Biomimetically Engineered Demi‐Bacteria Potentiate Vaccination against Cancer

Abstract Failure in enhancing antigen immunogenicity has limited the development of cancer vaccine. Inspired by effective immune responses toward microorganisms, demi‐bacteria (DB) from Bacillus are engineered as carriers for cancer vaccines. The explored hydrothermal treatment enables the Bacillus to preserve optimal pathogen morphology with intrinsic mannose receptor agonist. Meanwhile, the treated Bacillus can be further endowed with ideal hollow/porous structure for efficient accommodation of antigen and adjuvant, such as CpG. Therefore, this optimal engineered nanoarchitecture allows multiple immunostimulatory elements integrate in a pattern closely resembling that of bacterial pathogens. Such pathogen mimicry greatly enhances antigen uptake and cross‐presentation, resulting in stronger immune activation suitable for cancer vaccines. Indeed, DB‐based biomimetic vaccination in mice induces synergistic cellular and humoral immune responses, achieving potent therapeutic and preventive effects against cancer. Application of microorganism‐sourced materials thus presents new opportunities for potent cancer therapy.


Experimental Section
Characterization of Surface Properties.
Zeta potential of the 4 kinds of hydrothermal treated bacteria was determined by ZetaSizer (Nanoseries, Malvern) in pure water and the surface contact angle was measured by using contact angle meter (OCA-20, DataPhysics).

Evaluation of Loading Efficiency and Release behaviour for OVA and CpG.
For the encapsulation studies, CpG or OVA at certain concentrations was mixed with 50 μg DB in 100 μL PBS at 4 o C for 12 h after vacuum negative pressure treated. Free CpG or OVA was washed away by three centrifugation-redispersion cycles, with all supernatants reserved. The loading amount of CpG or OVA was determined by mass balance between the initial incubation solution and the reserved supernatant. For in vitro release studies, OVA or CpG-loaded DB was incubated in 1 mL PBS (pH 7.4) under agitation at 37 °C. Supernatants were periodically collected by centrifugation and replaced with fresh buffer of equal volume.
The CpG concentration was determined using an Infinite M200 microplate spectrophotometer and a NanoQuant Plate (Tecan), and the OVA concentration was determined by a BCA protein determination kit (Thermo).

Culture of DCs.
DCs were generated by flushing tibia and femurs of 6-8 weeks old male C57BL/6 mice.

Observation of Capture of DB at the DC Membrane.
For SEM imaging, DCs were adhered to poly-L-lysine-coated glass slides, and 2.5 μg mL -1 of DB was then added for 1 h of incubation. The cells were washed thoroughly immediately afterward to remove free DB. Next, the samples were fixed in 2.5% (v/v) glutaraldehyde, dehydrated with gradient methyl cyanide solutions, and gold-sprayed for imaging under a field emission SEM (JEOL JSM6700F).

Intracellular Trafficking of DB.
DCs were incubated with DB for 1 h at 37 o C and washed to remove free DB, followed by further incubation for another 6 h to ensure the synchronisation of intracellular processes.
For early endosome staining, the cells were fixed in 4% (v/v) formaldehyde and permeabilized via a 10 min incubation with 0.2% (v/v) Triton X-100 at room temperature.
Subsequently, 5% (v/v) goat serum was added and incubated for 30 min to block nonspecific adsorption. Staining with a polyclonal rabbit anti-murine EEA1 antibody was performed at 4 °C for 24 h, followed by 30 min of incubation with Texas Red-goat anti-rabbit IgG at room temperature. For lysosome staining, DCs were directly stained with LysoTracker Red for 30 min at 37 o C, without cell fixation and permeabilization. The colocalization ratios (CRs) of DB with early endosomes or liposomes were calculated using Leica LAS Colocalization software.

Preparation and Characterization of Traditional Particulate Adjuvants.
Liposomes were prepared by thin-film hydration method. [1] Briefly, a mixture of 1palmitoyl-2-oleoylphosphatidylcholine (POPC) and cholesterol (in molecular ratio 1:2) were dissolved in chloroform, and a thin-film was formed on the inner side of the round bottom flask by evaporating the solvent under vacuum using a rotavapor. The film was then hydrated using 10 mL of PBS containing 50 ng mL -1 water-soluble Cy5. The formed liposomes were then sonicated using a probe sonicator for 2 min, and then sized by extrusion through a 0.4 μm pore-sized polycarbonate membranes repeatedly. The un-encapsulated Cy5 was then removed by dialysis. PLA NPs and Chitosan NPs were prepared using the Shirasu Porous Glass (SPG) membrane emulsification technique. [2] By selecting a SPG membrane with a certain pore size, the size of the emulsion droplets and subsequent NPs can be well controlled around 500 nm.
To prepare the fluorescent NPs, oil-soluble Cy5 was added together with PLA to the oil phase before the emulsification, free Cy5 was then removed by utterly washed with water. The zeta potential and size distribution were determined by Zetasizer (Nanoseries, Malvrn).

