Nanocarrier design for pathogen-inspired innate immune agonist delivery

Abstract

In complex diseases such as cancer, modulating cytokine signatures of disease using innate immune agonists holds therapeutic promise. Novel multi-agonist treatments offer tunable control of the immune system because they are uniquely “pathogen-inspired”, eliciting robust anti-tumor responses by promoting synergistic cytokine responses. However, the chief strategic hurdle is ensuring multi-agonist delivery to the same target cells, highlighting the importance of using nanomaterials-based carriers. Here, we place nanocarriers in center stage and review the delivery hurdles related to the varying extra- and intracellular localizations of innate immune receptors. We discuss a range of nanomaterials used for multi-agonist delivery, highlighting their respective benefits and drawbacks. Our overarching stance is that rational nanocarrier design is crucial for developing pathogen-inspired multi-agonist immunotherapies.

The case for pathogen-inspired delivery technologies in cancer immunotherapy

In cancer and other complex diseases, transformative therapeutic advantages exist for novel tools that can modulate the cytokine makeup of the diseased microenvironment. The tumor microenvironment (TME) of aggressive cancers is notoriously immunosuppressive or ‘cold’, rendering any local or infiltrating anti-tumor immune cells dysfunctional. Cold tumors urgently warrant the development of new classes of therapies that remodel the TME towards proinflammatory or ‘hot’, to sustain immune activity and promote durable tumor clearance. Innate antigen-presenting cells (APCs), including dendritic cells (DCs) and macrophages, are considered ideal target cells for TME remodeling approaches, since they are robust producers of proinflammatory cytokines and can prime and activate adaptive T and B cells. Importantly, treatment with innate immune agonists can be more self-amplifying compared to cytokine therapy, because agonists are smaller molecules that can better target intracellular pathways [1]. Furthermore, owing to their smaller size, agonists can be more readily manipulated for packaging into nanocarriers for delivery.

Emerging “pathogen-inspired” immune agonist therapies reflect the unique ability of microbes to ferry multiple agonists into a target cell (e.g., bacterial cell wall components, viral capsid proteins, and microbial nucleic acids); together, these act in concert to promote the robust and synergistic production of proinflammatory cytokines [2]. Notably, such cytokine responses have both breadth and depth. Nanomaterials technologies have provided unique engineering solutions to these design considerations, enabling the co-encapsulation of multiple synergistic agonists within the same nanocarrier or nanoparticle (NP) system and allowing for added built-in capabilities to direct NPs to target sites, cells, and pathways. Here, we review the varying localizations of innate immune sensing receptors within a cell, such as the plasma membrane, endosomal membranes, and cytosol, and the delivery challenges associated with harnessing them. We discuss the utility of specific classes of materials for agonist delivery and highlight the engineering advantages for multi-agonist delivery (see Highlights and Significance). Our stance is that effective multi-agonist delivery technologies rely heavily on the rational design of their nanocarriers because delivery to their cognate receptors depends on how effectively they are ferried into target cells. As a class of next-generation cancer immunotherapy, the novelty of multi-agonist NPs centers on leveraging engineering strategies to harness the powerful innate immune-activating attributes of pathogens for durable anti-tumor control, without any risk of infection.

Highlights Box.

• Pathogen-inspired nanocarriers take lessons from microbes and ferry multiple innate immune agonists to the same target cell; together, they activate multiple innate immune sensing pathways and orchestrate robust and synergistic cytokine responses.

• Nanomaterials engineering design strategies must take into consideration the varying extra- and intracellular localizations of innate immune sensing receptors to optimize agonist delivery to specific cognate receptors.

• Different classes of lipids, polymers, metals, and hybrid combinations of materials have been and are being tested for multi-agonist delivery. This underscores the central importance of rational nanocarrier design to achieve therapeutic efficacy.

Significance Box.

Cytokine remodeling therapies that utilize innate immune agonists may offer a shift in or complementary treatment paradigms for cancer and other complex diseases. Rational nanocarrier engineering design considerations are strongly warranted, particularly from the materials perspective. These pathogen-inspired, multi-agonist delivery technologies emerge as a significant class of next-generation cancer therapeutic strategy.

Multi-compartmental localization of innate immune sensing receptors and clinical limitations of free agonists

Innate immune sensing receptors or pathogen recognition receptors (PRRs) are distributed across cellular compartments of multiple eukaryotic species (mammalian and non-mammalian), primarily on APCs, such as DCs and macrophages [3]. PRRs are activated by highly conserved motifs unique to microbes or pathogen-associated molecular patterns (PAMPs), such as bacterial cell wall components, viral capsid proteins, and microbial nucleic acids (Fig. 1) [4]. Certain immune agonists utilized for therapy, such as agonists of Toll-like receptor (TLR) and Stimulator of Interferon Genes (STING) pathways, are designed from known PAMP structures. TLRs, originally identified in Drosophila melanogaster, and C-type lectin receptors (CLRs) reside on plasma and endosomal membranes [5]. TLR1, −2, −4, −5, and −6 are expressed on the plasma membrane, whereas TLR3, −4, −7, −8, and −9 are found on endosomal membranes of vertebrates and invertebrates [6]. PRRs such as retinoic acid-inducible gene I (RIG-I)-like receptors (RLRs) and nucleotide-binding oligomerization domain (NOD)-like receptors (NLRs), as well as DNA sensors, such as the STING pathway, are located in the cytosol and the endoplasmic reticulum [7–10]. Activation of innate immune sensing receptors, such as TLRs, can induce downstream immune activation resulting in processes such as PANoptosis, a mode of inflammatory cell death linked to the inflammasome response, measured via proinflammatory IL-1β and IL-18 production [11]. Induction of PANoptosis has already shown promise in reducing tumorigenesis as demonstrated in vivo in B16F10 melanoma tumor-bearing mice [12].

Figure 1. Examples of pathogen-derived innate agonists for nanocarrier design.

These pathogen-derived agonists are shown as they naturally occur in bacteria, viruses, and fungi, depicting how they can inspire nanocarrier design. ssDNA: single-stranded deoxyribonucleic acid; ssRNA: single-stranded ribonucleic acid Figure created with BioRender.com.

