Polymeric particle-based therapies for acute inflammatory diseases

Most clinical treatments for acute inflammatory diseases centre on infection control through antibiotics, pain management and supportive care, involving steroidal and nonsteroidal anti-inflammatory therapies 21 , 25 – 28 . However, these approaches have limitations, including increased risk of secondary infections with prolonged use owing to their lack of specificity, coupled with their systemic (oral or intravenous) mode of delivery 27 , 29 – 34 . Alternatively, intravenous cellular and monoclonal antibody therapies have emerged as promising agents to target and treat the damaging effects of inflammation 35 – 39 (Fig.  1b ). Stem cell therapies against inflammation typically involve transplantation of autologous mesenchymal stem cells (MSCs), which are multipotent cells derived from bone marrow, adipose tissue, dental pulp or umbilical cord tissue. Following implantation, MSCs secrete immune-suppressive cytokines (IL-10 and transforming growth factor-β (TGFβ)) to promote regulatory T (T reg ) cell production and inhibit differentiation of T helper 1 (T H 1) and T helper 2 (T H 2) cells, aiding in containing the pathological inflammation 40 – 43 . Finally, MSCs can suppress neutrophil recruitment and activation through secretion of superoxide dismutase 3 (refs. 41 , 43 , 44 ). Several clinical trials using MSCs have shown promising results in improving the symptoms of inflammatory disorders, including rheumatoid arthritis, multiple sclerosis and acute respiratory distress syndrome (ARDS) 45 . For example, intravenously injected human umbilical cord MSCs considerably reduced the concentration of inflammatory cytokines in a COVID-19 patient with ARDS, leading to the patient’s recovery 46 , with similar results shown in COVID-19 patients with pneumonia (ChiCTR2000029990) 47 , 48 . Despite showing promise, stem cell therapies have intrinsic limitations; isolating and expanding MSCs for each donor is time-consuming, making them unfit for large-scale manufacturing as would be needed for widespread, acute inflammatory illnesses such as COVID-19. MSCs as inflammatory therapeutics also require many viable cells (10 million per dose, with multiple doses required), which can prove challenging to access for autologous transplantation 49 . To address these limitations, allogeneic MSCs can be sourced in high numbers from donors 50 . Alternatively, exosomes secreted from MSCs can be leveraged as alternatives to live-cell therapeutics as they carry several immunotherapeutic components of the parent MSCs, including cytokines, signalling lipids and mRNA 51 . Proteins can be designed to target cytokines, such as TNF, IL-1 and IL-6, or to dampen systemic inflammation in patients with immune disorders, by blocking antibodies or by inhibiting cytokines through receptor binding 52 . Cytokine inhibitors have shown excellent efficacy in treating inflammation in clinical trials 53 ; for example, the anti-TNF antibody Humira in the treatment of rheumatoid arthritis ( NCT00049751 ) 54 . Antibodies can be deployed for targeted vasculature blocking to prevent the transmigration of immune cells and thus stop inflammatory activation 55 , 56 . Bimosiamose, for example, is a small-molecule inhibitor of adhesion proteins L-selectin, E-selectin and P-selectin, which showed a promising reduction of inflammation in chronic obstructive pulmonary disease, psoriasis and asthma patients during phase II clinical trials 55 , 56 . However, cytokine-targeting therapeutics lack specificity 52 , with reduced haematopoiesis as one of the most common side effects, ultimately leading to increased infections owing to reduced WBC populations 52 . Immune cells can further be prevented from interacting with inflammation by specific immune cell blocking. Obstructing complement receptors (C5aR) 57 , cytokine receptors (such as IL-1R and IL-6R) 52 , B cell activation receptors (CD20) 58 , 59 and adhesion receptors (PSGL1, α4β7 integrin, αEβ7 integrin and very late antigen 4 (VLA4)) 55 , 56 , 60 on the surface of immune cells can selectively inhibit inflammatory activation. For example, the small molecule CCX168 is a C5aR inhibitor that blocks binding to the activating complement protein C5a on neutrophils and has demonstrated efficacy in reducing inflammation in a phase III clinical trial ( NCT02994927 ) for vasculitis (inflammation of blood vessels) 57 , 61 . Similar to cytokine targeting, targeted immune cell blocking can have systemic consequences owing to redundancies within the immunity cascade 62 . Furthermore, complexities associated to ligand-receptor-mediated responses can cause unintended downstream signalling effects 62 . Cellular and antibody-based approaches, however, remain limited owing to the complexity of the inflammatory cascade, and they do not target the root cause of acute inflammation 63 . Alternatively, polymeric antibody-based approaches can mitigate overzealous inflammatory responses in acute lung injury (ALI), ARDS, sepsis, asthma, rheumatoid arthritis, type I diabetes, coagulopathic diseases and neurodegenerative diseases.

Polymeric particle-based therapeutics are extremely versatile and are therefore an excellent tool for designing treatment strategies against inflammatory diseases. By designing the material, size and shape of particles, sites and cell subtypes can be specifically targeted to distinct inflammatory diseases. Particle-based therapeutics have substantially improved the clinical efficacy of a variety of therapies, including therapies for endometriosis, cancer, growth failure, gum disease and mood disorders 123 . Despite a range of clinically available polymeric-particle-based therapeutics and the plethora of literature on the topic, only 12 PLGA particle-based formulations have been approved by the FDA over the past 30 years 123 , 189 . This stark contrast between research and clinical approval stems mainly from translational inconsistencies between animal models and humans. Although our understanding of inflammatory pathways is constantly evolving, the exact relation between particle design and subsequent inflammatory responses remains to be investigated. For example, the immune cells work in concert with complement pathways to respond to all invading foreign materials, including particles designed as immunotherapeutics 24 . Despite this known involvement, in vitro assays are limited owing to the difficulties of recapitulating inflammatory signalling pathways in a test tube 24 . Additionally, complement reactions vary across species, making it challenging to develop reliable in vivo assays for clinical translation 24 . Pigs have high CARPA reactions to particle-based therapeutics and have become an expensive but reliable model for a variety of applications 24 , 190 . Slower infusion rates, coupled with optimized surface properties, can help to prevent CARPA reactions 190 . However, screening assays need to be developed to investigated how to prevent CARPA reactions. The design of new particle-based therapeutics further poses challenges in terms of regulatory approval. Unlike systemic, carrier-free therapeutics, particle-based medicines are typically composed of polymeric vehicles, therapeutic agents and surface modifications 191 – 193 . A slight change in any of these components can considerably alter particle function, biodistribution and toxicity, which makes regulatory evaluation challenging. For this reason, the National Cancer Institute instigated the establishment of the Nanotechnology Characterization Laboratory (NCL) to develop standardized assays to characterize particle-based therapeutics and related toxicities 191 . The NCL was established to streamline the clinical trial and FDA approval process; however, these procedures are designed for cancer therapeutics and not for generalized therapies 191 – 193 . Importantly, NCL guidelines, such as prolonged evasion of the mononuclear phagocyte system, inherently exclude particle-based therapeutics that are designed to target circulating phagocytes 193 . Despite these translational hurdles, the consistent progress made by the scientific community and a streamlined approval process could revolutionize particle-based therapeutics. Particle technologies are practical solutions for several severe inflammatory diseases. The customizable nature and the physical and chemical attributes of particles fit the demand for innovative clinical applications, including the treatment of system-wide inflammation and vaccine development 64 , 117 , 194 – 196 . The synthesis of polymers can now be fine-tuned; however, the clinical translation of polymeric-based particles remains limited owing to a lack of scaling-up technologies for the fabrication of non-spherical particles in large quantities. The physicochemical parameters governing laboratory-batch particle fabrication are often very complex, and so large-scale production workflows need to be developed. Fortunately, promising large-scale processes are currently being explored for complex-shaped particle fabrication, including lithography-based and microfluidics technologies 197 , 198 . Although more work is needed to overcome these obstacles, the rapid expansion of particle-based medicine can offer state-of-the-art solutions to global problems 199 .

