Cracking the Blood–Brain Barrier Code: Rational Nanomaterial Design for Next-Generation Neurological Therapies

1. Introduction

Neurological disorders and brain tumors remain among the most devastating and hard-to-treat health challenges worldwide. Neurodegenerative diseases like Alzheimer’s disease and Parkinson’s disease affect tens of millions (over 57 million people lived with dementia globally in 2021)​ [1], incurring enormous social and economic costs (~$1.3 trillion annual global cost for dementia in 2019)​ [2]. Malignant CNS cancers such as glioblastoma multiforme (GBM) are relatively rarer but extremely lethal, with a median survival of only ~12–15 months even with aggressive therapy​. GBM is the most deadly brain tumor, causing about 200,000 deaths per year worldwide​ [3]. A unifying obstacle in these conditions, from neurodegeneration to brain cancer, is the difficulty of delivering therapeutic agents into the brain. The culprit is the blood–brain barrier (BBB), a specialized network of cerebral microvascular endothelium that usually preserves the neural microenvironment but also strictly limits drug access to the central nervous system (CNS).

The BBB shields the brain from toxins, pathogens, and fluctuations in blood composition through tight junctions between endothelial cells, pericytes, and astrocytic end-feet, forming a highly selective barrier [4]. Its tight junctions (~1.4 nm pore size) and efflux transporters block nearly all large molecules and ~98% of small-molecule drugs (F. Wang et al., 2024). Consequently, systemic therapeutics rarely penetrate effectively; monoclonal antibodies typically achieve <0.1% brain uptake, and even FDA-approved anti-amyloid antibodies for AD reach just ~0.01–0.1%​ [5]. Similarly, most small-molecule neurotherapeutics require high lipophilicity, yet remain limited by efflux pumps, size, and polarity. Thus, the BBB remains a significant obstacle for CNS drug delivery​ [6].

This barrier-imposed delivery problem leads to several critical challenges in current therapy. First, many efficacious drugs in vitro or in peripheral tissues fail in CNS trials because adequate concentrations in the brain cannot be attained​. For instance, the chemotherapy paclitaxel is potent against glioma cells but is a substrate of P-glycoprotein; when given systemically, it accumulates to only a minute fraction of plasma levels in the brain, rendering it ineffective unless the BBB is modulated​ [7]. Second, limited BBB permeability forces clinicians to use extreme measures for CNS delivery. High systemic doses are often required to push a small amount of drug across the BBB, which can cause significant off-target toxicity in the rest of the body (e.g., dose-limiting cardiotoxicity of doxorubicin or peripheral immunosuppression by high-dose cytokines). In other cases, drugs must be delivered directly or locally to the CNS, for example, intrathecal chemotherapy or convection-enhanced delivery in brain tumors to bypass the BBB. Such invasive approaches are risky, impractical for chronic conditions, and poorly tolerated by patients [8]. Similarly, methods to transiently open the BBB (such as focused ultrasound with microbubbles) can enable higher drug entry but also carry risks of tissue damage or infection by allowing toxins or pathogens to breach. Owing to these limitations, many CNS diseases remain inadequately treated. In GBM, for instance, standard chemoradiation yields a two-year survival under 20%​ in part because drugs like temozolomide, while somewhat BBB-permeable, still do not uniformly reach all tumor cells behind an intact barrier. There is an urgent need for new delivery strategies that can safely and effectively ferry therapeutic molecules across the BBB [9].

Nanotechnology has emerged as a promising avenue to overcome the BBB’s filtering mechanism and improve drug delivery to the brain​ [10,11,12]. Nanoscale drug carriers, such as polymeric NPs, liposomes, micelles, dendrimers, and protein-based nanovectors, can be engineered to evade or exploit the BBB’s defenses in ways conventional drug formulations cannot. A key advantage is that nanocarriers can be “rationally” designed with specific physicochemical properties (size, shape, surface charge) and functional ligands to engage transport pathways of the BBB actively. In parallel, advances in materials science have led to “smarter” nanobiomaterials that respond to the brain microenvironment. For instance, stimuli-responsive nanocarriers that release their drug cargo upon sensing pH or enzymatic triggers in the brain, or neutrophil-mimetic NPs that can migrate across an inflamed BBB. These approaches collectively aim to maximize CNS delivery while minimizing systemic exposure. Indeed, targeted nano-delivery can reduce off-target toxicity. In one study, an intranasally administered ferritin NP was shown to concentrate in glioma tissue while largely sparing healthy organs, suggesting a safer profile than intravenous chemotherapy​ [13,14].

This review provides a robust conceptual and mechanistic framework for rationally designing nanobiotechnological systems capable of effectively crossing the BBB. Rather than merely listing existing delivery methods, we emphasize understanding nanoparticle (NP)-barrier interactions and identifying critical factors underlying successful translocation. We systematically examine key NP transport mechanisms, drawing from recent cellular and in vivo studies, and detail essential physicochemical parameters such as NP size, shape, surface chemistry, targeting ligands, and stimulus-responsive properties. Exemplary systems illustrate successful strategies, including functionalized polymeric NPs, engineered exosomes, and BBB-targeted lipid nanoparticles. Therapeutic applications across various CNS disorders are critically evaluated, showcasing preclinical data demonstrating improved drug delivery, enhanced brain bioavailability, extended survival, and neurological recovery. This integrative review serves as a practical guide for researchers developing NP-based therapies, assisting in the informed selection of design parameters and identifying appropriate validation models. Our framework supports the clinical translation of CNS nanotherapeutics by linking fundamental BBB biology to advanced nanotechnological engineering and addressing essential regulatory considerations. Ultimately, strategic and evidence-based NP design represents a powerful approach to effectively overcome the BBB, enabling targeted and efficacious brain therapies.

2. The BBB: Physiology and Restrictive Mechanisms

2.1. Composition and Function of the BBB

The BBB's composition and function comprise brain endothelial cells sealed by tight junctions, supported by pericytes, astrocytic end-feet, and a basement membrane. This neurovascular unit provides selective permeability, maintaining CNS homeostasis. Human brain capillaries extend extensively (~20 m² surface area), ensuring close neuronal proximity to the blood supply. BBB endothelial cells lack fenestrations and exhibit minimal vesicular transport, restricting transport pathways [15,16,17].