The efficacy of delivery of innate immune agonists in free form is currently being tested in multiple clinical trials in patients with several types of cancer. A synthetic form of TLR7/8 agonist imidazoquinoline and a safe derivative of TLR4 agonist lipopolysaccharide (LPS), monophosphoryl lipid A (MPLA), have already been approved for use by the US Food and Drug Administration (FDA) in humans in several vaccine formulations, including treatments for human papillomavirus (HPV) and shingles [13]. STING agonist ADU-S100, which showed anti-tumor efficacy when delivered intratumorally in B16F10 melanoma tumor-bearing C57BL/6 mice, underwent Phase I clinical trials in combination with immune checkpoint blockade (ICB) for lymphoma, melanoma, breast cancer, colorectal cancer, Merkel cell carcinoma, and sarcoma. However, both trials were terminated early due to a lack of anti-tumor activity and/or sponsor decision. [14, 15] (NCT02675439)^I, (NCT03172936)^II. STING agonist E7766 has been tested by itself in clinical trials for the treatment of advanced solid tumors; this study was also prematurely terminated by the sponsor without any clinical data collected [16] (NCT04144140)^III. TLR4 agonist G100 has been tested as a stable emulsion in patients; notably, in combination with both ICB (pembrolizumab) and radiotherapy for patients with low grade Non-Hodgkin’s lymphoma, highlighting the utility of designing multi-faceted treatments for augmented efficacy (NCT02501473)^IV. TLR7 agonist CAN1012 has proven effective in SCC7 squamous cell carcinoma, CT26 colon carcinoma, and MC38 colon carcinoma in mouse tumor models and is currently being tested in Phase I trials in patients via intratumoral injection with unresectable or metastatic advanced solid tumors (NCT04987112)^V [17].

Taken together, these trials are limited in their dependence on intratumoral (i.t.) injection and delivery of agonists in free form without a nanocarrier. Intratumoral (i.t.) delivery does not allow administration into tumors located in inaccessible areas, such as organs in the peritoneal cavity. Furthermore, with i.t. delivery, there is no mechanistic basis for targeting or directing agonists to tissue-resident immune cells nor can any safeguards be built in to ensure that agonists do not escape the TME and cause off-target toxicity. In an ongoing Phase I/II study, this consideration was partly addressed by utilizing a hydrogel-based solution that allowed the controlled release of a prodrug of TLR7 agonist, resiquimod, using i.t. injection administered in CT26 colon carcinoma-bearing mice; the treatment resulted in reduced long-term tumor burden when used in combination with ICB [18] (NCT04799054)^VI. The multi-compartmental localization of innate immune sensing receptors has therefore posed a formidable challenge for developing innate immune agonist therapies, with specific hurdles when aiming to activate multiple innate immune pathways in a true pathogen-inspired fashion.

Harnessing materials engineering for agonist delivery

The multi-compartmental localization of PRRs has clear evolutionary advantages for protection against infections, providing a multi-layered defense system. However, this organization poses a formidable engineering challenge for pathogen-inspired nanomaterials platforms since delivery of agonists to specific cellular compartments is difficult (Fig. 2, Key Figure). Furthermore, activation of innate immune pathway receptors is highly specific to certain cell types. For instance, TLR activation in B cells promotes mitosis, but this effect is not seen in other cell types such as splenic T cells [19]. TLR4 activation in human monocytes induces IL1β production, whereas monocyte-derived macrophages do not show this response [20]. Signaling intensities in specific cells can also influence apoptotic behavior, such increased STING expression in T cells following STING activation, leading to apoptosis (observed in vitro using cell viability assays in primary mouse cells [fibroblasts, bone marrow-derived macrophages (BMDMs) or DCs (BMDCs)]) [21]. Additionally, TLR expression varies across cell types, with human myeloid DCs (mDCs) expressing all TLRs except TLR9, while human plasmacytoid DCs (pDCs) predominantly express TLR9 [22]. Collectively, these findings suggest that effective therapies must incorporate engineering strategies that deliver agonists to specific target cells. In the context of cancer, even though T cells are the ultimate effector cells of choice due to their direct killing capabilities, innate immune sensing therapies may likely be effective only if they target APCs.

Key Figure, Figure 2. Mechanisms by which pathogen-inspired nanocarriers promote agonist engagement in host cells at different subcellular locations.

LPS: lipopolysaccharide; iE-DAP: γ-D-Glu-mDAP; MDP: muramyl dipeptide; cdGMP: cyclic dimeric guanosine monophosphate; NF-κB: nuclear factor kappa-light-chain-enhancer of activated B cells; IRF: interferon regulatory factor; IKK: inhibitor of NF-κB kinase; TLR: Toll-like receptor; RIG: retinoic acid-inducible gene; NOD: nucleotide oligomerization domain; STING: stimulator of interferon genes; TNFα: tumor necrosis factor-alpha; IFN: interferon; dsRNA: double-stranded ribonucleic acid; ssRNA: single-stranded ribonucleic acid. Figure created with BioRender.com.

Because of their readily accessible interface as early points-of-contact, plasma and endosomal membrane PRRs are widely targeted by nanocarrier-based strategies. For these purposes, the multi-faceted functionalities of nanomaterials (e.g., size, charge, tunability of syntheses, hydrophobicity versus polarity, compatibility with physically distinct agonists and materials) are being harnessed for engineering design. Lipid-protamine (LP) NPs have been designed with anionic liposomal cores to electrostatically bind to and encapsulate synthetic TLR3 agonist polyinosinic:polycytidylic acid (polyI:C), which is cationic [23]. CD40 antibodies were conjugated to surfaces of LP NPs to facilitate targeting and early activation of APCs in a B16F10 melanoma mouse model [23]. Lipid-based outer membrane vesicles (OMVs) have been designed to activate a variety of PRRs; their outer shells are derived from membranes of gram-negative bacteria containing structures such as LPS and other agonists of extracellular TLRs, as shown in studies of primary splenic and BMDCs) from transgenic OT-I and OT-II TCR, Batf3-, and MyD88-deficient mice on the C57BL6 background [24]. Besides nanocarriers, immune-stimulating antibody conjugates can activate TLR7/8 [25]. Reduced tumor volume was observed when TLR7/8 agonist T785 was delivered systemically, bound to a tumor-targeting antibody for HER2 (trastuzumab) and CD20 (rituximab) in human hepatocellular carcinoma HCC1954 xenograft mouse models. Immune activation was likely due to the upregulation of genes associated with interferon regulatory factor 7 (IRF7) and NF-κB because both pathways have been implicated in the downstream activation of TLR7/8 [26]. While immune-stimulating antibody conjugates offer a versatile approach for co-delivery of agonists and antibodies, they are limited in that they do not offer inherent “stealth” characteristics or non-immunogenicity that can be offered by nanocarriers. Furthermore, combining multiple agonists using antibody technology poses a complex design hurdle.