Evaluation of DC Activation and Antigen Presentation.
In total, 1×10 6 cell mL -1 DCs were incubated with different formulations for 24 h in 24well plates. At the end of the incubation, the DCs were collected by centrifugation (500 g, 3 min) and stained with antibody for 30 min at 4 o C. The following fluorophore-conjugated antibody reagents (purchased from BioLegend and eBioscience) were used: APC-Cy7-CD11c, PE-CD40, FITC-CD80, PE-CD86, PerCP-eFluor 450-MHC II, and APC-SIINFEKL-MHC I.
Subsequently, the cells were washed with PBS, and the levels of surface costimulatory molecules and recognition signals for T cell activation were analysed by FC. Otherwise, to monitor the cytokine profile of the DCs, cell culture supernatants were collected, and the secretion levels of IL-6, IL-12, TNF-α, MCP-1, and IFN- were detected using a CBA Mouse Inflammation Kit (BD) to quantify. [3] Briefly, capture beads and phycoerythrin (PE) detection reagent were incubated with standard samples or test samples for 2 hours, washed in wash buffer. After acquisition of sample data using the flow cytometer, the sample results were tabulated and graphed using the BD CBA Analysis Software.

Analysis of PRR mRNA Expression.
RNA was isolated with TotalRNAExtractor (Sangon Biotech, Shanghai, China) from naive DCs and DC treated by DB, CpG, or DB:CpG for 24 h. RNA samples were then DNase I digested, reverse transcribed with iScript cDNA Synthesis Kit (Bio-Rad Laboratories, Hercules, Canada), and analysed by qPCR with Ssofast EVI Green Supermix (Bio-Rad Laboratories, Hercules, Canada). Data was displayed as relative expression compared to GAPDH. PCR products were sequence verified. The sequences of primers for target genes and internal control gene GAPDH are listed in the table below.

In-vivo Fate of Antigen.
To by Vectra platform and the recruited cells were quantitated by using the inform software (Caliper Life Sciences, Hopkinton, USA). [4] Analysis of DC Migration to Lymph Nodes.
To track in vivo DC migration from injection site towards lymph nodes, the draining lymph nodes (LNs) were harvested at 24 h after immunisation. Cell suspensions from LNs were prepared by mechanical disruption and pressing of the tissue through 70 μm cell strainers. The CD11c + SIINFEKL-MHC I + cell numbers among CD11c + cells (DCs) were examined after staining by APC-Cy7-CD11c and APC-SIINFEKL-MHC I mAbs.

Preparation of Splenocytes and CD8 T Cells.
Spleen of 6-8 weeks old male C57BL/6 or OT-1 mice was harvested and ground to prepare single-cell suspension. After red blood cell lysis, cells were washed and cultured in complete RPMI 1640 medium containing 10% (v/v) heat-inactivated FBS. For CD8 T cell enriching, splenocytes were sorted by Dynabeads (Life technologies) to deplete CD8-negative leucocytes, and CD8T with purity around 90% could be harvested.

Evaluation of Health Condition.
To examine the T cell-subset distribution after the treatment of developed tumors, mice

Quantitative Determination of CTL and Treg Cell in Tumor.
To  For all groups, the initial particle number in culture medium was normalized to that of DB (2.5 μg mL -1 ).
(c) Comparative internalization profiles of free, DB-mixed, and DB-loaded CpG and OVA. DC uptake of both guest agents was found to be greatly improved by encapsulation in DB, enhanced intracellular delivery for both CpG and OVA could be achieved simultaneously.
(d) Effect of the MRs on the uptake of DB by DCs. Relative DC uptake of DB prepared via different hydrothermal treatment times was shown in left, to determine the proportion of MRmediated endocytosis, mannose (200 μg mL -1 ) was added to block the MRs 1 h before coincubation. The data of MR-blocked group in were normalized to the internalization in the corresponding control group (left). These results confirmed that the MR ligand on 6hDB was sufficient to significantly promote the interaction with APCs. Data in all groups represent the mean ± s.d. of three independent experiments with n=3. *p<0.05, **p<0.01.    Mice with established tumors were treated with different formulations, blood samples were collected before scarification on day 23 (for developed tumor therapy), and the serum AST, ALT, BUN, LDH, and ALP levels were determined. The data are represented as the mean ± s.d. with n=7. Among these parameters, AST and ALT are specific indicators of hepatic toxicity, and LDH is a parameter related to injury of major organs, including the liver, heart, and kidney. In contrast with the abnormally enhanced levels in the other groups, the serum AST and ALT concentrations in the 2×DB:CpG/OVA group returned to normal ranges. A similar result was observed for the LDH levels, which indicated that treatment with the biomimetic vaccine had effectively protected the mice from hepatic or other organ damage. Together, all these data showed that mice that received the biomimetic vaccine were in much better health and had a better prognosis.