CORD-19:123db7cdec9fc63aa09782ea9f4267603be8f74d JSONTXT 8 Projects

DNA-scaffolded biomaterials enable modular and tunable control of cell-based cancer immunotherapies Abstract Advanced biomaterials provide versatile ways to spatially and temporally control immune cell activity, potentially enhancing their therapeutic potency and safety. Precise cell modulation demands multi-modal display of functional proteins with controlled densities on biomaterials. Here, we develop an artificial immune cell engager (AICE) platform -biodegradable particles onto which multiple proteins are densely loaded with ratiometric control via short nucleic acid tethers. We demonstrate the impact of AICE with varying ratios of anti-CD3 and anti-CD28 antibodies on ex vivo expansion of human primary T cells. We also show that AICE can be used to control the activity of engineered T cells in vivo. AICE injected intratumorally can provide a local priming signal for systemically administered AND-gate chimeric antigen receptor T cells, driving local tumor clearance while sparing uninjected tumors that model potentially cross-reactive healthy tissues. This modularly functionalized biomaterial thus provides a flexible platform to achieve sophisticated control over cell-based immunotherapies. Functionalized biomaterials can work synergistically with natural or engineered immune cells for cancer immunotherapy 1-15 . Adoptive cell therapy (ACT) using chimeric antigen receptor (CAR) engineered T cells has shown success and clinical approval for the treatment of B cell cancers 16 . However, for CAR T cells to fulfill their promising potential, particularly for targeting solid tumors 17, 18 , important challenges must be overcome to improve both efficacy 17-20 and safety 21-24 . For engineered anti-tumor T cells to achieve durable tumor remission in patients, a demanding manufacturing process is required to ensure the quantity and quality of the cell product for therapeutic use 1, 25-27 . To increase the tumor targeting specificity and All rights reserved. No reuse allowed without permission. avoid "on-target, off-tumor" toxicity in bystander healthy tissues 28 , CAR-T cells have been engineered with combinatorial antigen AND-gate activation control that requires sensing two antigens on a target cell to initiate killing. The clinical application of this strategy, however, requires systematic identification of tumor specific antigen combinations and would also benefit from better control of cell activity 27, 29 . Biomaterials functionalized with modulatory biomolecules at pre-specified densities have been shown to communicate with and control therapeutic immune cells both ex vivo and in vivo [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] . For example, synthetic materials have been functionalized with agonistic antibodies for CD3 and CD28 to drive ex vivo T cell expansion 25, 26 . However, biodegradable materials 25 with optimized material structure/composition and controlled surface moiety loading can provide unique opportunities to improve the ease of manufacturing and the quantity and quality of T cells produced. There is also a growing interest in the use of biomaterials for local modulation of engineered T cell activity in vivo during the course of treatment 1 . Nonetheless, the use of biomaterials to precisely modulate immune cells still faces key challenges. There is a need for robust chemical conjugation strategies to surfacefunctionalize biodegradable materials with multiple signals (e.g. proteins/antibodies) at high densities and precisely controlled ratios [30] [31] [32] [33] . PEG (polyethylene glycol) is commonly used as a linker or scaffold for the surface conjugation of biomolecules 31 , but this method is limited by the inefficient presentation of functional groups for attaching biomolecules due to PEG's flexibility 34 . Synthetic short oligonucleotidesnatural polymers with controllable sequence and structure -have been used as surface scaffolds on metallic particles for siRNA delivery 34-37 , but have yet to be fully utilized for modular payload assembly on biodegradable materials. In this work, we developed short synthetic DNA scaffolds for the efficient and versatile functionalization of proteins or antibodies on the surface of biodegradable particles (Fig. 1a) . A series of optimization studies were carried out to achieve maximum loading density, ratiometric control of moiety loading, adaptability to different particle size/composition, and feasibility for in vivo use. Micron-sized artificial immune cell engagers (AICE) were made from biocompatible poly(lactic-coglycolic acid) (PLGA) polymer 38 (Fig. 1a) . We demonstrated that: i) we can load a range of immune modulators (e.g. anti-CD3, anti-CD28, and IL-2) on AICE at controlled ratios and densities; ii) the ratiometric control of costimulatory ligands is essential for optimal ex vivo T cell