Mahmoud A. Younis
abc,
Yusuke Sato
abd,
Seigo Kimura
e and
Hideyoshi Harashima
*ab
aLaboratory of Innovative Nanomedicine, Faculty of Pharmaceutical Sciences, Hokkaido University, Sapporo 060-0812, Japan. E-mail: harasima@pharm.hokudai.ac.jp
bInstitute of Vaccine Research and Development (IVReD), Hokkaido University, Kita 21, Nishi 11, Kita-ku, Sapporo 001-0021, Japan
cDepartment of Industrial Pharmacy, Faculty of Pharmacy, Assiut University, Assiut 71526, Egypt
dLaboratory for Molecular Design of Pharmaceutics, Faculty of Pharmaceutical Sciences, Hokkaido University, Sapporo 060-0812, Japan
eIntegrated Research Consortium on Chemical Science, Graduate School of Science, Nagoya University, Nagoya 464-8602, Japan
First published on 22nd July 2025
Applying lipid nanoparticle (LNP) technology to ribonucleic acid (RNA) nanomedicines was integral to the success of mRNA vaccines against COVID-19. To expand the power of LNP technology, extrahepatic delivery systems have been developed using specific ligands that target the cells in question. However, recent increases in evidence support targeting without the need to attach specific ligands to nanocarriers. In this review, we focused on protein corona-mediated extrahepatic delivery of nanoparticles as an alternative to classic ligand-mediated active targeting. First, the interaction of LNPs with biological components and the impact that the physicochemical properties of LNPs exert on their biological fate are discussed. Then, we highlight a new system that targets activated hepatic stellate cells (aHSCs) as a successful model achieved through intensive optimization of LNPs based on an ionizable cationic lipid library. We also discuss cumulative evidence that support the ligand-free extrahepatic delivery of nanoparticles to a broad diversity of tissues, such as the spleen, lungs, brain, tumors, kidneys, placenta, pancreas, and bone marrow. In conclusion, we propose protein corona-mediated extrahepatic delivery as a new strategy of active targeting for RNA nanomedicines and inspire the future directions in this area.
Active targeting has been the main strategy to achieve cell-specific delivery of nanomedicines during the past two decades.6,7 By exploiting the differential expressions of certain receptors or unique features in the microenvironments of target cells (e.g. pH, redox potential, enzymatic activity, etc.), nanocarriers are equipped with targeting moieties that either bind to these receptors or respond to such specific features to increase their affinity to the cells in question.8 A wide variety of targeting moieties have been investigated, ranging from small ligands to antibodies and chemical linkers.9,10 Despite their promising potential in vitro, the in vivo application of ligand-based delivery systems is challenged by multiple obstacles, including their poor stability in biological fluids, improper pharmacokinetic performance, immunogenicity, and difficulties in controlling ligand–receptor binding in the dynamic three-dimensional in vivo environment, which is much more complicated than the static two-dimensional cell culture models. Moreover, the scale-up of such delivery systems is limited by their intricacy and multiple preparation steps. Collectively, the abovementioned factors reduce the clinical translatability of ligand-based nanocarriers.11 Thus, innovative targeting approaches are necessary to cope with the aforementioned challenges.
Recently, the role of the “protein corona” has been recognized as an important factor in the in vivo distribution of LNPs12 since LNPs are covered with plasma/serum proteins once they enter blood circulation. Hajipour M. J. et al. reported on 4022 unique proteins that were identified on nanoparticles based on a literature search using physicochemical properties, sizes, and protein sources.13 There is increasing information available on the relationship between certain components of the protein corona and tissue distribution of nanoparticles, such as that for the liver (apolipoprotein E; ApoE), lung (serum albumin and ApoE), spleen (complements and immunoglobulins such as opsonins), and brain (ApoE4 and ApoB-100).14 Therefore, a new strategy has emerged for tissue-selective targeting based on the protein corona as an endogenous ligand without the external introduction of artificial ligands.15
In this review, we first explore the molecular mechanisms of interactions between nanoparticles and biological components from the viewpoint of using a library of ionizable cationic lipids. Subsequently, we highlight a successful case study in which an active targeting system to the activated hepatic stellate cells (aHSCs), a model of a specific minor cellular population, was developed for potential clinical applicability in the treatment of liver fibrosis. Furthermore, we summarize recent strategies for protein corona-mediated delivery to various extrahepatic tissues, such as the spleen, the lungs, the brain, tumors, kidneys, the placenta, the pancreas, and bone marrow. Finally, we inspire future perspectives in this research area.
LNPs are predominantly composed of lipids. Although various models have been proposed for an improved internal structure, these generally represent a structure with a large number of lipid molecules inside.30 Although the specific gravity of LNPs becomes slightly higher when nucleic acids are encapsulated, these forms are much lighter than that of nanoparticles made of harder materials.31 Therefore, LNPs exhibit specific gravity similar to that of the various lipoproteins and extracellular vesicles present in blood. In addition, the particle sizes of these endogenous vesicles and LNPs are within a similar range. Therefore, it is assumed to be difficult to highly purify corona-formed LNPs from biological fluids by using a single separation method, such as size exclusion chromatography (SEC) or centrifugation. The conventional approach dictates that either a combination or the optimization of these methods is necessary.32 At least two studies have reported on affinity purification methods using antibodies against polyethylene glycol (PEG); these methods typically use polymers to control the particle size of LNPs and improve their stability in blood.33,34 In other reports, Francia V. et al. demonstrated that magnetic iron-oxide-loaded LNPs (IOLNP) allow for separation of LNP-corona complexes from biological media through magnetic separation,35 and Baimanov D. et al. developed a rapid affinity-based technique achieved by chemically immobilizing LNPs via carbodiimide cross-linking on the surface of biosensors.36 The development of such new purification approaches is expected to facilitate future elucidation of the biomolecular corona of LNPs. In addition, a number of studies have reported analyses of the biomolecular corona via LNP purification from biological fluid using only simple centrifugation or SEC.37–39 Criteria assuring the quality of the samples measured are considered to be future issues. From another perspective, it has also been noted that the results of analyses differ depending on the facility under which the proteome analysis is performed, and comparisons between independent reports should understandably proceed with caution.40
Phospholipids are the structural lipids in LNPs, and are preferentially localized at the interface of LNPs due to their relatively bulky hydrophilic head group.41 Therefore, their chemical structure and content has a significant impact on the interfacial properties of the LNPs. Chander et al. found that when the phospholipids in LNPs were replaced with egg sphingomyelin (ESM), their content could be increased by 40 mol%, which resulted in decreases and increases in functional mRNA delivery in the liver and secondary lymphatic tissues (spleen and bone marrow), respectively, compared with LNPs containing the typical 10 mol% of 1,2-distearoyl-3-sn-glycero-phosphocholine (DSPC).42 The ESM improved both blood stability and blood half-life, which is attributed to increases in the opportunities for mRNA delivery into extrahepatic tissues. Hashiba et al. compared DSPC content from 5 to 25 mol% and found that LNPs containing 25 mol% DSPC improved the functional mRNA delivery to extrahepatic tissues, including the spleen, the lungs, and kidneys.43 Substitution of DSPC with DSPG biocentrically tilts the selectivity of LNPs towards reticuloendothelial system (RES) cells, which includes liver sinusoidal endothelial cells (LSECs) and Kupffer cells.44 Since phospholipids are also involved in the pH-change-induced reorganization of LNPs and have significant impact on the level of functional mRNA delivery,45 the role of phospholipids in enhancing function and diversifying applicability of the LNPs will be significant in the future.