For the treatment of inflammatory conditions, the material, size, shape, surface chemistry and deformability of polymeric particles can be modified to ensure efficient interaction with immune cells (Fig.  4 ). Fig. 4 Particle optimization. a | Materials can be designed to be biodegradable, anti-inflammatory, biomimetic or stimuli-responsive. b | Microparticles are phagocytosed and target the vascular wall, whereas nanoparticles permeate the vasculature. c | Polymeric materials can be formulated into a variety of shapes to target specific cell populations. d | Functional groups on polymeric materials can be conjugated with targeting ligands or stealth polymeric chains. In addition, the polymer can be positively or negatively charged. e | Particle rigidity can be modified to selectively target specific cell populations and endothelium in the blood. Soft particles are ideal for marginating along the endothelium in blood flow compared to rigid particles. RBC, red blood cell; PEG, polyethylene glycol; MMP, matrix metalloproteinase. a | Materials can be designed to be biodegradable, anti-inflammatory, biomimetic or stimuli-responsive. b | Microparticles are phagocytosed and target the vascular wall, whereas nanoparticles permeate the vasculature. c | Polymeric materials can be formulated into a variety of shapes to target specific cell populations. d | Functional groups on polymeric materials can be conjugated with targeting ligands or stealth polymeric chains. In addition, the polymer can be positively or negatively charged. e | Particle rigidity can be modified to selectively target specific cell populations and endothelium in the blood. Soft particles are ideal for marginating along the endothelium in blood flow compared to rigid particles. RBC, red blood cell; PEG, polyethylene glycol; MMP, matrix metalloproteinase. Material selection affects the release of byproducts and particle accumulation, thereby influencing immune cell modulation and inflammatory signals (Fig.  4a ). For example, phagocytosis of non-degradable polystyrene by neutrophils can induce inflammatory neutrophil phenotypes 108 . By contrast, degradable PLG particles confer anti-inflammatory traits to DCs and monocytes 121 . Degradable polymeric materials are ideal for particle-based therapeutics of inflammatory diseases, because they are easily modified, optimized, produced in large quantities, and most importantly, do not accumulate in the body. PLGA has been widely explored for polymeric nanoparticle design in the context of inflammation owing to its ease of synthesis, manipulation and biocompatible degradation products (lactic and glycolic acids) which can be metabolized in vivo 122 . PLGA has been incorporated into several US Food and Drug Administration (FDA)-approved particle-based therapeutics and is thus considered a safe material 123 . In small quantities, the primary degradation byproduct of PLGA, lactic acid, has anti-inflammatory properties on both macrophages and DCs 114 , 121 . Depending on the molecular weight and thus degradation rate of the specific PLGA polymer, these anti-inflammatory properties may vary 121 . For example, low-molecular-weight PLGA (10 kDa) that degrades quickly will have inherently anti-inflammatory properties, whereas a high-molecular-weight PLGA polymeric particle will take longer to degrade and may have initial inflammatory properties prior to degradation 121 , 124 , 125 . However, PLGA particles can be used to deliver anti-inflammatory therapeutics such as nonsteroidal anti-inflammatory drugs (NSAIDs) to further reduce inflammation and to overcome any immediate side effects of PLGA degradation 126 . In treating neurological inflammation, particles comprised of polymerized phosphatidylserine, a marker for apoptotic cells, reduced inflammation of activated microglial cells and macrophages in vitro 127 . Although in vitro studies of these biomimetic particles have proved promising, in vivo experiments have only been completed in a myocardial infarction mouse model using phosphatidylserine loaded liposomes, resulting in improved angiogenesis and scar formation 128 . Polymeric particles can also be fabricated from polymerized anti-inflammatory compounds. Degradable polymers can be functionalized with a range of anti-inflammatory agents, including aspirin, naproxen and ibuprofen 129 , 130 . The resulting compound can then be synthesized into a particle by single oil–water emulsion 117 , 131 . Intravenous injection of the resulting polymerized salicylic acid (PolyAspirin) particles have shown to alleviate lung inflammation in an endotoxin and a bacterial mouse model of ALI and ARDS, respectively 117 . PolyAspirin particles more efficiently diverted neutrophils from the inflamed lungs and further reduced inflammatory cytokines compared to non-treated polystyrene and PLGA particles 117 . One suggested mechanism is that interactions between neutrophils and PolyAspirin particles prevent the initial accumulation of neutrophils in the lungs, and the degradation of PolyAspirin may ameliorate inflammation 117 . Degradable polymeric particles can also be designed to degrade at the site of inflammation, typically within the inflamed tissue space, by incorporating stimuli-responsive properties. Site-dependent degradation at low pH or through cleavage by matrix metalloproteinases (MMPs) can be implemented in particles, allowing the stimulus-triggered release of incorporated anti-inflammatory drugs 132 . Furthermore, vanillyl alcohol, an antioxidant and anti-inflammatory agent, can be incorporated into copolyoxalate through hydrogen-peroxide-sensitive peroxalate ester bonds, resulting in a polymer that can easily be formulated into a particle-based therapeutic 125 . The particle then degrades through hydrogen peroxide scavenging when exposed to nitric oxide, a molecule heavily expressed by the endothelium during inflammation 125 . Similarly, naproxen, an anti-inflammatory drug, can be modified with the ROS scavenging linker, phenylboronic acid (PBA), and then conjugated onto dextran 133 . The modified PBA–dextran can then be formulated into nanoparticles together with a pH-sensitive acetylated dextran, resulting in significantly reduced cytokine release from stimulated macrophages in vitro 133 . Crosslinked poly-amino acid-conjugated polyethylene glycol (PAAP) is another example of a degradable polymer that breaks down at low pH and can be loaded with anti-inflammatory proteins, such as DNAse1, to degrade NETs 134 . ROS-responsive poly(propylene sulfide) (PPS) microparticles loaded with curcumin have shown in vivo efficacy after being intramuscularly injected at the site of ischaemia in a diabetic mouse model 135 . When exposed to ROS, the PPS particles degraded, leading to ROS scavenging and curcumin release. Here, the curcumin functioned synergistically with PPS particles; however, non-loaded PPS particles also had therapeutic properties, suggesting that ROS scavenging may be enough to reduce inflammation 135 . Overall, versatile stimuli-responsive materials have great potential for treating inflammatory diseases as well as for immunotherapeutic applications 136 . The size customizability of particulates makes them particularly interesting for immunotherapeutics, as they can be designed to mimic pathogens and airborne particles while driving specific immune responses 124 , 137 (Fig.  4b ). Upon particle administration, microparticles ( ≥ 1 μm) are rapidly cleared through phagocytosis by macrophages, including rat alveolar macrophages, murine peritoneal macrophages and human spleen macrophages 138 – 140 . By contrast, nanoparticle uptake can occur through multiple mechanisms, including phagocytosis, pinocytosis, caveolae-, clathrin- and scavenger receptor-mediated endocytosis 139 , 140 . Nanoparticles have a low risk of capillary occlusion, and they travel passively across permeable vasculature, which is a common feature of inflammation 141 . The passive transport of nanoparticles makes them ideal candidates for accumulation within the inflamed tissue. For example, PEGylated polyalkylcyanoacrylate nanoparticles injected into an experimental autoimmune encephalomyelitis rat model, can accumulate in