adjuvant) into the footpad of WT C57BL/6 mice as illustrated in Fig. 2b and collected the popliteal LNs at different time-points 3 h, 24 h, 48 h, 96 h and 216 h to assess the persistence of the labelled nanoparticles by flow cytometry. The results demonstrate that CuMV_TT-p33 injected in free form disappears from the popliteal LN in ~ 4 days while formulating the nano-vaccine with the micron-sized MCT adjuvant causes slower but prolonged release of the nanoparticles over 9 days (Fig. 2c). These findings were also supported when imaging the popliteal LNs by confocal microscopy (Fig. 2d and S2–S5).Fig2Fig. 2

The micron-sized MCT adjuvant displays depot effect when combined with CuMV_TT-p33 nano-vaccine. a Confocal microscopy imaging of AF488 CuMV_TT-p33 nano-vaccine following formulation with the micron-sized MCT adjuvant, 1) GFP signal of AF488 CuMV_TT-p33 nano-vaccine 2) bright light field 3) an overlay 4–6 3D images with bright light field 4) an MCT crystal decorated with CuMV_TT-p33 nano-vaccine particles. b A sketch illustrates the two prepared nano-vaccines, the 1st group consists of AF488 CuMV_TT-p33 nano-vaccine and the 2nd consists of AF488 CuMV_TT-p33 nano-vaccine formulated with MCT adjuvant. d Total number of AF488 CuMV_TT-p33 nanoparticles in the popliteal LNs collected 3 h, 24 h, 48 h, 96 h and 216 h post-injection of the two prepared nano-vaccine groups in mice footpad. Statistical analysis by unpaired Student’s t test. c Confocal microscopy images of popliteal LNs 24 h and 216 h post-injection of the two prepared nano-vaccine groups in mice footpad, GFP signal was detected in LNs, whole mount view of z-stacks was acquired. One representative experiment of 3 similar experiments is shown

We then tested whether formulating CuMV_TT-p33 nano-vaccine with MCT would enhance the specific T cell response in vivo. Therefore, six vaccines and formulations were prepared as outlined in Fig. 3a. CpG 1668 and Alum were independently formulated with CuMV_TT-p33 nano-vaccine to benchmark the potency of the micron-sized MCT adjuvant. Actin coupled to CuMV_TT-VLPs was used as a non-specific control peptide to show that the obtained response is specific to p33 peptide. The different vaccine formulations were injected once s.c. in WT C57BL/6 mice and spleens were collected seven days later for tetramer and intra-cellular cytokine staining. The H2-D^b allele p33 (KAVYNFATM) tetramers have been used to enable direct visualization and quantification of p33 specific T cells. The results showed that admixing CuMV_TT-p33 nano-vaccine with CpGs 1668 or formulating it with the micron-sized MCT adjuvant induced the highest percentage of p33 specific T cells upon single injection (Fig. 3b and f). We then assessed the cytokine secretion in each group, mainly IFN-γ and TNF-α. The secretion of both cytokines was enhanced when combining the nano-vaccine with CpGs 1668 or MCT adjuvant (Fig. 3c and d). Formulating CuMV_TT-p33 nano-vaccine with Alum did not enhance the production of p33 specific T cells nor the secretion of IFN-γ or TNF-α. Furthermore, when analyzing the dual secretion of IFN-γ and TNF-α in polyfunctional T cells, a large percentage of the cytokine-producing T cells was found to be polyfunctional in the groups admixed with CpGs 1668 or MCT (Fig. 3e and g).Fig3Fig. 3

Formulating CuMV_TT-p33 nano-vaccine with the micron-sized MCT adjuvant induces significant p33 specific T cell response and enhances cytokines secretion. a Vaccination scheme for six vaccine groups, CuMV_TT-VLPs, CuMV_TT-Actin, CuMV_TT-p33, CuMV_TT-p33 admixed with CpG 1668, CuMV_TT-p33 formulated with MCT and CuMV_TT-p33 formulated with Alum. b Percentage of CD8⁺ Tetramer⁺ CTLs (means ± SEM) in the spleen in each vaccinated group. c Percentage of CD8⁺ IFN-γ⁺ secreting cells (means ± SEM) in the spleen in each vaccinated group. d Percentage of CD8⁺ TNF-α⁺ cells (means ± SEM) in the spleen in each vaccinated group. e Percentage of CD8⁺ IFN-γ⁺/TNF-α⁺ secreting cells (means ± SEM) in the spleen in each vaccinated group. Statistical analysis by Oneway ANOVA (Turkey’s Multiple Comparison Test). f Representative flow cytometry dot plots showing the percentage of CD8⁺ Tetramer⁺ CTLs in each vaccinated group. g Representative flow cytometry dot plots showing the percentage of CD8⁺ IFN-γ⁺/TNF-α⁺ secreting cells in each vaccinated group. (3 mice per group), one representative of 3 similar experiments is shown