Recently, Fei et al. proposed a SELECT platform (simplified LNP with engineered mRNA for cell-type targeting), which is different from the usual LNP without phospholipids, that can still deliver mRNA to the lung effectively.46 They started from a 5-component system (SM-102, DOTAP, DSPC, Chol, PEG) and found no effect of DSPC on the transfection activity in the lung. They examined a 4-component system (SM-102, DOTAP, Chol, PEG) excluding DSPC, and found no effect of cholesterol. Finally, they found that a 3-component system (SM-102, DOTAP, PEG) can exert the highest transfection activity in the lung compared to those of the 4- and 5-component systems. Therefore, the precise role of phospholipids in the transfection activity of mRNA should be carefully examined in each tissue of concern with specified ionizable lipids.
The apparent value for the acid dissociation constant (pKa) of LNPs could be controlled depending on the chemical structure of the ionizable lipids.47,48 The pKa is an important property for functional RNA delivery, and has an optimal range of 6.2–6.5 in the liver.49 Sato et al. have developed various ionizable lipid libraries with a wide range of pKa values, and reported the effects on both the cell and tissue selectivity of LNPs. Comparisons of the intrahepatic distribution of LNPs with different pKa values have shown that LNPs with a pKa below 6.0 selectively accumulated in liver parenchymal cells, and showed a significant change in selectivity towards liver sinusoidal endothelial cells (LSECs) and even Kupffer cells with increases in the pKa.47 Modulation of the pKa by a combination of two ionizable lipids with different pKa values results in similar LSEC-specific accumulation and gene silencing after siRNA delivery, suggesting that the LSEC-tropic properties of the LNPs are pKa-dependent rather than ionizable lipid chemical structure-dependent.50 Compared with the aforementioned DSPG-LNPs, the signs of the charges are opposite, which suggests different accumulation mechanisms. By using our original ionizable lipid library, Younis et al. showed that activated hepatic stellate cells (aHSC) efficiently take up LNPs with a pKa of around 7.5,51 This pKa range is similar to what is optimal for LSECs, but the use of 1,2-dioleoyl-3-sn-glycero-phosphoethanolamine (DOPE) as a phospholipid was also important. Interestingly, the same LNPs were also shown to be spleen-tropic in healthy mice, but aHSC-tropic only in liver-fibrosis models. This phenomenon could be due to changes in the biomolecular corona associated with plasma proteome changes during pathological conditions (i.e., changes in the biological identity of LNPs), as described above. Furthermore, Hashiba et al. developed an ionizable-lipid library that focuses on branched scaffold structures, which includes the total carbon number and symmetry. The results found in this library suggest that the symmetry of branching significantly contributes to physical stability during storage and fusogenicity. Interestingly, these results suggest that the structure of ionizable lipids, specifically the total carbon number of branched scaffolds, significantly changes the tissue tropism of functional mRNA delivery between the liver and the spleen.52 The fact that the head structure of ionizable lipids is constant and that there is no significant difference in the pKa of LNPs suggests that any change in biological identity could be due to changes in membrane fluidity or to other physicochemical properties.
The particle size of LNPs has a significant impact on functional RNA delivery, and reducing the diameter of LNPs to below ∼50 nm significantly reduces the level of functional RNA delivery, which is only observed in the presence of serum.47,53,54 While this could be explained as membrane destabilization due to sparse lipid packing and increased interfacial energy with decreasing particle size, differences in the biomolecular corona are also suggested. Indeed, the effect of particle size on the biomolecular corona of nanoparticles made of various materials has been clarified.55,56 In particular, the larger the particle size, the greater the amount of protein adsorbed and the thicker the corona layer.55 A correlation between the amount of protein and cellular uptake has also been reported. Adsorption of immunoglobulin also results in a complement reaction on the nanoparticle interface, which promotes uptake into phagocytes (i.e., clearance from the blood) via the complement pathway.57,58 We have shown that the particle size of RNA-loaded LNPs could be largely controlled by adjusting either the amount of PEG modification or the salt concentration during the manufacturing process, which leads to increased levels of functional RNA delivery to peritoneal macrophages or to splenic dendritic cells following either intraperitoneal or intravenous administration, respectively.59–61 Kranz et al. have shown that mRNA-lipoplexes with diameters of approximately 200 nm, and without ligand decoration, efficiently delivered mRNA into splenic dendritic cells, and clinical trials are underway by BioNTech as mRNA cancer vaccines for melanoma patients.62,63 The particle size of mRNA-LNPs also appears to influence the induction of antigen-specific antibody expression following intramuscular administration.64
Thus, the chemical structure and physical properties are positively correlated with the in vivo fate of nanoparticles. An understanding of the biomolecular corona reveals these correlations. The chemical space of nanoparticles is practically infinite, and currently only a small fraction of it has been explored. Although this indicates the enormous potential of nanoparticles, it would be difficult to randomly explore all of them. It is important to search for nanoparticle formulations that show efficient and useful properties via biocentric approaches and machine learning based on inspiration from researchers.