the central nervous system 142 . The passive passage of nanoparticles through the blood–brain barrier is attributed to an increase in cerebrovascular permeability, characteristic of such animal model of brain inflammation. Another example is the spleen, where nanoparticle accumulation is greater for 500 nm and 100 nm than for 20 nm polystyrene particles, as shown in an endotoxin-induced systemic inflammation mouse model 143 . Optimum particle size is also crucial for particle retention to inflammation sites that do not intend to reach the vasculature or lymphatic system. For example, intra-articular injection of poly( d , l )-lactic (PLA) particles in a mouse model of arthritis showed that 300 nm and 3 μm particles could rapidly diffuse out of the inflamed joint, hindering long-term accumulation and thus, therapeutic benefit 144 . Particle accumulation and retention at inflammation sites through vascular immunotargeting is also dependent on the carrier size. For example, vascular-targeted nanoparticles often lack high targeting efficiency owing to poor vascular wall localization 110 – 113 , 145 . Although micrometre-sized particles can exhibit higher vascular targeting, they are more susceptible to immune cell uptake and dangerous capillary occlusion compared to nanoparticles 110 , 145 , 146 . Particle association with inflammatory cells could also provide therapeutic benefits for unresolved inflammatory conditions through immune cell rerouting 109 . For intravenously injected therapies, particle–cell association can be improved by exploring microspheres as therapeutic carriers. Micrometre-sized carriers display enhanced migration out of the red blood cell core in blood flow, consequently co-localizing these particles with leukocytes that are also enriched near the blood vessel wall 110 , 112 . Polymeric microparticles successfully modulated inflammation through cell–particle interaction in multiple inflammatory mouse models 109 , 147 ; daily intravenous injection of drug-free polymeric microparticles target inflammatory monocytes in the circulation and redirect their migration out of the injured site 147 . Microparticles can further prevent neutrophil adhesion to inflamed tissue in vitro, where selectin-targeted polystyrene (≥2 μm) particles reduced neutrophil adhesion to activated human umbilical vein endothelial cells (HUVECs) more efficiently compared to nanoparticles at equal concentration 108 . Therefore, therapeutic use of particles requires an optimum size range design, depending on the type of inflammatory condition, desired immune response and route of particle administration. Phagocytic cells internalize pathogens and airborne particles of various size and shape (Fig.  4c ). Despite considerable progress in understanding the mechanisms of cellular recognition of conventional spherical carriers, our knowledge about the effect of the carrier’s shape on phagocytosis and the subsequent immune response remains limited. Early reports describe that the particle axis of elongated polystyrene particles modulates the mechanism of macrophage uptake 148 . Macrophages proceed with phagocytosis only if the first point of contact is at the minor axis (that is, the smaller side) of elongated particles. Such axis-dependent uptake was attributed to actin remodelling, which is necessary for engulfing particles, indicating that the minor axis of elongated particles favoured actin cup formation rather than cell spreading on the particles 148 . Thus, owing to the high energy requirement for actin remodelling, high-aspect-ratio carriers such as elongated particles show reduced phagocytosis by macrophages compared to spherical carriers, ultimately increasing carrier residence time at sites of particle delivery 149 – 151 . When macrophages are exploited as therapeutic targets, other particle shapes can be explored; for example, low-aspect-ratio spherical and disk-shaped polystyrene particles are phagocytosed by macrophages at a faster rate than are elongated particles 152 . Unlike macrophages, both primary human and mouse neutrophils preferentially internalize rod-shaped particles over spherically shaped ones. Here, the selective particle-neutrophil uptake is independent of material type, and increasing aspect ratios of the particles increase phagocytosis 120 . The observed higher internalization of rod-shaped particles by neutrophils was associated with possible neutrophil-specific phagocytic mechanisms 120 , probably linked to the role of neutrophils as the primary human defence against bacterial infection, many of whom have elongated shapes. Such a shape-dependent internalization is an excellent opportunity to engineer particle-based therapies for neutrophilic inflammatory disorders. The advantages of non-spherical polymeric carriers for anti-inflammatory therapies go beyond their morphology-dependent and cell-type-specific uptake. In particular, rod-shaped and disk-shaped particles demonstrate greater particle margination within the blood compared to spherical carriers 153 – 155 . These geometries partially counteract hydrodynamic forces in the bloodstream, enabling a large contact area between carriers and cells, thus being attractive for vascular-targeted drug delivery. For example, rod-shaped and spherical polystyrene particles coated with anti-VCAM1 were designed to evaluate the effect of particle shape on binding affinity. Targeted elongated particles showed greater targeting efficiency than targeted spheres in inflamed brain endothelial cells in vitro 156 . ICAM1-targeted rod-shaped polystyrene nanoparticles also showed preferential accumulation in the endothelium of the brain and lungs of healthy mice compared to targeted spherical particles, providing opportunities to enhance selective organ targeting using shape effects 157 . Benefits of non-spherical particles for inflammatory therapies also include overcoming cardiopulmonary reactions caused by both complement activation and responsive macrophages. Unlike with spherical particles, rod-shaped and disk-shaped polystyrene particles did not induce cardiopulmonary distress post-intravenous injection in pigs 158 . Hence, shape modification of polymeric particles is an appealing strategy to leverage particle-based therapies for desired immune responses, and to circumvent adverse effects of such treatments. Surface chemistry and coatings of particulate systems substantially affect the carrier’s interaction with immune cells, including particle clearance and therapeutic effect. Intravenously injected particles are inevitably tagged by plasma proteins that form a protein corona. Particle parameters such as surface charge and hydrophobicity have an essential role in protein corona formation and composition, which dictates subsequent cellular interactions (Fig.  4d ). In general, hydrophobic particles showcase higher protein absorption than hydrophilic ones. Likewise, surface charge affects the level of absorption of plasma proteins. For example, increasing the negative surface charge of polystyrene nanoparticles boosts protein absorption, but not protein corona species 159 . The composition of protein corona can vary among particles with different levels of surface hydrophobicity and cationic or anionic surface charges, ultimately governing uptake by phagocytes 160 , 161 . Typically, proteins adsorbed onto particles behave as opsonins, enhancing particle internalization 161 . Rapid particle clearance is a major challenge for designing particle-based immunotherapies that aim to reach the vascular wall or inflamed or damaged tissue. The cellular uptake of carriers can be mitigated by modifying their surface with a hydrophilic polymer, such as PEG 161 – 163 . PEGylated particle formulations can evade uptake by immune cells, extending their blood circulation time 161 – 163 (Fig.  3a ). Besides intravenous delivery, PEGylated particles also showcase longer residence time through other routes of administration, such as the pulmonary route. Pulmonary delivery of non-spherical polymeric hydrogels functionalized with PEG coatings reduced mouse alveolar macrophage uptake in vitro and in vivo. Accordingly, PEGylated particles showed increased retention