Pericytes, which are abundant in CNS vessels (covering more than 90% of vessels in mice), structurally support the blood-brain barrier and induce endothelial tightness. Astrocytic end-feet regulate BBB function, while the basement membrane (50–100 nm thick) restricts permeability and anchors cells [18,19]. These features yield high transendothelial electrical resistance (TEER) and low permeability, protecting the CNS from blood-derived fluctuations and toxins.

2.2. Tight Junctions and Paracellular Restriction

Tight junctions between endothelial cells, composed primarily of claudin-5, occludin, and junctional adhesion molecules linked to actin via ZO-1, create the primary physical barrier. These junctions minimize the paracellular space, allowing only tiny hydrophilic molecules (<1.8 nm diameter) to enter with negligible entry [6,20]. Claudin-5 deletion demonstrates selective permeability to molecules <800 Da, reinforcing the tight junctions’ role in size-dependent exclusion [21,22,23]. Consequently, more than 98% of small-molecule drugs fail to cross the intact blood-brain barrier (BBB) passively [4,24]. The integrity of tight junctions can be disrupted by inflammation or hypoxia, transiently increasing permeability, whereas the tightening of tight junctions correlates with BBB maturation [25]. Conversely, in development, the tightening of tight junctions (especially claudin-5 enrichment) correlates with the maturation of BBB function​ [23,26]. Overall, the tight junctions network endows the BBB with a restrictive paracellular gate that, in healthy states, admits to virtually no polar molecules larger than a few hundred daltons.

The BBB also features biochemical defenses: efflux transporters and metabolic enzymes. P-glycoprotein (P-gp) and breast cancer resistance protein (BCRP), abundant on endothelial membranes (P-gp: ~4–7 pmol/g, BCRP: ~8–16 pmol/g), actively extrude xenobiotics, significantly restricting CNS entry (Storelli et al., 2021). Impaired efflux activity in Alzheimer’s disease reduces clearance of neurotoxic peptides despite normal transporter levels [27]. Brain endothelial cells also express enzymes (γ-glutamyl transpeptidase, alkaline phosphatase, MAO, and cytochrome P450s) that degrade circulating molecules, thereby preventing CNS entry. Thus, tight junctions, efflux transporters, and enzymatic activity collectively ensure the BBB effectively excludes nearly all drugs and biomolecules from passive CNS entry [28].

2.3. Major Transport Mechanisms Across the BBB

Despite its stringent selectivity, the BBB allows regulated molecular transport via several distinct mechanisms. Passive diffusion enables the transport of a limited range of small (<400–500 Da), lipophilic molecules, including gases (O₂, CO₂) and certain drugs (nicotine, ethanol). However, only approximately 2–6% of small-molecule therapeutics penetrate the BBB passively due to stringent physicochemical constraints​ [28,29]. Tight junctions (~1–2 nm pore size) severely restrict paracellular transport between endothelial cells, permitting negligible passage of water and small ions. Pathological disruption of tight junction integrity can temporarily increase permeability to molecules that are otherwise restricted [6,21,22]. Carrier-mediated transport (CMT) involves specialized solute transporters that facilitate the uptake of essential nutrients, such as glucose (GLUT1, ~0.5 μmol/g/min), amino acids (LAT1), and vitamins. Notably, alterations in transporter expression occur in diseases such as Alzheimer's [27]. Receptor-mediated transcytosis (RMT) utilizes specific endothelial receptors (e.g., TfR, LDLR, LRP-1, insulin receptor) to shuttle essential peptides, proteins, and nutrients across the BBB. Although effective, RMT typically delivers limited amounts (~0.1–2% of the injected dose for therapeutic antibodies), sufficient for potent drugs [30]. Adsorptive-mediated transcytosis (AMT) involves non-specific electrostatic interactions with cationic or amphipathic molecules, significantly enhancing uptake (e.g., ~10–20-fold increase for cationic albumin). Despite its non-selectivity, AMT is exploited therapeutically, though potential cytotoxicity warrants careful control [31]. Table 1 summarizes the key features of each central transport mechanism across the BBB, highlighting representative molecules, size limitations, typical efficiencies, and relevant examples.

Given the complexity and strictness of BBB physiology, selecting appropriate experimental models is critical for evaluating nanocarrier transport, biodistribution, and therapeutic efficacy. Table 2 provides a comparative overview of in vitro and in vivo models for studying nanotechnology transport across the BBB. Each model offers physiological relevance, experimental control, and throughput balance. In vitro models range from simple immortalized cell monolayers with high scalability but low barrier tightness to advanced hiPSC-derived BBB-on-chip systems that closely recapitulate human BBB properties (with TEER values approaching in vivo levels) [32,33]. Incorporating astrocytes, pericytes, or flow typically enhances barrier integrity and transporter functionality, albeit with increased complexity. In vivo rodent models remain indispensable for demonstrating NP delivery in an intact organism; they provide the full neurovascular unit context and pharmacokinetic realism. However, species differences (e.g., rodent vs. human P-gp expression) must be accounted for [34]. Non-human primates offer the most human-like BBB for late-stage validation, although practical and ethical constraints limit their use. Notably, emerging zebrafish models can expedite early screening, but results should be extrapolated to mammals with caution. In summary, no single model is “best” for all purposes – researchers should choose the model that best fits their experimental goals. Early high-throughput screens can be done in simpler systems (cell lines or zebrafish), medium-complexity models (primary co-cultures, iPSC chips) are ideal for mechanistic insights, and critical translational studies are reserved for rodent and primate models. This strategic, multi-model approach maximizes discovery efficiency and clinical relevance, guiding nanomedicine development across the BBB toward successful human applications.