Liver fibrosis is a chronic disorder that can proceed into irreversible and life-threatening complications. Activated hepatic stellate cells (aHSCs) are the fundamental players in the development and progression of liver fibrosis. In response to a chronic liver injury, the hepatic stellate cells transform from a quiescent phenotype (qHSCs) into activated myofibroblasts (aHSCs), with a subsequent deposition of extracellular matrix (ECM).68,69 While gene therapy sounds promising for the treatment of liver diseases, the gene delivery to aHSCs is a challenging process owing to the stroma-rich microenvironment of a fibrotic liver that restricts the access of nanocarriers to their target cells.70 Moreover, the hard-to-transfect nature of fibroblast-based aHSCs is a formidable barrier to efficient gene transfer.71 Furthermore, the vast majority of the reported drug-delivery systems to aHSCs have relied on ligand-based approaches (e.g., targeting peptides, retinoid derivatives, or antibodies) that exert a negative impact on the in vivo stability and pharmacokinetic performance of such systems.71,72
We introduced a novel concept for the ligand-free targeting of aHSCs via the use of a protein corona-based approach for the gene therapy of liver fibrosis, which is outlined in Fig. 2. To develop this approach, a library of molecularly diverse ionizable lipids was recruited for the delivery of either mRNA or siRNA to aHSCs, both in vitro and in vivo. A microfluidic device, which is referred to as an invasive lipid nanoparticle production device (iLiNP), was applied to achieve scalable and precise control of the physico-chemical properties of the produced LNPs. Interestingly, the apparent acid dissociation constant (pKa) of the LNPs was found to play a pivotal role in determining the intrahepatic tropism of the LNPs,73 where a semi-neutral pKa (∼7.25) favored aHSCs, but an acidic pKa (5.8–6.7) was more favorable for hepatocytes. Subsequently, two ionizable lipids, CL15A6 and CL15H6, were recognized for their high selectivity to aHSCs. In addition, the particle size of the LNPs was a key factor in determining the RNA delivery efficiency to aHSCs in vivo, where a sub-hundred nm size was essential for an efficient penetration through the stroma barrier of the fibrotic liver, which emphasized the value of recruiting microfluidics technology in the preparation of LNPs that encapsulate RNA therapeutics.5,51 Moreover, the nitrogen/phosphate (N/P) molar ratio, by which ionizable lipids and RNA are mixed, had a substantial impact on the RNA delivery efficiency. An optimum N/P ratio was required to achieve a balance between the RNA encapsulation efficiency, the endosomal escape capability, and bioavailability of the RNA cargo. Low N/P ratios resulted in poor packaging of the RNA cargo in question, particularly that of mRNA. Furthermore, LNPs with low N/P ratios demonstrated a poor level of functional RNA delivery efficiency, probably because of their low endosomal escape capabilities, which is the critical factor that determines the fate of a nucleic acid payload following the cellular uptake process. High N/P ratios also had a negative impact on the RNA delivery efficiency, owing to the extensive packaging of the anionic RNA, which limits its intracellular release and availability.74 The results revealed that an N/P ratio of 8–12 is optimum for mRNA delivery to aHSCs. Meanwhile, a lower N/P range of 4–8 is favorable for siRNA delivery. The discrepancies of the optimum parameters between mRNA and siRNA could be attributed to the different molecular size, which affects their packaging into the LNPs and their subsequent intracellular release from them. It is also noteworthy that the mRNA-loaded LNPs had a larger particle size (∼80 nm) compared with that of their siRNA-loaded counterparts (∼50 nm), which subsequently affects the surface area of LNPs, and potentially affects the nature of the protein corona adsorbed to them.5,51 Eventually, the helper lipids incorporated into the LNPs dramatically affect their in vivo performance. LNPs incorporating phosphoethanol amine derivatives of either 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine (DOPE) or 1,2-dipalmitoyl-sn-glycero-3-phosphoethanolamine (DPPE) were superior to those with a phosphatidyl choline moiety. This probably could be attributed to their ability to undergo conformational transitions from lamellar to inverted hexagonal orientations, which promotes interactions with the endosomal membrane and subsequently the capability for endosomal escape.75,76 Interestingly, incorporating phosphatidyl cholines into the LNPs shifted their tropism from aHSCs to hepatocytes, where the unsaturated derivative 1,2-dioleoyl-sn-glycero-3-phosphocholine (DOPC) demonstrated a higher efficiency compared with that of the saturated derivative 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC).51
![]() | ||
Fig. 2 A schematic of a novel strategy to reprogram aHSCs and reverse liver fibrosis using self-homing LNPs. The composition of LNPs and their physico-chemical properties were tweaked using a combination of ionizable lipids and helper lipids. Following intravenous administration, the LNPs harness endogenous serum proteins to generate a protein corona that houses aHSCs in a fibrotic liver. Tuning the particle size using a microfluidic device allows LNPs to penetrate the stroma-rich microenvironment and access their target. Subsequently, the LNPs deliver a cocktail of siRNAs to reprogram aHSCs into qHSCs and reverse liver fibrosis. The figure is adapted from Younis et al.,5,51 with permission from Elsevier (Copyright 2023, Elsevier). |
Following the systematic optimization described above, self-homing LNPs were generated with high selectivity and efficient RNA delivery to aHSCs. Mechanistic investigations have suggested that such LNPs are delivered to the liver via apolipoprotein E (ApoE)-independent machinery, and are potentially recognized by endogenous platelet-derived growth factor (PDGF), which is elevated in the serum in the case of fibrotic diseases. Subsequently, the LNPs are taken up by aHSCs through overexpressed platelet-derived growth factor receptor beta (PDGFRβ) via clathrin-mediated endocytosis. The administration of monoclonal antibodies targeting PDGFRβ prior to the administration of LNPs reduced the RNA functional delivery efficiency by approximately 90%, which supports the above hypothesis.5
Finally, the abovementioned LNPs were loaded with a cocktail of siRNAs targeting smoothened homologues (SMO) and transforming growth factor beta 1 (TGFβ1) signaling in aHSCs. Therapeutic assessment in mice undergoing thioacetamide (TAA)-induced liver fibrosis revealed a significant reversal of liver fibrosis and restoration of normal liver functions. Analysis of the glioma-associated transcription factors (GLI) suggested the reprogramming of aHSCs into qHSCs by the simultaneous knockdown of Hedgehog (Hh) and TGFβ1 cascades, which are the key pathways involved in the activation of hepatic stellate cells.77 Meanwhile, single-target monotherapies showed partial amelioration of the fibrotic status, but failed to accomplish the therapeutic goals or restore the healthy status of the liver.51