in the lungs and minimal inflammatory response for at least a month 164 . Although PEG is often recognized as immunologically safe and allergies caused by this compound are rare, some PEGylated drug formulations can trigger complement activation and, in a small portion of patients, lead to severe anaphylaxis 165 – 169 . Additionally, the widespread use of PEGylation in pharmaceutical research has led to the discovery of PEG-specific antibodies that compromise its potential efficacy 67 , 170 . For example, increased clearance of PEGylated particles from blood circulation in mouse and rats have been reported after repeated doses, particularly for liposome carriers 67 , 170 – 172 . Thus, PEG alternatives are being explored, including biodegradable polymers such as poly(glutamic acid) (PGA) and ionic liquid coatings 173 , 174 . The immune evasive effects of particle PEGylation have mainly been explored for macrophage and monocyte uptake and less for neutrophils, despite their important role in immune response modulation and pathogen removal 163 , 175 . Surprisingly, PEGylation of carriers had the opposite effect on particle uptake in human blood; PEGylated polystyrene or PLGA particles showed increased uptake by human neutrophils compared to their non-PEGylated counterparts 175 . It was determined that factors present in the human plasma contribute to the lost immune evasive properties of PEGylated particles 175 . Surface modification further includes decorating particles with vascular-targeted ligands to localize and accumulate carriers at sites of inflammation or injury. Defining the optimal ligand surface density is essential to prevent suboptimal targeting of the endothelium or nonspecific targeting effects. In general, high ligand density on the particle surface increases the probability of encountering the specific binding partners and reduces particle-detaching forces owing to increased multivalent interactions 107 , 176 , 177 . However, excessive ligand density may inhibit optimal carrier binding to target cells owing to antibody steric hindrances or overcrowding. Optimal ligand surface density is also crucial in controlling selective binding of vascular-targeted carriers to pathological vasculature while minimizing binding to healthy tissue sites 107 . For example, carriers targeted with excessively high anti-ICAM1 surface density face a high off-target risk owing to the ubiquitous basal expression of ICAM1 on the vascular wall in healthy tissues 102 . In a mouse model of ALI, low density of ICAM1 on poly(4-vinylphenol) (PVPh) nanoparticles increased selective binding of these nanoparticles to inflamed pulmonary tissue relative to healthy vasculature 177 . By contrast, high ICAM1 density resulted in nanoparticle binding to both healthy and injured endothelium 177 . In vitro, in vivo and in silico binding assays showed that a low ligand density minimizes binding to areas with low receptor expression but maximizes binding to surfaces with highly expressed receptors 107 . In the case of inflammation, receptor expression increases at the vascular wall, improving the likelihood that particles decorated with a low density of ligands finding the receptors to enable adhesive and multivalent interaction 107 . In summary, the examples above illustrate the heterogeneous nature of particle surface chemistry on broad aspects of blood circulation and particle uptake. Specifically, the differences observed in immune cell populations affect the design of particle-based immunotherapies. Likewise, it showcases the importance of designing safe formulations to minimize exacerbation of inflammatory and allergic responses (Table  2 ). Table 2 Particle coatings to promote stealth characteristics Stealth coating Mechanism of action Clinical applications Advantages Disadvantages Polyethylene glycol (PEG) Hydrophilic polymer generates steric repulsion, reducing protein adsorption 202 Chronic inflammatory diseases (multiple sclerosis, arthritis, Crohn’s disease), gout, haemophilia, chronic kidney disease, prostate cancer, leukaemia, acromegaly, and hepatitis B and C 203 Biocompatible 202 FDA approval for human use 123 Tuneability: effective PEGylation depends on chain length, PEG chain architecture, grafting density 202 Does not completely eliminate protein adsorption 202 Does not protect polymer particle from phagocytosis by neutrophils in human blood 175 Chitosan Polysaccharide primary amino groups yield cationic properties 204 No FDA-approved particle-based formats, but has been evaluated in a clinical study in a nasal spray formulation of fentanyl chitosan 205 : chitosan nanoparticles enhanced bioavailability and systemic exposure 205 Biocompatible, biodegradable, non-toxic, stable 206 Fine tuning of properties by tuning molecular weight 206 Mucoadhesive 204 , 206 Antimicrobial 206 Controlled drug release 204 , 206 Weak non-fouling properties 204 Cell membrane Natural cell membranes are collected and coated onto synthetic particles 207 Polymeric nanoparticles coated with prostate-specific membrane antigens enhanced particle accumulation within prostate tumours 207 Prolonged circulation 207 Enhanced targeting capabilities 207 Ability to directly modulate immunity 207 Biomimetic Batch-to-batch variation 207 Zwitterion Contains both positive and negative moieties, creating overall neutral charge 202 ; both moieties interact with water molecules so that the hydration layer prevents opsonization 202 ; the anti-fouling properties increase as the distance between oppositely charged moieties decreases 202 No FDA-approved product Non-haemolytic 202 Reduced nonspecific protien adhesion 202 Cannot be used for active targeting 202 Difficult to tune surface properties 202 Cellular uptake is not inhibited 202 Ionic liquids Particles can be suspended in ionic liquid emulsions or covalently bonded with ionic liquids 208 ; intramolecular and intermolecular interactions between the ionic liquid and particle/loaded drug determine the particle properties 208 No FDA-approved product Tuneable 208 Stealth 208 Antimicrobial 208 Stable 208 Mechanism of degradation is unknown 208 , 209 Particle coatings to promote stealth characteristics Polyethylene glycol (PEG) Biocompatible 202 FDA approval for human use 123 Tuneability: effective PEGylation depends on chain length, PEG chain architecture, grafting density 202 Does not completely eliminate protein adsorption 202 Does not protect polymer particle from phagocytosis by neutrophils in human blood 175 Chitosan Biocompatible, biodegradable, non-toxic, stable 206 Fine tuning of properties by tuning molecular weight 206 Mucoadhesive 204 , 206 Antimicrobial 206 Controlled drug release 204 , 206 Cell membrane Prolonged circulation 207 Enhanced targeting capabilities 207 Ability to directly modulate immunity 207 Biomimetic Zwitterion Non-haemolytic 202 Reduced nonspecific protien adhesion 202 Cannot be used for active targeting 202 Difficult to tune surface properties 202 Cellular uptake is not inhibited 202 Ionic liquids Tuneable 208 Stealth 208 Antimicrobial 208 Stable 208 Particle deformability provides a tuneable factor in particle-based therapeutic design. By adjusting the polymer content in the polymer precursor solution or the functionality of the polymer building block, the degree of crosslinking can be modulated 178 . Therefore, the particle’s elasticity and flexibility can be tuned to improve leukocyte avoidance, vascular localization, vascular navigation and biodegradation 178 – 180 . Thus, recent work has sought to understand the role of particle deformability in designing particle-based drug carriers. Studies investigating the effect of particle elasticity with regards to drug carrier design work across a range of moduli, typically from around 10 kPa up to approximately 1,000 kPa (ref. 178 ). Within this range, a softer particle benefits from a longer circulation time, thereby avoiding clearance by leukocytes. Conversely, comparatively stiffer particles exhibit a shorter circulation time and increased phagocytosis 181 – 183 . For example, stiff (3,000 kPa) PEG-based nanoparticles are engulfed at a faster rate by J774 macrophages compared to softer (10 kPa) ones in vitro. Soft