Enhanced Vaccine Immunogenicity Enabled by Targeted Cytosolic Delivery of Tumor Antigens into Dendritic Cells

Molecular vaccines comprising antigen peptides and inflammatory cues make up a class of therapeutics that promote immunity against cancer and pathogenic diseases but often exhibit limited efficacy. Here, we engineered an antigen peptide delivery system to enhance vaccine efficacy by targeting dendritic cells and mediating cytosolic delivery. The delivery system consists of the nontoxic anthrax protein, protective antigen (PA), and a single-chain variable fragment (scFv) that recognizes the XCR1 receptor on dendritic cells (DCs). Combining these proteins enabled selective delivery of the N-terminus of lethal factor (LFN) into XCR1-positive cross-presenting DCs. Incorporating immunogenic epitope sequences into LFN showed selective protein translocation in vitro and enhanced the priming of antigen-specific T cells in vivo. Administering DC-targeted constructs with tumor antigens (Trp1/gp100) into mice bearing aggressive B16–F10 melanomas improved mouse outcomes when compared to free antigen, including suppressed tumor growth up to 58% at 16 days post tumor induction (P < 0.0001) and increased survival (P = 0.03). These studies demonstrate that harnessing DC-targeting anthrax proteins for cytosolic antigen delivery significantly enhances the immunogenicity and antitumor efficacy of cancer vaccines.

Amino acid and genetic sequence of DC-Targeting scFv.S4 Table S2.
Summary of long antigen peptide sequences.S5 Table S3.
Flow cytometry analysis of scFv binding to CHO cells.S12 Fig. S8.
 For peptide 1a, bromoacetic acid was coupled to the side chain of lysine using standard SPPS conditions as specified above. For peptide 1b, AZDye 647 dye was coupled to the side chain of lysine by through gentle agitation of the resin, AZDye 647 NHS Ester (0.9 equiv) and DIPEA (1 equiv) in DMF for 16 h.

Development of XCR1-targeting scFv.
Plasmid construction.The scFv plasmid was constructed from an anti-XCR1 IgG (clone: MARX10), comprising amino acid sequence from the IgG variable heavy (VH) and light (VL) chains, a hydrophilic spacer region (G4S)4 between the two chains, and a sortase recognition tag (LPSTGG) at the C-terminus.The codon-optimized gene (Integrated DNA Technologies) encoding the scFv was cloned into a pET-SUMO vector (Invitrogen), according to the manufacturer's instructions.The mixture was transformed into Mach1™ T1 phage-resistant E. coli, streaked on an LB Agar plate (30 µg/mL Kanamycin), and incubated for 12 h at 37 o C. Individual colonies were selected and expanded in 10-mL cultures of LB Broth containing 30 µg/mL Kanamycin.After incubating for 12 h at 37 o C and 180 rpm, plasmids were recovered using a miniprep kit (QIAGEN) according to the manufacturer's instructions.Incorporation of the gene insert was established using Sanger sequencing.
Recombinant expression.The scFv protein was expressed as a SUMO fusion protein with SHuffle® T7 Express Competent E. coli (New England Biolabs).The plasmid was transformed into E. coli, streaked onto an LB Agar plate (30 µg/mL Kanamycin), and incubated at 30 o C.After 15-24 h, individual colonies were selected, expanded into larger cultures with Terrific Broth (0.4% v/v glycerol, 30 µg/mL Kanamycin), and incubated at 30 o C and 180 rpm.After OD600 reached ~0.6-0.8, protein expression was induced with 0.4 mM isopropyl β-d-1-thiogalactopyranoside (IPTG) and incubated at 18 o C and 130 rpm.After 24 h, the cultures were harvested by centrifugation (10,000 × g for 10 min.)and stored at -80 o C.

Anthrax protein expression and purification.
Protective antigen.PA, mPAC, and mPAC[F427A] variants were expressed in BL21(DE3) E. coli (New England Biolabs) at the New England Regional Center of Excellence/Biodefense and Emerging Infectious Diseases (NERCE), as described previously (4).After the expression, PA was isolated from the E. coli pellets through osmotic shock by suspending the cells in a sucrose buffer (20 mM Tris pH 8.5, 1 mM EDTA, 20% sucrose), followed by pelleting and resuspending the cells in a 5 mM MgSO4 buffer.PA was purified by pelleting the cell lysate, filtering the supernatant through a 0.2 µm filter, and purifying the filtrant by anion exchange chromatography.