![]() | ||
Fig. 3 The mechanism of SORT LNP tissue targeting, including the formation of unique protein coronas. Dilliard et al. investigated protein coronas formed on the SORT LNPs. β2-GPI and vitronectin were determined to be the most abundant of proteins, and act as endogenous ligands for targeting the spleen and lungs, respectively. (A) A proposed three-step endogenous targeting mechanism for tissue-specific mRNA delivery by SORT LNPs in which 1) PEG lipid desorption 2) enables distinct plasma proteins to bind SORT LNPs, 3) resulting in cellular internalization in the target tissues by receptor-mediated uptake. (B) Ex vivo bioluminescence of major organs excised from C57BL/6 mice IV injected with liver, spleen, and lung SORT LNPs incorporating either sheddable PEG lipids (C14-PEG2K) or less sheddable PEG lipids (C18-PEG2K) (0.1 mg FLuc mRNA/kg body weight, 6 h). Total luminescence produced by each organ is reduced when less sheddable PEG lipid is used, suggesting that PEG lipid desorption is a key process for efficacious mRNA delivery by SORT LNPs. (C) Quantification of total luminescence produced by functional protein translated from FLuc mRNA in target organs of C57BL/6 mice IV injected with liver, spleen, and lung SORT LNPs incorporating either C14- or C18-PEG2K (0.1 mg FLuc mRNA/kg body weight, 6 h). (D) ELISA quantification of serum hEPO in C57BL/6 mice treated with liver, spleen, or lung SORT LNPs encapsulating hEPO mRNA (0.1 mg hEPO mRNA/kg body weight, 6 h). Using a less-sheddable PEG reduces SORT LNP potency. (E) SDS–PAGE of the plasma proteins adsorbed to the surface of mDLNP, liver SORT, spleen SORT, and lung SORT LNPs. LNPs with different organ-targeting properties bind distinct plasma proteins. (F) The average abundance of proteins with distinct biological functions in the protein coronas of mDLNP and liver, spleen, and lung SORT LNPs. The choice of SORT molecule leads to large-scale differences in the functional ensemble of plasma proteins which bind the LNP. (G) Isoelectric point distribution for the most enriched proteins which constitute 80% of the protein corona of the LNPs. A SORT molecule’s headgroup structure influences the pI distribution of the protein corona. (H) The top five most abundant plasma proteins that bind different SORT LNPs (n = 3). The chemical structure of SORT molecule affects the number one plasma protein that is most highly enriched on the surface of SORT LNPs. Data are shown as mean ± SEM. Statistical significance was determined using an unpaired two-tailed Student’s t test (*P < 0.05). The figure is reproduced from Dilliard et al.143 with permission from Natl. Acad. Sci. USA (Copyright 2021, Natl. Acad. Sci. USA). |
Based on these results, recent approaches to LNP-based spleen targeting has highly depended on the surface charge, which includes several conditions such as the use of ionizable lipid with a lower apparent pKa value, the formation of RNA-rich lipoplexes to achieve a negative net charge, and the incorporation of negatively charged phospholipids. However, an excessive negative charge decreases the functional delivery of mRNA due to electrostatic interactions with cationically charged lipids, which leads to an inhibition of endosomal escape. Therefore, with the exception of surface charge, an alternative targeting strategy is desirable. Fenton et al. succeeded in developing an ionizable lipid, OF-Deg-Lin, that is capable of delivering mRNA into splenic B lymphocytes.80 Interestingly, intravenous injection of the OF-Deg-Lin LNPs resulted in spleen-specific gene expression, while mainly accumulating in liver tissue. The LNPs were taken up into splenic B lymphocytes at a level consistent with that of splenic monocytes/macrophages. The OF-Deg-Lin LNPs showed an apparent pKa value of 5.7, which is lower than the typical value suitable for liver targeting. However, detailed targeting mechanisms such as an endogenous ligand/receptor were not elucidated. By contrast, Suzuki et al. recently developed an ionizable tri-oleoyl-Tris (iTOT) library. iTOT lipids consist of bulky unsaturated scaffolds similar to that of OF-Deg-Lin, and are more hydrophobic than commercially available ionizable lipids. The authors discovered that increasing the molar ratio of DSPC dramatically suppressed and improved the functional delivery of mRNA in the liver and spleen, respectively. Although this was commonly observed when using both the iTOT and commercially available ionizable lipids, TOT-5, which is one of the iTOT lipids and showed a lower pKa, achieved maximal spleen-specificity. This suggests a greater level of hydrophobic and near-neutral surface properties, which is important. It is noteworthy that the DSPC-rich formulation resulted in greater levels of both hydrophobicity and microviscosity, which resulted in LNP interface-adsorbed proteins that are more conducive to initiating complement pathways compared with the action of apolipoproteins. As a result, the DSPC-rich LNPs functionally delivered mRNA into splenic B (particularly marginal zone B and MZB) cells via the C3b-CD21/35 pathway, which induced MZB-mediated antigen-specific anti-tumor immunity following intravenous injection and enabled the development of safer intramuscularly administered mRNA vaccines.81,82 In the same context, Younis et al. reported that DSPC-rich (20 mol%) LNPs based on an ionizable cationic lipid CL15H6, with a nearly-neutral surface charge, tend to travel to the splenic dendritic cells post-intravenous administration. This creates a potential application as an anticancer vaccine that could demonstrate performance superior to the clinically relevant formulation, RNA-lipoplex, at the same dose.83 Although the precise mechanism of selectivity remains under investigation, the complement pathway is expected to contribute to the delivery of such LNPs, owing to the presentation of DSPC at the surface of LNPs.41
Diverse cell types in the spleen and lymph nodes play different and important roles in the complicated immune cascade, and little is known about target selectivity for diverse cell types. While selective targeting to various immune cell types is important, the formulation of design strategies that focus only on relatively simple properties such as surface charge will face limitations. The discovery of physicochemical properties that could overcome these limitations and the development of novel chemical spaces, as well as advances in high-throughput screening technologies, are highly desirable.