In order to test the immunogenicity and efficacy of combining nanoparticles with micron-sized adjuvants in a melanoma model, we have adapted a challenging melanoma murine model based on transplanting ~2mm³ of B16F10p33 tumour fragment into the flank of WT C57BL/6 mice. The tumour was allowed to grow for 5 days more following transplantation before the vaccination regimen started (Fig. 4a). Three groups were prepared as illustrated in Fig. 4b, CuMV_TT-VLPs as a control, CuMV_TT-p33 nano-vaccine alone and CuMV_TT-p33 nano-vaccine formulated with MCT. Tumours were collected for analysis 14 days after tumour transplantation as the control group reached the ethically allowed maximal size of ~1000mm³. The obtained results revealed that formulating CuMV_TT-p33 nano-vaccine with the micron-sized MCT could significantly hinder B16F10p33 tumour progression when compared to the control group (p < 0.0001) or to the group vaccinated with CuMV_TT-p33 nano-vaccine alone (p 0.0055) (Fig. 4c and d). Tumour-infiltrating lymphocytes (TILs) represent a prognostic factor for effective immune responses especially in melanoma [21, 22]. Therefore, we measured the total number of the infiltrated CD8⁺ T cells and p33 specific CTLs (Fig. 4e) in TILs and calculated the density of these cells in each vaccinated group (number of cells divided by tumour volume). Formulating CuMV_TT-p33 nano-vaccine with MCT significantly increased the density of total CD8⁺ T cells (p. 0.0024) (Fig. 4f) as well as the density of p33 specific CTL measured by tetramers (p. 0.0093) (Fig. 4g) in comparison to mice vaccinated with CuMV_TT-p33 nano-vaccine alone. IFN-γ production was also enhanced when formulating CuMV_TT-p33 with MCT (Fig. 4h). Thus, formulating the a nano-vaccine in MCT enhanced infiltration by specific T cells as well as anti-tumor protection.Fig4Fig. 4

Formulating CuMV_TT-p33 nano-vaccine with the micron-sized MCT adjuvant causes tumour regression and enhances CD8⁺ and p33 specific CTL infiltration into B16F10p33 tumours. a A diagram illustrating the adapted tumour experimental method based on injecting ~ 1 × 10⁶ B16F10p33 melanoma cell line into the flank of RAG2 ^−/− deficient C57BL/6 mice. Twelve to thirteen days later the growing tumours were collected and processed in ~2mm³ for transplantation into the flank of WT C57BL/6. b Vaccination scheme for three vaccine groups, CuMV_TT-VLPs, CuMV_TT-p33 and CuMV_TT-p33 formulated with MCT. c Tumour growth curve of subcutaneous B16F10p33 melanomas in each vaccinated group, mice were euthanized when the tumour reached ~1000mm³, arrows indicate start of treatment. d Tumour volume mm³ (mean ± SEM) measured at day 14 post tumour collection in each vaccinated group, each dot represents a tumour. e Representative flow cytometry dot plots showing the total number of CD8⁺ Tetramer⁺ CTLs in each vaccinated group. f Density of CD8⁺ T cells (means ± SEM) in each vaccinated group, “measured by dividing the total number of CD8⁺ cells in TILs by the tumour volume^”. g Density of CD8⁺ Tetramer⁺ CTLs (means ± SEM) in each vaccinated group, “measured by dividing the total number of CD8⁺ Tetramer⁺ CTL by tumour volume^”. h Percentage of CD8⁺ IFN-γ ⁺ secreting cells (means ± SEM) in each vaccinated group. Statistical analysis by Student’s t test. (5 mice per group), one representative of 3 similar experiments is shown