2.4. Key NP Properties for Crossing the BBB

Delivering therapeutics across the BBB requires NPs with finely tuned physicochemical properties. Size, surface charge, shape, and composition determine a nanocarrier’s ability to traverse the brain endothelium while evading clearance [14,49]. Recent preclinical studies (2020–2025) provide quantitative insights that guide the rational design of BBB-penetrant nanotechnologies, as summarized in Figure 2.

a)

Size

NP size profoundly influences BBB transport. Optimal diameters are generally in the tens of nanometers, balancing efficient transcytosis with minimal sequestration by clearance organs. In a mouse model, 20 nm insulin-coated gold NPs (AuNPs) achieved the highest brain accumulation (at 2 hours post-injection) compared to identical 50 nm and 70 nm formulations. The 20 nm AuNPs showed the widest biodistribution in brain tissue. In contrast, larger AuNPs were less effective​​ [50]. Similarly, an in vitro BBB model found that 25 nm PEGylated silica NPs had superior transport efficiency relative to 50 nm and 100 nm particles [51]. In vivo, 25 nm silica NPs (with slight positive surface charge) remained in circulation for over 24 hours and yielded a 6-fold higher brain drug concentration than free drug administration [52]. These small NPs likely exploit transcellular pathways and narrow paracellular gaps between endothelial cells​, whereas tight junctions mostly exclude larger particles [53,54].

There are practical size limits at both extremes. Particles <10 nm are rapidly filtered by the kidneys and can diffuse out of the brain almost as quickly as they enter​. For instance, ultrasmall ~2 nm gold nanoclusters readily cross the BBB but show non-negligible neurotoxicity and rapid clearance​ [55,56]. On the other hand, NPs >200 nm are prone to opsonization and sequestration by the reticuloendothelial system (RES) in the liver and spleen, leaving an insufficient fraction to reach the brain. Indeed, 200–250 nm gold NPs were found to accumulate less in the brain (and in systemic organs) than 10–15 nm gold NPs, which distributed more broadly (albeit with higher potential toxicity)​​ [56,57]. These studies indicate an optimal NP size window (roughly 10–100 nm) that maximizes BBB penetration. Within this window, 50–150 nm particles often achieve longer plasma half-lives than tiny ones yet are still small enough to extravasate; for example, ~50 nm polymeric NPs showed deeper brain tissue penetration in a brain injury model than 200 or 800 nm analogues​.

Additionally, particle size strongly influences the protein corona composition in circulation, impacting BBB receptor-mediated uptake. For instance, lipid NPs (~30 nm) preferentially bind apolipoprotein E, enhancing LDLR- and LRP-1-mediated transcytosis compared to larger counterparts. Notably, optimal size windows may shift under pathological conditions; ~100 nm polymeric NPs demonstrated superior accumulation in GBM models versus smaller or larger particles [58,59]. Thus, the ideal size range of ~10–100 nm balances efficient receptor targeting, prolonged systemic circulation, and enhanced brain delivery.

b)

Surface charge (zeta potential)

The BBB endothelium carries a negatively charged glycocalyx, meaning NP surface charge critically mediates electrostatic interactions with the barrier. Mildly cationic NPs can undergo enhanced adsorptive uptake by BBB cells, but strongly positive charge also risks disrupting the tight junction integrity. Quantitative perfusion studies have shown that neutral or slightly anionic surface charges are optimal for safe BBB transit​. Lockman et al. found that neutral NPs and low-concentration anionic formulations preserved rat BBB integrity and achieved brain delivery, whereas highly cationic NPs caused immediate BBB leakage and toxicity​. Notably, the brain uptake rate of anionic NPs (at non-disruptive concentrations) exceeded that of neutral or cationic versions in the same model​ [60]. This suggests a trade-off: positively charged NPs bind avidly to the negatively charged cell membranes, promoting internalization, but excessive positive charge triggers unwanted barrier opening and clearance by proteoglycan-rich regions of the endothelium.

Recent mechanistic studies support a moderate cationic charge for BBB crossing. A 2020 mathematical model incorporating endothelial surface charge predicted that a positively charged NP experiences enhanced transcellular permeability by electrostatic attraction to the negatively charged membrane [61]. Experimentally, cationic liposomes (+30 mV ζ-potential) show higher uptake by brain endothelial cells than equivalent neutral or −30 mV liposomes [62,63]. In one comparative study across eight cell lines, positively charged NPs had the fastest uptake rates vs. negative or neutral NPs​ [64]. However, high uptake does not guarantee deep brain delivery. Strong cationic particles may stick to the luminal surface or become entrapped in endothelial lysosomes. In contrast, negatively charged NPs (e.g. ~−35 mV) diffuse more freely and can penetrate further into brain parenchyma [62,65]. For example, one report noted that although +30 mV gold NPs enter BBB cells efficiently, -36 mV gold NPs penetrated brain tissue due to greater mobility through the endothelium [63]. The downside is that negative NPs (and uncharged PEGylated NPs) tend to be less readily internalized. In vivo, neutral or PEG-coated NPs often exhibit the longest circulation times; for instance, neutral PEGylated liposomes can circulate for hours. In contrast, highly cationic NPs are rapidly opsonized and cleared from the bloodstream [66]. Slightly negative or near-neutral surfaces thus appear to best balance BBB uptake and systemic stealth. In summary, a zeta potential in the range of –10 to +10 mV (or slightly beyond) is often favorable for BBB-targeted nanocarriers: it provides some affinity for the negatively charged endothelium without causing cytotoxicity or immediate clearance​ [60]. Fine-tuning surface charge (for example, by incorporating zwitterionic or charge-shielding coatings that become positive only near the membrane) is a strategic approach to maximize brain delivery.

c)

Shape

Whether spherical, rod-like, discoidal, or filamentous, NP shape has emerged as a critical design parameter for BBB transit. Shape affects how particles navigate blood flow, how they interface with cell membranes, and how phagocytes recognize them. Recent studies show that non-spherical shapes can improve brain delivery under certain conditions. For instance, polymeric nanorods targeted to transferrin receptors on BBB endothelium exhibited a seven-fold higher cellular uptake than equivalent spheres in an in vitro BBB model​ [67]. Although that dramatic 7-fold uptake was reported in an earlier study with functionalized rods, even untargeted rods demonstrate advantages in transcytosis efficiency. Nowak et al. (2020) used a human microfluidic BBB model to compare 200 nm spheres vs. rod-shaped particles: while spherical NPs adhered more to the endothelial surface, the rod-shaped NPs translocated across the endothelium about twice as efficiently per cell-bound particle. The authors hypothesized that rods enter endothelial cells via a distinct pathway or orientation that favors vesicular transport​​ [57,68]. In essence, a rod or filamentous shape can “partition” into the cell membrane differently than a sphere, potentially exploiting elongated endocytic pits or aligning with membrane invaginations.