Fehring et al. have reported on the development of a lung-tropic siRNA-lipoplex (DACC lipoplex) composed of their original cationic lipid AtuFECT01 (β-L-arginyl-2,3-L-diaminopropionic acid-N-palmityl-N-oleyl-amide trihydrochloride), cholesterol, and mPEG2000-DSPE (1,2-distearoyl-sn-glycero-3-phosphoethanol amine-N (methoxy (polyethylene glycol)-2000)) at a molar ratio of 70:
29
:
1.84 The DACC lipoplex showed a highly cationic property (zeta potential of 40–50 mV). When the DACC lipoplex formulation was used to deliver siRNA to lungs in an injected dose of 40% per gram of lungs, a reduction of greater than 80% in Tie-2 mRNA was documented following consecutive tail vein injections. Improved survival of a lung metastasis model mouse was achieved via CD31 inhibition of the lung tissue. An incorporation of 50 mol% of permanently cationic lipid, either 1,2-dioleoyl-3-trimethylammonium-propane (DOTAP), dimethyldioctadecylammonium (DDAB) or 1,2-dimyristoyl-sn-glycero-3-ethylphosphocholine (EPC), into ionizable lipid-based LNPs was prepared to deliver mRNA into lungs (referred to as lung-targeting SORT LNPs).78 The lung-targeting SORT LNPs had a higher (>9) apparent pKa. Proteomic analysis of a protein corona revealed an enrichment of vitronectin (Vtn) on the lung-targeting SORT LNPs (Fig. 3).38 Vtn has a RGD motif and is known as an endogenous ligand against αvβ3 integrin. Further investigation revealed that variations in the chemical structure of both the hydrophobic scaffold and the hydrophilic headgroup structures impact the quality of functional mRNA delivery to the lungs.85 The lung-targeting SORT LNPs were able to access lung epithelial cells, which are located in the deeper tissues. This indicates the potential for diverse therapeutic applications such as therapeutic protein expression and genome editing. The lung-selective delivery of mRNA encoding broadly neutralizing antibodies against SARS-CoV-2 variants by utilizing the lung-targeting SORT technology has achieved significant expression of the encoded human monoclonal antibody 8-9D in the lungs and in bronchoalveolar lavage fluid (BALF).86 Standard liver-tropic LNPs induce high levels of 8-9D antibodies in serum, but have failed to achieve a sufficient concentration of antibodies in BALF. This is because IgG itself lacks the ability to translocate from the bloodstream to mucosal tissues. Importantly, the lung-targeting LNPs exhibited nearly complete protection against beta and Omicron BA.2 strain challenges, whereas their liver-targeting counterparts failed, clearly suggesting the importance of controlled and efficient expression of antibodies in the lungs. As such, the incorporation of permanently cationic lipids into ionizable-based LNPs is a promising strategy for targeting the lungs. However, Omo-Lamai et al. reported that the lung-targeting highly cationic LNPs induce massive thrombosis in the lungs and other organs.87 The thrombosis occurs through the binding of LNPs and changes the confirmation of fibrinogen, activating both platelets and thrombin. The fibrinogen-mediated dangerous clotting could be addressed by pre-treatment/conjugation of the direct thrombin inhibitor bivalirudin and a reduction in the LNP size. Qiu et al. discovered that lipidoids with amide bonds (N-series) can achieve lung targeting. However, those with ester bonds (O-series) target the liver.39 In this system, permanently cationic lipids are not required for lung targeting. Proteomic analysis revealed enrichment of fibrinogen and fibronectin on the N-series LNPs, suggesting these enriched proteins contribute to lung targeting. However, as mentioned above, fibrinogen binding potentially leads to blood clotting. Indeed, the N-series LNPs increase the plasma levels of thrombin–antithrombin (TAT), which is a marker of recent clotting in the same manner as the lung-targeting SORT LNPs.87 However, the N-series LNPs achieved induction of genome editing in lungs after co-delivery of Cas9 mRNA and single-guide RNA, which suggests its potential as a radical treatment for lung-related genetic diseases. Xue et al. screened combinatorially synthesized cationic degradable lipids for lung targeting in vivo utilizing DNA barcoding technology, and identified top-performing lipid CAD9.88 The LNP-CAD9 exhibited superior functional delivery of Cre mRNA to the lungs compared with the MC3/DOTAP-based SORT LNPs. No information was available in this study regarding either endogenous ligands or apparent pKa values of the LNP-CAD9.
Targeting the brain for drug delivery is particularly challenging due to several physiological and biochemical barriers, with the blood–brain barrier (BBB) being the most formidable. Regarding macromolecular brain delivery, innovative therapeutic modalities for various neurodegenerative diseases and brain-related genetic disorders have been developed. These include systemically administered AAV vector, Zolgensma,91 intrathecal administered antisense oligonucleotides, Nusinersen,92 the anti-amyloid beta antibody drug, aducanumab,93 and the transferrin (Tf) receptor-targeted antibody–protein conjugate Pabinafusp Alfa.94 Strategies using nanoparticles containing LNPs have also been studied, but most of them involve active targeting in which ligand molecules such as antibodies or sugars are presented on the nanoparticle surface.95,96 These have not been implemented as drugs, and one of the reasons is that the formation of a protein corona inhibits their targeting ability. For example, Tf-conjugated silica nanoparticles bound to serum proteins mask the surface Tf, which reduces the targeting of Tf receptors.97 Another study reports that polymeric nanoparticles modified with the HIV-1 trans-activating trans-activator peptide and/or alpha neural/glial antigen 2—which are known to cross the blood–brain barrier (BBB) and target oligodendrocyte precursor cells, respectively—are unable to cross the BBB, likely due to the formation of a protein corona.98 Conversely, the formation of a protein corona is a double-edged sword with respect to targeting, and factors that function positively in brain delivery have also been reported. Certain surfactant- or peptide-modified nanoparticles (NPs), such as poly(ethylene glycol)-, polysorbate-, or amyloid β-protein (Aβ)-CN peptide-modified NPs, can absorb apolipoproteins (Apos) like ApoE or ApoB, forming Apo-rich protein coronas. These absorbed proteins interact with lipoprotein receptors on the blood–brain barrier (BBB), thereby facilitating NP entry into the brain.99–101
The following is an example of protein corona-mediated brain targeting using systemically administered lipid-based carriers. We also discuss the delivery of small molecules and macromolecules through structural optimization of the constituent lipids, although there is no mention of the impact of a protein corona. One example of protein corona-mediated targeting involves modifying the binding pattern of apolipoproteins on the surface of liposomes, which has been shown to enhance accumulation in the brain and demonstrate therapeutic effectiveness in a mouse model of glioma.100 This report details how the liposomal surface was altered using a short, non-toxic peptide derived from beta-amyloid (Aβ1–42). This modification specifically targets the lipid-binding domain of apolipoproteins to control their adsorption patterns. This engineered liposomal system enables brain-targeting proteins to associate within the bloodstream, effectively exposing their receptor-binding domains on the liposomal surface as they circulate. The second example highlights a study on the lipid derivatization of the neurotransmitter (NT) tryptamine. LNPs containing these lipid derivatives were used for the delivery of small molecules, antisense oligonucleotides (ASO), and proteins to the brain via systemic administration.102 The authors selected three types of NTs—dimethyltryptamine, phenethylamine, and phenylethanolamine—and synthesized NT-derived lipidoids (NT1: dimethyltryptamine, NT2: phenethylamine, NT3: phenylethanolamine). They examined brain accumulation of these lipidoid-incorporated liposome-like nanoparticles using fluorescent labeling, which showed that NT1-lipidoids enhanced brain accumulation more effectively than either NT2- or NT3-lipidoids. NT1-lipidoid