By comparison, materials-based strategies used for delivery to cytosolic PRRs often rely on built-in engineering capabilities that promote agonist release in endosomes. Degradable poly(beta amino ester) (PBAE) NPs were used to deliver STING agonist ML-317, conjugated via cationic arginine polypeptides [27]. B16F10 melanoma-bearing mice treated with these NPs demonstrated increased the expression of Type I IFNs and other proinflammatory cytokines via cytokine analysis of tumors and spleens. PEG-lipid nanodiscs [28] and virus self-assemblies [29] also reduced tumor burden and prolonged survival in mouse cancer models upon delivery of agonists of STING or RLR pathways. Unlike lipid-based NPs, lipid nanodiscs are unable to encapsulate aqueous agonists due to their flexible shape, but it is this attribute that allows them to drain effectively to tumors. Viral self-assemblies can be advantageous due to their structural similarity to natural viruses that are highly proficient at breaching cell membrane defenses and funneling their own uptake [30]. Nanobiologics containing muramyl dipeptide (MDP) or muramyl tripeptide phosphatidylethanolamine (MTP-PE), known activators of NOD2 receptors, have been designed for delivery via intravenous (i.v). treatment; these agonists were effective in reducing tumor growth in a B16F10 melanoma mouse model with increased production of proinflammatory cytokines TNFα, IL6, and IFNγ compared to controls [31].

Notably, while the vast majority of biological carriers are NP systems, enabling the tailoring of their extra- and intracellular trafficking capabilities, a small subset have been formulated as stationary implants. While not always feasible depending on the target tissue, effective implantable carriers have no requirements for extracellular trafficking, either locally in the tissue, or systemically in blood circulation; this carries important implications for the readily translational capacity of these systems. Implantable subcutaneous devices such as ‘NanoLymph’ dual-reservoirs have been engineered to enable the sustained release of TLR7/8 agonist resiquimod and demonstrated local DC recruitment as well as antigen-specific CD8⁺ T cell activation in an ovalbumin (OVA)-expressing murine model [32]. 3D-printed scaffolds containing shell and core layers have been designed to integrate non-nucleotide STING agonist SR-717 in addition to AKT inhibitor MK-2206 for synergistic STING activation; these promoted the recruitment and enrichment of activated DCs and M1-like polarized macrophages in mouse models of B16F10 melanoma and H22 hepatocellular carcinoma, helping to mount an effective anti-tumor immune response [33]. Nanofluidic implants have also been engineered to deliver CD40 and PD-L1 monoclonal antibodies as well as TLR7/8 agonist resiquimod and a STING agonist, showing anti-tumor efficacy in the EMT6 triple-negative breast cancer mouse model [34]. Besides cancer, lipid-based NPs received spotlight attention during the recent COVID-19 pandemic, being formulated to stably encapsulate mRNA encoding antigens specific to the SARS-CoV-2 virus. Further developments of this design using data from B16F10 melanoma and TC-1 cervical cancer mouse models suggested that unique heterocyclic lipid headgroups of NPs strongly agonized the STING pathway as novel adjuvants, promoting long-term protection against tumor growth [35] Taken together, materials engineering strategies can be readily harnessed to deliver innate immune agonists to target cells. We can build on this complexity to design rational systems for multi-agonist delivery.

Multi-agonist delivery and synergistic cytokine responses

Due to the complexity and diversity of responses promoted by innate immune agonists, engineering site- and cell-specific drug delivery systems is of great importance. Such systems often function via incorporation of ligands on the carrier (i.e., active targeting or directing) or by tailoring physical characteristics such as size, charge, and shape (i.e., passive targeting or directing), in addition to exploiting tumor-specific changes in vasculature (i.e., “leaky” endothelium, overexpression of targetable biomarkers) that are reviewed extensively elsewhere [36, 37] Co-encapsulation of multiple innate immune agonists with shared downstream signaling cascades can drive synergistic responses in proinflammatory cytokine production, significantly enhancing the efficacy of immunotherapies beyond additive effects, (as evidenced from studies using primary APCs and other cell lines, as well as from multiple mouse tumor models) [38–41]. Materials-based delivery approaches offer the ability to control and amplify crosstalk between immune pathways, tailor release kinetics, and optimize spatial distribution, thus fine-tuning the immune response for improved therapeutic outcomes (Fig. 3). These approaches are essential to facilitate safe and orchestrated amplification of immune responses, mitigation of toxicities by minimizing excessive reliance on reticuloendothelial system (RES) clearance, and expansion of the therapeutic windows for synergistic combinations.

Figure 3. Examples of materials widely used for nanocarrier design.

These materials include polymers [50–52], lipids [39,40,44], metals [60–62], and hybrid materials [72,73]. Figure created with BioRender.com.

