Nanodrug delivery systems for regulating microglial polarization in ischemic stroke treatment: A review

Abstract

The incidence of ischemic stroke (IS) is rising in tandem with the global aging population. There is an urgent need to delve deeper into the pathological mechanisms and develop new neuroprotective strategies. In the present review, we discuss the latest advancements and research on various nanodrug delivery systems (NDDSs) for targeting microglial polarization in IS treatment. Furthermore, we critically discuss the different strategies. NDDSs have demonstrated exceptional qualities to effectively permeate the blood–brain barrier, aggregate at the site of ischemic injury, and target specific cell types within the brain when appropriately modified. Consequently, NDDSs have considerable potential for reshaping the polarization phenotype of microglia and could be a prospective therapeutic strategy for IS. The treatment of IS remains a challenge. However, this review provides a new perspective on neuro-nanomedicine for IS therapies centered on microglial polarization, thereby inspiring new research ideas and directions.

Introduction

Stroke is a disease of the central nervous system (CNS), marked by high morbidity, disability, and mortality, and it has emerged as a major global public health issue that greatly threatens human well-being. ¹ Ischemic stroke (IS) is the predominant form of stroke, constituting approximately 87% of all stroke occurrences. ² Currently, the primary goal of clinical treatment of IS is to perform thrombolysis as soon as possible to restore blood flow in the ischemic area and save neurons in the ischemic penumbra. ³ However, because of the limited time available for thrombolysis and the numerous complications involved, only a few patients with IS can benefit from thrombolysis within the “golden hour.” Additionally, revascularization inevitably leads to secondary ischemia-reperfusion injury, exacerbating reactive oxygen species (ROS) production and inflammation, which may ultimately cause hemorrhagic transformation or malignant edema. ⁴ Efforts have been made recently to develop neuroprotective agents; however, their clinical applications are limited owing to difficulty in crossing the blood–brain barrier (BBB), low solubility, and short half-life. ⁵ Moreover, the presence of individual variations among patients in terms of their disease condition, sex, age, and other factors, poses a significant challenge, as many neuroprotective agents undergoing clinical trials are particularly vulnerable to unfavorable outcomes. ⁶

With the advent of precision medicine, researchers are realizing that only by focusing on pathological mechanisms can new therapeutic targets be discovered and traditional treatments revolutionized. Initially, local hypoperfusion caused by arterial occlusion leads to neuronal energy depletion, imbalanced Na⁺ and K⁺ homeostasis, glutamate efflux, and other intense reactions, resulting in the activation of cell death mechanisms, such as apoptosis, necrosis, autophagic cell death, and ferroptosis. ⁷ A novel cell death pathway, cuproptosis, has been discovered to potentially contribute to the pathology of IS.^8,9 Existing evidence suggests a correlation between abnormal Cu²⁺ metabolism and mitochondrial dysfunction. ¹⁰ Furthermore, there is a positive correlation between plasma copper levels and the risk of developing IS. ¹¹ After IS, damaged cells release damage-associated molecular patterns (DAMPs, e.g., heat shock proteins and adenosine triphosphate) that activate immune responses, resulting in the recruitment of neutrophils and monocyte-derived macrophages.^5,12 Alongside the release of inflammatory mediators, matrix metalloproteinases, and ROS, the BBB is eventually disrupted, exacerbating the accumulation of neurotoxic substances. ¹³ This forms an intractable vicious cycle, and generally, these cascade reactions can cause irreversible brain tissue damage. Among them, microglia-mediated neuroinflammatory response is an essential progression for the development of IS. ¹⁴ As one of the first responding cell types after IS, microglia are rapidly activated upon receipt of endogenous danger signals (i.e. DAMPs), peaking in the damaged area within 2–3 days and lasting for several weeks. ¹⁵ Influenced by the inflammatory microenvironment, activated microglia are polarized into pro-inflammatory M1 and anti-inflammatory M2 types, with the former being involved in the initiation and maintenance of the inflammatory response, whereas the latter exerts an anti-inflammatory response and promotes the repair of damaged tissues. Based on the high plasticity of microglial polarization phenotypes, inhibition of the M1 type and promotion of the M2 type have emerged as research hotspots with broad perspectives for modulating neuroinflammatory responses to ameliorate IS. ¹⁴

Rapid advances in nanotechnology have provided opportunities for developing therapeutic strategies to modulate microglial polarization. Nanodrug delivery systems (NDDSs) for the treatment of IS have been extensively developed and possess excellent bioavailability and biocompatibility.^16–19 Owing to their unique particle size and versatile engineering features, NDDSs have demonstrated efficacy in enhancing drug stability, prolonging drug half-life in the body, and reducing non-specific distribution. ²⁰ Unlike the restrictive treatment of traditional drugs, numerous nanomaterials, including inorganic, organic, and metallic materials, provide more opportunities for the construction of NDDSs-based strategies. ²⁰ Specifically, liposomes and polymer nanoparticles (NPs) have been approved by the Food and Drug Administration. ²¹ NDDSs can simultaneously carry multiple drugs and gene therapeutic agents to achieve precise regulation by targeting microglial polarization-related pathways.^22,23 Specific nanomaterials, such as iron oxide NPs, ²⁴ magneto-plasmonic NPs, ²⁵ and fluorescein-coupled nanomedicine ²⁶ enable cerebral imaging and localization. This information is of immense importance for the clinical diagnosis and treatment of neurological disorders. For example, bovine serum albumin-MnO₂ NPs exhibit outstanding imaging capabilities for BBB permeability in stroke, and they are expected to be a favorable biocompatible magnetic resonance contrast agent for BBB imaging in humans. ²⁷ Particularly, NDDSs have a unique advantage in penetrating the BBB, allowing non-invasive treatment of brain diseases through passive diffusion into the vascular endothelium, charge adsorption-mediated endocytosis, and receptor-mediated active targeting. ²⁸

In this review, we first briefly overview microglia and their dynamics under IS and then focus on the latest research and progress in NDDSs to modulate microglial polarization for treating IS, with the aim of providing further impetus to the basic research and clinical translation of nanotherapeutic strategies (Figure 1).

Figure 1.

NDDSs for regulating M1/M2 microglial polarization in treating IS. After IS onset, a “brawl” between M1/M2 microglia occurs at the lesion site. In round 1, the victory of M1 microglia results in neurotoxic effects and negative impact on neuronal repair and prognosis. In round 2, with the aid of “foreign aid” nanodrug delivery systems, M2 microglia gain the upper hand and exert neuroprotective effects.

NDDSs: nanodrug delivery systems; TNF-α: tumor necrosis factor-α; IL: interleukin; TGF-β: transforming growth factor-β; ROS: reactive oxygen species; siRNA: small interfering RNA.

