In situ cellular hitchhiking of nanoparticles for drug delivery.

The design of cellular hitchhiking systems contributes greatly to their success and clinical translation. Some key factors that should be carefully considered for designing cellular hitchhiking systems include the efficient loading of therapeutics, avoidance of premature release, migration of cells to the target site, efficient hand-off between carrier and target cells, and stability and clearance of the therapeutic. The choice of cells in a cellular hitchhiking system has a major influence on the efficacy and application of the system and is primarily governed by the nature of the disease, the intrinsic homing ability of the cells toward the diseased tissue, and how efficiently these cells deliver cargoes. For example, immune cells are cells of interest when considering carriers to target tumors. In addition to their primary phagocytic ability which favors cargo transportation, there is a characteristic infiltration of leukocytes in tumors, making up to 50% of the tumor mass[ 77 ], particularly macrophages- tumor-associated macrophages (TAMs) differentiated from monocytes which are recruited from the local circulation. These immune cells play an important role in both hindering and sometimes promoting tumor progression [ 77 ]. Interestingly, despite the absence of well-developed blood vessels in hypoxic tumor cores, there is a significant presence of immune cells mainly driven by cytokine release by tumors. Thus, cold tumors that are traditionally inaccessible to drugs could benefit from hitchhiking NPs to immune cells (e.g., monocytes, macrophages, neutrophils, and lymphocytes) that can penetrate these low-accessible regions [ 65 , 67 ]. Antigen-specific T-cells home from secondary lymphoid organs where they are primed to tumors or other antigen-rich areas where they exert their cytotoxic effects which is the underlying principle behind adoptive T-cell transfer [ 67 ]. Thus, antigen-specific T-cells can also be exploited for targeted delivery of drugs. In a study by Cole et al., [ 78 ] antigen-specific T-cells were designed by coupling therapeutic viral vectors with T-cells, and the adoptive transfer of these T-cells carrying viral vectors was reported to cure an established metastatic tumor more effectively than traditional adoptive T-cell transfer. On the other hand, leukocytes seem to be an ideal carrier for targeting liquid tumors in the blood due to the colocalization of leukocytes and cancer cells at the endothelium due to their similar size [ 79 ]. E-selectin present on the endothelial wall of blood vessels causes the tethering and rolling of dislodged cancer cells. This enables their transmigration to other tissues[ 79 , 80 ]. In a similar process, these selectins also cause the margination of leukocytes and subsequently result in firm adhesion to the endothelium[ 79 ]. Hence colocalization of therapeutic-bearing leukocytes with migrating tumor cells in the blood vessel can provide a means to control metastasis. The chemotaxis of neutrophils to infection and inflammatory sites could make them attractive carriers when targeting disease states characterized by inflammation[ 73 ]. Neutrophils make up the largest portion of white blood cells and have the ability to cross various epithelial and endothelial barriers encountered by NPs [ 81 , 82 ]. This participation of neutrophils in innate immunity has been exploited to increase NP accumulation in tumors [ 83 – 85 ], improve the delivery of antibacterial drugs to infection sites [ 81 , 86 ], and improve the performance of medical imaging[ 81 ]. In many neurological diseases like Alzheimer’s and Parkinson’s diseases and hepatic encephalopathy, there is an inflammatory component that results in the recruitment of mononuclear phagocytes including dendritic cells, macrophages, and monocytes. These cells can penetrate the blood-brain barrier (BBB) which is otherwise difficult for most therapeutics, thus such states could benefit from therapeutics hitchhiked to these cells [ 61 , 87 ]. Aside from neurological disease, monocytes and macrophages can also be applied to other inflammatory disease states as they have innate inflammation-targeting ability [ 77 , 88 , 89 ]. The lymphatic system regulates immune responses by providing sites for antigen presentation and immune activation. However, their function extends from this to maintaining fluid homeostasis, lipid transportation, and clearing of components from peripheral tissues into the systemic circulation. The diversity in the physiological function of the lymphatic system underscores its involvement in a wide range of diseases and makes it a valuable target [ 90 , 91 ]. Hitchhiking of therapeutics onto albumin is a useful tool in targeting not only tumors as described earlier, but also diseases related to the lymphatic systems including infectious diseases, cancers and metastasis, atherosclerosis, and autoimmune diseases among others. Albumin in its natural course of traffic returns from the tissues into the blood via the lymphatic system and accumulates in draining lymph nodes [ 31 ]. Lymphatic targeted drug delivery systems have been reported to be effective in treating some of these diseases, thus hitchhiking drugs to albumin resident in the subcutaneous tissues via subcutaneous injection holds great therapeutic potential [ 30 ]. This strategy has been employed in lymph node identification and imaging [ 92 – 96 ], suppressing viral replication [ 97 , 98 ], and can be clinically extended to other areas [ 30 ]. One innovative use of this strategy is in the development of vaccines for lymph node-targeted immunomodulation. The clinical use of immunomodulators such as imidazoquinoline is limited by systemic inflammation. However, by conjugating immunomodulators to lipid-polymer amphiphile, immune activation can be achieved in and restricted to the lymph nodes by virtue of in situ albumin hitchhiking [ 99 – 101 ]. The enhanced permeation and retention (EPR) effect, a pathophysiological effect of disorganized angiogenesis resulting in leaky tumor vasculature and increased permeability, has been exploited to improve tumor drug accumulation[ 23 , 102 ]. However, contrary to expectations, the delivery of NPs to these areas remains inefficient, and more recent studies suggest a major dependence of NPs entry into tumors on active transport rather than the EPR effect[ 103 – 106 ]. RBC hitchhiking thus has the potential to improve NP accumulation in solid tumors as the peritumoral zones are characterized by irregular perfusion and slower blood flow to meet nutrient and oxygen demands of the