activation and expansion with higher yield and less exhaustion; iii) we can locally administer antigen-presenting AICE to control ANDgate CAR-T cell activation and tumor clearance in vivo (Fig. 1a) . This modular materials-based strategy can provide versatile and precise synthetic control of natural or engineered immune cells for cancer immunotherapy 1, 12, 29 . Synthetic short DNA strands were immobilized on the surface of polymeric particles to create an adaptable scaffold for loading bioactive molecules. The emulsion protocol of particle synthesis 38 was modified to add conjugated polymer-DNA amphiphilic molecules during the fabrication process such that the hydrophilic DNA segment distributes around the hydrophobic core upon sonication (Fig. 1b) . Polymer-DNA molecules were first synthesized through a conjugation of thiol-modified DNA to Maleimide (Mal)-modified polymer. We optimized the polymer, DNA length, solvent, and reaction conditions to generate optimal polymer-DNA molecules to form stable PLGA (50:50, MW 38-54k, Sigma) particles with dense DNA scaffolds (Fig. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint Supplementary 1a-d). PLGA10k-PEG5k-Mal was identified as an important polymer-DNA component to ultimately form microparticles of the desired size (1-2 μm in diameter), so we kept the polymer constant while varying the excess of DNA input during the conjugation reaction and observed an increase in polymer coupling efficiency (Fig. 1c) . The direct incorporation of polymer-DNA reaction mixtures with different DNA:polymer ratios into the particle fabrication protocol yielded a significant increase of surface payload-attachable DNA scaffold density, as determined by a fluorescence-based hybridization analysis (Fig. 1d) . Strikingly, the highest average surface loading density on particles (~5 million DNA duplexes per particle, Fig. Supplementary 1e-g) was roughly analogous to the theoretical limit (at ~4 million by footprint calculation, based on ~2 nm diameter of DNA duplex) of a spherical particle with a 2 μm diameter. In particular, this hybridization-guided loading was about 27fold more efficient than that from the traditional method 33 , in which thiol-modified DNA molecules were conjugated to surface-exposed Mal groups after particle fabrication using equal input amount of PLGA10k-PEG5k-Mal (Fig. 1d) . The hybridization-guided biomolecule assembly protocol was optimized to a 30-minute incubation at 37 o C ( Fig. Supplementary 1h) , and functionalized particles can be further lyophilized for storage and transportation (Fig. Supplementary 1i) . To accommodate diverse applications from intracellular payload delivery to extracellular signal transduction, DNA-scaffolded particles can be synthesized at various sizes and also loaded with biomolecules in their core. Particle size can be controlled by tuning key emulsion parameters ( Fig. 1e and Supplementary Fig. 1j ). We observed a concave shape on one side for some larger particles (~2 μm in diameter) but not for smaller particles (~500 nm) (Fig. 1f) possibly due to the high All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint surface hydrophilicity with DNA scaffolds leading to higher surface to volume ratio. Additionally, we showed that the core of these DNA-scaffolded microparticles could also be loaded with biomolecules including oligonucleotides and peptides through a double-emulsion protocol ( Fig. 1g and Supplementary Fig. 1k ). This strategy of fabricating DNA-scaffolded particles was also successfully replicated using other polymers such as polylactic acid (PLA) (Fig. 1h and Fig. Supplementary 1l,m) . Biomolecules can be loaded on the surface of DNA-scaffolded particles using one of two strategies: a) linking functional groups of the scaffolds to biomolecules through the surface step-by-step conjugation via a bifunctional linker; and b) through the direct hybridization of complementary DNA-biomolecule conjugates to the scaffold ( Fig. 2a) . In the case of a fluorescently labeled human IgG, the surface conjugation strategy resulted in saturated protein loading even with increased densities of available DNA linkers on surface (Fig 2b, Fig 1d and Fig. Supplementary 2a) . In contrast, the hybridization-guided assembly strategy showed a dramatically higher level of IgG loading for particles with denser DNA linkers, and the increase in IgG loading corresponded to the scaffold density (Fig. 2b) . The highest level of IgG loading achieved (at ~0.6 million per particle) was again comparable to the theoretical footprint limit of a spherical particle at 2 μm in diameter (~0.64 million per particle, based on ~5 nm diameter of IgG) (Fig. 2b) . For generating DNA-antibody conjugates with the minimum damage to the antibody activity, different chemistries ( Fig. Supplementary 2b,c) were attempted and the "TCEP" strategy 6 was identified as optimal, whereby the antibody