Overview of microglia: Origin, function, and polarization

Microglia are resident macrophages and the sole myeloid cells found in the CNS, with their initial discovery dating back to 1899. ²⁹ The developmental origin of microglia has been ambiguous for many years. It was not until 2010, when Ginhoux et al. ³⁰ used fate mapping analysis, that this particular cell population was confirmed to originate from yolk sac-primitive macrophages. Microglia begin to colonize the developing brain as early as day 9.5 of the mouse embryonic stage of formation. ³¹ Under the BBB, they are protected from bone marrow-derived circulating factors and undergo self-proliferation and renewal. However, the presence or absence of progenitor cells in the adult mammalian CNS has generated considerable controversy. Elmore et al. ³² suggested that microglia that regenerate and recolonize the brain (i.e. repopulated microglia) may differentiate from nestin-expressing proliferating progenitors. However, this perspective was later debunked by Huang et al., ³³ whose study revealed that repopulated microglia actually originate exclusively from the self-proliferation of less than 1% of the microglia that remain in the brain, rather than from microglial progenitor cells. They also discovered a high degree of functional similarity between the repopulated and primary microglia using whole-brain RNA sequencing. ³³

In the adult brain, microglia account for approximately 10% of all brain cells and are critical for maintaining brain homeostasis. ³⁴ Throughout human early embryonic formation, microglia are not only involved in neuronal differentiation and synapse formation, but also in the removal of dead cellular debris, pathogens, and aberrant proteins through phagocytosis mediated by surface receptors, for example, Toll-like receptors (TLR), NOD-like receptors, and CD36. ³⁵ These functions highlight the essential role of microglia in both healthy and diseased organisms. With the continuous advancement of cutting-edge imaging techniques, such as super-resolution fluorescence microscopy, ³⁶ time-lapse two-photon microscopy, ³⁷ and serial block face scanning electron microscopy, ³⁸ additional functions of microglia in CNS development are anticipated to be gradually unveiled.

In healthy humans, resting, branching M0 microglia are not completely quiescent; they are highly sensitive to environmental variations and act as “sentinels” to monitor changes in the brain. ³⁹ While under abnormal conditions, microglia are rapidly reactive and characterized by morphological transformation from branching to amoeboid shape to better cover limited areas to exercise phagocytosis. ³⁵ Additionally, M0 microglia undergo a change in phenotype, differentiating into two types: the classically activated M1 type and the alternatively activated M2 type, that is, microglial polarization. Specifically, M1 microglia are amoeboid and express surface markers, for example, inducible nitric oxide synthase (iNOS) and CD16/32. Their cytosolic key transcription factors, the nuclear factor-kappaB (NF-κB), signal transducers and activators of transcription (STAT) 1 and 3, and interferon regulatory factor 3 (IRF3), up-regulate the expression of M1-like genes, which promotes cellular secretion of pro-inflammatory cytokines (tumor necrosis factor [TNF]-α, interleukin [IL]-12, and IL-1β). ⁴⁰ There is a marked increase in the branching of the M2 type, whose main surface markers are arginase (Arg)-1 and CD206. M2 microglia release anti-inflammatory cytokines, such as transforming growth factor-β, IL-4, and IL-10 through the expression of transcription factors including STAT6 and peroxisome proliferator-activated receptor-gamma. ⁴⁰ Depending on the activating factor, the M2 type can be further subdivided into M2a, M2b, and M2c subtypes. ⁴¹ The function of these different subtypes remains to be further investigated.

Roles of microglia in IS: From a “double-edged sword” to a therapeutic target

The dual effect of microglia in IS has been recognized and is closely related to the temporal and spatial natures of their distinct polarization phenotypes, which determine their functional and distributional characteristics. ⁴² At early onset, M2 microglia rapidly infiltrate into the ischemic core to relieve neuroinflammation and attenuate neuronal apoptosis by secreting anti-inflammatory substances and neurotrophic factors, for example, nerve growth factor, brain-derived neurotrophic factor, and glial cell line-derived neurotrophic factor. ⁴³ In the following 3 days, M2 microglia are gradually polarized toward M1 type and migrate toward the ischemic penumbra, with the M1 microglia count dominating on day 7. ⁴³ After 14 days of ischemia, M1 microglia remain high, whereas M2 microglia return to pre-injury levels. ⁴⁴ Contrary to M2 microglia, M1 microglia release numerous neurotoxic agents, including pro-inflammatory cytokines, proteases (matrix metalloproteinases 3 and 9), and ROS, which constantly activate the inflammatory cascade response, resulting in the breakdown of the BBB and aggravating the tissue damage. ⁴⁵ This adverse effect may persist into the chronic phase. Hence, microglia present a “double-edged sword” during the progression of IS and other neuroinflammation-related diseases.

Uncovering the mechanisms and epigenetic modifications in microglia will enable researchers to exploit this “double-edged sword” in IS therapies. Beyond the previously mentioned nuclear transcription factors, for more detailed mechanisms of M1/M2-type polarization, we recommend previous reviews.^46,47 Not surprisingly, damage due to M1 microglial infiltration is antithetical to the therapeutic goal of the timely clinical salvage of ischemic neurons, especially in the ischemic penumbra. This partially explains why hindering the polarization of microglia toward the M1 type could effectively improve cerebral ischemic injury. Moreover, the microglial phenotype trends more toward the M1 type as age increases. ⁴⁸ This suggests that in the era of increasing aging and yearly increasing incidence of IS, more attention should be paid to the impact of accelerated polarization of microglia to the M1 type, an intrinsic pathological change influenced by age factors, on patients’ prognosis.

The modulation of microglial polarization to enhance IS therapy is an attractive treatment, and many drugs that have emerged in the last 3 years have shown a certain degree of therapeutic potential (Table 1). Candesartan (an angiotensin II type I receptors antagonist), a clinically available drug, has the potential to reverse the M1 type-induced neurotoxic effects by regulating the polarization of M1 microglia to M2 type in a TLR4/NF-κB-dependent manner. ⁴⁹ The small molecule drug, synthetic peptide LP17, blocks the activation of microglial triggering receptor, thereby inhibiting M1-type polarization and neutrophil infiltration, alleviating the inflammatory response. ⁵⁰ Some microRNAs (miRs), such as miR-377, miR-29b, and miR-183, also have the potential to reshape microglial M1/M2 phenotype polarization. ⁵¹ However, the BBB prevents most drugs from reaching the brain to perform their effects. The emergence of NDDSs allows single or multiple drugs to easily cross the BBB, enabling synergistic and precisely targeted therapies at the site of cerebral injury. ⁵² The drugs and regulatory mechanisms listed in Table 1 can provide researchers with optional loaded-drug sources and therapeutic targets for constructing NDDSs or other advanced drug delivery systems.

Table 1.

Targeting microglial polarization for the treatment of IS.

