Nanoparticle-Mediated In Situ Molecular Reprogramming of Immune Checkpoint Interactions for Cancer Immunotherapy

Immune checkpoint blockade involves targeting immune regulatory molecules with antibodies. Preclinically, complex multiantibody regimes of both inhibitory and stimulatory targets are a promising candidate for the next generation of immunotherapy. However, in this setting, the antibody platform may be limited due to excessive toxicity caused by off target effects as a result of systemic administration. RNA can be used as an alternate to antibodies as it can both downregulate immunosuppressive checkpoints (siRNA) or induce expression of immunostimulatory checkpoints (mRNA). In this study, we demonstrate that the combination of both siRNA and mRNA in a single formulation can simultaneously knockdown and induce expression of immune checkpoint targets, thereby reprogramming the tumor microenvironment from immunosuppressive to immunostimulatory phenotype. To achieve this, RNA constructs were synthesized and formulated into stable nucleic acid lipid nanoparticles (SNALPs); the SNALPs produced were 140–150 nm in size with >80% loading efficiency. SNALPs could transfect macrophages and B16F10 cells in vitro resulting in 75% knockdown of inhibitory checkpoint (PDL1) expression and simultaneously express high levels of stimulatory checkpoint (OX40L) with minimal toxicity. Intratumoral treatment with the proposed formulation resulted in statistically reduced tumor growth, a greater density of CD4+ and CD8+ infiltrates in the tumor, and immune activation within tumor-draining lymph nodes. These data suggest that a single RNA-based formulation can successfully reprogram multiple immune checkpoint interactions on a cellular level. Such a candidate may be able to replace future immune checkpoint therapeutic regimes composed of both stimulatory- and inhibitory-receptor-targeting antibodies.


Supplementary Methods
In vitro transfection of B16F10 using PEI transfection reagent B16F10 cells were seeded into a 12-well plate at a density of 120xK/well in fully supplemented RPMI-1640 medium 24xh before the transfection and incubated at 37°C, 5% CO2. PEI transfection was performed according to manufacturer's protocol. The PEI reagent (3 μL/condition) was mixed with siRNA (1.18 μg/ condition) and/or pDNA (0.8 μg/ condition) in the jetPRIME buffer (100 μL/condition).
The mixture was incubated for 15 min at room temperature to form PEI/siRNA-pDNA complexes before being added to cells in 1 mL of complete media and incubated for 48h at 37°C, 5% CO2.

In vivo target validation using PEI complexes
B16F10 cells 10 6 were implanted subcutaneously into the lower flank of C57BL/6 mice in 100 μL PBS.
PEI/ siRNA-pDNA complexes were prepared by mixing 1.6 μg plasmid (pOX40L or pNeg) and/or 1.18 μg siRNA (siPDL1 or siNeg) with 3 μg PEI and jetPRIME buffer, diluted to a total volume of 50 μL, and incubated for 15 mins. On day 5 post implantation, once palpable, tumours were i.t. injected with 50 μL complex formulation per tumour. Additional doses were administered at days 7 and 11 post implantation. Tumour size and mouse weight were monitored until mice reached their humane end points.

In vitro SNALP mLuc transfection of B16F10
SNALPs containing mLuc were prepared as previously described (see: Preparation of SNALPs). B16F10 cells were cultured and transfected as described in the relevant methods section (see: In vitro SNALP transfection of B16F10 cell lines). To establish the effect of serum during transfection, cells were transfected in the presence or absence of 10% v/v FCS for 4 h. An additional, untransfected group, was included to normalize the background signal. After elapsed time, conditions lacking serum were supplemented with FCS to final concentration 10% v/v; conditions with serum received complete media in equivalent volume. Following 48h incubation, luciferase assay was performed as described in manufacturers' data sheet. In brief: culture media was removed, and cells were gently washed with PBS, 200 µL of 1x reporter lysis buffer was added to each well before being freeze thawed 3 times.
Obtained lysate was clarified by centrifugation (14000 rpm, 5 mins). The quantity of luciferase was detected by adding 100 µL of luciferase to 20 µL of cell lysate and being read immediately using a plate reader (BMG LABTECH, FLUOstar® Omega). Data is presented as relative light units (RLU) for each condition with the background signal obtained from untransfected wells subtracted.

Supplementary Data
Supplementary   B16F10 cells were cultured until 90% confluent before being (A) transfected with either empty plasmid (pNeg) or plasmid containing OX40L ORF (pOX40L) or (B) with either nonspecific siRNA (siNeg) or siRNA specific for PD-L1 (siPDL1) in PEI. Cells were left untransfected (control) as negative controls. In all conditions siRNA was fixed at 88nM and pDNA fixed at 0.8 µg.
Following 48 hours cells were harvested and stained with fluorescently labelled anti mouse PDL1 or anti mouse OX40L monoclonal antibodies and acquired by flow cytometry, cells were first gated on their FSC/SSC profile before PD-L1 or OX40L expression was analysed. For PD-L1 silencing values are expressed as MFI percentage of control which was normalised to 100%. In each case mean and SEM is shown (n=3 experimental repeats). (C,D) C57/Bl6 mice were subcutaneously implanted with B16F10 tumour cells (1x10 6 /per tumour) (n=8). Tumours were allowed to develop until palpable (day 5) at which point each tumour was injected with either Buffer, siNeg/pNeg, siPDL1 (siPDL1), pOX40L (pOX40L) or siPDL1/pOX40L. Tumour growth was monitored and subsequent doses were administered at days 7 and 11 in accordance with (C). The tumour growth rate is shown in (D). Mean and SEM is shown in each case, statistical analysis was carried out using students T test *p<0.05, **p<0.005, ns non-significant. SNALPs were formulated with siRNA, mRNA or a combination of both RNA molecules as previously described. SNALPs were dropcasted on to a graphene grid and imaging was performed using a Tecnai Osiris transmission electron microscope. Images represent either a single event or a region of interest representative of the wider field. SNALPs have shown an irregular structure with evidence of internal concentric rings. Once the tumours reached c.5 mm in diameter (day 8) and at days 10 and 14 post implantation mice were injected i.t. with SNALPs containing both mOX40L and siPDL1 (13 µg, siPDL1-mOX40L), or injected with buffer (Control). The tumour growth was recorded until the humane end points were met. For each group the data is presented as a spaghetti plot for individual mice (A-B). Mouse survival over the course of the experiment is plotted as a Kaplan-Meier plot (C). Tumour size data is presented as mean ± SEM. Survival data was analysed using a Mantel-Cox test, ns nonsignificant.