As a counter example, delivery of TLR9 agonist CpG and TLR4 agonist MPLA without a nanocarrier platform was recently investigated [42]. In initial experiments, murine BMDCs demonstrated the synergistic expression of mRNAs encoding cytokines such as IL-12β following dual TLR9/TLR4 agonist treatment in vitro; the results showed distinctive differences in gene ontology (GO) biological processes, including activation of innate immune responses and cell cycle regulation, with optimal enrichment in cells treated with combined agonists (79 unique GO processes)[42] . When combined in a vaccine against a murine model of H1N1 pdm09 influenza harboring a recombinant hemagglutinin (HA) protein, this combination of agonists functioned efficiently as an adjuvant, based on proinflammatory cytokine production, antiviral gene expression, and reduced animal weight compared to controls [42]. However, since these formulations lacked a nanocarrier, co-delivery of both agonists and the HA protein were not ensured. While it remains to be tested, this design attribute (or lack thereof) may be the impetus for its failure to outperform existing PR8 vaccine formulations.

Here, we frame our stance by highlighting the novel design considerations of specific materials, including lipids, polymers, metals, and hybrid systems, that provide unique advantages for delivering multiple innate immune agonists to appropriate physiological and subcellular locations. Notably, to our knowledge, there are no reports of the use of multi-agonist treatments in clinical trials to date. As these novel treatments are translated, our stance is that a nanocarrier must be included and rationally designed to be used for delivery (see Clinician’s Corner for a discussion of translational considerations).

Clinician’s Corner I: Safety, clearance, and delivery route considerations.

• Off-target effects in patients treated with cytokine remodeling therapies and strategies to mitigate them. Off-target effects of agonists can include the uninhibited production of proinflammatory cytokines defined as a “cytokine storm”. This effect has been demonstrated in immunotherapies involving CAR T cells, ICB, and cytokines or agonists [74–77]. A cytokine storm can lead to fatigue and severe weight loss as well as multi-organ failure [78, 79]. Nanocarriers are being engineered with tailoring advantages to optimize their size, shape, and material to optimize targeted delivery and limit off-target toxicity [80].

• Materials clearance considerations. Concerns about clearance and long-term toxicity of nanomaterials persist. In particular, studies assessing intraperitoneal (i.p.) injection of solid lipid NPs showed their accumulation in several organs (liver, spleen, kidneys), driving potentially toxic effects over prolonged use [81, 82]. Polymer-based systems can also present toxicity issues. Long-term biodistribution and toxicity studies in mice and humans indicated that PLGA NPs can also accumulate in the liver and spleen, leading to potential toxicity with repeated administration [83, 84]. Metal-based systems exhibit significant cytotoxicity and long-term persistence in the body. Their clearance is often slow, leading to prolonged exposure of tissues to potentially harmful metal ions. For example, gold NPs can induce oxidative stress and inflammatory responses in the liver and kidneys. Iron oxide NPs (used primarily in imaging and therapy), can also cause oxidative damage and interfere with normal cellular functions via reactive oxygen species (ROS) generation and downstream activation of cell signaling pathways, leading to apoptosis or necrosis. Toxicities of metal-based NPs have been reviewed extensively elsewhere [85].

• Delivery route considerations. Direct i.t. delivery ensures high local concentrations of nanocarriers, maximizing effectiveness while minimizing systemic exposure and mitigating undesirable side effects. I.t. delivery has shown promising results in preclinical and early-phase clinical trials, particularly by enhancing local immune responses [86, 87]. However, i.t. therapy is often limited by the accessibility of tumors and may not be suitable for metastases and deeply situated cancers. Systemic delivery of nanomaterials aims to target tumors through i.v. administration, leveraging the enhanced permeability and retention (EPR) effect (and/or other biological phenomena [36, 88, 89]) to accumulate in tumor tissues. This approach allows the simultaneous treatment of primary tumors and metastatic lesions. However, clinical outcomes of systemic delivery have been mixed. The success of systemic delivery is highly dependent on the tumor type, vasculature, and ability of the nanocarrier system to evade immune clearance. Strategies that combine both delivery strategies may offer synergistic benefits, optimizing concentrations of immune agonists at the tumor site while also addressing metastatic disease.

Lipid-based systems

Lipid-based systems are favored for their biocompatibility and ability to encapsulate hydrophilic and hydrophobic payloads that are otherwise physically distinct and incompatible together in free form [43]. Like other materials-based systems, lipid-based NPs can be designed to also protect cargo from degradation and off-target release and toxicity, extend circulation time via surface PEGylation, and enhance cellular uptake [44]. Compared to other materials, lipids offer high biocompatibility, making them ideal candidates for systemic immunotherapies that seek to activate anti-tumor immunity in not only primary tumors but also widely disseminated metastatic sites. Disadvantages of lipid-based systems include potential leakage of agonists into circulation due to flexible membrane makeup, as well as off-target uptake by RES organs, which can result in systemic toxicity [45].

Specifically, using neutral lipids and a PEG surface, dual-agonist lipid-based NPs were used to co-encapsulate STING agonist cyclic di-guanosine monophosphate (cdGMP) in the inner aqueous core and TLR4 agonist monophosphoryl lipid A (MPLA) in the external bilayer of lipid-based NPs [39, 40] (Fig. 4). This design enabled MPLA to engage TLR4 upon NP uptake in endosomes; upon endosome acidification and lipid disintegration, cdGMP was released into the cytosol to engage STING machinery. Following systemic delivery in the blood and tumor accumulation in murine models of triple-negative breast cancer and melanoma, dual STING/TLR4 activation promoted synergistic production of Type I IFNs and CD8⁺ T cell-mediated tumor clearance. This was demonstrated by improved animal survival, decreased tumor burden, and flow cytometry analysis showed several immune cell subsets in multiple mouse tumor models, including 4T1 triple-negative breast cancer, B16F10 melanoma, and Panc-02 pancreatic ductal adenocarcinoma. Combined with immune ICB, this approach led to extended survival in these tumor models and curative responses when used as a neoadjuvant pre-surgical treatment, driving de novo epitope recognition by T cells and humoral immunological memory [38]. When combined with the RAS inhibitors tremetinib and palbociclib, these NPs promoted upregulation of senescence markers in tumor cells and augmented CD8⁺ T cell-mediated clearance in genetic mouse models of pancreatic cancer; this highlighted the potential of using the novel combination of innate immune agonists with RAS inhibitors [46]. In this study, the authors demonstrated a strong safety profile based on minimal systemic immune activation and detection of liver enzymes alanine transaminase (ALT) and aspartate transaminase (AST) in systemic circulation; this showed that liver uptake of NPs elicited only minimal and transient animal weight loss.