Particle flexibility often goes hand-in-hand with shape. Filamentous micelles and worm-like filomicelles (long and flexible) have demonstrated extraordinarily long circulation and increased accumulation in some tissues, likely by evading phagocytosis. In the brain context, one benefit of discoidal or rod-shaped NPs is improved margination in capillaries; these shapes drift toward the endothelial wall under flow more than spheres, increasing the odds of BBB contact​. A 2024 study by Sierri et al. directly compared ~100 nm lipid NPs of identical chemistry but different shapes (spherical vs. discoid vs. deformable “soft” particles) in a human BBB model. Strikingly, discoidal NPs had about double the endothelial permeability of spheres (permeability coefficient ~1.3×10⁻⁵ cm/min for discoids vs ~6–7×10⁻⁶ for spheres). The deformable (ellipsoidal) NPs were intermediate in BBB crossing. The discoids also traversed intercellular tunneling nanotubes more efficiently, suggesting that shape influences initial BBB crossing and subsequent spread in brain tissue. The superior BBB transit of discoids was attributed to their larger surface area in contact with the cell membrane and favorable orientation during trafficking [69]. Consistent with this, Fu et al. (2022) found that rod-shaped polymeric NPs outperformed spherical NPs in crossing an in vitro brain microvascular model​ [70]. In summary, elongated or flattened shapes (rods, disks) promote BBB interaction and transcytosis, whereas spherical NPs often remain in circulation or are taken up by the RES. That said, extremes of shape can affect biodistribution: high aspect-ratio filaments might avoid uptake by liver/spleen but could be too large to traverse tight endothelial junctions if very long. Thus, an aspect ratio of roughly 2–5 (standard for short rods or discoids) appears optimal for BBB delivery​ [57]. By tailoring NP shape, researchers can influence circulation half-life (with filaments lasting longer in blood and enhance margination to brain capillaries, thereby improving the chances of BBB penetration.

d)

Stiffness

NP stiffness significantly influences BBB permeability and transcellular transport efficiency. Generally, stiffer NPs exhibit greater endothelial uptake and enhanced BBB penetration compared to softer particles, as evidenced by rigid polystyrene spheres (bulk modulus ~3000 MPa) demonstrating approximately tenfold higher translocation than softer PEG-based hydrogels (~3 MPa) [57,71,72,73].

In vivo GBM models confirm greater accumulation of stiffer NPs within brain tumors than their deformable counterparts [63]. Mechanistically, stiffer particles are more readily endocytosed due to lower membrane deformation requirements, although softer NPs may undergo more efficient intracellular trafficking and exocytosis, facilitating transcytosis. Stiffness also influences vascular margination: rigid particles migrate toward vessel walls, increasing endothelial interactions without disrupting tight junction integrity [57]. Thus, NP stiffness modulation represents a promising strategy for optimizing drug delivery across the BBB.

e)

Composition and material class

The composition of NPs (lipidic, polymeric, protein-based, inorganic) significantly influences their BBB permeability and biocompatibility.

Lipid-based NPs: Liposomes and solid lipid NPs (SLNPs) effectively cross the BBB, particularly when functionalized. PEGylated liposomes (~100 nm) have limited penetration alone, but ligand decoration markedly improves uptake. For instance, VCAM-1 peptide-targeted liposomes achieved ~6% injected dose/g brain versus ~1.2% untargeted [74]. Thiamine-functionalized SLNs also increased brain drug concentrations 1.4-fold [75]. Lipid NPs excel in genetic cargo delivery, showing efficient mRNA transfection with reduced off-target accumulation [66].

Polymeric NPs: Biodegradable polymers (PLGA, PLA, dendrimers) commonly enhance BBB transport via PEGylation or targeting ligands. LDL receptor-targeted PLGA NPs achieved ~4% ID/g brain compared to ~1% untargeted [18,76]. Polymeric nanocarriers offer tunable release kinetics and flexible shapes for enhanced penetration and tumor suppression [57].

Protein-based NPs: Ferritin and virus-like particles (VLPs) naturally exploit receptor-mediated pathways. Human H-ferritin effectively delivers antibodies into brain tumors [77]. Peptide-functionalized protein NPs (e.g., RVG29, transferrin, angiopep-2) further enhance receptor-specific uptake [78].

Inorganic NPs: Gold, iron oxide, and silica NPs provide stable cores suitable for functionalization and imaging. Targeted gold NPs (~50 nm) significantly outperform untargeted ones (0.23% vs. 0.04% ID/g) [56]. Iron oxide nanoparticles (SPIONs, <20 nm) achieve brain accumulation suitable for MRI imaging when suitably coated [18]. Moreover, an “optimal” iron oxide size of around 10–50 nm has been identified to maximize circulation time and minimize phagocytic uptake​ [79]. Silica NPs (~25 nm) form beneficial protein coronas, enhancing BBB penetration [52].

Table 3 presents an overview of recent experimental findings from original research on NP performance in brain delivery. It includes key physicochemical attributes (size, charge, composition), quantitative metrics on brain accumulation (% ID/g), circulation half-life, and therapeutic outcomes across both healthy and pathological BBB models. By contrasting NP types, lipid-based, polymeric, protein-based, and inorganic, under varied surface modifications and targeting strategies, this table is a critical reference for guiding the rational design of nanocarriers optimized for BBB penetration. Overall conclusions highlight that NP functionalization using BBB receptor-specific antibodies or appropriate polymer coatings significantly enhances brain uptake compared to untargeted particles. Optimal performance is typically achieved with NP sizes below 100 nm, carrying neutral or slightly negative surface charges, which facilitates longer circulation half-lives and improved brain retention. However, functionalization alone does not universally guarantee therapeutic efficacy; thus, surface modifications and particle composition must be carefully tailored to maximize clinical outcomes.

The BBB severely restricts brain uptake of systemically administered therapeutics. Innovative nanotechnologies are being developed to ferry drugs across the BBB in meaningful quantities (Figure 3). Recent studies emphasize human-relevant models to assess translational potential, from induced pluripotent stem cell (iPSC)-derived BBB cultures to non-human primates and early clinical trials. Below, we review key nanotechnological strategies for BBB crossing, highlighting quantitative brain delivery metrics (e.g., percentage of injected dose per gram brain, brain:plasma ratios) and therapeutic outcomes (tumor suppression, cognitive improvements) where available.