Therapeutic agents	Treatment administration	Animal models	Targeting pathways	Outcomings	References
Adenovirus expressing sh-ribosomal protein S3	Intracerebroventricularly injection	tMCAO SD rats	Activated NLRP3 inflammasome via SIRT1	Induced M2-type polarization, alleviated neuronal apoptosis, and injury	Zhou et al. 53
Baicalein	Gastric gavage	MCAO C57BL/6j mice	Inhibited the TLR4/NF-κB pathway, downregulated p-STAT1	Inhibited pro-inflammatory markers expression and increased anti-inflammatory markers expression both in vivo and in vitro	Ran et al. 54
Hesperidin	Gastric gavage	MCAO C57BL/6 mice	Inhibited the TLR4/NF-κB pathway	Inhibited M1-type polarization, and improved neurological deficits	Zhang et al. 55
Inosine	Intraperitoneal injection	MCAO SD rats	Reduced NLRP3 and NLRP1 inflammasome	Induced M2-type polarization and improved motor coordination and cognition	Datta et al. 56
JXL001	Vena caudalis injection	MCAO C57BL/6j mice	Compromised NLRP3 inflammasome activation via the NF-κB pathway	Inhibited M1-type polarization and decreased the expression of pro-inflammatory cytokines in the ischemic penumbra region	Bian et al. 57
Lipoxin A4	Lateral ventricle injection	MCAO SD rats	Inhibited the Notch pathway	Promoted M2-type polarization, which could be weakened by Notch inhibitors	Li et al. 21
Melatonin	Intraperitoneal injection	tMCAO/R C57BL/6 mice	Upregulated p-STAT6 and downregulated p-STAT1	Facilitated the M2-type switch, protected against neuronal apoptosis	Li et al. 58
Minocycline	Intraperitoneal injection	MCAO/R C57BL/6 mice	Upregulated p-STAT6 and downregulated p-STAT1	Promoted M2-type polarization and neuronal survival, inhibited reactive gliosis	Lu et al. 59
Muscone	Intraperitoneal injection	dMCAO C57BL/6j mice	Activated PPAR-γ pathway	Intensified microglia transformation into M2 type, increased levels of anti-inflammatory cytokines	Liu et al. 60
Non-mitogenic fibroblast growth factor 1	Intranasal administration	Photothrombosis stroke C57BL/6 mice	Activated Nfr2 pathway	Enhanced anti-inflammatory microglia, promoted functional recovery	Dordoe et al. 61
Recombinant human fibroblast growth factor 21	Intraperitoneal injection	MCAO C57BL/6 mice	Suppressed NF-κB and upregulated PPAR-γ	Alleviated neuroinflammation by modulating spatiotemporal dynamics of microglia	Wang et al. 62
Salvianolic Acids for Injection	Intraperitoneal injection	MCAO/R SD rats	Inhibited NLRP3 inflammasome/pyroptosis axis	Reshaped M1-type polarization patterns to M2 type, reduced neuronal apoptosis	Ma et al. 63
Pseudoginsenoside-F11	Oral administration	tMCAO SD rats	Regulated in a jmjd3-dependent way	Promoted M2 peak advances from day 14 to day 3 post-injury, allowing the damaged brain to repair earlier as soon as possible	Hou et al. 64

MCAO: middle cerebral artery occlusion; tMCAO: transient middle cerebral artery occlusion; MCAO/R: middle cerebral artery occlusion reperfusion; dMCAO: distal middle cerebral artery occlusion; NLRP1/3: NOD-like receptor thermal protein domain associated protein 1/3; SIRT1/6: silent mating type information regulation 2 homolog 1/6; p-: phosphorylated; Nfr2: nuclear factor erythroid 2-related factor 2; jmjd3: jumonji domain-containing protein 3.

NDDSs targeting the microglial polarization phenotype for IS therapies

In this section, we primarily delve into how different types of NDDSs contribute to improving IS by influencing the signaling pathways associated with microglial polarization. The four major categories include: (1) NDDSs inspired by microenvironments with high levels of ROS, (2) NDDSs inspired by biomimetic materials, (3) NDDSs inspired by natural biological particles, and (4) NDDSs inspired by small interfering RNA (siRNA). Considering the limited research in this field, the classification presented here does not strictly adhere to categorizing nanomaterials or microglial polarization-related mechanisms individually, but rather combines both factors to provide a more comprehensive understanding of the pathological progression of IS. This approach offers the advantage of unraveling personalized mechanism-based NDDSs and serves to facilitate the exploration of promising therapeutic agents (such as small molecules that can modify polarization-related processes) for further investigation.

NDDSs inspired by microenvironments of high-level ROS

Following ischemic and reperfusion events, owing to the mitochondrial dysfunction and the disruption of the balance of the redox system, a large amount of ROSs, such as hydroxyl radical (^•OH), superoxide anion (O₂^•–), and hydrogen peroxide (H₂O₂) accumulate heavily in the damaged area. ⁶⁵ In delicate brain tissues, which are extremely sensitive to oxidative stress, ROS interacts with the respective biomolecules to exacerbate tissue damage by inducing apoptotic cascades, damaging DNA, and initiating lipid peroxidation. ⁶⁶ Although the mechanisms of microglial polarization are not well elucidated, many studies have demonstrated that ROS, as an essential member of the DAMPs, can upregulate pro-inflammatory cytokine expression, triggering aberrant activation of the M1 type.^67,68 Therefore, ROS-scavenging/responsive NDDSs are ideal therapeutic tools to effectively drive microglial polarization toward the anti-inflammatory M2 type and rescue ischemic neurons.

ROS-scavenging NDDSs

ROS-scavenging NDDSs provide a sizable pathway for improving the efficiency of the delivery of antioxidant neuroprotectants to the brain. Dihydrolipoic acid (DHLA) is a reduced form of lipoic acid that potently scavenges ROS and promotes the regeneration of endogenous antioxidants. ⁶⁹ Xiao et al. ⁷⁰ prepared DHLA-loaded gold nanoclusters (DHLA-AuNCs) by mixing DHLA and HAuCl₄ with the reducing agent, NaBH₄. The hydrodynamic diameter of DHLA-AuNCs was 1.87 ± 0.14 nm, and under conditions where the concentration was lower than 5 μg/mL, the potential toxicity and metabolic effects on the immortalized microglial cell line, BV-2, in in vitro culture were determined to be negligible. Flow cytometry analysis showed that DHLA-AuNC treatment significantly reduced intracellular ROS levels in BV-2 cells in a dose-dependent manner. The expression of M1-type markers, including major compatibility complex II, CD86, and iNOS, as well as pro-inflammatory cytokines IL-1β, IL-6, and TNF-α, exhibited a notable down-regulation. In contrast, the expressions of the M2-type markers Arg-1 and CD206, along with the anti-inflammatory cytokine IL-10, was upregulated. Although DHLA alone also inhibited the M1 polarization phenotype, it was not as effective as DHLA-AuNCs. In an ex-vivo brain slice model subjected to oxygen-glucose deprivation/reoxygenation (OGD/R) conditions, the authors did not observe specific localization of DHLA-AuNCs in microglia, indicating the necessity for further exploration using in vivo experiments.