Figure 4. Representative approaches for nanoparticle-based multi-agonist delivery and downstream synergistic activation of cytokine pathways.

DOPC: dipalmitoylphosphatidylcholine; DSPC: distearoylphosphatidylcholine; mPEG-DSPE: methoxy-poly(ethylene glycol)-1,2-distearoyl-sn-glycero-phosphoethanolamine; MPLA: monophosphoryl lipid A; cdGMP: cyclic dimeric guanosine monophosphate [38]; DEAEMA: diethylaminoethyl methacrylate; BMA: butyl methacrylate; cGAMP: cyclic guanosine monophosphate-adenosine monophosphate; TLR: Toll-like receptor; TRAM: TRIF-related adaptor molecule; TRIF: TIR domain-containing adaptor inducing interferon β; TBK: tank binding kinase; STING: stimulator of interferon genes; IFNAR: interferon alpha receptor; IFN: interferon [55]; PLGA: poly(lactic-co-glycolic) acid; PEI: polyethylenimine; TRAF: tumor necrosis factor receptor-associated factor; MyD88: myeloid differentiation primary response 88 [41]; DMXAA: dimethylxanthenone acetic acid; MOF-801: metal-organic framework 801; NF-κB: nuclear factor kappa-light-chain-enhancer of activated B cells [63]. Portions of Fig. 4 have been adapted from the indicated references. Figure created with BioRender.com.

Nanoporous lipid-based microparticles (MPs) with neutral lipids were developed as a therapeutic cancer vaccine system co-loaded with STING agonist 2’3’-cyclic GMP-AMP (cGAMP) and TLR9 agonist cytosine-guanine (CpG) (TLR9 agonist) and combined with tumor antigenic peptides, such as Trp2 [47]. Co-uptake of both agonists on the same MPs by the same target cells drove the synergistic production of Type I IFNs and other proinflammatory cytokines (TNFα, CCL5) in BMDCs. Importantly, they showed efficacy in murine models of lung metastatic melanoma, HER2 breast cancer, and colorectal cancer via reduced tumor burden and extended survival, highlighting this combination as a potential platform approach in cancers with known tumor neoantigens [47].

While ionizable lipids have demonstrated high utility for delivering nucleic acid therapeutics based on their ability to promote endosome escape, their use for the delivery of innate immune agonist cargoes is less explored, partly because many target PRRs are located on plasma or endosomal membranes. However, for cytosolic targets such as STING and MDA5, pH-responsive polymers have enhanced therapeutic responses. They promoted endosomal escape and drove improved expression of antigen presentation genes, compared to free agonists, in RAW macrophages, THP-1 monocytes, and A549 epithelial cells. They reduced tumor burden and prolonged survival in EG7.OVA mouse lymphoma models compared to free antigens and agonists [48]. This suggested that ionizable lipid counterparts may play a valuable role as components of engineered nanocarriers of STING agonists.

Polymeric systems

Polymer-based systems, including polymeric NPs and micelles, also provide versatility in agonist encapsulation and release kinetics [49]. These systems are biodegradable and offer tunable physicochemical properties that enable controlled release of drugs over extended periods of time while minimizing toxicity [50]. Similar to lipid-based systems, specific surface modifications of polymeric systems can facilitate targeted delivery and evasion of RES organs, improving accumulation at a target site, based on a large number of studies conducted in vivo, and has been extensively reviewed elsewhere [51]. As mentioned, pH-responsive polymer systems that enable endosomal escape have also emerged in recent years as unique tools for the delivery of nucleic acid cargoes and agonists targeted to cytosolic PRRs [52]. Compared to other materials, polymers offer improved stability and highly controlled release kinetics, making them strong candidates for delivery of versatile payloads and of particular interest in the extended/prolonged delivery of immune agonists to primary tumors via intratumoral injection. Disadvantages of polymeric systems include increased risk of particle aggregation with potential resulting systemic toxicity.

Polymer-based pathogen-like particles (PLPs) [NPs or microparticles (MPs)], consisting of poly(lactic-co-glycolic acid) (PLGA) with branched polyethyleneimine (PEI) on the surface, co-loaded with TLR4 agonist MPLA and TLR9 agonist CpG have been engineered to examine signaling pathways responsible for multi-agonist immune responses [41] (Fig. 4). The biophysical properties of PLP carriers (e.g., size, surface vs. encapsulated agonists, low vs. high agonist density) drove the synergistic production of Type I IFNβ and IL-12p70 in bone marrow-derived APCs (shown via ELISA). PLPs carrying TLR7 agonist R848 and RIG-I agonist poly-U/UC RNA were also designed and tested in murine influenza virus infection models in vitro and in vivo in BALB/c mice [53]. PLGA was also used to build a degradable NP system co-loaded with TLR9 agonist CpG, STING agonist 3’3-cGAMP, and RIG-I agonist 5’ppp-dsRNA, along with antigenic peptides, to drive antitumor immunity [54]. Here, NPs suppressed tumor growth in an orthotopic B16F10 murine melanoma model when delivered systemically, but not when delivered intratumorally, highlighting the importance of systemic immune responses and delivery to achieve efficient tumor control, including for metastasis.

Recently, polymer-based NPs were engineered with multiple immune agonists to use as cancer vaccine platforms [55] (Fig. 4). Specifically, co-encapsulation of STING agonist cGAMP and TLR4 agonist MPLA within PEGylated di-block polymersomes containing pH-responsive poly(2-diethylaminoethyl methacrylate) (PDEAEMA) synergistically enhanced the activation of APCs in vitro and the expression of DC costimulatory marker CD86; in addition, they drove the expansion of polyfunctional antigen-specific T cell populations co-expressing proinflammatory cytokines IFNγ and TNFα, as well as peptide antigen cross-presentation in vivo. This NP vaccine showed therapeutic efficacy via reduced tumor burden and prolonged survival in the EG7.OVA murine model when administered with the model SIINFEKL antigenic peptide. Modest survival benefit was also observed in the MC38 colon carcinoma mouse model when combined with ICB [55]. These findings have strong implications for the use of polymeric NP-based peptide vaccines.