N-terminus of Lethal factor (LFN).
LFN and LFN-DTA variants were expressed as SUMO fusion proteins in a Champion pET-SUMO vector: SUMO-LFN-LPSTGG-H6 and SUMO-LFN-DTA.After expression, the proteins were isolated from E. coli pellets by suspension in Tris buffer (20 mM Tris, 150 mM NaCl, pH 8.5), lysis by sonication, and purification with a HisTrap FF Ni-NTA column.The proteins were eluted with Tris buffer containing 0.5 M imidazole, then buffer exchanged to remove imidazole.SUMO was cleaved using SUMO protease.Afterward, LFN was used without further purification for protein ligation reactions with SrtA*.LFN-DTA was further purified by size-exclusion chromatography.Purified LFN and LFN-DTA were analyzed by SDS-PAGE at 165 V for 36 min on an Invitrogen Bolt™ 4-12% Bis-Tris Plus Gel with Bolt™ MES SDS Running Buffer (1x).Gels were visualized by SimplyBlue™ SafeStain (Coomassie).Clean fractions were pooled, concentrated with an Amicon® Ultra-15 Centrifugal Filter Unit, and analyzed by LC-MS.

Cell viability analysis.
Cells were maintained at 37 o C and 5% CO2.CHO-K1 (ATCC) cells were maintained in F-12K media supplemented with 10% (v/v) fetal bovine serum and 1% penicillin/streptomycin. XCR1-CHO (Creative Bioarray) were maintained in RPMI-1640 media supplemented with 10% (v/v) fetal bovine serum and 10 µg/mL puromycin.Prior to an experiment, cells were plated at 1×10 3 cells/well in a 96-well flat-bottom plate and allowed to adhere overnight.Protein treatments were prepared in the corresponding cell culture medium, which included PA, scFv-mPAC, or scFv-mPAC[F427A] and serial dilutions of LFN-DTA.The cells were resuspended with the protein treatments, followed by incubation for 72 h at 37 o C and 5% CO2.Cell viability was measured by the CellTiter-Glo luminescent assay (Promega) according to the manufacturer's procedure.Relative viability was inferred from the luminescence values.All counts were normalized to untreated cells and all experiments were done in triplicate.
Animal studies.All experiments were performed under an institute-approved IACUC protocol following federal, state, and local guidelines for the care and use of animals.C57BL6/J mice were procured from The Jackson Laboratory.6-to 12-week-old female mice were used for these studies and were housed with free access to food and water ad libitum over the experiment.For the subcutaneous (s.c.) injections, the animals were shaved at the right flank and the tail base.In vivo NIRF imaging, shaving, injections, and tumor measurements were performed under anesthesia using 2-3% isoflurane inhalation along with O2.The animals were sacrificed by CO2 inhalation.In vivo NIRF imaging was performed using the In Vivo Imaging System IVIS (Perkin Elmer, USA).All compounds injected in mice were sterile or filtered using 0.22 µm.
Intracellular cytokine staining.Mice (C57Bl/6, 5-7 weeks) were vaccinated with s.c.injections, in which half of the dose given on each side of the tail base.Peripheral blood was analyzed 7 d after each boost by intracellular cytokine staining of IFN-γ and TNF-α.Peripheral blood was collected, red blood cells were lysed, and lymphocytes were plated with 10 μg/mL of the epitope peptide (Table S3, OVA257-264).After 2 h incubation at 37 °C and 5% CO2, brefeldin A was added and samples were returned to incubation for an additional 4 h.Samples were stained with a fixable viability dye followed by extracellular staining.Using a fixation/permeabilization kit (BD Biosciences), samples were prepared for intracellular antibody staining and assessed by flow cytometry.
MHC tetramer staining.Mice (C57Bl/6, 5-7 weeks) were vaccinated with s.c.injections, in which half of the dose given on each side of the tail base.Peripheral blood was analyzed 7 d after each boost by MHC tetramer staining to detect antigen-specific CD8 + T cells.Peripheral blood was collected, red blood cells were lysed, and lymphocytes were plated with H-2Kb/SIINFEKL tetramer stain and 50 nM dasatinib.After 30 min incubation at 37 °C and 5% CO2, samples were prepared for extracellular antibody staining and assessed by flow cytometry.
Tumor therapy.Thirty female C57Bl/6J mice, 5-7 weeks-old, 18-20 g were s.c injected in the flank with 3 × 10 5 B16-F10 cells suspended in 100 µL of complete medium.At Day 4 post tumor induction, the animals were sorted into groups (n = 8-10 per group (Fig. A).At Day 4, 10, 16 and 22 the mice were s.c vaccinated at both sides of the tail base with 2 × 25 µL of c-di-GMP (25 µg) combined with either Trp1 + gp100 (50 pmol each) or LFN-Trp1-gp100 (50 pmol) + scFv mPAC (10 pmol), then intraperitoneally (i.p) injected with 200 µg (100µL) of anti-PD1 CD279 antibody (BioXCell, NH, USA).Animals in the naive group were i.p injected with anti-PD1 and s.c.injected with saline solution.Animals were weighed every 2 days and sacrificed if they lost ≥ 20% of their weight.To determine the survival, the tumor surface (L × W) was measured daily using a caliper, and the animals were then sacrificed when the tumor reached 100 mm 2 .Statistical analysis.Calculations were made using GraphPad Prism 8.For splenocytes quantifications, comparisons were done using one-way ANOVA with Fisher's least significance difference test.For in vivo tumor growth monitoring, calculations were performed both using a two-way ANOVA with Dunnett's multiple comparison versus the naive group (until Day 16 for Day-to-Day comparisons) or two-way ANOVA with Tuckey comparison (from Day 0 to the end for curve comparisons) (*p< 0.05; **p< 0.005; ***p<0.001and ****p< 0.0001).Kaplan-Meier percent survival curves were compared using the log-rank Mantel-Cox test.