hinge region was selectively reduced to expose the thiol group for 3'NH 2 -modified complimentary DNA to attach through a MAL-dPEG4-NHS linker (Quanta Biodesign). For example, anti-PD-L1 antibody All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint conjugated with DNA through this strategy showed intact binding activity ( Fig. Supplementary 2c) , and its loading on PLGA microparticles resulted in high binding specificity for PD-L1 positive cells (Fig. 2c) . For DNA tethering of proteins other than antibodies, other options are available to yield products with intact activity (Fig. Supplementary 2d,e) . Another advantageous feature of using DNA as a surface scaffold is that the uniqueness of nucleotide sequences enables the independent control of loading multiple cargos on the same material surface (Fig 2d) . Polymer-DNA conjugates with three distinct DNA sequences (namely R, G, and B strand) that do not anneal with each other (Fig. Supplementary 2f) were synthesized and incorporated into particle emulsion with reaction mixtures at different ratios (3:1:1, 1:1:1 and 1:1:3). After surface hybridization with a mixture of their respective dye-labeled complementary strands (Fig. 2d) , we found that the ratios of hybridized DNA scaffolds of different sequences on PLGA microparticles were consistent with the polymer-DNA conjugate input, evidenced by both confocal fluorescence imaging (Fig. 2e) and a fluorescencebased quantification assay (Fig. 2f) . The ratiometric control of dye-labeled DNA strand loading was equally efficient for PLA microparticles (Fig. Supplementary 2g ,h). Based on this, we co-loaded three different proteins (GFP and two antibodies), each individually attached with one of the three complementary DNA strands, onto PLGA microparticles with varying ratios of the DNA surface scaffold sequences above. Similar distribution of each protein cargo compared with the scaffold population was observed (Fig. 2g) , demonstrating the robust ratiometric control of surface protein loading using this particle synthesis strategy. To enable in vivo use, we explored the stability of the DNA scaffolds on the surface of PLGA microparticles in the presence of a large excess of DNase. (Fig. 2h) . We All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint observed an almost complete (~87%) degradation of DNA scaffolds after DNase treatment; however, this dropped down to ~20% with IgG molecules conjugated to the scaffolds through the surface step-by-step conjugation (~1/4 of the highest IgG loading through hybridization-guided assembly) (Fig. 2b,i) . This suggests that the surface-loaded proteins are able to protect the DNA scaffolds. In human serum, we did not observe as significant DNA degradation for the naked particles due to the lower DNase concentration in serum than used in Fig.2i ; however, the degree of DNA degradation similarly decreased as surface payload proteins were attached (Fig. 2j) . Macrophage uptake is a major obstacle 39 for particle-based drug delivery in vivo, therefore we incorporated a CD47-mimic "self"-peptide 40 on PLGA microparticles ( Fig. 2k) . Fluorescein-labeled PLGA microparticles with and without surface-loaded "self"-peptide were co-incubated with a mouse macrophage line J774A.1 in vitro. The particle uptake assay showed that the "self"-peptide decoration significantly reduced the uptake of particles by macrophages ( Fig. 2k and Supplementary Fig 2i,j) . This "self"-peptide functionalization provides a possible strategy to reduce the clearance of particles in vivo. Given the ability to synthesize particles with ratiometric control of multiple surface cargos, we loaded anti-CD3 and anti-CD28 antibodies on AICE (PLGA microparticles) at varying ratios from 1:5 to 5:1. Anti-CD3/CD28 AICE were compared to commercial anti-CD3/CD28 Dynabeads at the same particle to cell ratio for the ability to expand human primary T cells with minimized differentiation and exhaustion (Fig. 3a) . Functionalized AICE activated T cells (Fig. 3b) and yielded higher or equivalent T cell expansion compared to Dynabeads expansion across three All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint human T cell donors (Fig. 3c,d) . Although there were large donor-to-donor differences (also observed for Dynabeads), we observed a linear increase of cell yield from AICE-[1:5, anti-CD3:anti-CD28] to AICE-[3:1] for both CD4+ and CD8+ cells at day 14 ( Fig. 3c and Fig. Supplementary 3a) . The phenotype of expanded T cells at day 14 was then explored by measuring expression of CD45RA and CCR7 surface markers (Fig. 3e) . Interestingly, the population distribution of the 4 differentiation states (naïve, central memory [CM], effector memory [EM] and effector memory RA [EMRA]), displayed a pattern ( Fig. 3f and Fig. Supplementary 3b) corresponding to the cell expansion trend among AICE with different anti-CD3 to anti-CD28 ratios ( Fig. 