nanoparticles combined with various lipids were used to deliver classic polyene antifungal drugs (amphotericin B), tau-ASO, and GFP-Cre fusion proteins to the brain. A possible mechanism of BBB permeation by serotonin receptors and other receptors expressed in brain vascular endothelial cells has been considered, but the details of this mechanism are unknown. Other reports also have screened for brain-targeted LNPs using ionizable lipid libraries conjugated with neuroprotective factors (vinpocetine, berberine) or small molecule ligand structures (L-DOPA, D-serine, temozolomide, tryptamine, cinnamic acid, MK-0752) that are able to pass through the BBB.103–105 These strategies are distinct in their simplicity of design compared with that of more complicated methods of modifying antibodies and other ligand molecules on the surface of nanoparticles. Furthermore, mRNA is known to be delivered to cerebral vascular endothelial cells without the use of targeting ligands, simply by altering the type and composition of lipids. However, the role of the protein corona is not discussed, and details of the mechanism remain unknown.106 The influence of the protein corona on brain delivery and the BBB permeation mechanism are still relatively new research areas, but are being vigorously studied, with future progress expected.107–109
A classic understanding of active targeting has attributed the mechanism of nanoparticle delivery to the tumor solely to the binding of the targeting ligand to its target receptor on the cancer cells. However, we recently demonstrated a different perspective. Upon decorating an LNP formulation with SP94 peptide, which has a strong selectivity to hepatocellular carcinoma cells (HCC), the LNPs demonstrated a highly selective delivery of siRNA to HCC in vitro.113 On the contrary, such a formulation failed to induce significant gene silencing in the tumor tissue in vivo. Our systematic optimization revealed that the targeting ligand is not the sole player that affects the in vivo fate of the nanocarriers. Indeed, the composition and physico-chemical properties of the nanocarriers may exert more important roles. In a stroma-rich tumor model like HCC, ultra-small lipid nanoparticles (usLNPs) with an average particle diameter of 60 nm exerted powerful gene silencing activity in comparison with larger particles, owing to their high capability of penetrating the tumorous microenvironment. In addition, the inclusion of a low ratio of PEG-lipid derivative into usLNPs filled the gaps in the curved surface of the small-sized nanoparticles, which subsequently masked their recognition by apolipoprotein E (ApoE) and reduced their off-target delivery to healthy liver tissues. Furthermore, the ratio of the ionizable lipid and the nature of the helper lipid incorporated into the LNPs had a dramatic impact on their tumor-specific siRNA delivery efficiency.75
Another approach in the same context involves the synthesis and screening of huge libraries of molecularly diverse biomaterials, through which the structure–activity relationship (SAR) of such materials could be understood, and materials with intrinsic tumor-homing properties could be identified. Siegwart's group synthesized a library of 1500 biodegradable materials via sequential orthogonal reactions, where biodegradability was imparted through the introduction of metabolically labile ester bonds, and molecular diversity was applied in the cores and peripheries. A candidate, 5A2-SC8, succeeded in achieving a potent siRNA delivery in mice bearing an aggressive MYC-driven HCC model. Moreover, it specifically delivered a tumor-suppressor microRNA, let-7 g, which attenuated tumor growth and improved the survival rate of the mice in question.114
Furthermore, functionalization of the nanocarriers with bioinspired or biomimetic materials could generate an “artificial” protein corona that affects their in vivo fate. Huang et al. functionalized nanoparticles with clusterin (Apo J) to serve as an artificial protein corona, which reduced the hepatic and splenic distribution of the nanoparticles and improved their tumor accumulation.115 In another study, Pan et al. reported that the modification of nanocarriers with bovine serum albumin (BSA) increases their homing to the tumor endothelial cells via the albumin receptor, gp60.116 Cell-derived coatings have also been investigated. For example, in their natural setting, red blood cells (RBCs) evade RES through the activation of an inhibitory molecule on the surface of macrophages, which is referred to as “signal regulatory protein alpha (SIRPα)”. Therefore, functionalization of the nanoparticles with RBC membranes is known to prolong their circulation time by 10-fold compared with the classic PEG coatings.10,117 Similarly, Zhuang et al. harnessed a platelet membrane coating to selectively deliver siRNA against survivin genes in a mouse model of breast cancer.118
![]() | ||
Fig. 4 Impact of the particle size on the distribution of nanoparticles in renal compartments. Nanoparticles smaller than 2 nm are entrapped in endothelial glycocalyx, GBM, and podocytes, while those sized 5–130 nm pass through the glomerular filtration barrier. Smaller nanoparticles (<100 nm) tend to enter the urinary space, whereas larger ones are retained in the mesangium. Glomerular deposition diminishes for nanoparticles >100 nm. Mesoscale nanoparticles (350–400 nm) can enter tubular epithelial cells. Larger nanoparticles (>500 nm) face cellular uptake limitations, and micrometer-scale nanoparticles are trapped in pulmonary capillaries. The figure is reproduced from Cheng et al.,119 with permission from John Wiley and Sons (Copyright 2024, John Wiley and Sons). |
The surface charge of nanoparticles is also an important parameter in glomerular filtration. The filtration barrier has a strong negative charge from heparan sulfate in the glycocalyx and anionic proteoglycans on the GBM. These exert a strong repulsive force on the negatively charged particles, which prevents the glomerular passage of negatively charged particles.128
Shape is another important factor that helps determine the efficiency of glomerular filtration.128 Disk or rod-like nanoparticles have longer circulation times compared with that of spherical nanoparticles because it is easier for macrophages to internalize spherical shapes due to the higher probability of interactions with the high curvature of membranes. Single-walled carbon nanotubes with a molecular weight of ∼350–500 kDa, a length of 100–500 nm, and a diameter of 1.2 nm were found to be similarly cleared via glomerular filtration.129 As long as nanoparticle width is below the size cutoff (∼10 nm), different lengths follow an order of magnitude for clearance via glomerular filtration.
In an interesting recent study, Whitehead's group reported a successful strategy for the selective delivery of mRNA to the insulin-producing pancreatic β cells via the use of intra-peritoneally (IP)-administered LNPs. These LNPs were formulated using a combination of structurally diverse ionizable lipids (lipidoids) and helper lipids, where the incorporation of a cationic helper lipid, DOTAP, improved the efficiency of mRNA delivery to the pancreas. Interestingly, mechanistic investigations suggested that LNPs were first taken up by peritoneal macrophages, which subsequently facilitate the pancreatic delivery of mRNA through an exosome-mediated horizontal gene transfer.137
In another study, Shen and co-workers reported on the delivery of interleukin-12 (IL-12)-encoding mRNA to an orthotopic model of pancreatic ductal adenocarcinoma following the IP administration of LNPs based on a cationic lipid, P6CIT, and a mixture of cholesterol, DSPC, and DMG-PEG 2000 as helper lipids. The composition of LNPs was optimized using the Design-of-Experiments (DoE) approach. The developed therapy succeeded in the immunological reprogramming of the tumor microenvironment from a “cold” to a “hot” microenvironment, with a subsequent efficient eradication of the tumor.138 Although the precise mechanism of the LNP delivery to the spleen was not revealed in this study, peritoneal macrophages could have played a role in mediating such a delivery, in a manner similar to that of the above-mentioned study.