Notably, certain nanomaterials possess excellent antioxidant properties that regulate microglial polarization. For example, the two oxidation states of Ceria NPs (CeNPs), Ce³⁺ and Ce⁴⁺, serve as superior ROS scavengers, with the former scavenging •OH and O₂^•– and the latter being responsible for removing H₂O₂. ⁷¹ On the basis of these properties, Zeng et al. ⁷² prepared CeNPs coated with polyethylene glycol (PEG) through thermal decomposition and ultrasound-assisted methods. An electron spin resonance technique showed that CeNP-PEG (diameter 9.3 nm) exhibited time- and concentration-dependent scavenging ability toward various ROS (^•OH, O₂^•–, and H₂O₂). In OGD/R-insulted microglia BV-2 cells, pretreatment with CeNP-PEG reduced ROS fluorescence intensity by more than 80%, blocked the NF-κB inflammatory pathway, and promoted the reversal of M1 to M2 types. Further, Jia et al. ⁷³ used a small cyclic peptide called LXW7, which has a specific binding affinity to integrin αvβ3, and this peptide was coupled to the surface of CeO₂ NPs. Underpinning the effective attenuation of oxidative stress, LXW7 reduced focal adhesion kinase (FAK)/STAT3 phosphorylation by blocking the integrin pathway, leading to the inhibition of lipopolysaccharide (LPS)-induced inflammation in BV-2 cells. Moreover, manganese oxide, iron oxide, and copper oxide NPs have shown the potential to alleviate oxidative stress in neurological disorders and could be integrated with therapeutic agents to enhance efficacy. ⁷⁴

ROS-responsive NDDSs

ROS-responsive NDDSs allow for the controlled release of target drugs in microglial microenvironments with elevated levels of extracellular or intracellular ROS, which feature certain targeting characteristics. Polylactic acid (PLA) coating can be degraded by ROS and thus act as “gatekeepers” for NDDSs. Shen et al. ⁷⁵ developed PLA-coated low-density lipoprotein receptor (LDLR) ligand peptide-modified mesoporous silica NPs for resveratrol delivery. Resveratrol, a secondary metabolite found in some plant species and natural foods, has anti-oxidative stress, anti-inflammatory, and neuroprotective properties. ⁷⁶ With PLA encapsulated, the release of resveratrol in PBS was only 5% after 5 days (compared to 90% without PLA); however, rapid release was observed in the presence of superoxide. LDLR attached to the outer layer via 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide/N-hydroxysuccinimide (EDC/NHS) chemistry can enhance the uptake of NDDSs by the BBB through receptor-ligand interactions. In an in vitro BBB model, these NDDSs effectively crossed the BBB, and in an environment of high ROS concentrations produced by microglia, the PLA coating was degraded to achieve controlled release of resveratrol, which ultimately reduced LPS-induced microglial activation toward the pro-inflammatory phenotype. To achieve precise delivery of NDDSs to the mitochondria—the main ROS-producing subcellular organelle—Wang et al. ⁷⁷ formulated a stepwise targeting of NDDSs by mixing a solution of cyclo(Arg-Gly-Asp-d-Tyr-Lys) peptide, (cRGD)-PEG-Acetal-PEG-polycaprolactone (PCL)-triphenylphosphine (TPP) polymer, and PEG-Acetal-PCL-PEG polymer with resveratrol using solvent evaporation methods (Figure 2). First, the resveratrol-loaded NDDSs effectively penetrated the BBB through the receptor-ligand interaction of cRGD with α_υβ₃ integrin, which is highly expressed in cerebrovascular endothelial cells. Subsequently, NDDSs are phagocytosed by microglia at the site of cerebral ischemia. With TPP-mediated targeting of the inner mitochondrial membrane, resveratrol was released directly into the mitochondria, removing excess ROS and modulating microglial polarization toward the M2 type. In the OGD/R-insulted BV-2 cells, the M2/M1 ratio increased from 0.2 to 3.4 after treatment with these NDDSs. Additionally, both neurological deficits and survival were remarkably improved in NDDS-treated transient middle cerebral artery occlusion (tMCAO) mice. In addition to antioxidant drugs, ROS-responsive NDDSs can also be used to piggyback inhibitors of key proteins that directly regulate microglial polarization to M1 and deliver to cerebral ischemia areas with high levels of ROS, thereby promoting conversion to the M2 phenotype. ⁷⁸

Figure 2.

Therapeutic application of cRGD/TPP@Res micelles. (a) Schematic illustration, internal metabolism and mechanisms for treating IS. In vivo therapeutic efficacy of tMCAO mice: (b) Two-photon imaging revealing the process of micelles penetration into the brain parenchyma. IR780, red; FITC, green. (c) CLSM imaging showing the colocalization of micelles in the NeuN⁺ neurons, GFAP⁺ astrocytes, Iba-1⁺ microglia. cRGD and cRGD/TPP, red; NeuN/GFAP/Iba-1, green; Nucleus, blue; Mitochondria, stained with TOMM20. In vitro therapeutic efficacy of OGD/R-treated BV-2 cells: (d) Ratio of M2 (CD206⁺) and M1 (CD16/32⁺) type. (e) Quantitative analysis of ROS fluorescence. Reproduced by permission of Wang et al. ⁷⁷

cRGD: cyclo(Arg-Gly-Asp-d-Tyr-Lys) peptide; OGD/R: oxygen-glucose deprivation/reoxygenation; Res: resveratrol; TPP: triphenylphosphine.

Oxidative stress is a common pathological mechanism in almost all CNS disorders. To date, many antioxidant neuroprotective agents have shown promising results in IS, especially in clinical research. Among them, edaravone, as a potent free radical scavenger, has been one of the most extensively studied antioxidants since its launch in Japan in 2001. ⁷⁹ Seven randomized controlled trials involving 2069 patients with IS suggested that edaravone was effective in improving neurological function, regardless of treatment course and patients’ age. ⁸⁰ In a phase III trial (ClinicalTrials.gov identifier: NCT02430350), edaravone dexborneol showed a favorable 90-day functional prognosis after treatment of patients with acute IS within 48 h, which was superior to that of edaravone alone. ⁸¹ The beneficial role of other antioxidant neuroprotectants, such as uric acid ⁸² and butylphthalide, ⁸³ in improving the prognosis of patients with IS has also been tentatively explored. These outcomes reveal the broad therapeutic promise of NDDSs inspired by ROS or oxidative stress microenvironments. Notably, ROSs are one of the most important signaling mediators in the human body. Premature or late interference with ROSs, as well as incomplete drug release, may impact the effectiveness of NDDSs. ⁸⁴ The development of more intelligent and refined NDDSs for IS treatment remains an ongoing pursuit.