1. Receptor-Mediated Transcytosis (RMT)

RMT exploits endogenous nutrient/carrier receptors on brain endothelium to shuttle NPs or biologics into the brain. Binding to receptors such as the transferrin receptor (TfR), insulin receptor (INSR), low-density lipoprotein receptor-related protein 1 (LRP1), lactoferrin receptor, or folate receptor triggers internalization and vesicular transport across endothelial cells. This strategy has dramatically improved BBB delivery in multiple models:

1a. Transferrin Receptor (TfR)

The TfR is abundantly expressed on brain microvessels (especially during development) and is a prime RMT target​. Classical studies in rats showed that an anti-TfR antibody (OX26) achieved ~0.3% of the injected dose (ID) in the brain, over 10-fold higher than a non-specific IgG (0.03% ID). Recent work has refined TfR-targeting by adjusting antibody affinity/valency. Johnsen et al. demonstrated that gold NPs conjugated with a low-affinity, monovalent anti-TfR delivered 0.23% ID/g into brain parenchyma, versus only 0.04–0.08% ID/g with high-affinity bivalent variants​ [18]. Likewise, Denali Therapeutics engineered a TfR-binding “Enzyme Transport Vehicle” (ETV) with moderate affinity; in mice, an ETV–IDS (iduronate-2-sulfatase) fusion achieved broad CNS distribution and enzyme uptake by neurons, lowering pathological glycosaminoglycans in brain by 49–76%, compared to only 25–43% reduction using a high-affinity bivalent IgG format​. Notably, monovalent TfR carriers avoid “trapping” in endothelial endosomes, yielding higher transcytosis efficiency [91]. In non-human primates, systemically delivered TfR-targeted biologics have demonstrated distributed brain delivery without significant peripheral side effects, validating TfR RMT as a clinically viable route. For example, a bispecific antibody with one TfR-binding arm and one therapeutic arm reached ~0.5–1.1% ID/g in capillaries and significantly enhanced parenchymal delivery [18]. These advances underscore that binding affinity and avidity are critical: too high affinity can impede release into brain parenchyma, whereas optimized moderate affinity promotes deeper delivery.

1b. Insulin receptor (INSR)

The INSR is widely expressed on the BBB endothelium and can mediate brain uptake of insulin and analogues [92]. High-affinity human INSR antibodies (e.g., 83-14) have been used as “Trojan horses” to ferry drugs across the BBB. In a clinical trial for Hunter syndrome, an INSR-targeted enzyme (valanafusp alpha) showed CNS activity, indicating successful BBB penetration​ [82]. NPs have also leveraged INSR: for instance, HSA (human serum albumin) NPs conjugated with an insulin mimetic (29B4 antibody) penetrated the BBB and elicited CNS therapeutic effects in rodents. Although INSR targeting can risk hypoglycemic signaling, careful engineering (e.g., insulin agonist antibodies that activate transport without strong insulin-like effects) has enabled significant brain delivery [93,94].

1c. LRP1 (Low-Density Lipoprotein Receptor-Related Protein 1)

LRP1 is highly expressed on brain endothelium and shuttles ligands like ApoE and lactoferrin. The peptide Angiopep-2 (Ang-2) binds LRP1 and has been widely used to decorate NPs. Angiopep-functionalization triggers robust transcytosis; an Ang–2–drug conjugate crossed the BBB and accumulated in brain tumors in clinical trials (ANG1005 for glioma). Recent docking studies have optimized Ang-2 analogues for tighter LRP1 binding​ [95,96]; an Ang-2–drug conjugate crossed the BBB and accumulated in brain tumors in clinical trials (ANG1005 for glioma). Recent docking studies have optimized Ang-2 analogues for tighter LRP1 binding [97]. In one study, an artificial LRP1-binding peptide “L57” showed efficient uptake in primary human brain microvascular endothelial cells, indicating it may serve as a new BBB shuttle [98]. Tubular transcytosis pathways have been observed for LRP1, suggesting unique mechanisms that can be modulated by peptide “shuttles” [99].

1d. Lactoferrin receptor

Lactoferrin (Lf), an iron-binding glycoprotein, crosses the BBB via its receptor on brain endothelium [100,101]. NPs coated with lactoferrin exploit this route. For example, Youssef et al. (2025) coated lipid nanocapsules with Lf to deliver the antioxidant drug apocynin (APO) and lavender oil in a rat epilepsy model. The Lf-coated nanocapsules achieved greater brain accumulation of APO and significantly reduced seizure severity, yielding a Racine score of ~0.67 (near absence of seizures) versus much higher scores in uncoated NP or free drug groups. Lf-NPs also prolonged seizure latency and lowered neuroinflammatory markers, indicating effective therapeutic delivery to the brain​ [101]. Similarly, oral Lf-decorated gold NPs have been studied for GBM, given that LfR is upregulated in many tumors​ [102]. These results highlight lactoferrin’s potential as an endogenous targeting ligand to improve drug BBB permeability​.

1e. Folate receptor

Folate receptor-α is minimally expressed on the intact BBB, but it is present in the choroid plexus and often overexpressed in brain tumor cells (gliomas) [103]. Folic acid is a popular ligand targeting NPs​ (Y. Wu et al., 2023), especially targeting metastatic or leptomeningeal tumor deposits. In glioma-bearing models, folate-targeted NPs accumulate in tumors via folate receptor-mediated endocytosis​ [104]. For instance, folate-coated gold NPs loaded with indomethacin showed enhanced uptake in FR-positive glioma cells and extended survival in tumor-bearing mice [105]. While folate RMT is not a significant route into the normal brain, it provides a tumor-specific BBB bypass, improving drug delivery to FR-rich brain tumors without affecting normal brain tissue.

To maximize RMT delivery, researchers combine ligands (e.g., TfR antibody plus a cell-penetrating peptide or RVG peptide) on the same NP [78]. These multi-target NPs aim to engage multiple uptake pathways synergistically. Overall, RMT-based nanocarriers have achieved measurable brain uptake on the order of 0.1–0.5% ID/g in rodents [106], which, while modest in absolute terms, represents an order-of-magnitude improvement over untargeted delivery. More importantly, these carriers have shown therapeutic efficacy in disease models, from enzyme replacement in lysosomal disease (reducing CNS pathology) to drug delivery in brain tumors (inhibiting growth), underscoring the translational promise of RMT nanotechnologies [91].