particles also have a higher persistence in the blood for up to four hours, after which this difference is substantially reduced 181 . Similarly, micrometre-sized, rigid polyacrylamide beads having a threefold-higher modulus have a greater propensity of being phagocytosed by bone-marrow-derived macrophages in vitro compared to softer beads 182 . Bone-marrow-derived monocytes exhibit similar behaviour, with up to threefold-reduced uptake of soft (1.3 kPa) disk-shaped particles compared to their rigid counterparts (15 kPa) 183 . These studies suggest that deformability is an important parameter to consider, especially in avoiding leukocyte clearance for vascular-targeted approaches to immunomodulation. In addition to leukocyte–particle interactions, the role of elasticity in modulating particle accumulation at specific sites has been explored 179 , 180 . Soft PEG-based particles (20–100 kPa) outperform stiffer particles (300–500 kPa) in an in vitro blood flow system under various shear rates; at low rates (500 s −1 or less), softer particles adhere to the endothelium at the same or greater rate compared to their rigid counterparts 179 , 180 . This trend is reversed at high-shear (1,000 s −1 or greater) conditions 179 . Additionally, soft hydrogel microparticles can shuttle nanoparticles to the vascular wall 184 ; for example, intravenous delivery of nanoparticles to the endothelium of mice can be enhanced by loading them into deformable microparticles 184 . Deformable particles in particular are better suited for immunomodulatory approaches that do not rely on cellular uptake for activity, especially in the case of loading and delivering smaller particles to a site of inflammation such as in vascular-targeted approaches. Deformable particles can also be used to mimic cells, such as platelets, for therapeutic applications. For example, despite showing great potential for treating coagulopathic diseases (in which clotting does not occur fast enough) 185 , 186 , platelet transfusions can still result in immunogenic side effects 187 . Poly( N -isopropylacrylamide- co -acrylic-acid) microgel particles (1 µm) conjugated to a fibrin antibody can mimic the size, morphology and fibrin binding of platelets 187 . These platelet-like particles increase clot formation and stability in traumatic brain injuries, preventing post-traumatic neuroinflammation. Additionally, they have a longer shelf life compared to natural platelets, with potential applicability for treating other haemorrhagic bleeding disorders that lead to downstream inflammation 187 . Deformability mainly influences particle circulation and uptake; however, it could also be tuned to achieve stimuli-responsive properties. For example, deformable materials designed to degrade at specific sites with low pH or high MMP concentrations enable selective protein or drug release 188 .

Drug delivery systems can be applied to avoid adverse effects related to systemic treatments of inflammation. For example, lipid nanoparticles and lipid-based drug delivery systems, such as liposomes, show high biocompatibility and cargo stability, as demonstrated by the recent success of lipid-nanoparticle-based mRNA vaccines 64 . These carriers have mainly been deployed for priming immune cells in developing cancer vaccines and, recently, for vaccine against SARS-CoV-2 viral infections 64 – 66 . However, lipid systems are limited by poor structural stability, a limited range of potential cargos, low drug loading and poor circulation time 67 – 69 . Alternatively, synthetic polymer-based drug carriers provide spatiotemporal control over release and reproducibility compared to other drug delivery formulations 70 , 71 . The choice of LNPs versus polymeric carriers depends on the use and type of drug cargo, the cellular target and the in vivo delivery route 72 . Liposomal particles were first used as drug delivery vehicles in cancer therapy to improve drug delivery efficiency to solid tumours; however, immunotoxicity remains a serious side effect for intravenous formulations 73 . Surface modifications of particle-based drug carriers, such as the grafting of polyethylene glycol (PEG) chains on particles, often referred to as PEGylation, were intended to help particles evade the immune system, while also increasing efficiency and reducing side effects of cancer therapeutics 74 . However, immunotoxicity issues persist in intravenous (as opposed to intramuscular) injections, leading researchers to re-evaluate the therapeutic potential of immune cell‒particle interactions 75 – 77 . The tendency of particulate therapies to interact with immune cells can be independently exploited as a therapeutic strategy. For example, in an in vivo model of melanoma, mice treated with poly(lactic- co -glycolic acid) (PLGA) particles loaded with a chemotherapeutic drug (paclitaxel) in combination with the CXCL1 chemokine, had significant reduction in tumour size compared to all control groups 78 . Interestingly, neutrophil uptake of the particles followed by chemotaxis to the CXCL1-treated tumour was responsible for the decrease in tumour burden, pointing to the potential applicability of particle-based therapeutics in immune-cell-related diseases 78 . Therefore, particle-based therapeutics have been explored for tissue-localized and systemic treatment of inflammation (Fig.  2 and Table  1 ). Fig. 2 Targeting immune cells by particle-based therapeutics. a | Tissue-resident immune cells, such as macrophages, dendritic cells, memory T cells and B cells, are directly targeted through local tissue immunomodulation. Particle administration can passively reprogram inflammatory cells for immunosuppressive purposes. b | Polymeric particles are designed with ligands or antibodies that target leukocyte adhesion molecules overexpressed in inflamed vasculature. The ligand-mediated anchoring of particles to the endothelium permits accumulation of drug carriers to pathological areas, halting inflammation. Targeted particles can also competitively block binding sites used in immune cell migration, preventing immune cell accumulation at sites of inflammation. c | Polymeric particles can divert or reroute circulating blood-resident cells, such as neutrophils and monocytes, from inflammation sites through particle–cell association and cell phenotypic changes after particle uptake. WBC, white blood cell. Table 1 Targeting cell populations in different locations Targeting strategy Targeted cells Route of administration Platform Size and shape Modifications Notes Refs. Tissue-specific DCs Subcutaneous PLGA ~1 µm and ~30 µm spheres Loading of TGFβ1, GM-CSF, vitamin D 3 , type II collagen and insulin Larger particles recruit and condition DCs through release of GM-CSF and TGFβ1. Simultaneously, smaller loaded particles are phagocytosed by local DCs at the injection site for reprogramming and migrate to lymph node 85 , 86 ~800 nm spheres TGFβ surface modifications and loading of OVA 323–339 peptide Co-stimulatory particles are phagocytosed by local DCs for immune reprogramming 88 In vitro study Polystyrene 150 nm and 2 µm spheres, 3× stretched rods from 150 nm and 2 µm spheres Physical absorption of poly I:C or CL264 Spherical particles show stronger DC activation than rod-shaped particles. Nanospheres promote the strongest activation 200 Lymph-node-resident immune cells (such as DCs or T cells) Intranodal PLGA ~5 µm and ~ 300 nm spheres Loading of poly I:C PLGA particles reach the lymph node through direct injection. Microparticles release poly I:C at the site of injection for sustained DCs activation. By contrast, nanoparticles are rapidly phagocytosed by lymph-node-resident DCs and macrophages 201 ~3–4 µm spheres Loading of MOG peptide and rapamycin Intranodal injection of microparticles to promote polarization of T cells 89 Local phagocytic immune cells (such as macrophages) In vitro study PU ~35 nm and ~63 nm spheres Negative and positive surface charge Inhibition of M1 macrophage polarization after uptake of negatively charged nanoparticles 92 Ac-DEX ~829 nm spheres Loading of rapamycin Particles are phagocytosed by activated macrophages, reducing production of pro-inflammatory molecules through