NDDSs inspired by biomimetic material

Although ligand modification increases the affinity of NDDSs for cells expressing the corresponding target molecules, it is not efficient in enhancing the vascular infiltration of such NDDSs into tissues. ⁸⁵ Besides, owing to individual heterogeneity, the degree and duration of BBB opening varies widely among patients with IS of different severities. To fully harness the targeted delivery capabilities of NDDSs, innovative biomimetic NDDSs derived from cell membrane-based components or endogenous/exogenous substances have emerged, thereby providing a novel direction for IS treatment. ⁸⁶

Cell membrane-coated biomimetic NDDSs are highly biocompatible and stable, and the “cell membrane-microenvironment” mediator empowers these NDDSs to interact with the specific pathological microenvironment. ⁸⁷ After IS, erythrocytes, leukocytes, and platelets cross the BBB either spontaneously or by inflammatory factor chemotaxis, and participate in complex pathological processes, for example, oxidative stress, neuroinflammation, and thrombosis in the damage region, so a wide range of cells present in the circulation would serve as an optional source of cell membrane coating. Extensive research has been conducted on cell membrane-coated biomimetic nanomedicine for treating IS, revealing their diverse advantages in drug loading, targeting, and biodegradability with distinct cell membrane coatings. ⁸⁸

Erythrocyte membrane-coated biomimetic NDDSs

Erythrocyte membranes as coatings for encapsulating nanomedicine have the benefits of evading immune clearance, a long in vitro half-life, and high drug-loading capacity. ⁸⁹ More importantly, erythrocyte membrane-coated biomimetic NDDSs reach the brain via the vascular circulation and are able to pass through the BBB almost unimpeded. ⁹⁰ However, NDDSs solely modified with erythrocyte membranes face challenges in actively targeting the lesion site, necessitating the incorporation of targeting ligands in the outer layer to improve brain delivery efficiency.^91,92 Lv et al. ⁹¹ coupled CLEVSRKNC in erythrocyte membrane-coated biomimetic NDDSs. CLEVSRKNC is a stroke-homing peptide that mediates the targeting of this NDDS with a long circulating lifespan (>48 h in vivo) to neurons in the ischemic penumbra. The released neuroprotectant NR2B9C interacts with postsynaptic density proteins on the ischemic neuron membrane, thus blocking the overproduction of nitric oxide. Serological and histological analyses showed that the biomimetic NDDSs have high biocompatibility and safety in vivo. After treating middle cerebral artery occlusion (MCAO)/reperfusion rats, the brain infarct volume and neurological function deficits were notably improved. This study did not investigate the detailed molecular mechanisms by which NDDSs exert neuroprotective effects and modulate microglial polarization.

Nanomaterials for erythrocytes have great therapeutic potential. Recently, Yin et al. ⁹³ developed a novel nanoerythrocyte (NEMR) with a dual-target chain modification of RVG29 and MG1 peptides for enhanced targeting of the BBB and precise recognition of M1 microglia by physical extrusion. NEMR (diameter 189.7 ± 12.3 nm) exhibited a spherical and uniform vesicular structure, with good stability in terms of size and morphology in PBS and cell culture medium. Upon uptake by M1 microglia, NEMR stimulated the neurogenic locus notch homolog protein 1/hairy and enhancer of split-1/STAT3 signaling pathway by releasing heme oxygenase-1 (HO-1) intracellularly, which further inhibited NF-κB p65 nuclear translocation and eventually remodeled M1 microglia to the M2 phenotype. After NEMR treatment in MCAO rats, the expression levels of M1 microglia markers—iNOS, IL-6, and TNF-α—decreased in the ischemic brain tissue. In contrast, the expression of M2 markers—CD206, Arg-1, and IL-10—increased. By promoting the protective M2-like transformation of microglial phenotype, NEMR inhibits inflammation caused by IS and exerts a beneficial neuroprotective effect. Interestingly, HO-1 can metabolize heme to produce carbon monoxide and bilirubin, which are powerful antioxidants that scavenge excess ROS. Encouraged by its potent neuroprotective effects, the authors used NEMR to deliver edaravone to treat MCAO rats. The results revealed that edaravone-loaded NEMR had favorable biosafety in vivo and could precisely regulate inflammatory microglia in the focal region, promote anti-inflammatory cytokine expression, reduce BBB permeability, and improve prognosis. Additionally, NEMR showed superior efficacy in mice with experimental autoimmune encephalomyelitis. These findings suggest that NEMR functions as a safe and effective NDDS to modulate microenvironmental immune homeostasis, with high clinical translational potential in the treatment of microglia-mediated inflammation-associated CNS disorders.

Leukocyte membrane-coated biomimetic NDDSs

Unlike the low targeting of bare erythrocyte membrane-coated NDDSs, leukocyte membrane-coated biomimetic NDDSs can penetrate the brain parenchyma and deeper into the pathologic microenvironment via receptor-mediated cytophagy and inflammatory chemotaxis. This is due to the fact that after IS, pro-inflammatory cerebrovascular endothelial cells overexpressing intercellular cell adhesion molecule-1, vascular cell adhesion molecule-1, and p-selectin could effectively recruit leukocytes (mainly neutrophils and macrophages) expressing the corresponding ligands (CD44 and CD11b) from peripheral blood to infiltrate the lesion site. ⁹⁴

Neutrophils are the most sensitive subpopulation of leukocytes interacting with inflammatory cerebrovascular endothelial cells. ⁹⁵ Feng et al. ⁹⁶ created a leukocyte membrane-coated mesoporous Prussian blue nanozyme (MPBzyme@NCM) labeled with fluorescein isothiocyanate (FITC). In physiological environments, including physiological saline and serum, MPBzyme@NCM can maintain long-term stability, indicating its potential use for systemic circulation therapy. After intravenous injection into tMCAO mice, immunofluorescence results showed that these NDDSs exhibited remarkable targeting capabilities toward the brain tissue on the damaged lateral side and could be specifically phagocytosed and ingested by microglia, as opposed to neurons or astrocytes, upon crossing the cerebral vascular endothelium. This selective uptake by microglia leads to a reduction in the proportion of CD16/32⁺ cells and an increase in CD206⁺ cells, promoting microglial polarization toward the M2 phenotype, thereby alleviating the inflammatory response caused by IS. After the 28-day treatment period, the group treated with MPBzyme@NCM exhibited a significantly higher survival rate (approximately 91%), whereas the control group showed a significantly lower survival rate (approximately 50%). To explore the molecular mechanisms involved, the authors demonstrated, in vitro, that MPBzyme@NCM scavenged ROS and activated the STAT6 pathway to promote M2-type polarization. Moreover, the expression of CD206 decreased after the addition of a STAT6 inhibitor, whereas that of CD16/32 increased.