2. Adsorptive-Mediated Transcytosis (AMT)

Adsorptive-mediated transcytosis relies on electrostatic attraction of cationic molecules to the negatively charged endothelial membrane, inducing nonspecific endocytosis. Unlike RMT, AMT does not require specific receptors, making it broadly applicable. Various cationic coatings and cell-penetrating peptides (CPPs) have been used to trigger AMT for brain delivery:

2a. Cationic cell-penetrating peptides

Poly-arginine peptides (such as R9), the TAT peptide (derived from the HIV transactivator protein), and penetratin (from Antennapedia) carry multiple positive charges that promote adsorption to the BBB surface. For example, NPs functionalized with PepH3, a 7-amino-acid cationic peptide derived from dengue virus capsid, showed greatly enhanced uptake in BBB models. PepH3-tagged NPs exhibited active transcytosis across both rat and human BBB cell monolayers​ [107,108]. In vivo, radiolabeled PepH3 derivatives achieved high brain uptake with low accumulation in liver, lung, and kidney, indicating a degree of selectivity [107,109]. Notably, making the endothelial surface less negative (by enzymatically removing glycocalyx sialic acids) reduced PepH3-NP uptake, confirming that electrostatic interactions are the primary driver. These findings show that cationic shuttles can exploit AMT to cross the BBB, and some (like PepH3) may do so with minimal off-target deposition​ [108]. Other CPPs, such as penetratin and poly(arginine) -8, have similarly increased NP translocation in vitro and in situ (e.g., brain perfusion models) by several-fold compared to unmodified NPs. However, in vivo quantitation is less common.

2b. Cationic polymers (e.g., chitosan)

Chitosan, a polycationic polysaccharide, has been widely used to decorate NPs for BBB delivery as recently reviewed by [110]. Its positive charge and mucoadhesive properties facilitate both AMT and modulation of tight junctions. Khan et al. (2023) demonstrated that DNA-loaded chitosan nanoparticles (NPs, ~260 nm, with a positive ζ potential) could effectively transfect brain cells in vivo after systemic administration. In their study, GFP-encoded chitosan NPs injected intraperitoneally in mice showed GFP expression in the brain, confirming that the NPs crossed the BBB. The brain delivery was achieved without the use of chemical targeting ligands; the chitosan’s inherent AMT property was sufficient. Notably, the chitosan vectors showed low cytotoxicity and immunogenicity in vitro (U87 glioma cells had ~85% viability) and no apparent toxicity in vivo​ [111]. These data position chitosan NPs as safe, efficient gene delivery vehicles to the CNS, leveraging adsorptive uptake. Other cationic polymers (e.g., polyethyleneimine, cationic dendrimers) also promote BBB transit, though toxicity must be carefully managed.

2c. Cationic surface coatings

Even without distinct peptides or polymers, tuning an NP’s surface charge can impact BBB uptake. Slightly positive or even “near-neutral” (mildly negative) zeta potentials favor BBB transcytosis [106]. For instance, one study found that rod-shaped polymeric NPs with a mildly negative surface had optimal uptake in brain endothelial cells (7-fold higher than neutral NPs) [67,112]. In practice, some researchers coat NPs with cell membranes (from e.g., leukocytes or platelets) to confer biological identity and charge that enhance BBB passage via a combination of AMT and other mechanisms. The general principle is that increasing NP affinity for the endothelial membrane (through charge or hydrophobic patches) can initiate vesicular transport across the BBB.

Overall, AMT-based approaches often lack the absolute specificity of RMT, but they offer versatility. They are beneficial for delivering large macromolecular complexes or gene vectors that might not fit into a single receptor pathway. By combining AMT peptides with targeting ligands (dual-function NPs), researchers aim to achieve both high uptake and specific delivery. The main quantitative limitation of AMT is potential sequestration in endothelial cells or perivascular spaces; however, evidence from studies like the chitosan NP study shows that a meaningful fraction of the dose can reach the parenchyma to exert biological effects.

3. Magnetically guided NPs

Magnetic targeting employs external magnetic fields to direct superparamagnetic iron oxide NPs (SPIONs) across the BBB, concentrating them in specific brain regions such as tumors. SPIONs coated with surfactants (e.g., Tween) have successfully crossed the intact BBB in rats when guided by magnets, with negligible uptake observed without magnetic fields​ [113,114,115].

3a. Magnetic liposomes

In glioma therapy, magnetic liposomes containing temozolomide and SPIONs demonstrated significant tumor accumulation, slowed tumor growth, and modestly improved survival under magnetic guidance compared to controls. These NPs also acted as MRI contrast agents without observed toxicity​​ [114].

3b. Magnetic hyperthermia combo

Combining magnetic targeting with hyperthermia via alternating magnetic fields (AMF) further enhances therapeutic outcomes. A recent study utilized Fe₃O₄@Chitosan@ZIF-8@RVG29 nanoparticles, resulting in effective glioma cell killing and tumor apoptosis [116].

Magnetic targeting significantly increases drug concentration and spatial precision in brain tumors. Despite requiring specialized equipment and facing limitations in magnetic field penetration, ongoing clinical trials suggest this approach as a promising strategy for overcoming the BBB.

4. Virotechnological strategies

Viruses naturally evolved to penetrate biological barriers, including the BBB (some neurotropic viruses cross via receptor-mediated mechanisms). Engineered viral vectors and virus-like particles leverage this capability for drug and gene delivery:

4a. Adeno-Associated Viruses (AAV)