3c) . Through staining for T cell exhaustion markers LAG-3, PD-1 and TIM-3, we found that the population of exhausted cells at the optimal condition AICE- [3:1] was generally less than those activated by Dynabeads among the three donors ( Fig. Supplementary 3c,d) . Taken together, the data demonstrates that the surface ratio control of functional moieties on synthetic materials is important for the quantity and phenotype of cells yielded from ex vivo T cell expansion. As an alternative to the standard protocol of supplementing free IL-2 in the media for ex vivo T cell culture 25 , we loaded IL-2 on AICE through the surface presentation via its antibody clone #5355 (Fig. Supplementary 3e) . This particular anti-IL-2 antibody was engineered to facilitate the binding of IL-2 to its β and γ receptor on T cells thus promoting the proliferation of non-Treg T cells 41 (Fig. 3g and Fig. Supplementary 3f ). Using the optimal condition of AICE[3:1] (anti-CD3:anti-CD28) determined above, we compared primary T cell expansion by supplementing equal amount of free IL-2 versus surface bound IL-2 ( Fig. Supplementary 3e) . Surface IL-2 loading enhanced CD4+ and CD8+ T cell expansion, and particularly improved expansion of CD4+ primary T cells after day 8 (Fig. 3h,i and The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint 3a,g). Although with higher level of cell yield, the populations of LAG-3 and PD-1 positive CD4+ T cells still decreased in the two tested donors (Fig. Supplementary 3h ), demonstrating the advantages of AICE-surface presentation of IL-2 for ex vivo expansion. We also explored the use of AICE coated with an orthogonal antigen (GFP) to prime AND-gate CAR-T cell tumor recognition circuits, with the goal of preventing "ontarget off-tumor" toxicity by restricting cell killing to tumors locally-injected with AICE microparticles (Fig 4a) . These AND-gate T cells utilize a modular synthetic Notch (synNotch) receptor with an extracellular domain to recognize a target antigen, and an intracellular transcriptional activator (TF) domain to control expression of a CAR targeting a second antigen 28 . Killing is only induced when both antigens are presented to T cells, with one activating the synNotch receptor to release the TF domain for CAR expression and the other activating CAR-mediated cytotoxicity 28 . Herein, human primary T cells were transduced with a gene encoding a constitutivelyexpressed synNotch receptor with an extracellular anti-GFP nanobody and an intracellular Gal4DBD-VP64 synthetic transcription factor (TF) domain, together with a TF-inducible CAR gene targeting HER2 (human epithelial growth factor receptor 2) 28 (Fig. 4a) . In this system, T cells should only express the anti-HER2 CAR and kill HER2 positive target cells after priming via the anti-GFP synNotch binding AICE-presented GFP. Human primary CD8+ T cells engineered with the above circuit were co-cultured for 24 hours with single antigen target cells (human melanoma cell line A375 with HER2 overexpressed) in the presence or absence of GFP+ AICE microparticles. SynNotch/CAR T cells showed selective elevation of the early activation marker All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint CD69 in the presence of both HER2+ A375 cells and GFP+ AICE, and the activation was similar to T cells exposed to dual antigen target cells (human leukemia cell line K562 with GFP and HER2 co-expressed) (Fig. 4b) . For both CD4+ and CD8+ engineered primary T cells, we observed selective secretion of cytokines (IL-2 and IFN-γ, respectively) after 48 hours of co-culture with both HER2+ A375 cells and GFP+ AICE (Fig. 4c) . Increasing AICE GFP density led to elevated levels of cytokine secretion (Fig. 4c, Fig. Supplementary 4a) . Noticeably, T cells secreted significantly more cytokines with AICE presenting synNotch-priming GFP than with target cells presenting both synNotch and CAR antigens (A375 with GFP and HER2 co-expressed) (Fig. Supplementary 4b) . Human primary CD8+ T cells containing the same circuit as above also showed AND-gate killing behavior, selectively killing HER2+ A375 cells in the presence of GFP+ AICE microparticles. Co-incubation of GFP+ AICE and CD8+ human primary T cells led to some killing of co-cultured HER2+ A375 cells (with 1:1 ratio) at 24 hours (Fig. 4d,e) , and the killing became more apparent after 48 hours with higher densities of priming signals on AICE ( Fig. 4f and Fig. Supplementary 4c-f ). In contrast, initial attempts using particles synthesized by traditional conjugation chemistry and with far lower GFP density failed to activate AND-gate CAR-T cells, likely due to the inadequate ligand density for robust synNotch activation. As a demonstration of adaptability for translational use, a synthetic short peptide (PNE peptide) computationally predicted to be non-immunogenic 42 was engineered on AICE and co-incubated with its corresponding anti-PNE