Except for neutrophil membrane-coated NDDSs, macrophage membrane-coated NDDSs have achieved promising efficacy in various inflammatory diseases, including IS.^97,98 Li et al. ⁹⁹ prepared macrophage membrane-encapsulated MnO₂ nanoparticles loaded with fingolimod (FTY), named Ma@(MnO₂+FTY), to ameliorate IS by modulating microglial polarization and decreasing oxidative stress in the following three aspects (Figure 3): (1) increasing the targeting to the inflammatory microenvironment of the brain via macrophage membranes; (2) scavenging excess ROS through MnO₂, thereby inhibiting the NF-κB signaling pathway, while reacting with H₂O₂ to generate O₂ in situ to salvage damaged neurons; ⁷² and (3) activating STAT3, a key factor for M2 type microglia polarization, by using FTY. ¹⁰⁰ In particular, paramagnetic Mn²⁺ produced by Ma@(MnO₂+FTY) can be used as a contrast agent for enhanced magnetic resonance imaging of brain tissue after ischemia-reperfusion injury, with great promise for both therapy and diagnosis.

Figure 3.

Therapeutic application of Ma@(MnO₂+FTY). (a and b) Schematic illustration of internal metabolism and mechanisms for treating IS. In vitro therapeutic efficacy of BV-2 cells: (c) Quantitative analysis of relative fluorescence intensity and (d) immunofluorescence analysis of the M2 (CD206⁺) and M1 (CD16/32⁺) type. CD206, green; CD16/32, red. In vivo therapeutic efficacy of tMCAO/R rat: (e) Fluorescence of the probe in ischemic at 1, 4, and 24 h. (f) Representative images of TTC staining. Reproduced by permission of Li et al. ⁹⁹

Other novel membrane-coated biomimetic NDDSs

On the basis of the potential of M2 microglia to induce the repolarization of M1 microglia to the M2 type, Duan et al. ¹⁰¹ innovatively produced M2 microglial membrane-coated, catalase-loaded tannic acid NPs (TPC@M2 NPs) by thoroughly mixing the M2 microglial membrane with TPC NPs (Figure 4). Compared to the free state, the enzyme stability of catalase encapsulated in TPC@M2 NPs is greatly improved. In BV-2 cells, treatment with TPC@M2 NPs resulted in decreased expression of the M1-type polarization marker, CD16/32, and the pro-inflammatory cytokine, IL-12, and increased expression of the M2-type polarization marker, CD206, and the anti-inflammatory cytokine, IL-10. They speculated that TPC@M2 NPs inhibited microglial polarization toward the M1 type by effectively eliminating excess ROS and blocking the subsequently triggered NF-κB signaling pathway. Compared with naked TPC NPs (without membrane modification), near-infrared fluorescence imaging revealed a marked enhancement in the fluorescence signal within the brains of MCAO rats treated with TPC@M2 NPs, suggesting that M2 microglial membranes favored inflammatory chemotaxis and targeting ability to ischemic tissues. Treatment with TPC@M2 NPs resulted in a significant increase in the number of CD206⁺ M2 microglia and a decrease in the infarct volume from 50% to 15%, together with a clear improvement in neurological deficits.

Figure 4.

Therapeutic application of TPC@M2 NPs. (a) Schematic illustration, internal metabolism and mechanisms for treating IS. In vivo therapeutic efficacy of MACO mice: (b) Diagram process. (c) Representative images of TTC staining. (d) Immunofluorescence analysis and MFI of the M2 (CD206⁺) and M1 (CD16/32⁺) type. CD16/32, green; CD206, red. (e) Immunohistochemical analysis using H&E and TUNEL staining. (f) Quantitative analysis of neurological score. Reproduced by permission of Duan et al. ¹⁰¹

CAT: catalase; TA: tannic acid; MFI: mean fluorescence intensity.

Tumor cell membranes are also expected to construct biomimetic NDDSs for IS treatment, which is closely related to the ability of tumor cells to migrate toward the BBB by adhering to the cerebral blood vessels through the medium of leukocytes and platelets. ¹⁰² A study using 4T1 breast cancer cell membrane-based NDDSs for the treatment of IS revealed that biomimetic NDDSs could preferentially accumulate on the ischemic hemispheric side, with a 4.79-fold delivery efficiency compared with that on the normal hemisphere, and effectively reduce the cerebral infarct volume. ¹⁰³

Cell membrane coatings act as “lubricants” that allow NDDSs to traverse the BBB, thereby overcoming the limitations associated with conventional drug delivery strategies and facilitating specific targeting of the brain. However, cell membrane-coated biomimetic NDDSs remain in their infancy, and their mechanisms of functioning in vivo are still unclear, despite potential safety hazards. For example, leukocyte membrane-based NDDSs may activate unnecessary immune responses, and membranes derived from tumor cells with incompletely cleared genetic substances may carry the risk of inducing cancer after delivery to the body. ¹⁰⁴ Due to the diversity of technologies and methods, an optimal manufacturing process for cell membrane-coated biomimetic NDDSs has not yet been identified. In membrane engineering, both covalent and non-covalent modifications present their respective advantages and disadvantages, respectively sacrificing the protective role on membrane protein activity and the firmness of the connection. ¹⁰⁵ The primary approach for determining the success of membrane modification is through assessing membrane potential, particle size, morphology, and other factors. ¹⁰⁶ However, the reliability of these methods is heavily dependent on experimental conditions and subjective factors. Therefore, there is an urgent demand for precise, visual identification methods. Extensive research is required before cell membrane-coated biomimetic NDDSs can be applied in clinical settings.

NDDSs inspired by natural biological particles

Natural nanoscale biological components present in the human body (represented by exosomes) can serve as drug delivery carriers and are often classified as biomimetic NDDSs. We emphasize here that such natural NDDSs differ from cell membrane-coated NDDSs by no longer relying on synthetic nanomaterials encapsulated within the cell membrane. Most studies evaluating the safety and efficacy of NPs have been based on acute-phase animal disease models, with limited discussion on the subacute and chronic phases. ¹⁰⁷ The potential long-term toxicity of some NPs in vivo has been recognized. For instance, silica NPs may lead to neuronal apoptosis through the activation of endoplasmic reticulum stress signaling pathways. ¹⁰⁸ Notable variations in the inflammatory response, tissue toxicity, and blood composition changes induced by intravenous injection of silica NPs at 10 and 180 days have been reported. ¹⁰⁹ To summarize, nature-inspired NDDSs largely circumvent the potential safety issues associated with nanomaterials, and they feature purer naturalness and biogenicity instead of a greater potential for clinical translation.

Exosomes-based NDDSs

Exosomes, which are nanoscale vesicles ranging from 40 to 160 nm in diameter, play pivotal roles in various biological processes. For example, cellular metabolism, immune response, and neural regeneration. ¹¹⁰ As a pathway for intercellular communication, exosomes work as natural nanocarriers to deliver a wide range of bioactive substances (including proteins, lipids, and nucleic acid molecules) to regulate the function of recipient cells. ¹¹⁰ Specifically, exosome-based NDDSs have multiple advantages such as ease of crossing the BBB, high safety, and low immunogenicity, and thus are promising approaches for treating CNS diseases.

Exosomes derived from diverse cell sources have been demonstrated to hold the potential to modulate microglial polarization for IS treatment.^111,112 As the mechanism is thoroughly scrutinized, researchers have found that small-molecule therapeutic substances enriched in exosomes play a crucial role. MiRs account for approximately 41.72% of exosome RNAs and are the most widely studied nucleic acids with the highest content. ¹¹³ MiRs can post-transcriptionally regulate M1/M2 type polarization-related genes to remodel microglia phenotype. Mesenchymal stem cell-derived exosomes highly expressing miR-223-3p promote M1 to M2 type conversion by down-regulating the expression of cysteinyl leukotriene receptor 2. ¹¹⁴ MiR-30d-5p-enriched exosomes derived from adipose-derived stem cells (ADSCs) inhibit autophagy-mediated M1-type polarization and relieve cerebral ischemic injury by suppressing the expression of autophagy-related genes (Beclin-1 and Atg5). ¹¹⁵ Exosomes with a high level of miR-145d1 could downregulate forkhead box protein O1 and inhibit M1-type polarization. ¹¹⁶ Circular RNAs (circRNAs) in exosomes act as miRNA sponges; that is, they restore or enhance downstream gene expression by competitively binding to miRNAs. ¹¹⁷ ADSC exosomes containing circ-Rps5 induce silent mating-type information regulation 2 homolog-7 expression by sponging miR-124-3p, which in turn mediates microglial polarization from the M1 to M2 type. ¹¹⁸ Similarly, circRNA-Ptpn4-modified ADSC exosomes mediate M2-type polarization to repair IS-induced neurological injury by inhibiting miR-153-3p expression and enhancing nuclear factor erythroid 2-related factor 2 expression. ¹¹⁹

The emergence of engineered exosomes has further advanced the therapeutic application of exosomes in IS. ¹²⁰ The term “engineering” includes two main aspects of processing and upgrading. The first aspect is the internal engineering of naked exosomes, that is, enhancing the efficacy of exosome-based NDDSs by loading them with therapeutic agents. Li et al. ¹²¹ developed edaravone-loaded macrophage-derived exosomes (Exo+Edv) to explore their therapeutic performance against permanent middle cerebral artery occlusion (PMCAO). The Exo+Edv, with a particle size of 68.06 ± 1.94 nm, was collected using ultra-speed centrifugation methods and exhibited higher maximum plasma concentrations, lower clearance, and a longer half-life relative to free edaravone, suggesting that edaravone bioavailability was notably improved by exosome delivery. In the PMCAO model, co-localization of Exo+Edv with Iba1-labeled microglia reprogrammed M1/M2-type polarization. Notably, the 7-day mortality rate of Exo+Edv-treated PMCAO rats was reduced to 0% (40% in the blank group and 20% in the edaravone group), suggesting the superior neuroprotective properties of the engineered exosomes against PMCAO. The second aspect is the external engineering of naked exosomes, that is, surface molecular modifications to increase the targeting ability of exosome-based NDDSs. Building on the mannose receptors on the surface of microglia, mannose can serve as an exosome guide for precise drug delivery. Mannose-modified luteolin (an antioxidative stress drug)-loaded platelet-derived exosomes created by Liu et al. ¹²² not only inhibited the activation of the M1 type and promoted the transformation of the M2 type but also maintained BBB stability and suppressed astrocyte-mediated inflammation. Overall, their study showed that regulating the microglial polarization phenotype effectively improved neurovascular unit dysfunction and played a powerful neuroprotective role (Figure 5).

Figure 5.

Therapeutic application of lut/man-pEXO. (a) Schematic illustration, internal metabolism and mechanisms for treating IS. In vivo therapeutic efficacy of MCAO/R rats: (b) Representative images and quantitative analysis of evans blue extravasation. (c) Representative images of TTC staining. Reproduced by permission of Liu et al. ¹²²

Lut: luteolin; Man: mannose; pEXO: platelet-derived exosomes.

Ferritin-based NDDSs

Ferritin is a protein naturally found in humans, characterized by a spherical cage-like structure formed by the self-assembly of 24 subunits consisting of two types: heavy-chain ferritin (HFn) and light-chain ferritin. ¹²³ HFn forms NPs autologously, which are outstandingly biocompatible and degradable without eliciting undesired immune reactions or toxicity in vivo. ¹²⁴ More critically, HFn-based NDDSs hold great promise as they can autonomously recognize and bind to transferrin receptor 1 expressed on cerebrovascular endothelial cells, enabling them to penetrate the BBB without the need for additional modifications. ¹²⁵

In a recent study by Liu et al., ¹²⁶ CsA@HFn was synthesized by attaching cyclosporine A (CsA) to HFn NPs (Figure 6). With a mean particle size of 26 nm (similar to that of HFn), CsA@HFn possesses sensitive responsiveness to ROS and pH, enabling the controlled release of the loaded drug, CsA. On the contrary, CsA@HFn exhibits high stability within 5 days in neutral solutions, such as distilled water and PBS. CsA prevents the opening of the mitochondrial permeability transition pore, thereby mitigating mitochondrial dysfunction-associated ROS and cytochrome C release, and it promotes microglial survival and M2-type polarization. ¹²⁷ In the in-vitro BBB model, the BBB transport efficiency of CsA@HFn was distinctly better than that of free CsA, and this high permeability was proven to be inhibited by ferritin pretreatment. CsA@HFn maintained mitochondrial membrane potential stability, reduced ROS-mediated oxidative stress and cytochrome c-mediated apoptosis, and protected SH-SY5Y cells from OGD/R injury. Immunofluorescence analysis revealed a reduction in the population of Iba-1 and CD16/32 co-positive cells and an increase in the population of Iba-1 and CD206 co-positive cells in the CsA@HFn-treated MCAO mice. In the area of cerebral infarction, the expression pattern of tissue inflammatory factors tended to be of the anti-inflammatory M2 type, and the decrease in ROS levels showed a dose-dependent relationship with the NDDSs.

Figure 6.

Therapeutic application of CsA@HFn. (a) Internal metabolism and mechanisms for treating IS. In vivo therapeutic efficacy of MACO mice: (b) Number of the M2 (CD206⁺) and M1 (CD16/32⁺) type. (c) The IL-12, (d) TNF-α, and (e) IL-10 level. Reproduced by permission of Liu et al. ¹²⁶

CsA: cyclosporine A; HFn: heavy chain ferritin; mPTP, mitochondrial permeability transition pore; TfR1: transferrin receptor 1.