pH-dependent release of rapamycin from particle matrix 93 Polystyrene 0.5–3 µm spheres; major axis 0.35–2.5 μm, minor axis 0.2–2 μm rods; major axis 0.35–2.5 μm, minor axis 0.2–2 μm disks – Disk-shaped and spherical particles show enhanced macrophage uptake compared to elongated particles 152 Vasculature Activated endothelial cells Intravenous Polystyrene 500 nm and 2 μm spheres; 500 nm ESD (AR = 6) and 2 μm ESD (AR = 4) rods sLe a and anti-VCAM1 surface modification Targeted rod-shaped microparticles adhere at a higher rate than targeted microspheres to inflamed aortic segments and plaque 154 PLGA ~200 nm spheres γ3 peptide surface modification and loading of sparfloxacin and tacrolimus Targeted nanoparticles concentrate antibacterial and anti-inflammatory drugs at site of inflammation (lungs) 95 PAE 100 nm spheres Anti-ICAM1 surface modification and loading of TPCA-1 Targeted nanoparticles concentrate in the inflamed lungs and release anti-inflammatory drug from pH-responsive polymer matrix 97 In vitro study Polystyrene, silica and titania 500 nm spheres sLe a surface modification Dense nanoparticles adhere to inflamed HUVEC at a higher rate than neutrally buoyant nanoparticles 113 Circulating white blood cells Circulating phagocytes Intravenous Polystyrene, PLG, HPPS 2 µm, 500 nm and 15 nm spheres Unloaded or drug-loaded particles Particles passively target phagocytes in the bloodstream to divert them from sites of inflammation 109 , 115 , 116 Intraperitoneal or intravenous PLGA and PLA ~400 nm spheres Varied surfactants and molecular weight of polymer for fabrication Physiochemical properties of the particles influenced immunomodulatory effects 114 In vitro study Polystyrene 0.5–2 µm spheres Carboxylated, PEGylated or sLe a -coated particles Collisions in blood flow, particle binding to endothelium, and particle phagocytosis were found to reduce leukocyte adhesion to inflamed endothelium in blood flow 175 Neutrophils Intravenous PolyA 1 µm spheres Polymerized salicylic acid PolyA particle treatment in ALI and ARDS reduces inflammatory damage in lungs and enhance survival compared to PLGA and polystyrene particles 117 In vitro study PLGA 1–3 µm spheres or 1.5 µm (long axis) rods – Physical properties of particles preferentially target neutrophils through larger size or rod shape 119 , 120 Ac-DEX, acetalated dextran; ALI, acute lung injury; AR, aspect ratio; ARDS, acute respiratory distress syndrome; CL264, adenine analogue; DC, dendritic cell; ESD, equivalent spherical diameter; GM-CSF, granulocyte–macrophage colony-stimulating factor; HPPS, high-density lipoprotein-mimicking peptide-phospholipid scaffold; HUVEC, human umbilical vein endothelial cell; ICAM1, intracellular cell adhesion molecule 1; MOG, myelin oligodendrocyte glycoprotein; OVA, ovalbumin; PAE, poly(β-amino ester); PLG, poly(lactide- co -glycolide); PLGA, poly(lactic- co -glycolic acid); PolyA, PolyAspirin; poly I:C, poly(inosinic:cytidylic acid); PU, polyurethane; sLe a , sialyl Lewis A; TGFβ1, transforming growth factor β1; TPCA-1, 2-[(aminocarbonyl)-amino]-5-(4-fluorophenyl)-3-thiophenecarboxamide; VCAM1, vascular cell adhesion molecule 1. a | Tissue-resident immune cells, such as macrophages, dendritic cells, memory T cells and B cells, are directly targeted through local tissue immunomodulation. Particle administration can passively reprogram inflammatory cells for immunosuppressive purposes. b | Polymeric particles are designed with ligands or antibodies that target leukocyte adhesion molecules overexpressed in inflamed vasculature. The ligand-mediated anchoring of particles to the endothelium permits accumulation of drug carriers to pathological areas, halting inflammation. Targeted particles can also competitively block binding sites used in immune cell migration, preventing immune cell accumulation at sites of inflammation. c | Polymeric particles can divert or reroute circulating blood-resident cells, such as neutrophils and monocytes, from inflammation sites through particle–cell association and cell phenotypic changes after particle uptake. WBC, white blood cell. Targeting cell populations in different locations Ac-DEX, acetalated dextran; ALI, acute lung injury; AR, aspect ratio; ARDS, acute respiratory distress syndrome; CL264, adenine analogue; DC, dendritic cell; ESD, equivalent spherical diameter; GM-CSF, granulocyte–macrophage colony-stimulating factor; HPPS, high-density lipoprotein-mimicking peptide-phospholipid scaffold; HUVEC, human umbilical vein endothelial cell; ICAM1, intracellular cell adhesion molecule 1; MOG, myelin oligodendrocyte glycoprotein; OVA, ovalbumin; PAE, poly(β-amino ester); PLG, poly(lactide- co -glycolide); PLGA, poly(lactic- co -glycolic acid); PolyA, PolyAspirin; poly I:C, poly(inosinic:cytidylic acid); PU, polyurethane; sLe a , sialyl Lewis A; TGFβ1, transforming growth factor β1; TPCA-1, 2-[(aminocarbonyl)-amino]-5-(4-fluorophenyl)-3-thiophenecarboxamide; VCAM1, vascular cell adhesion molecule 1. Local tissue immunomodulation can be achieved by stimulating or activating tissue-resident immune cells, for example, by vaccination 79 – 84 . However, particle-based vaccine therapeutics can also be designed to target and passively reprogram inflammatory DCs within a draining lymph node, acting as regulatory vaccines for immune suppression in autoimmune diseases 81 , 85 , 86 (Fig.  2a ). DC-based immunomodulatory therapeutics mostly rely on autologous transplantation of exogenously tolerized DCs, which can be costly and plagued by variability of the transplanted DC populations 85 . Injecting specialized polymeric particle-based therapeutics circumvents the limitations of autologous transplantation by directly conditioning and reprogramming DC populations within the host’s lymphoid organs 87 . One approach by which polymeric particles modulate tissue relies on tissue draining; for example, a steroid hormone can be co-delivered with immunosuppressive cytokines loaded into a PLGA nanoparticle for the treatment of rheumatoid arthritis 85 . This immunosuppressive approach reprogrammed DCs within the draining lymph node, effectively regressing rheumatoid arthritis symptoms in mice 85 . Similar particle-based treatments containing immunosuppressive cytokines and autoantigens within PLGA particles were shown to retrain DCs in mouse models of type 1 diabetes and multiple sclerosis, reducing disease presentation and progression 86 , 88 – 90 . More specifically, co-delivery of autoantigens and cytokines through a PLGA-based particle therapeutic provided the antigen-specific immune tolerance necessary to reduce disease presentation and progression in a mouse model of multiple sclerosis 88 . The three main routes for therapeutic particle delivery are intravenous, mucosal and direct tissue injection (subcutaneous or lymphoid) (Fig.  3 ). Subcutaneously injected particles traffic to lymph nodes through tissue-resident antigen-presenting cells and ultimately induce tolerance by DCs, following natural peripheral tolerance pathways 88 . Conversely, intravenously injected particles affect tolerance through liver-resident antigen-presenting cells (Kupffer cells) and macrophage scavengers in the spleen 88 . Similar to subcutaneous injection, particle-based therapeutics directly injected into the lymph node stimulate an increase in the expansion of antigen-specific effector T cells 89 (Box  1 ). Fig. 3 Route of administration for immunomodulation. The effectiveness of particle-based therapeutics strongly depends on the route of administration. a | Intravenous routes allow systemic delivery, but pose challenges including quick clearance and complement activation. Grafting of polyethylene glycol (PEGylation) reduces protein corona formation allowing for improved targeting and reduced clearance. b | Mucosal drug delivery allows direct targeting of inflamed tissue. Here, particles are designed to target or travel through mucous membranes of diseased tissue. c | Direct tissue drug delivery can be applied to target lymphoid tissue by intramuscular, subcutaneous or intranodal injections. VTC, vascular-targeted carrier. The effectiveness of particle-based therapeutics strongly depends on the route of