Dahlman's group applied the concept of protein corona-based targeting to deliver either siRNA for gene silencing or single-guide RNA (sgRNA) for genome editing to BMECs post systemic administration. A barcoding strategy was introduced, in which LNPs were labelled with specific short DNA barcodes to enable high-throughput screening of more than 100 LNPs in a single mouse. Subsequently, next-generation sequencing (NGS) was used to track the in vivo fate of the administered LNPs. Through an in vivo-directed evolution strategy, an enriched LNP (referred to as BM1) was identified. BM1 had a simple composition of an ionizable lipid, 7C1, cholesterol, and C18PEG2K. Interestingly, the particle size of LNPs did not affect their in vivo tropism. By contrast, the chemical composition of LNPs, mainly the cholesterol ratio and the length of the PEG-lipid tail, had a dramatic impact on the in vivo tropism toward BMECs. The authors hypothesized that the selectivity to BMECs could have been attributed to a combination of two potential factors. First, the length of the PEG-lipid tail could have affected the pharmacokinetics of LNPs and shielded them from the reticulo-endothelial system (RES), with a subsequent escape from hepatic and splenic accumulations. Second, the introduction of cholesterol could have affected the binding of serum proteins to the surface of LNPs, which subsequently would have affected the composition of the formed protein corona.140
Target tissue/cells | Formulation name | Particle size (nm) | PDI | Z-Potential (mV) | pKa | N/P | EE (%) | Lipid composition | Endogenous ligand/receptor | Physiological condition | Strategy to find | Chemistry | Ref. |
---|---|---|---|---|---|---|---|---|---|---|---|---|---|
Abbreviations: EE, encapsulation efficiency; DSPG, 1,2-distearoyl-sn-glycero-3-phosphoglycerol; Chol, cholesterol; DMG-PEG2K, 1,2-dimyristoyl-rac-glycero-3-methoxypolyethylene glycol-2000; DSG-PEG 2000, distearoyl-rac-glycerol-PEG 2000; DoE, design of experiments; DOTMA, 1,2-di-O-octadecenyl-3-trimethylammonium propane; DOPE, 1,2-dioleoyl-sn-glycero-3-phosphoethanolamine; 18PA, 1,2-distearoyl-sn-glycero-3-phosphate; DSPC, 1,2-distearoyl-sn-glycero-3-phosphocholine; DOPS, 1,2-di-(9Z-octadecenoyl)-sn-glycero-3-phospho-L-serine; PS, phosphatidyl serine; DSPE-PEG 2000, 1,2-distearoyl-sn-glycero-3-phosphoethanolamine-N-[amino(polyethylene glycol)-2000]; DOPC, 1,2-dioleoyl-sn-glycero-3-phosphocholine; HCC, β2-GPI, β2-glycoprotein I; hepatocellular carcinoma; DODAP, 1,2-dioleoyl-3-dimethylammonium-propane. | |||||||||||||
Spleen/immune cells | srLNP | 66.6 (CryoTEM) | — | −22 | — | 6 | 93 | MC3![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Stabilin-2 (receptor) | Healthy | Replacing neutral phospholipids with anionic phospholipids | Ionizable lipid and anionic phospholipid | 44 |
CL15F6-4 LNP | 229.3 | 0.316 | −5.87 | 6.9 | 6 | 66.7 | CL15F6-4![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Not determined | Healthy | Systematic examination of the tail structure of ionizable lipids | Ionizable lipids with branched tails of different total carbon length | 144 | |
A-11-LNP | 547 | 0.19 | −1.4 | — | 7.2 | 89.2 | CL4H6![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Not determined | Healthy | Formulation optimization by DoE | Ionizable lipid | 145 | |
RNA-LPX | ∼270 | ∼0.24 | ∼−32 | — | 0.65 | ∼80 | DOTMA![]() ![]() ![]() ![]() |
Not determined | Healthy | Optimization of the mRNA/lipid charge ratio | Cationic lipid | 62 | |
Spleen SORT | 167.8 | 0.144 | −1.84 | 3.97 | — | — | 5A2-SC8![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
β2-GPI | Healthy | Systematic examination of the fifth lipid component | Ionizable lipid and anionic phospholipid | 143 | |
PS-LNP | 95.46 | 0.132 | −2.58 | 6.31 | 6 | >85 | MC3![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Not determined | Healthy | Biomimetic use of PS that recognized by phagocytes | Ionizable lipid and PS derivative | 146 | |
OF-Deg-Lin LNPs | 75 | 0.197 | — | 5.7 | — | 59 | OF-Deg-Lin![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
C3b-CD21/35 (receptor) | Healthy | — | Ionizable lipid with degradable linkages | 80 | |
15% DSPC LNP | 109 | 0.11 | 1.8 | 6.2 | 10.2 | 90.3 | TOT-5![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
C3b-CD21/35 (receptor) | Healthy | Increase in the hydrophobicity and microviscosity of the LNP interface | Ionizable lipid and helper phospholipid | 82 | |
CL15H6 LNPs | ∼80 | <0.1 | −0.5 –(2.4) | 7.25 | 8 | >90 | CL15H6![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Not determined | Healthy | Step-wise formulation optimization | Ionizable lipid and helper phospholipid | 83 | |
Lung | DACC lipoplex | 74 | 0.25–0.35 | 46.4 | — | 8.4 | — | AtuFECT01![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Not determined | Healthy | In vivo screening of biodistribution | Cationic lipid | 84 |
Lung SORT | 114.8 | 0.159 | −0.89 | >11 | — | — | 5A2-SC8![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Vitronectin | Healthy | Systematic examination of the fifth lipid component | Permanently-cationic lipid | 143 | |
306-N16 LNP | 81 | 0.2 | −5 | — | — | — | 306-N16![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Fibrinogen | Healthy | Examination of the impact of linker chemistry on organ targeting | Lipidoid with an amide linker in hydrophobic tails (N-series) | 147 | |
LNP-CAD9 | 150.1 | 0.21 | 4.5 | — | — | 71.6 | 3-A2-7b![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Not determined | Healthy | In vivo screening via DNA barcoding | Combinatorial cationic degradable (CAD) lipid | 148 | |
Brain | SLN (cetyl palmitate NP) | 211.3 | — | −10 | — | — | — | Cetyl palmitate, polysorbate 80 | Artificial ApoE4 corona | Healthy | Protein corona decoration (ApoE4 adsorption) | Cetyl palmitate (C16) solid core stabilized by polysorbate 80 surfactant | 99 |