The immunogenicity of the CuMV_TT-p33 nano-vaccine formulated with the micron-sized MCT adjuvant was then compared to the immune-stimulatory B type CpGs and the widely used adjuvant Alum using the same aggressive B16F10p33 tumour model. Four groups were prepared as shown in Fig. 5a. The results again revealed that formulating CuMV_TT-p33 nano-vaccine with CpGs 1668 or MCT would significantly (p 0.0072, 0.0129 respectively) hinder B16F10p33 tumour progression, which was not the case when formulating CuMV_TT-p33 with Alum (p 0.4188) (Fig. 5b). In a next step, we measured the total number of infiltrated CD8⁺ T cells (Fig. 5c) and p33 specific CTLs (Fig. 5d) in the tumour and calculated the density. There was a general increase in the groups mixed with CpGs 1668 or formulated with MCT or Alum (Fig. 5e and f). Formulating CuMV_TT-p33 nano-vaccine with the micron-sized MCT adjuvant showed comparable results to CpGs 1668, the gold standard adjuvant in mice. Furthermore, MCT adjuvant was more potent at increasing the infiltration of total CD8⁺ and p33-specific T cells into the tumour microenvironment when compared to formulating the vaccine with Alum.Fig5Fig. 5

The micron-sized MCT adjuvant shows comparable activity to B type CpGs and is superior to Alum in driving protection against B16F10p33 melanoma. a vaccination scheme for four vaccine groups, CuMV_TT-VLPs, CuMV_TT-p33 admixed with CpG 1668, CuMV_TT-p33 formulated with MCT and CuMV_TT-p33 formulated with Alum. b Tumour growth curve of subcutaneous B16F10p33 melanomas in each vaccinated group, mice were euthanized when the tumour reached ~1000mm³, arrows indicate start of treatment. c Representative flow cytometry dot plots showing the total number of CD8⁺ T cells in each vaccinated group, gated on TILs. d Representative flow cytometry dot plots showing the total number of CD8⁺ Tetramer⁺ CTLs in each vaccinated group, gated on CD8⁺ T cells. e Density of CD8⁺ T cells (means ± SEM) in each vaccinated group, “measured by dividing the total number of CD8⁺ cells in TILs by the tumour volume^”. f Density of CD8⁺ Tetramer⁺ CTLs (means ± SEM) in each vaccinated group, “measured by dividing the total number of p33 tetramer⁺ CTLs by tumour volume^”. Statistical analysis by Student’s t test. (4 mice per group), one representative of 3 similar experiments is shown

Production of IFN-γ and TNF-a cytokines by TILs was also assessed upon immunization with CuMV_TT-p33 admixed with B type CpGs or formulated with MCT or Alum. Production of IFN-γ (Fig. 6a and b) and TNF-a (Fig. 6c and d) was assessed separately or as dual cytokine production in polyfunational T cells (Fig. 6e and f). There was no significant difference in the production of single IFN-γ, TNF-a or dual IFN-γ/TNF cytokines between the groups admixed with CpGs 1668 or formulated with MCT adjuvant (p 0.3986, 0.3433 and 0.4120 respectively). However, there was a significant difference when comparing the group formulated with MCT to the one formulated with Alum (p 0.0179, 0.0187 and 0.006, respectively).Fig6Fig. 6

The micron-sized MCT adjuvant shows comparable production of cytokines to B type CpGs and is superior to Alum in B16F10p33 melanoma. a Percentage of CD8⁺ IFN-γ ⁺ secreting cells (means ± SEM) in each vaccinated group. b Representative flow cytometry dot plots showing the frequency of CD8⁺ IFN-γ⁺ secreting cells in each vaccinated group. c Percentage of CD8⁺ TNF-⍺⁺ secreting cells (means ± SEM) in each vaccinated group. d Representative flow cytometry dot plots showing the frequency of CD8⁺ TNF-⍺⁺ secreting cells in each vaccinated group. e Percentage of dual IFN-γ⁺ and TNF-⍺⁺ secreting cells (means ± SEM) in each vaccinated tumour, gated on CD8⁺ cells in TILs. f Representative flow cytometry dot plots showing the frequency of dual IFN-γ⁺ and TNF-⍺⁺ secreting cells in each group, gated on CD8⁺ cells in TILs. Statistical analysis by Student’s t test. (4 mice per group), one representative of 3 similar experiments is shown