administration. a | Intravenous routes allow systemic delivery, but pose challenges including quick clearance and complement activation. Grafting of polyethylene glycol (PEGylation) reduces protein corona formation allowing for improved targeting and reduced clearance. b | Mucosal drug delivery allows direct targeting of inflamed tissue. Here, particles are designed to target or travel through mucous membranes of diseased tissue. c | Direct tissue drug delivery can be applied to target lymphoid tissue by intramuscular, subcutaneous or intranodal injections. VTC, vascular-targeted carrier. Polymeric particles can not only be designed to reprogram immune cells, but can also be exclusively immunosuppressive and directly target sites of inflammation, for example, by being loaded with immunosuppressive drugs 81 , 91 – 93 . This approach is beneficial given that most immunosuppressants used to treat inflammatory diseases can oversuppress the immune system, causing several side effects. For example, rapamycin acts as an immunosuppressant, anti-inflammatory and anti-proliferative agent that can lead to diarrhoea, headaches, myelosuppression and hyperlipidaemia 93 . Encapsulating rapamycin and similar drugs in polymeric particles could prevent undesirable side effects, allowing efficient drug delivery at sites of inflammation in rheumatoid arthritis, multiple sclerosis and graft-versus-host disease 93 , 94 . Particle-based therapeutics with surface decoration of targeting ligands can improve targeting of tissue-specific inflammation sites following intravenous administration. For example, localized infection-related inflammation in the lungs of mice can be treated with intravenously injected PLGA particles that bind specifically to inflamed cells (tumour necrosis factor receptor 1 (TNFR1) on macrophages and intercellular adhesion molecule 1 (ICAM1) on endothelium) at the diseased site 95 – 97 . To target antigen-presenting cells such as lung-resident macrophages, mesoporous silica nanoparticles can be coated with ligands, including TNFR1, to induce endocytosis 96 . As the particles degrade, they slowly release immunosuppressant agents to reduce inflammatory signals 96 . This approach can be used for a range of diseases, including autoimmune and acute inflammatory diseases 95 , 96 , 98 , 99 . The therapeutic efficacy of particle-based immunomodulation therapies is strongly dependent on the route of administration. Intravenous, local, oral and inhaled delivery each require distinct particle designs 210 , 211 . Intravenous drug delivery . Intravenously delivered particles can access the entire body and provide immediate drug effects. Systemic administration of therapeutics is typically hindered by rapid capture and clearance of particulates by circulating immune cells, ultimately preventing prolonged accumulation at sites of inflammation 212 , 213 . The circulation time of intravenously injected therapies can be improved by modifying the physiochemical properties of particles (for example, hydrophobicity) through functionalization with stealth polymers, such as polyethylene glycol (PEG). PEGylation of particles has become standard in the design of long-circulating particle formulations to improve systemic administration 170 , 212 . The size and charge of particles can be optimized further to improve blood residence time and promote specific interactions with circulating white blood cells (WBCs) 213 , 214 . However, challenges remain for intravenously injected particle-based formulations, including particle-induced complement activation and development of anti-PEG antibodies 23 , 165 , 212 , 215 , 216 . Mucosal drug delivery . Oral and inhaled administration allows the accumulation of particles in pathological tissue of the lungs and gastrointestinal tract through mucosal tissue targeting. The therapeutic efficacy of oral and inhaled administration is strongly dependent on tissue-resident cells and organ-specific barriers, such as the mucus layer. The airway and the gastrointestinal tract are common target sites for asthma, cystic fibrosis and inflammatory bowel disease therapies 217 , 218 . The mucus layers coating these organs pose a barrier, preventing particles from reaching the underlying epithelial surfaces. Mucosal networks vary in different diseases, making customization of drug delivery systems challenging 217 – 219 . For example, in airway diseases, mucus overproduction and a decrease in mesh pore size hinder the diffusion of pulmonary drug therapeutics 220 . Unlike the dense mucus network in lung diseases, the mucus layer in active ulcerative colitis shows abnormal penetrability, which is advantageous for the design of colon-targeted oral particle formulations, preventing inflammatory episodes and allowing particle uptake and retention within the colon region 221 . The surface and charge of particles can be modulated to increase muco-penetration and mucoadhesive properties of particles, for example, by coating carriers with PEG and mucolytic enzymes 222 – 225 . Accumulation and retention time of particles in the mucus can be improved by reducing the size of particles to the nanoscale, particularly in the inflamed colon 221 , 226 , 227 . Epithelia below mucosal surfaces can also be targeted by pH-dependent and functionalized nanoscale delivery carriers 221 , 228 – 230 . Once trapped in the mucosal lining, particles can interact with epithelium and ultimately reach the mucosa-associated lymphoid tissues 231 , 232 , where they can interact with a variety of immune cells. Direct tissue drug delivery . Particle-based therapeutics designed for tissue drug delivery are typically delivered directly into tissue by subcutaneous, intramuscular or intranodal injection. Tissue-specific injection routes are often used for vaccines to tolerize dendritic cell populations in lymph nodes. Intranodal injections can be the most effective owing to the ensured delivery to lymphoid tissue; however, getting a needle and syringe directly into a lymph node requires specific training and ultrasound technology 89 , 201 , 210 . These hurdles can be avoided by injecting therapeutics subcutaneously or intramuscularly, where they will drain to the lymph node 87 , 233 – 235 . However, efficacy may be affected owing to inefficiencies of particle transport to the lymph node. Vascular endothelial cells are major participants and regulators of inflammatory reactions. During acute inflammation, the endothelium rapidly changes its phenotype to support various stages of the inflammatory response. Primarily, activated endothelial cells facilitate the capture and extravasation of leukocytes to infected or damaged tissue 100 . During the inflammatory response, leukocyte adhesion molecules are overexpressed at the injured or inflamed endothelium 100 , 101 . Thus, particles can be designed to target the remodelled endothelium and underlying signalling pathways, for example, by surface modification with antibodies or cell surface proteins against leukocyte adhesion molecules, including selectins, ICAM1 and vascular cell adhesion molecule (VCAM1), which are known to be involved in leukocyte recognition, adhesion and extravasation 100 , 101 . Vascular-targeted carriers are advantageous for their ability to localize and accumulate at specific disease sites throughout the vasculature, providing controlled release of therapeutics and preventing systemic side effects 102 , 103 (Fig.  2b ). For example, PLGA and poly(lactic acid)-poly(ethylene glycol) (PLA-PEG) spheres coated with biotinylated antibodies against the selectins VCAM1 and ICAM1 (refs. 104 , 105 ) (Fig.  3a ) can target inflammation markers and exhibit selective adhesion towards inflamed endothelium both in vitro and in vivo 104 , 105 . Likewise, vascular-targeted carriers are often designed with dual adhesion receptors, simulating the multistep adhesion process of WBCs and improving particle adhesion properties 106 , 107 . The ability of vascular-targeted carriers to bind to leukocyte adhesion molecules can also be leveraged to competitively block excessive and unregulated immune cell trafficking, as occurs in conditions such as ARDS 108 . Although such an approach seems