Using polymers designed with multi-faceted arms [charged, hydrophobic, and pH-responsive due to poly(2-diisopropylamino) ethyl methacrylate (PDPA)], polymeric systems were developed to co-encapsulate TLR7/8 agonist R848 and TLR9 agonist CpG in addition to antigenic peptides [56]. This ‘nanovaccine’ drove significant improvement in antigen presentation in vitro in DC2.4 cells compared to control groups; upon subcutaneous (s.c.) delivery, they showed efficient lymph node accumulation in healthy mice. In the EG7.OVA murine model, durable and potent SIINFEKL-specific CD8⁺ T cell responses were observed via flow cytometry. To highlight versatility in this platform approach, nanovaccine treatment also resulted in anti-tumor modulation and enhanced survival when combined with ICBs in vivo in MC38 adenocarcinoma and GL261 glioblastoma murine models. Here, the ADPGK neoantigen was used, which despite being widely reported and used in MC38 syngeneic models, was recently reported as undetectable via sequencing studies [57]. Experiments aimed at elucidating these conundrums and identifying particularly immunogenic antigens in immunosuppressive cancers are certainly needed, with strong implications for optimizing the selection of neoantigens in cancer therapy. In earlier studies, polymer-based systems using poly(lactic acid) (PLA) NPs were used to co-encapsulate multiple NOD agonists, demonstrating immune activation in vitro in human monocyte-derived DCs [58].

With a different approach, recently, poly(2-acrylamido-2-methyl-1-propanesulfonic acid) (PAMPS) synthesized as a di-block copolymer with pH-responsive PDEAEMA was reported as a novel system that activated TLR9 and STING pathways by itself in vitro using immortalized DCs (as evidenced from ELISA and immunofluorescent imaging), and did not require encapsulated agonists [59]. This polymer was conjugated to protein antigens to generate a delivery system that drove robust immune responses in healthy BALB/c mice in proof-of-concept prophylactic studies [59]. Future experiments could incorporate PEGylated lipids to improve biocompatibility and could seek to incorporate additional agonists to drive synergistic proinflammatory responses when combined with TLR9 or STING activation (e.g., TLR4 agonists) [39].

Metal-based systems

Metal-based systems offer unique advantages including very high stability and tunable surface chemistry. These systems are of high interest for multimodal imaging and therapy. Inherent properties of metals can be exploited for image-guided therapy because they enable real-time monitoring of drug delivery and therapeutic responses [60]. Their use in systems that deliver single innate immune agonists (particularly STING agonists) has been recently reported, where they have reduced tumor burden and extended survival in MC38 colon carcinoma (zinc NPs), B16F10 melanoma (iron NPs) and CT27 colon carcinoma (manganese NPs) mouse models [61–63]; however, these metals as a general class of materials exhibit limited reach in versatile delivery of multiple agonists, largely because their functionalities are not multi-faceted and can often accommodate only one physical class of agonists. Nonetheless, compared to other materials, metals offer great theranostic capabilities by combining diagnostic capabilities with therapy; however, the long-term safety profiles remain a concern in the context of drug delivery. Disadvantages of metal-based systems in the delivery of innate agonists include slow clearance times and long-term deposition in RES organs, which may result in accumulated toxicity.

In a unique report, a self-assembling metal-organic framework 801 (MOF-801) containing fumaric acid and Zr-O clusters was designed to co-deliver TLR9 agonist CpG and STING agonist DMXAA to cells for TME remodeling via tumor-associated macrophages (TAMs) and DCs upon systemic delivery in circulation [64] (Fig. 4). Here, treatment showed improved promotion of M1-like polarization in RAW264.7 macrophages and BMDC maturation in vitro. Moreover, in the orthotopic Hepa1–6 murine hepatoma model, cure rates of 80% were observed, although this efficacy required eight doses based on the data presented [64]. Further optimization of the carrier system may be the key to sufficiently tailoring this TLR9/STING combination for improved efficacy with less frequent dosing, and should be carefully considered in future studies.

Hybrid materials systems

Hybrid NP formulations are composed of rationally selected combinations of large materials classes, including lipids, polymers, and metals. Hybrid systems have emerged in recent years as potent and promising carriers, but have yet to realize their full potential with respect to delivery and/or co-delivery of innate immune agonists. In particular, polymeric-lipid nanoparticles (PLNs) are of high interest because they combine enhanced stability and biocompatibility associated with lipid-based systems with the controlled release kinetics of polymeric systems. PLNs can be classified as polymer core-lipid shell NPs [65], lipid bilayer-coated NPs [66], polymer-caged nanobins [67], monolithic lipid-polymer NPs [68], and lipid-polymer-lipid NPs [69]. These systems are extensively reviewed elsewhere [70]. Unfortunately, to date, they have exhibited little reach in the realm of innate immune agonist delivery [71]. Integration of metal-based NPs (in particular, gold) with polymer- [72] or lipid-based [73] materials in various nanostructures is enabling novel applications for these systems, including delivery of more versatile payloads; indeed, they overcome certain traditional limitations of gold NPs and have high utility for multi-agonist delivery. Therefore, multi-agonist therapies heavily rely on the makeup of their materials. Careful consideration of the utility of specific materials components is thus warranted in the rational design of NPs (see Clinician’s Corner I and II boxes).

Clinician’s Corner II: PEG antibodies and combination therapies.

• Implications for using PEG antibodies. Poly(ethylene glycol) (PEG) is extensively used in nanomedicine due to its favorable properties such as biocompatibility and water solubility. In lipid-based systems, PEG lipids such as PEG-DMG (diffusible) and PEG-DSG (persistent) are selected based on desired pharmacokinetic profiles (migration to NP surface for enhanced transfection efficiency, prolonged circulation, etc) [90]. Despite PEG’s high lethal dose 50 (LD50, or dose at which lethality is observed in 50% of test group) in rodents (2000 mg/kg) and its historical safety profile [91], recent data show adverse reactions, and the existence of PEG-specific antibodies suggests potential toxicity and immune activation in animal models and humans [92, 93]. Theoretically, antibodies should not form against the ethylene glycol units of PEG because they are small, inert, and lack immunogenic epitopes. However, the immune system could generate anti-PEG antibodies through several hypothesized mechanisms, including the formation of immunogenic PEG-protein conjugates, PEGylated NPs acting as adjuvants, and epitope spreading. Additionally, certain amino acids (serine, threonine, glutamine, asparagine) may share structural or functional similarities with PEG, explaining cross-reactivity.