Exosomes, ferritin, or the previously mentioned erythrocytes and nature-inspired NDDSs exhibited a certain degree of heterogeneity, which brings unpredictability to the treatment. At present, it remains a tremendous challenge to successfully load drugs into these nature-inspired NDDSs while preserving their integrity. Co-encapsulation-based passive loading methods exhibit low loading efficiency, while active loading methods, such as electroporation, ultrasound, and extrusion, may cause damage. ¹²⁸ Achieving standardized control and quality production of natural NPs is an issue that warrants further investigation. A recent study found that miR-3613-3p-enriched brain microvascular endothelial cell exosomes could promote M1 type microglia activation and exacerbate neurotoxicity. ¹²⁹ This implies that researchers should also pay attention to the potential negative impact of certain “undesirable NDDSs” within the human body on treatment outcomes.

NDDSs inspired by siRNA therapeutics

siRNA is a potent nucleic acid therapy that silences gene expression. The impressive potential of siRNA in the treatment of CNS diseases stems from its notable attributes, including high specificity for disease-causing targets, ease of development and preparation, and low-dose drug administration. ¹³⁰ Currently, siRNA is well established for basic research on remodeling microglial phenotypes to improve IS.^131–133 However, beyond the drawback of difficulty in crossing the BBB, naked siRNA faces many problems, such as susceptibility to nuclease interference, excessive in vivo transrenal clearance, and low targeting of specific cells. Fortunately, the rapid development of modern nanotechnology offers unprecedented opportunities for safe and efficient siRNA delivery to treat brain diseases. ¹³⁴ In particular, NDDSs can piggyback on naked siRNAs to achieve targeted transport toward specific cell types at damaged sites.

The TLR4-related signaling pathway has emerged as a prominent and extensively investigated mechanism involved in the modulation of microglial polarization toward the M1 phenotype. Early studies have found that defects in TLR4 reduce the expression of TNF-α, iNOS, and other M1-like genes, contributing to the reduction of cerebral infarct volume and inflammatory response.^135,136 Ganbold et al. ²² reported a novel TLR4 siRNA-loaded lipid-NPs (LNP) with surface modification of biocompatible short peptidomimetics for increasing microglia internalization (Figure 7). After being encapsulated in LPN, siRNA can effectively avoid degradation by nucleases, thereby enhancing its accumulation in target cells. LNP inhibited the expression of TLR4 messenger RNA (mRNA) and protein in primary microglia under OGD conditions and promoted a shift in the secreted cytokines toward an anti-inflammatory M2-type pattern. Strong fluorescent signals of FITC-labeled siRNA were detected in the peri-infarct region of the LNP-treated tMCAO mice and co-localized with Iba1⁺ microglia.

Figure 7.

Therapeutic application of siTLR4/DoGo310 LNPs. (a) Schematic illustration, internal metabolism and mechanisms for treating IS. In vivo therapeutic efficacy of tMCAO mice: (b) H&E staining, (c) RT-qPCR, and (d) ELISA analysis of the TNF-α, IL-1β, IL-4, and IL-10 levels. Reproduced by permission of Ganbold et al. ²²

After IS, stimulation of TLR4 further triggers the downstream NF-κB pathway, and NF-κB p50/p65 form a heterodimer complex that enters the nucleus and initiates the expression of M1-like genes, thereby exacerbating the inflammatory response. ¹³⁷ Along similar lines, Ganbold et al. ²³ developed microglia-targeting NDDSs composed of a curdlan-based inner core and a mannose-modified outer shell for loading NF-κB p65 siRNA. In primary microglia, mannose-functionalized NDDSs are specifically taken up via receptor-mediated endocytosis to deliver siRNAs. In the peri-infarct region of NDDS-treated tMCAO mice, notable findings included a prominent reduction in both NF-κB p65 mRNA and protein levels, an increase in neuronal density, and a reduction in pyknosis and edema. This study introduced a non-invasive targeted approach for modulating microglial polarization, ultimately leading to the down-regulation of pro-inflammatory M1-like genes through inhibition of the NF-κB signaling pathway. This approach demonstrated substantial improvements in neuroinflammation and aided in the preservation of ischemic neurons.

There are a paucity of reports on the modulation of microglial polarization phenotypes in IS based on siRNA-loaded NDDSs. Polarization-related mechanisms are complex and diverse, and interference with multiple pathways/targets may be more effective than that with a single pathway. In the future, targeting multiple M1-type polarization-related proteins by the simultaneous delivery of more than two siRNAs via NDDSs could be considered as a research direction worth exploring. In addition to siRNAs, other gene therapies (e.g. ZFNs, TALENs, and CRISPR/Cas9 gene editing technology) and small-molecule drugs are also expected to be used in nanotherapeutic strategies, on the condition that off-target events can be meaningfully avoided.

Conclusion and outlook

According to a recent predictive analytics study, IS will account for up to 4.9 million deaths worldwide by 2030. ¹³⁸ The treatment of IS remains a challenge. Elucidation of the pathogenesis of this disease will facilitate the discovery of novel therapeutic targets and drug development. The high degree of plasticity of microglial polarization phenotypes in complex pathologies has attracted considerable attention. As aforementioned, this “double-edged sword” has been skillfully crafted into a therapeutic “blade.” The pioneering intersection of nanomedicine and neuroscience has spawned a new field of neuro-nanomedicine. ¹³⁹ Through this novel perspective, precision therapy for IS has been further revolutionized, particularly with mechanism-based nanostrategies. NDDSs, with high BBB penetration, high therapeutic efficacy, and high individualization, have shown great value in the treatment of IS by targeting single or multiple links in the intricate “microenvironment-cell/subcell-protein/gene” system associated with microglial polarization.

However, the complex interplay between the function and state of microglia and their surroundings also introduces some uncertainties, including whether the dichotomic categorizations of the M1/M2 or pro-/anti-inflammatory phenotype is truly applicable, and how much of the research data obtained in animal models can be considered as a snapshot of humanity. ¹⁴⁰ Novel integration models based on epigenetics, transcriptomics, and proteomics have been proposed to characterize the multidimensional state of microglia. ¹⁴⁰ In addition to revolutionizing phenotypic classification, the role of peripheral blood-derived immune cells, gut microbiota, and other underlying factors involved in the regulation of microglial intracellular homeostasis deserves exploration.

Most studies of NDDSs for IS treatment are in the preclinical phase. Moreover, there is a lack of therapeutic evidence in aged models because the primary focus is on non-aged animals. However, further clinical data are required to validate the long-term efficacy of this treatment strategy. Technical improvements are urgently required to ensure the in vitro safety of NDDSs. An in-depth exploration of the molecular mechanisms underlying microglial polarization would be beneficial for accelerating the development of NDDSs. Furthermore, novel therapeutic approaches are expected to synergize with NDDSs to regulate microglial polarization phenotypes. For example, alterations in brain rhythms induced by visual non-invasive flickering sensory stimulation have been found to affect microglia morphology and the expression of inflammatory factors through modulation of the NF-κb signaling pathway. ¹⁴¹ With the breakthroughs and rapid development in the field of life sciences, we believe that NDDSs will usher in a new era, paving the way for the diagnosis and treatment of IS and other microglia polarization-related CNS diseases.