synNotch/anti-HER2 CAR CD8+ T cells and HER2+ target cells, and killing was significantly enhanced by PNE+ AICE addition (Fig. Supplementary 4g) . The co-incubation of target cells with synNotch CAR-T cells in the absence of AICE also showed moderate All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint cytotoxicity ( Fig. 4f and Fig. Supplementary 4c-g) , which is likely due to the leakiness of inducible CAR expression also observed from previous reports and can be improved through further optimization of synNotch receptors. To determine the feasibility of using local particle injections to send spatially-defined signals to therapeutic cells in vivo, we first monitored the distribution of intratumorally-injected AICE microparticles. We labeled surface DNA and polymer core with two distinct infrared dyes (Quasar705 and IR800CW respectively, We hypothesized that AICE made of GFP decorated PLGA microparticles could be injected intratumorally as a local activator for systemically administered anti-GFP synNotch/anti-HER2 CAR T cells (Fig. 5c) . NSG mice were implanted subcutaneously with the same HER2-overexpressed K562 xenograft tumors in bilateral flanks as a model for a tumor and healthy tissue both expressing the CAR antigen. After a week, mice were administered synNotch CAR-T cells once All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint intravenously in combination with AICE intratumorally injected for 4 doses (Fig. 5c) . The size of AICE-injected ipsilateral tumors decreased over time, in contrast to the contralateral tumors within the same mice without AICE injection and tumors in mice injected with AICE plus untransduced primary T cells ( Fig. 5d-g) . We also performed fluorescence microscopy on fixed tumor samples of mice sacrificed at an early timepoint and observed selective T cell infiltration in the AICE-injected tumor ( Fig. 5h and Supplementary 5f). These results demonstrate that AICE can provide a spatially controlled signal in vivo for the local activation of synNotch CAR-T cells and induction of precision tumor clearance with limited risk of cross-reaction against healthy tissues. In this work, we demonstrate a unique platform using synthetic short DNA oligonucleotides as surface scaffolds on biodegradable materials for the precise and controlled loading of multiple biomolecules at specific ratio and density. These advanced biomaterials possess the ability to regulate immune cells in the context of ex vivo and in vivo activation of both natural and engineered T cells. Through this highly modular platform, biocompatible materials can be employed as substrates for a wide spectrum of modulatory biomolecules, including cytokines 5 , antigens 10 , checkpoint inhibitors 43 , agonistic or antagonistic antibodies 44 , adjuvants 9, 11 , etc., to regulate the local environment and enhance the efficacy of immune cells in cancer immunotherapy 8 . Synthetic materials with tunable properties can enhance the pharmacokinetics of regulatory biomolecules in vivo 1, 6, 8 , and sometimes provide a necessary surface substrate for biomolecules to activate the cognate cell signaling 10, 25, 28 . Biomaterials fabricated with controlled size, shape, and composition -for example, microspheres, nanowires, and porous scaffolds -can also serve as a local adjuvant 2, 3, All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint 7, 13 for biomolecules or engineered immune cells to exert tumor targeted toxicity without systemic diffusion, thus minimizing the off-target side effects. This can make a particular difference for solid tumor treatment, where functionalized biomaterials could be administered locally or engrafted post-surgery 3, 17, 45 . Here we explored using a natural polymer, DNA, as the surface scaffold for loading biomolecules onto particles. We demonstrated, through near surface-saturated biomolecule loading and precise ratiometric control of moiety loading, that hybridization-based thermodynamics can overcome limitations of traditional surface conjugation such as reduced efficiency of multi-step reactions, steric hindrance, and decay of functional groups. The increased loading density of surface biomolecules was found to be necessary to make particles capable of presenting antigens to robustly activate synNotch receptors in primary human T cells, and may also significantly enhance immune cell signaling where ligand clustering is required 10, 46 . We also show the versatility of this strategy by fabricating particles of different sizes, as well as compositions with different degradation profiles. This method can be applied to other formats of biomaterial functionalization (e.g. scaffolds and nonspherical particles) where fragile biomolecules need to be presented at a material interface. The densely-packed surface DNA scaffolds and the payload attachments protected DNA linkers from enzymatic degradation, which shows great