  Studies have highlighted the generation of anti-PEG antibodies as key factors in hypersensitivity reactions to PEGylated nanomedicines. For example, intravenous injection of PEG-400 at 8.45 g/kg caused renal toxicity in canine models [94]. These reactions can include IgE, IgM, and IgG mediated Type I, II, and II hypersensitivities, as well as anaphylactoid reactions [95, 96]. A 2018 study showed the presence of anti-PEG constructs based on antibody detection assays in donor blood, and that these were not indicative of complement system activation, suggesting that detecting these antibodies might not predict subsequent antibody-mediated toxicity [97]. However, a recent study in 130 adults demonstrated that elevated concentrations of vaccine-induced PEG antibodies correlated with increased reactogenicity after multiple doses, implicating PEG as the likely cause of this increased immunogenic response [92]. This study indicated that PEG-specific antibodies could be boosted by lipid-based system vaccination, leading to increased NP-leukocyte association to immune cells in blood. Conflicting results across studies and the growing clinical use of PEG systems underscore the need for further exploration of these antibodies and their implications. Monitoring PEG-specific antibody titers in patients undergoing PEGylated therapy may be crucial for identifying those at higher risk for adverse reactions and guide the selection of appropriate nanocarrier formulations.

  To mitigate the impact of anti-PEG antibodies, one approach uses diffusible PEG-lipids, which may reduce immune responses by dissociating from the nanocarrier system [93]. Additionally, other water-soluble polymers, such as poly(2-oxazoline) (POx) [98], poly(N-(2-hydroxypropyl)methacrylamide) (PHPMA) [99], and zwitterionic polymers [100] are being explored as alternatives to PEG, potentially offering similar benefits without the immunogenic drawbacks, although these remain to be tested.

• Avenues for combination therapy. As the efficacy of ICB has been limited in part by the cold, immunosuppressive TME [101], innate immune agonists offer promising solutions when combined with ICB, aiding in the activation of innate immunity and subsequent priming and activation of adaptive immune cells to create a powerful proinflammatory milieu synergizing with ICB [38, 102, 103]. Materials-based carriers that can precisely control the spatiotemporal release and delivery of these combinations may further optimize synergies between innate immune activation and ICB, holding promise for durable and potent antitumor responses. Large-scale trials combining multiple innate immune agonists with or without ICB are warranted to fully evaluate the safety profiles of these therapy regimens.

Concluding remarks

What is next for pathogen-inspired nanocarrier technologies as next-generation therapies?

We have supported our stance for the key utility of nanocarriers in multi-agonist delivery platforms, discussed field advancements, and highlighted the challenges and limitations that remain to achieve efficacy. Specifically, these challenges are related to targeting specific tissues and cells, achieving intracellular trafficking towards cognate receptors as well as clearance. In the near future, we expect that rational design may have success using computational methodology, where we can weigh attributes of selected agonists against attributes of selected materials (see Outstanding Questions box). We expect that such in silico methods may provide a robust, quantitative design framework that is otherwise currently absent in purely empirical strategies. Furthermore, tumor sequencing data from patients should be considered in-depth prior to the designing delivery technologies; this can optimize parameters related to delivery as well as efficacy. For example, on a per-patient basis, we should consider whether a particular agonist of choice will be effective in terms of the number of target cells that are present in a patient’s tumor and their state of dysfunction. Finally, to design next-generation therapies, we must continue to pay close attention to systemic toxicities and make a judicious use of engineering tools to mitigate the former. Nevertheless, the future paints an exciting picture.

Outstanding Questions Box.

• Is there a universally optimal material for delivery of innate immune agonists that targets any cognate receptor(s)? Or will each system need to be designed and tailored independently based on the specific payload(s)? High-throughput preclinical studies assessing a multitude of formulations for each agonist will need to be conducted. To what extent can in silico design strategies help?

• Are there optimal agonist or PRR activation combinations for maximal synergistic proinflammatory effects? Are there unexplored combinations of innate immune agonists that offer potential for improved immune responses compared to currently existing single and combinatorial systems?

• Do certain cancer types respond to certain types of materials better than others? How can the agonist combination be tuned on a case-by-case basis per-patient?

• Can additional therapies (e.g., gene therapy) be employed to increase the expression of PRRs in target cells prior to innate immune agonist therapy? Would this increased receptor expression result in improved proinflammatory responses?

• To what extent will the synergistic immunostimulatory effects of nanocarrier systems containing multiple innate immune agonists that have been observed in mouse models translate to human patients?

• How can engineers improve or build on the design of existing carriers that were engineered initially to incorporate single agonists to co-encapsulate multiple agonists? Are entirely novel materials formulations fully necessary? Or will existing platforms with minor or major modifications suffice to potentiate the efficacy of these agonists?

• To what extent will the combination of nanomaterial-encapsulated agonists with traditional and in situ CAR T cell therapies foster the activation and persistence of these whole-cell therapies beyond what is currently observed in the clinic?

• To what extent is the synergistic proinflammatory cytokine production that incorporates multiple innate immune agonists on single carriers be attributed to intracellular downstream synergy of the respective PRR pathways? Mechanistic studies showing regulatory factor protein and gene expression can continue to provide insights.