counterintuitive, preventing immune cell migration to pathological areas has already shown promising results at preventing tissue damage and accumulation of inflammatory cytokines in preclinical studies in ALI mice 109 . Limitations of vascular-targeted particles remain in terms of performance and functionality. In particular, diseased blood conditions and the physical characteristics of particles, such as size and ligand density, can affect their ability to target the inflamed endothelium, ultimately hindering full therapeutic potential 110 – 113 . Given that most immune cells or cell precursors involved in pathologic inflammation circulate through the bloodstream, intravenously injected immunotherapeutics are among the most prevalent therapeutic options 27 , 29 , 39 . Polymeric particles can be used to block circulating WBCs from excessive tissue migration during severe inflammation, diverting these inflammatory cells away from the injured tissue 109 , 114 – 116 . The primary mechanism by which polymeric particles achieve immunomodulation in the bloodstream is by interaction with WBCs. Phagocytosis of particles by WBCs affects cell physiology, including cytokine release, surface protein expression and gene expression 108 , 114 , 115 , 117 , which leads to the alteration of cell trafficking and signalling (Fig.  2c ). Degradable particles alleviate the possibility of long-term particle accumulation, and their byproducts may further provide anti-inflammatory or therapeutic effects. PLG-based particles for example, degrade into lactate, which reduces the inflammatory signals of DCs and macrophages 118 , as shown by drug-free PLG- or PLA-based particles in mice models of spinal cord injury and sepsis 114 , 115 . Particles fabricated from high-density lipoprotein-mimicking peptide-phospholipid scaffolds (HPPS) can also be loaded with the anti-inflammatory drug curcumin and designed to specifically target and redirect monocytes 116 . The latter are targeted by scavenger receptor class B type 1, a receptor that strongly interacts with high-density lipoproteins. In a mouse model of multiple sclerosis, treatment with curcumin-loaded HPPS led to a reduction in monocyte accumulation and morbidity at the injury site, further highlighting the potential of targeting innate immune cells, such as monocytes, to treat inflammatory diseases 116 . Further optimization of the physicochemical properties of particles, known to affect phagocytosis of WBCs, will be required to ensure recognition of different subsets of immune cells within the blood 119 , 120 .

Inflammation plays an essential part in the immune response against harmful stimuli and injury through recognition and containment of invading pathogens and toxins. Overly responsive or uncontrolled inflammation can lead to tissue damage and organ dysfunction 1 – 3 , and is associated with numerous human disorders, such as acute lung and liver injury, sepsis, asthma, inflammatory bowel disease, rheumatoid arthritis and neurodegenerative diseases 2 , 3 . Acute inflammation, regulated by the innate immune system, is responsible for the initial recognition of an inflammatory stimulus and focuses on the rapid containment of the offending pathogen or injury. Such a response aims at accelerating inflammatory resolution and typically lasts on a scale of hours to days (Fig.  1a ). If the acute inflammation response is excessive or fails to contain the inflammatory stimulus, the response is shifted to a chronic (pathological) phase characterized by prolonged inflammatory episodes and can last on a scale of weeks to years 3 – 5 . Although chronic inflammation has mainly been associated with cells from the adaptive immune system, the innate arm of the immune system also has a role, because, in chronic inflammatory diseases, repeated acute inflammatory episodes propagate tissue damage 2 , 3 , 5 , 6 . Recognizing the persistent role of acute inflammation in chronic diseases has, in part, contributed to the growing interest in developing therapeutics aimed at suppressing this acute response. In this Review, we summarize crucial aspects of the acute inflammation response and discuss particle-based therapies developed towards modulating or resolving this process. Fig. 1 Inflammatory cascade and strategies for modulating inflammation. a | Neutrophil recruitment and functions in inflammation. Neutrophils slowly roll along the endothelium by expressing P-selectin glycoprotein ligand 1 (PSGL1) and L-selectin, which bind to their corresponding ligands on the endothelium. Once at the site of inflammation, lymphocyte function-associated antigen 1 (LFA1) expressed on the neutrophil locks with intercellular adhesion molecule 1 (ICAM1) on the endothelium to initiate transmigration to the infected or inflamed tissue space. At the site of infection, neutrophils phagocytose pathogens or release reactive oxygen species (ROS), granules or neutrophil extracellular traps (NETs) to prevent pathogenic spread. Neutrophils then commit to apoptosis, initiating migration of other immune cell types. Monocytes and macrophages clean up dead cellular materials (pathogens and neutrophils) by efferocytosis and ultimately migrate to the liver and lymph nodes to remove and process pathogenic materials. Additionally, post efferocytosis monocytes shift to a pro-resolution phenotype to promote tissue restoration. b | Non-particle-based therapeutics for modulating inflammation. Stem cell therapies: transplanted mesenchymal stem cells (MSCs) can secrete immunosuppressive cytokines to promote regulatory T (T reg ) cell production and inhibit T helper 1 (T H 1) and T helper 2 (T H 2) cell differentiation or suppress immune cell recruitment and activation. Cytokines and antibody-based therapies: blocking antibodies or decoy receptors bind to inflammatory cytokines to inhibit their activity and dampen systemic inflammation. Vasculature blocking: biological agents prevent transmigration of immune cells by blocking specific leukocyte adhesion molecules on vascular endothelial cells, halting the inflammatory response. Targeted immune cell blocking: therapeutics can directly inhibit immune cells from pathological activation by blocking specific receptors on their surface. mAb, monoclonal antibody. a | Neutrophil recruitment and functions in inflammation. Neutrophils slowly roll along the endothelium by expressing P-selectin glycoprotein ligand 1 (PSGL1) and L-selectin, which bind to their corresponding ligands on the endothelium. Once at the site of inflammation, lymphocyte function-associated antigen 1 (LFA1) expressed on the neutrophil locks with intercellular adhesion molecule 1 (ICAM1) on the endothelium to initiate transmigration to the infected or inflamed tissue space. At the site of infection, neutrophils phagocytose pathogens or release reactive oxygen species (ROS), granules or neutrophil extracellular traps (NETs) to prevent pathogenic spread. Neutrophils then commit to apoptosis, initiating migration of other immune cell types. Monocytes and macrophages clean up dead cellular materials (pathogens and neutrophils) by efferocytosis and ultimately migrate to the liver and lymph nodes to remove and process pathogenic materials. Additionally, post efferocytosis monocytes shift to a pro-resolution phenotype to promote tissue restoration. b | Non-particle-based therapeutics for modulating inflammation. Stem cell therapies: transplanted mesenchymal stem cells (MSCs) can secrete immunosuppressive cytokines to promote regulatory T (T reg ) cell production and inhibit T helper 1 (T H 1) and T helper 2 (T H 2) cell differentiation or suppress immune cell recruitment and activation. Cytokines and antibody-based therapies: blocking antibodies or decoy receptors bind to inflammatory cytokines to inhibit their activity and dampen systemic inflammation. Vasculature blocking: biological agents prevent transmigration of immune cells by blocking specific leukocyte adhesion molecules on vascular endothelial cells, halting the inflammatory response. Targeted immune cell blocking: therapeutics can directly inhibit immune cells from pathological activation by blocking specific receptors on their surface. mAb, monoclonal antibody.