sLip/SP-sLip | 152 | 0.06 | −39.5 | — | — | — | HSPC![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
ApoE, ApoA1, ApoJ | Intracranial glioblastoma (U87) in mice | Rational design (brain-targeting peptide to recruit ApoE) | Peptide (SP) on PEG-DSPE, HSPC | 100 | |
PTX loaded Aβ-CN-PMs (PTX/Aβ-CN-PMs) | 103.1 | 0.184 | 7.23 | — | — | 90.3 | 20 mg of mPEG2000-PLA1300, 1.8 mg of Aβ-CN-PEG2000-PLA1300, and 2 mg of PTX | ApoE (enriched corona via Aβ peptide) | Orthotopic U87 glioma in mice | Rational design (Aβ peptide for ApoE adsorption) | PEG-PLA polymer with an Aβ peptide graft | 101 | |
NT1 (tryptamine)-derived lipidoid LNPs | 100–800 (AmB-loaded LNP), 150–160 (ASO-loaded LNP), 100–250 (GFP-Cre loaded LNP) | 0.2–0.4 (AmB-loaded LNP), 0.2 (ASO-loaded LNP) | 20–30 (AmB-loaded LNP), −18 (ASO-loaded LNP) | — | — | — | NT1-O12B:PBA-Q76-O16B (AmB-loaded LNP), NT1-O14B:306-O12B-3:DSPE-PEG2K (ASO-loaded LNP), NT1-O14B:PBA-Q76-O16B (GFP-Cre-loaded LNP) | Not determined | Healthy | Rational design (neurotransmitter-derived lipidoid) | Tryptamine-derived ionizable lipid (NT1) | 102 | |
Vinpocetine-derived ionizable lipidoid LNPs | 93.7 | 0.167 | 1.7 | 6.2–6.5 | — | 91.1 | A5-B1-C4.2![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Not determined | Healthy/AD model mice | Library screening (structure–activity for BBB and CBF) | Cyclic tertiary amine (vincamine derivative) | 103 | |
Berberine-inspired ionizable lipid LNPs | 63.94 | 0.201 | −7 | 6.0–7.0 | — | — | A2-B13![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Not determined | Healthy/APP/PS1 mice | Library screening (alkaloid scaffold) | Tetrahydroisoquinoline (protoberberine) | 104 | |
MK16 BLNPs | 125 | 0.125 | 6 | 6.86 | BL/mRNA = 12.5/1 | 80 | MK16![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Not determined | Healthy/disease | Combinatorial screening (BLNP library) | Proprietary ionizable lipid scaffold | 105 | |
Cat-LNP | 76 | 0.13 | (Cationic) | ∼6 | ∼3/1 | 96 | cKK-E12 (51.4%)![]() ![]() ![]() ![]() |
Not determined | Healthy | Rational design (cationic helper lipid) | Pyridazine-based cationic lipid (cKK-E12) | 106 | |
Tumor | 5A2-SC8 | 64 | — | −0.74 | 6.59 | — | >90 | 5A2-SC8![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Not determined | HCC | High throughput screening of a huge library (1500 compounds) | Modular degradable dendrimers | 114 |
DODAP LNPs | ∼80 | ∼0.2 | −1.6 | ∼6.5 | 6 | >90 | DODAP![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Not determined | Breast cancer | Formulation screening and optimization | Ionizable lipid and helper phospholipid | 149 | |
Kidney | Polymeric mesoscale NPs | 386.7 (A-MNP), 402.8 (C-MNP) | — | −19.5 (A-MNP), 18.3 (C-MNP) | — | — | — | PEG/PLGA | Not determined | Healthy | Random | Poly(lactic-co-glycolic acid) conjugated to polyethylene glycol | 150 |
Placenta | LNP55 | 122 | 0.265 | 3.81 | 5.92/5.76 (pipette/microfluidic) | — | 79.4 | C14-494![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
β2-GPI | Pre-eclampsia | In vivo screening via DNA barcoding | Ionizable lipid | 133 |
Pancreas | Lipidoid LNP | 173.9 | 0.138 | — | 6.4 | — | 89.3 | Lipidoid 306Oi10![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Not determined | Healthy | Formulation screening and optimization | Branched ionizable cationic lipid | 137 |
Pantgt LNP | 167–208 | 0.12–00.31 | (−2)–(−5) | 7.17–8.34 | — | ∼80 | P6CIT![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Not determined | Pancreatic cancer | Formulation optimization by DoE | Polyethyleneimine (PEI)-based cationic lipid | 138 | |
Bone marrow | BM1 | 45.2 | 0.19 | — | 6.55 | — | — | 7C1![]() ![]() ![]() ![]() ![]() ![]() ![]() ![]() |
Not determined | Healthy | In vivo screening via DNA barcoding | Ionizable lipid and long-chained PEG-lipid | 140 |
Although ligand-based active targeting relies on a concrete knowledge of specific pathways and receptors, making it more reliable from the theoretical point of view, the clinical translation of such delivery systems is hampered by their low stability, complexity, and scalability issues.11 On the other hand, protein corona-based targeting simplifies the composition and production of nanoparticles, which can address the above challenges and contribute to the improvement of the clinical translatability of nanomedicines. Nevertheless, there are still multiple challenges that encounter such a promising approach, including the inter-subject variability in the levels of the endogenous ligands in question as well as their alterations depending on the physiological and pathological conditions, which subtracts from the reproducibility of the developed delivery systems.141 In addition, the limited precise information on the interactions between such delivery systems and the endogenous macromolecules and their physiological consequences raises some concerns on the biosafety of this emerging technology, and complicates its adoption from a regulatory point of view.142
In the present article, we highlighted a successful delivery system to activated hepatic stellate cells in fibrotic liver, which has been developed based on screening a library of ionizable cationic lipids with an intensive optimization of LNP formulation. This strategy can be extended to other tissues/cells to find a protein corona-mediated selective system to the spleen, the lungs, the brain, tumors, kidneys, etc. In the human body, there are well organized network systems via blood circulation for the cell-to-cell transport of macromolecules such as hormones (insulin, growth factors), lipid particles (HDL, LDL), and exosomes. We believe that by the end of the 21st century, researchers will have harnessed the power of the endogenous mechanisms of the human body for trafficking to achieve the highest possible level of drug targeting.
This journal is © The Royal Society of Chemistry 2025 |