Fig. S1 .
Fig.S1.LC-MS analysis of the anti-XCR1 scFv.The protein (50 ng) was loaded onto an Agilent Zorbax 5 μm 300SB-C3 column (2.1 × 150 mm) and was eluted with a gradient of 1-91% CH3CN in H2O with 0.1% FA and a flow rate of 0.5 mL/min.The protein was detected on an Agilent 6550 ESI-Q-TOF mass spectrometer.The TIC peak was integrated and the mass was deconvoluted using maximum entropy algorithm.

Fig. S2 .
Fig. S2.LC-MS analysis of linker peptides 1a (top) and 1b (bottom).The peptides (50 ng) were loaded onto an Agilent Zorbax 5 μm 300SB-C3 column (2.1 × 150 mm) and were eluted with a gradient of 1-91% CH3CN in H2O with 0.1% FA and a flow rate of 0.5 mL/min.The peptides were detected on an Agilent 6550 ESI-Q-TOF mass spectrometer.The TIC peak was integrated and the mass was deconvoluted using maximum entropy algorithm.

Fig. S4 .
Fig. S4.LC-MS analysis of G3-mPAC and G3-mPAC[F427A] after conjugating peptide 1b.The proteins (100 ng) were loaded onto an Agilent Zorbax 5 μm 300SB-C3 column (2.1 × 150 mm)and were eluted with a gradient of 1-91% CH3CN in H2O with 0.1% FA and a flow rate of 0.5 mL/min.The proteins were detected on an Agilent 6550 ESI-Q-TOF mass spectrometer.The TIC peak was integrated and the mass was deconvoluted using maximum entropy algorithm.

Fig. S5 .
Fig. S5.LC-MS analysis of scFv-mPAC and scFv-mPAC[F427A].The protein (100 ng) was loaded onto an Agilent Zorbax 5 μm 300SB-C3 column (2.1 × 150 mm) and were eluted with a gradient of 1-91% CH3CN in H2O with 0.1% FA and a flow rate of 0.5 mL/min.The proteins were detected on an Agilent 6550 ESI-Q-TOF mass spectrometer.The TIC peak was integrated and the mass was deconvoluted using maximum entropy algorithm.

Fig. S6 .
Fig.S6.LC-MS analysis of the AF647-labelled anti-XCR1 scFv after conjugating peptide 1a.The protein (50 ng) was loaded onto an Agilent Zorbax 5 μm 300SB-C3 column (2.1 × 150 mm) and was eluted with a gradient of 1-91% CH3CN in H2O with 0.1% FA and a flow rate of 0.5 mL/min.The protein was detected on an Agilent 6550 ESI-Q-TOF mass spectrometer.The TIC peak was integrated and the mass was deconvoluted using maximum entropy algorithm.

Fig. S7 .
Fig. S7.Flow cytometry analysis of scFv binding to CHO cells.Flow cytometry analysis of scFv binding to two CHO cell lines: (A) XCR1 + and (B) XCR1 -.Representative plots showing binding to cells after 30 min incubation with AF647-labelled anti-XCR1 scFv, anti-XCR1 IgG, and isotype IgG constructs.