promise for in vivo use. This organized packing of the particle surface with a protein corona may also minimize interactions with serum proteins to prevent macrophage clearance 47 . Furthermore, particle uptake by macrophages can be attenuated through the immobilization of a CD47 mimic "self"-peptide 40 . Safety concerns about immunogenicity from the DNA scaffolds can be addressed through optimization of sequences and special modifications 48 . All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint The precise ratio of agonistic antibodies (anti-CD3 and anti-CD28) for T cell activation on AICE impacts ex vivo T expansion efficiency, as well as T cell differentiation and exhaustion phenotypes. The ability to accurately control the ratio of multi-cargo particle loading should enable identification of novel, precise combinations of modulatory signals that lead to diversified and improved T cell behavior. The use of biodegradable materials also omits the need for synthetic material removal prior to cell implantation. Cumulatively, these properties suggest that AICE could aid in manufacturing for ACT by improving the ease-of-production and fine-tuning the yield and phenotype of the therapeutic cell product. We and other groups have developed strategies to engineer T cells with multiple receptors to detect combinations of antigens [27] [28] [29] . All prior combinatorial antigen recognition systems have targeted combinations of antigens expressed on tumor cells and/or tumor-related cells 27, 29 . Here, we demonstrate a method to engineer T cells to recognize combinations of endogenous and orthogonal antigens presented by tumor cells and biocompatible materials, respectively. By incorporating material-presented antigen into combinatorial recognition circuits, we enable real-time controlled local activation of these circuits that can be achieved through dynamic particle dosing in living organisms/patients. Local injection of our particles serves to safely control receptor activation and downstream behavior of therapeutic cells in vivo. We demonstrated that synNotch-based combinatorial antigen recognition circuits can incorporate material-presented antigen sensing, and it is likely that other combinatorial antigen recognition systems, for example, chimeric costimulatory receptors (CCRs) 49 in AND-gates or inhibitory CARs (iCARs) 50 in NOT-gates, can as well. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint We believe that the true promise of the AICE platform will be realized by using particles to present multiple signals within the tumor microenvironment. AICE particles could kick-start tumor killing by triggering CAR expression by synNotch-CAR T cells. T cell activity within the tumor could be further augmented by cofunctionalization of AICE particles with additional cargos that go beyond a synNotch ligand, for instance, a costimulatory ligand and a checkpoint inhibitor tailored to circumvent the specific immune evasion mechanisms of a particular tumor type 17, 45 . Ultimately, it would be ideal to utilize intratumoral injection of AICE particles to initiate antigen-specific local immunity that could be converted into antigen spreading and systemic immunity by incorporating other cues onto/into the particles to stimulate innate immunity 18 . Methods, including statements of data availability and any associated accession codes and references, are available in the online version of the paper. All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint by one-way ANOVA and Tukey's tests. (k) Representative confocal microscope images of macrophages treated with particles with and without "Self"-peptide presentation (n = 3 biologically independent samples per group). The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint Fabrication of polymeric particles with DNA scaffolds For the fabrication of PLGA particles with varying sizes (Fig. Supplementary 1j) , dried polymer-DNA All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint (LifeTein) and 50 nmol 5'Cy3-labeled DNA-21mer (Bioresearch Technologies) dissolved in 50 μL PBS (phosphate buffered saline, pH7.0) was mixed into 0.5 mL ethyl acetate with dissolved PLGA (Sigma #719900), and probe-sonicated at 7-8 W for 5 x 5 s with 10 s intervals on ice. Immediately after this, amphiphilic polymer-DNA and aqueous buffer was added following the same protocol described above. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint minutes, followed by two washes (centrifugation at 400 g x 5 minutes and resuspension in PBS) for flow cytometry with a BD LSR II. Anti-PD-L1-DNA conjugate linked through the "TCEP" strategy was loaded on DNA-scaffolded PLGA microparticles. PD-L1 overexpressed and wildtype K562 cells at 1 million/mL were respectively added with particles at 0.3 OD550 and incubated for 30 minutes at 37 o C, followed by cell nuclei staining with Hoechst 33342 (ThermoFisher #62249) and the imaging using the spinning disk confocal fluorescence microscope (Nikon Yokogawa CSU-22). His-tag GFP were expressed by Escherichia coli BL21 (DE3) (Novagen) transduced with pRSET-EmGFP vector (ThermoFisher, #V35320) in E. coli expression medium (MagicMedia, Invitrogen #K6803), and extracted using cell lysis reagent (Sigma, #B7435) followed by the purification using nickel-nitrilotriacetic acid affinity chromatography (Invitrogen #R90115). The MAL-PEG4-NHS linker (Quanta Biodesign) was mixed with GFP at 30-fold molar excess, and incubated at 37°C for 1 hour followed by the size-exclusion chromatography to remove the excess. 3'Thiolmodified complementary DNA was decapped using the protocol described above, and reacted with modified GFP at 1:10 molar ratio at 37°C for 1 hour followed by 4°C overnight. The next day, GFP-DNA conjugates are purified using nickelnitrilotriacetic acid affinity chromatography to remove unconjugated DNA. Protein-DNA conjugates were analyzed through SDS-PAGE gel electrophoresis (Genscript #M42012L) with SYPRO Ruby dye staining (ThermoFisher #S21900). Ratiometric control of surface DNA scaffolds with different sequences and the co-loading of versatile payloads 3'Thiolated DNA with different sequences were synthesized, namely R, G, and B (Fig. Supplementary 2f) . Following the conjugation with PLGA10k-PEG5k-Mal (Akina, #AI053), polymer-DNA molecules All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint from different sequences were mixed at varying ratios and incorporated into the particle fabrication process described above (100 nmol for each reactant). Complementary strands to the DNA scaffolds that were pre-conjugated with small molecules (e.g. fluorescent dye or biotin, at 1 μM/OD550) or proteins (e.g. GFP or antibodies, at 180 nM/OD550), were incubated with particles at 37 o C for 30 minutes in PBS buffer with 600 mM Na + and 0.01% Tween-20 supplemented for the surface hybridization of payloads. The excess was removed by three washes. For the coloading of multiple cargos, the input proportion of each individual one with a specific sequence was consistent with its DNA-polymer counterpart input during the particle fabrication. For the loading of biotinylated biomolecules on 3'-biotinylated complementary DNA hybridized particles, a large excess of streptavidin (Prozyme #SA10) was added at 1.1 mg/mL per OD550 and incubated at room temperature for 30 mins, followed by three washes. Biotinylated antibody, protein or peptide was added at 180 nM/OD550 and incubated at room temperature for 30 mins to bind with surface streptavidin followed by three washes. Loading efficiencies of different cargos on-surface or in-core were quantified through the fluorescence-based assay of dye-labeled DNA strands (5'Quasar570-compR, 5'Quasar705-compG, 5'Quasar670-compB, Fig. Supplementary 2f) , peptides (C-terminus FITC-labeled peptide, LifeTein, LLC) or proteins (FITC-labeled human IgG, Sigma #SLBW7799), after etching of particles by dimethyl sulfoxide (DMSO) and the dilution with 9x volume of water. The fluorescence signal was detected by a plate reader (Tecan Spark) and analyzed using a calibration curve with the normalization from OD550. Functionalized microparticles with fluorescent labels were imaged through a spinning disk confocal fluorescence microscope (Nikon All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint Yokogawa CSU-22), and the size distribution was quantified using ImageJ software analysis of images acquired. Fluorescent microparticles were also analyzed the concentration using a cell counter (Countess II FL AMQAF1000, ThermoFisher). Similarly, particles with and without IgG attachment together with those hybridized All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. Macrophage uptake assay Biotin-modified "self"-peptide (biotin-miniPEG-GNYTCEVTELTREGETIIELK[Lys(FITC)], LifeTein) was loaded onto streptavidincoated PLGA microparticles using the protocol described above. The streptavidin of the control particles without "self-peptide" were labeled with NHS-fluorescein (ThermoFisher #46409) at 6 μM/OD550 for 1 hour at room temperature, followed by three washes. Murine macrophage cell line J77A4.1 (ATCC) grown on glass bottom chamber slides (Thermo Scientific #154526) was treated with lipopolysaccharides (Sigma #L4391) at 100 ng/mL for overnight. On the next day, "self"-peptide loaded particles and control particles were added at 0.03 OD550 x 200 μL and incubated at 37 o C for 1 hour. Cells were then washed three times with PBS to remove uninternalized particles and fixed using 4% paraformaldehyde (Electron Microscopy Sciences #15710) for 20 minutes. After three washes, cells were imaged using the All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint spinning disk confocal fluorescence microscope (Nikon Yokogawa CSU-22). Images were analyzed for cell fluorescence signal using ImageJ software. were synthesized with PLGA microparticles functionalized with anti-CD3 antibody (Bio X Cell #BE0001-2, attached to 3'NH 2 -5'Qusar670-modified compB), and anti-CD28 antibody (Bio X Cell #BE0248, attached to 3'NH 2 -5'FAM-modified compR) at varying ratios from 1:5, 1:3, 1:1, 3:1 to 5:1 according to the method described above. After thawing and 24-hour recovery in culture, 1.4 x 10 5 CD4+/CD8+ human primary T cells in 200 μL medium were added with AICEs at 0.11 OD550 (approximately 2.5 All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint AICE to 1 cell), or CD3/CD28 Dynabeads at a 1:2.5 cell:bead ratio (Life Technologies #11131D) for 4 days. CD8+ T Cells were imaged through the spinning disk confocal microscope (Nikon Yokogawa CSU-22) one day after AICE addition. CD4+ and CD8+ T Cells numbers were quantified using a cell counter (Countess II FL AMQAF1000, ThermoFisher) every other day from day 6 to day 14 after AICE activation. Fresh IL-2 containing medium was supplemented to maintain the cell concentration between 0.5-1.5 million per mL. On day 14, T cell phenotype was studied by flow cytometry using the following antibodies: anti-CCR7 (BD #561271), anti-CD45RA (BD #562885), anti-LAG-3 (BD #565720), anti-TIM-3 (BD #565558), anti-PD-1 (Biolegend #329936). Anti-IL-2 clone #5355 (ThermoFisher #MA523696) was biotinylated using NHS- The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint mCherry expression to easily identify transduced T cells 28 . All constructs were cloned via InFusion cloning (Takara Bio #638910). Pantropic VSV-G pseudotyped lentivirus was produced via transfection of Lenti-X 293T cells with a pHR'SIN:CSW transgene expression vector and the viral packaging plasmids pCMVdR8.91 and pMD2.G using Fugene HD (Promega #E2312). Primary T cells were thawed the same day and, after 24 hours in culture, were stimulated with Human T-Activator CD3/CD28 Dynabeads (Life Technologies #11131D) at a 1:2.5 cell:bead ratio. At 48 hours, viral supernatant was harvested and the primary T cells were exposed to the virus for 24 hours. At day 5 after T cell stimulation, Dynabeads were removed and the T cells expanded until day 12 when they were rested and could be used in assays. T cells were sorted for assays with a FACs ARIA II on day 5 post T cell stimulation. or CD8+ T cells were co-cultured with target cancer cells at a 1:1 ratio, with the addition of AICE at 0.075 OD550 x 200 μL medium (100 μL T cell medium + 100 μL cancer cell medium) (unless otherwise specified). Dual antigen (GFP and HER2) positive target cells (A375 and K562) were used as positive controls. After mixing, cells were centrifuged for 2 min at 300 x g to initiate interaction of the cells. After 24 hours, the co-culture from CD8+ T cells were stained with anti-CD69 antibody (BD #562884) and analyzed by flow cytometry (BD LSR II), and also imaged through the spinning disk confocal microscope (Nikon Yokogawa CSU-22). After 48 hours, cytokine concentration in the supernatant was measured by IL-2 ELISA (for CD4+ T cells, eBiosciences #BMS2221HS) and Interferon-γ (IFN-γ) ELISA (for CD8+ T cells, ThermoFisher #KHC4021). To quantify target cell killing, 2.5 x 10 4 A375 cells were seeded on flat-bottom 96well tissue culture plate (Falcon #353072) and cultured for 8 hours, followed by the All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint addition of 2.5 x 10 4 CD8+ T cells and AICE at 0.075 OD550 x 200 μL final (unless otherwise specified). After 1-3 days, cells were gently washed with PBS for 2 times, and analyzed for cell viability using PrestoBlue Cell Viability Reagent (Invitrogen #A13262). In vivo tumor targeting NSG mice (Jackson Laboratory #005557, female, ~8-12weeks old) were implanted with two identical xenograft tumors -5 x 10 6 HER2+ K562 tumor cells subcutaneously on the left and right flank. Seven days after tumor implantation, human primary CD4+ (4 x 10 6 ) and CD8+ T cells (4 x 10 6 ) were injected intravenously into the tail vein of the mice. These T cells were either untransduced (control) or engineered with the anti-GFP synNotch Gal4VP64 receptor and the corresponding response elements regulating anti-HER2 4-1BBζ CAR expression. On the same day, AICE-GFP particles were injected intratumorally at one All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint side of the two flanks with 50 (or 10) OD550 x 50 μL per dose, leaving the other as the control. Three additional doses of AICE were injected into the same tumor every 4 days or starting as the tumor grew over 500 mm 3 Data availability. The authors declare that the data supporting the findings of this study are available within the paper (and its supplementary information files). All rights reserved. No reuse allowed without permission. The copyright holder for this preprint (which was not peer-reviewed) is the author/funder. . https://doi.org/10.1101/587105 doi: bioRxiv preprint