Table 1. Summary of materials used in nanocarrier design with their multi-agonist payloads and downstream effects in vitro or in animal models.

cdGMP: cyclic dimeric guanosine monophosphate; MPLA: monophosphoryl lipid A; STING: stimulator of interferon genes; IFN: interferon; TNF: tumor necrosis factor; poly(I:C): polyinosinic:polycytidylic acid; cGAMP: cyclic guanosine monophosphate adenosine monophosphate; IL: interleukin; R848: resiquimod; poly-U/UC: poly-uridine/uridine core; DMXAA: 5,6-dimethylxanthenone-4-acetic acid, CDA: Cyclic di-AMP; MSA: benzothiophene oxobutanoic acid; N/A: Not yet explored for delivery of innate immune agonists.

Material	Benefits	Limitations	Innate agonist (or payload) encapsulated	Immune activation phenotype	Treatment efficacy in vivo [survival benefit or C = cure rate (%)]	Ref
Lipids	• Biocompatibility • Encapsulation • Efficiency • Targeted delivery • Reduced toxicity	• Instability • Drug leakage • Liver uptake	cdGMP (STING) + MPLA (TLR4)	Type I IFNs, TNFα, T cell activation	Extended survival	[38,39]
			cGAMP (STING) + CpG (TLR9)	Type I IFNs, TNFα, CCL5, T cell activation	C = 40%	[47]
			Poly(IC) (TLR3)	Type I IFNs, expression of antigen presentation genes, T cell activation	Extended survival	[23]
Polymers	• Versatility • Highly controlled release	• Toxicity • complexity of manufacturing, reproducibility • Liver uptake	MPLA (TLR4) + CpG (TLR9)	Type I IFNs, IL-12p70	Not reported	[41]
			MPLA (TLR4) + cGAMP (STING)	Type I IFNs, TNFα, expression of antigen presentation genes	Extended survival	[55]
			R848 (TLR7/8) + CpG (TLR9)	antigen presentation, T cell activation	Extended survival	[56]
			CpG (TLR9) + cGAMP (STING) + 5’ppp-dsRNA (RIG-I)	Expression of antigen presentation genes, IL-10, IL-12, T cell activation	Extended survival, reduced metastasis	[54]
			R848 (TLR7) + poly-U/UC RNA (RIG-I)	Type I IFNs, TNFα, IL-1b, CXCL10, T cell activation	Not reported	[53]
Metals	• Surface modification • Dual capability for imaging and therapy	• Toxicity • Slow clearance • Limited versatility	DMXAA (STING) + CpG (TLR9)	Type I IFNs, promotion of M1-like polarization of TAMs	C = 80%	[64]
			CDA (cyclic dimeric adenosine monophosphate) (STING)	Type I IFNs, expression of antigen presentation genes	C = 80–100%	[61]
			MSA-2 (STING)	Type I IFNs, TNFα, IL-6, expression of antigen presentation genes, T cell activation	Extended survival	[62]
			CDA (STING)	Type I IFNs, expression of antigen presentation genes	C = 80%	[63]
Hybrid	• Combined properties of several materials	• Complex design and synthesis • Side effects from several materials	Enoxaparin	N/A	N/A	[65]
			Docetaxel	N/A	N/A	[69]
			CL264 (TLR7)	Type I/II IFNs	92% suppression of viral load	[71]
			Ammonium sulfate	N/A	N/A	[73]

Glossary

CAR T cells

T cells with an extracellular chimeric antigen receptor (CAR) reminiscent of the variable portions of an antibody with intracellular T cell receptor (TCR) signaling domains

Cytokine storm

excessive proinflammatory cytokine response that leads to off-target toxicity

Di-block polymersomes

polymeric vehicles with aqueous cores formulated with polymers made of two connected monomeric blocks

Enhanced permeability and retention

phenomenon by which blood vessels in tumors are “leaky” due to extended gaps between adjacent endothelial cells, enabling the non-discriminate sponge-like transit of many types of molecules into the TME compared to normal tissues

Gene ontology (GO) biological processes

resource generated from bioinformatics efforts connecting knowledge relating specific gene expression patterns to specific biological processes

Lipid bilayer-coated NPs

NPs synthesized with a lipid bilayer coating on the surface

Lipid-polymer-lipid NPs

NPs synthesized with three layers, with a polymer layer in between two lipid layers

Lipid-protamine

hybrid materials that combine lipids with protamine, a polycationic peptide, via electrostatic linkage

M1-like polarized macrophages

subset of macrophages described as proinflammatory, in contrast to M2-like macrophages, described as immunosuppressive/anti-inflammatory

Monolithic lipid-polymer NPs

hybrid lipid-polymer NPs that permit even (monolithic) distribution of a drug or agonist cargo

Nanocarrier

drug or agonist delivery vehicle that has nanoscale physical dimensions

Nanofluidic implants

stationary nanomaterials-based implant that can be refilled with therapeutic cargoes for sustained action

‘NanoLymph’ dual-reservoirs

nanoscale device reminiscent of a lymph node with two reservoirs containing immune agonists and antigens

Nanoparticle

nanocarrier that can be synthesized in particulate form, usually spherical in shape but can also be rod-shaped or planar

Nanoporous lipid-based microparticles

particulate lipid-based biomaterials with microscale physical dimensions; contain nano-sized pores for cargo loading.

Nanovaccine

synthetic vaccine formulation using nanomaterials

PANoptosis

unique form of cell death involving mechanisms of apoptosis, pyroptosis, and necroptosis pathways

Pathogen recognition receptors

innate immune sensing receptors recognizing agonist molecules expressed on microbes

Pathogen-associated molecular patterns

evolutionarily conserved motifs on innate immune agonists that bind to PRRs

PEG-lipid nanodiscs

disc-like NPs formulated from lipids and a surface PEG shell

Polymer core-lipid shell NPs

NPs with an inner polymer core and external lipid shell

Polymer-based pathogen-like particles

NPs incorporating multiple innate agonists within a polymeric material

Polymer-caged nanobins

nano-sized bin-shaped NPs that are encapsulated within a polymer cage or shell

Shell and core layers

synthesis methods by which external (shell) and internal (core) layers of materials or NPs can be modified

Surface PEGylation

coating of NP surfaces with PEG for longer circulation and “stealth” non-immunogenic properties

Virus self-assemblies

spontaneous formation (self-assembly) of compositional molecules of viruses, including capsid proteins and nucleic acids, into virus-like particles