AAV vectors (∼25 nm) are leading vehicles for gene delivery. Specific serotypes (e.g., AAV9) can cross the BBB, especially in neonatal mice or when the BBB is mildly perturbed. However, BBB transduction by natural AAVs is very limited in adult primates. Breakthrough work from 2020 to 2023 has resulted in engineered AAV capsids with enhanced blood-brain barrier (BBB) penetration. Chuapoco and colleagues developed AAV.CAP-Mac was selected by iterative screening for marmosets and macaques. In adult rhesus monkeys, AAV.CAP-Mac achieved widespread brain transgene expression after IV injection, transducing ~1.3% of cortical neurons, compared to only 0.5% with conventional AAV9.CAP-Mac showed broad tropism (neurons and some glia) and delivered functional genes (e.g., GCaMP for neural imaging) across multiple brain areas [117]. Another group created AAV-F (AAV.PHP.eB variants)* that crosses the BBB in mice by interacting with LY6A on endothelial cells, though these do not translate to primates. In 2023, an AAV capsid was reported that binds human TfR1 to ferry genes across the BBB. Such TfR-targeted AAVs increased mouse CNS gene delivery and showed uptake in ex vivo human BBB models [118]. The quantitative gains are striking; one variant (AAV.CAP-B10) gave 3–4 times higher neuronal transduction in macaques than AAV9, with much lower liver off-targeting. These advances suggest that IV AAV gene therapy for the brain is becoming feasible. Indeed, systemic AAV9 is already used in infants (e.g., onasemnogene for SMA) to deliver genes to spinal motor neurons, leveraging a partially immature BBB. Ongoing clinical studies are exploring AAV capsids for adults with Alzheimer’s and Parkinson’s, aiming to achieve a few percentage points of CNS targeting, enough for therapeutic benefit in many cases [117].

4b. Lentiviral and other viral vectors

Lentiviruses (e.g., modified HIV-1) and adenoviruses can transduce brain cells but generally do not cross the BBB efficiently when given IV. Hence, they have been used via direct injection into the brain or CSF. Recent innovations involve coating or retargeting these larger viruses. For example, a lentivirus pseudotyped with rabies glycoprotein (RVG) was shown to reach CNS neurons from the bloodstream in mice, using RVG’s ability to interact with nicotinic acetylcholine receptors on BBB endothelium [116]. Another approach uses exosomes to “piggyback” lentiviral vectors (see hybrid exosome section). While systemic lentiviral delivery to the brain isn’t yet routine, the field is moving toward producing virus-like particles decorated with BBB-targeting ligands.

4c. Virus-Like Particles (VLPs)

These are self-assembled protein cages derived from viruses, but without any viral genome, essentially “nanocontainers” that can carry drugs or genes. VLPs from various sources have been tested for brain delivery, including bacteriophage Qβ and MS2, plant viruses like Tobacco Mosaic Virus (TMV) and Cowpea Chlorotic Mottle Virus (CCMV), and human virus capsids (e.g., hepatitis B core). VLPs typically range 20–150 nm, a size amenable to crossing fenestrated barriers and, with modifications, potentially the BBB. For instance, TMV nanorods (300×18 nm) were albumin-coated to prolong circulation and successfully used to image and treat brain tumors in mice​ [68,119]. The filamentous shape of TMV may aid transport along endothelial cells. Qβ VLPs, which are ~30 nm icosahedra, have been functionalized with peptides such as angiopep or transferrin to engage RMT, effectively combining VLPs with the RMT approach. One study reports that exosomes loaded with a VLP carrying rhodamine dye crossed the BBB and increased brain drug delivery [120], illustrating a hybrid of VLP and exosome strategies. While still preclinical, VLPs offer high customizability (both genetic and chemical) to display targeting ligands and can be produced in high yields from plants or bacteria. They also tend to be biodegradable and less immunogenic than whole viruses. The challenge is achieving efficient BBB traversal; current VLPs require surface functionalization. Nonetheless, VLPs are a versatile platform, such as a ~120 nm hepatitis B VLP, which has been used to carry siRNA and target brain metastases, showing improved survival in mice (by homing to a tumor antigen and releasing siRNA upon endocytosis).

In summary, viral vectors have the advantage of active transport mechanisms (receptor- or cell fusion-mediated) that can be highly potent; for example, a single dose of AAV can transduce millions of neurons if it crosses the blood-brain barrier (BBB). The quantitative goal often cited is achieving a few percent of the total brain cell population transduced or a brain-to-serum drug ratio approaching 0.1; recent viral innovations are closing in on these targets [117]. Meanwhile, non-pathogenic VLPs offer a safer alternative, such as nanocarriers, that can utilize viral entry mechanisms without replicating them. Both are promising for diseases requiring gene therapy, enzyme replacement, or widespread neuromodulation.

5. Exosomes and extracellular vesicles

Exosomes (30–150 nm extracellular vesicles) naturally cross biological barriers and mediate intercellular communication, making them attractive endogenous delivery vehicles for the brain [121]. They possess cell-derived membrane proteins that can confer innate brain tropism; for instance, exosomes from neurons or macrophages may preferentially home to the brain. Key features of exosomes include biocompatibility, low immunogenicity, the ability to be loaded with drugs or RNA, and the flexibility to engineer their surface. Recent developments include:

5a. Intrinsic brain tropism

Certain exosomes inherently migrate to the brain. MSC (mesenchymal stem cell)-derived exosomes have been shown to cross the BBB in models of stroke and neuroinflammation. In a mouse stroke model, IV MSC exosomes enhanced functional recovery by improving neurogenesis and synaptic remodeling in the brain [122]. This therapeutic benefit implies significant vesicle delivery to brain tissue. Tissue distribution studies often find exosome uptake in the brain 2–5 fold higher than equivalent doses of free drug [120]. Moreover, exosomes can cross the BBB without disrupting it; they likely use native endocytosis/exocytosis pathways.

5b. Surface-engineered exosomes

To enhance targeting, researchers decorate exosome membranes with peptides or antibodies. A prominent example is the exosome displaying the RVG peptide (a rabies virus glycoprotein fragment that binds to nicotinic receptors). Cui et al. (2019) report that RVG-tagged exosomes, loaded with therapeutic cargo, showed enriched delivery to the cortex and hippocampus after IV injection, significantly improving learning and memory in Alzheimer’s mice. In that study, the RVG-exosomes (carrying siRNA to knockdown BACE1) restored cognitive function in a murine Alzheimer’s model, whereas non-targeted exosomes had minimal effect. Another group used vesicular stomatitis virus G protein (VSV-G) on exosomes to target brain endothelium, achieving successful gene delivery across the BBB in mice (pseudotyping exosomes with viral fusogens can create a hybrid between viral vector and natural vesicle)​ [123]. These approaches highlight that by displaying targeting ligands (such as RVG, RGD, and antibodies), exosomes can be directed to specific brain regions or cell types, much like synthetic nanoparticles, but with a biological cloak.