Fig. S8 .
Fig. S8.Protein translocation into CHO cells.Relative cell viability from translocated LFN-DTA into (top) XCR1 + and (bottom) XCR1 -CHO cells.Cells were incubated (72 h) with ten-fold serial dilutions of LFN-DTA in the presence of 20 nM PA, scFv-mPAC, or scFv-mPAC[F427A].Relative viability (% viable cells) was determined from a Cell Titer-Glo assay; viability was normalized to untreated cells.Data represent the mean of three replicate wells ± the standard deviation (s.d.).Data are representative of two independent experiments.

Fig. S9 .
Fig. S9.Flow cytometry analysis of scFv binding to murine splenocytes.(A) 5carboxyfluorescein (FAM) with a G3 linker peptide.(B) Sortase-mediated ligation of the anti-XCR1 scFv and FAM.(C) Fluorescent panel for identifying cross-presenting dendritic cells in murine splenocytes by flow cytometry.(G) Flow cytometry results showing binding to DCs, medullary macrophages, and T cells and B cells from murine splenocytes treated with FAMlabeled scFv and a diphtheria toxin protein (DTA, negative control).Data are representative of two independent experiments.

Fig. S10 .
Fig. S10.Flow cytometry gating on murine splenocytes (scFv-FAM).(A-J) Flow cytometry analysis of murine splenocyte populations after 30 min incubation with FAM-labeled scFv.(F, H, I) Analysis of FAM-positive populations.Data are representative of two independent experiments.

Fig. S11 .
Fig. S11.Flow cytometry gating on murine splenocytes (DTA-FAM).(A-J) Flow cytometry analysis of murine splenocyte populations after 30 min incubation with FAM-labeled DTA (negative control).(F, H, I) Analysis of FAM-positive populations.Data are representative of two independent experiments.

Fig. S12 .
Fig.S12.LC-MS analysis of the LFN-OVA252-270 (GGGLEQLESIINFEKLTEWTSS) conjugate.The protein (50 ng) was loaded onto an Agilent Zorbax 5 μm 300SB-C3 column (2.1 × 150 mm) and was eluted with a gradient of 1-91% CH3CN in H2O with 0.1% FA and a flow rate of 0.5 mL/min.The protein was detected on an Agilent 6550 ESI-Q-TOF mass spectrometer.The TIC peak was integrated and the mass was deconvoluted using maximum entropy algorithm.

Fig. S14 .
Fig. S14.Flow cytometry analysis of antigen presentation with DC2.4 cells.Flow cytometry analysis with DC 2.4 cells of MHC class I molecule Kb bound to the peptide SIINFEKL.Data was obtained at 6, 24, and 72 h post-treatment with either 20 nM PA, 20 nM PA + 1 µM LFN-SIINFEKL, or 1 µM SIINFEKL.

Fig. S18 .
Fig.S18.LC-MS analysis of LFN-Trp1-gp100 conjugate.The protein (50 ng) was loaded onto an Agilent Zorbax 5 μm 300SB-C3 column (2.1 × 150 mm) and was eluted with a gradient of 1-91% CH3CN in H2O with 0.1% FA and a flow rate of 0.5 mL/min.The protein was detected on an Agilent 6550 ESI-Q-TOF mass spectrometer.The TIC peak was integrated and the mass was deconvoluted using maximum entropy algorithm.

Fig. S21 .
Fig. S21.Phenotype of B16-F10 tumors.(A) Comparison between the control mice (naive group) and the mice in the vaccine group (LFN-Trp1-gp100/ ScFv-mPAC), after the 3 rd vaccine injection.(B) Observation of a representative mouse, among the 30% of mice from the vaccine group (LFN-Trp1-gp100/ ScFv-mPAC) with partially pigmented tumors, at Day 22 (before the 4 th vaccine injection) and Day 24 (after the 4 th vaccine injection).The white arrows indicate the localization of the tumor, the white circle indicates its circumference.

Table S1 .
. Amino acid sequence and codon optimized gene for the anti-XCR1 scFv.
1Hydrophilic spacer sequence (G 4 S) 4 is indicated by bold text with one line underneath, as shown here.2Sortaserecognition site is indicated by bold text with two lines underneath, as shown here.

Table S2 .
Summary of long antigen peptide sequences, containing N-terminal Gly residues for sortase-mediated ligation to LFN.

Table S3 .
Summary of short epitope peptide sequences.