5c. Drug delivery and hybrids

Exosomes have been loaded with chemotherapeutics, proteins, or siRNA for brain delivery. Haney et al. demonstrated that exosomes can transport the anti-cancer drug doxorubicin and even large antibodies across the blood-brain barrier (BBB). For glioblastoma, exosomes carrying the antibody cetuximab (targeting EGFR) plus doxorubicin significantly increased antibody brain accumulation and inhibited tumor growth compared to free cetuximab plus the drug. Exosomes delivered more than double the amount of antibody to brain tissue compared to direct antibody injection in that study, due to protected transport and possibly transferrable membrane proteins aiding in BBB transit [124]. Beyond using natural exosomes, “hybrid” NPs are being designed, for example, by coating polymeric NPs with exosome membranes (sometimes referred to as cell-derived nanovesicles) to endow them with exosomal homing abilities [120]. These hybrids have shown increased brain uptake and reduced clearance by the mononuclear phagocyte system. One study mentions glioma-targeted exosomes delivering siRNA to suppress oncogenes, crossing the BBB, and prolonging animal survival​ [125].

Exosome-based delivery is already moving to clinical testing. A first-in-human trial used autologous exosomes loaded with curcumin to treat patients with brain cancer and showed good safety at high doses (though efficacy is still under evaluation). The primary challenges are scalable production and efficient cargo loading. However, since exosomes can be derived from patient cells, they offer a personalized and highly biocompatible mode of therapy. They naturally avoid P-glycoprotein efflux and can even regulate BBB permeability (some brain endothelial uptake of exosomes triggers signaling that temporarily loosens tight junctions) [121] with their ability to cross the BBB and deliver functional payloads (e.g., restoring memory in AD models [123], exosomes are poised to become a powerful tool in the neurotherapeutic arsenal.

Active transport nanomotors represent an emerging approach for BBB traversal, with early evidence suggesting significantly improved drug penetration compared to passive methods. The continued development of these active, targeted, and responsive systems holds substantial promise for advancing BBB drug delivery.

Table 4 summarizes nanotechnological strategies for crossing the BBB, detailing delivery vehicles, targeting mechanisms, experimental models, precise quantitative brain uptake metrics, and therapeutic outcomes. Receptor-mediated transcytosis (RMT) is a clinically validated strategy, demonstrating robust brain delivery efficiencies (~0.1–1% ID/g) and significant therapeutic efficacy. Notably, transferrin receptor (TfR)-targeted NPs, including T7-functionalized PLGA NPs, achieve approximately sixfold increases in brain delivery, with measurable improvements in therapeutic outcomes in stroke and glioma models. Angiopep-2-functionalized lipid–silica NPs similarly doubled drug concentrations in the brain, highlighting the advantage of moderate receptor affinity ligands such as LRP1 for brain-targeted therapies.

2.5. Toxicological and Regulatory Considerations for Translational Advancement of BBB-Crossing Nanobiomaterials

A significant limitation in the clinical translation of BBB-crossing nanobiomaterials is the scarcity of robust clinical data. Translational medicine faces substantial barriers, primarily due to inadequate characterization of safety and efficacy profiles. A critical step forward involves implementing well-designed preclinical studies following a Quality by Design (QbD) approach, ensuring rigorous evaluation of critical attributes before clinical trials.

Before progressing to clinical application, CNS-targeted nanomedicines must undergo rigorous safety and regulatory assessments. To facilitate this essential stage, Table 5 provides an organized overview of critical pre-in vivo toxicological evaluations, including hemocompatibility, neurotoxicity, and long-term accumulation, along with specific regulatory references and recommended methodologies. This structured approach ensures comprehensive characterization and regulatory compliance for NPs designed to cross the BBB.

Hemocompatibility assessments (hemolysis <5%, minimal platelet activation, and low complement activation) ensure intravenous safety, mitigating risks of thrombotic and immune-mediated adverse events [152]. For neurotoxicity, CNS-targeted NPs should demonstrate no significant behavioral impairments or histopathological abnormalities at therapeutic doses, aligning with safety pharmacology guidelines [153,154]. Moreover, long-term accumulation studies must confirm that NPs are cleared from the CNS over time, ideally exhibiting >90% elimination within weeks post-administration, thereby minimizing delayed neuroinflammation or chronic toxicity [155].

Regulatory bodies (FDA and EMA) mandate comprehensive biodistribution and pharmacokinetic profiling, necessitating quantitative in vivo analyses using labeled nanomaterials to map NP fate, organ retention, and BBB crossing efficiency [156]. Likewise, immunotoxicity evaluations must preclude excessive immune activation or complement-mediated reactions (CARPA), verifying that anti-NP antibodies or cytokine storms are absent or manageable [157].

Finally, achieving manufacturing reproducibility during scale-up is pivotal. Establishing and maintaining Critical Quality Attributes (CQAs), such as particle size, polydispersity (PDI), zeta potential, and drug loading, across batches ensures consistent therapeutic outcomes and regulatory compliance [158]. Nanomedicine developers should proactively define and validate manufacturing processes to control batch-to-batch variability, leveraging GMP practices and possibly advanced techniques like microfluidics to maintain quality during scale transitions. Collectively, these data-driven recommendations form a robust foundation for safely advancing CNS-targeted NPs from preclinical validation towards clinical implementation.

3. Conclusions

This comprehensive guidance establishes a practical and rigorous framework for rationally designing and validating nanobiotechnological systems capable of effectively traversing the BBB. The summarized tables delineate key parameters for optimal NP physicochemical properties, targeting mechanisms, and validated experimental models, guiding researchers through informed and methodical design decisions. By integrating detailed considerations of NP size, charge, composition, and regulatory and toxicological standards, this guide enables precision engineering of nanocarriers that demonstrate improved CNS targeting, therapeutic efficacy, and safety profiles.

However, significant translational hurdles remain despite these structured approaches and technical advances. Future research must systematically address regulatory compliance, scale-up reproducibility, and rigorous safety validation, explicitly highlighted through recommended toxicological assays, immunogenicity assessments, and pharmacokinetic evaluations in the tables. By proactively addressing these practical considerations and rigorously validating nanoparticle performance in biologically relevant models, nanomedicine can more effectively overcome current barriers, accelerating the clinical realization of innovative therapies for neurological disorders.