tumor, and suppressing tumor growth in this region can also result in elimination[ 102 ]. However, the inability of RBC to extravasate into tissues as well as the lack of specificity is a limitation to its use. On the bright side, RBC hitchhiking can be a useful tool in targeting liquid tumors and metastatic fragments. The enucleated state of matured RBCs also plays a beneficial role in cellular hitchhiking as it overcomes the complication of phenotype switching observed when immunomodulatory drugs are loaded onto other cells like macrophages which leads to disease progression [ 48 ]. However, RBC-loaded NPs have been reported to dislodge from RBC at the first capillary bed downstream after administration (e.g., lungs if for intravenous administration) and thus could be better suited for pulmonary diseases [ 51 , 54 , 107 ]. Doxorubicin-loaded NPs hitchhiked on RBCs were reported to achieve better anti-metastatic efficacy when compared to free drug or just the NP with about 75% of the NPs shed off at a corresponding lung shear stress [ 108 ]. In disease conditions with risks of clot formation, hitchhiking of tissue-type plasminogen activators (tPA) to RBC improves its safety and efficacy and provides a strategy for prophylactic fibrinolysis[ 43 ]. Soluble plasminogen activators which are used to achieve thrombolysis and restore perfusion are limited to acute therapy of thrombosis due to their short half-life and inability to distinguish newly formed pathological clots from hemostatic clots. However, coupling of tPA to RBC increases their circulation by ten times more than free drug and enhances their ability to rapidly dissolve potentially occlusive clots. Thus, hitchhiking extends the application of fibrinolytics [ 109 , 110 ]. The above discussions go on to show that careful consideration of the cellular components at play in a disease state can be a useful guide in designing clinically applicable cellular hitchhiking systems. Table. 1 provides a summary of the different properties that could guide the selection of cells for drug delivery. While most studies report positive outcomes for the use of these cells, not much focus is placed on investigating the off-target accumulation of such drug-loaded cells. This is an important aspect to investigate to further drive these cell-based systems into the clinic. Ideally, loading of NPs into specific cells, targeting and accumulation of drug-loaded cells to disease sites, and more importantly limiting the release of cargo to the disease sites only, are the properties desired from such designs. Two major techniques have been used to deliver therapeutic cargo using cells including encapsulation of cargo within the cell and attachment of cargo to the cell surface (hitchhiking). Encapsulation presents a direct way of loading cargo into various cells. The natural transportation function of various cells provides an opportunity to enhance drug delivery and this strategy has been reported to protect cargo from the body and vice versa. While this is beneficial to protecting against clearance and non-specific tissue interactions, it could also limit the activity of both the drug and the carrier. For immune cells, the encapsulation of NPs is achieved by exploiting the endocytosis/phagocytic functions of these cells and can be easily tuned by controlling the size, charge, or functionalization of the NPs [ 69 , 111 ]. Many reports on the use of immune cells as drug capsules exist [ 64 , 112 , 113 ], and different methods including the use of NPs and prodrugs are employed to ensure cell viability. However, in such designs, drug release can be guided by lysosomal degradation or enzymatic activation of the drug and in some cases cell death [ 69 , 112 ]. While adoptively transferred cells for site-specific drug release can be sacrificed as such to achieve the primary aim, this restricts the application of this strategy to primarily ex vivo cell engineering. On the other hand, the simple structure and large internal compartments of RBCs make them ideal cell capsules for drugs. Encapsulation of drugs within RBCs affords protection of the drugs from phagocytosis and reduces non-specific interaction with organs/tissues [ 111 ]. This approach has been applied to achieve sustained drug release from RBC and for enzyme replacement therapy. The encapsulation of poorly permeable drugs within RBC such as dexamethasone 21-phosphate enables the long-term release of more permeable dexamethasone upon conversion by cytoplasmic enzymes and this is useful in managing inflammatory disorders that require high glucocorticoid doses [ 45 , 74 , 114 , 115 ]. In enzyme replacement therapy, the administration of RBC-loaded enzymes protects the enzyme from rapid degradation as they remain trapped in the cells. However, the therapeutic efficacy of the enzyme is dependent on the diffusion of the substrate into RBC via transporters and the products’ diffusion out of the cells as well [ 116 – 118 ]. Despite the large internal compartment, drug loading into RBCs is often low due to a high RBC: drug ratio often required during the loading process. Since RBCs lack endocytic functions and the entry of substances is tightly controlled by membrane transporters, the clinical application of drug loading by encapsulation is further limited. Generally, the clinical application of this strategy is hindered as current methods to achieve encapsulation such as osmotic swelling, use of cell-penetrating peptides, and fusion of target cells with drug-loaded liposomes are limited to ex vivo or in vitro preparation [ 48 ]. Other concerns about this strategy include the question of the effect of drug loading on cell function, limited loading efficiency, and variability in drug release [ 111 ]. The attachment of therapeutic cargo to the cell surface is considered to inflict less stress on the cells and takes advantage of the high surface-to-volume ratio. Cargos delivered this way are more accessible to the body. The ability of surface coupling to limit adverse effects associated with non-specific interaction with cognate receptors and the opportunity to modify the surface of the cargo to permit targeted delivery to sites future highlights the attractiveness of hitchhiking [ 48 , 49 ]. This attachment can be achieved using either non-specific adsorption, covalent binding, or targeting ligands. Adsorption of NPs to the cell surface is a simple technique that is widely used in cellular hitchhiking systems obtained by hydrophobic interaction, electrostatic force, hydrogen bonding, or Van der Waals interaction between the cargo and the cell [ 49 , 78 ]. The concept of “cellular backpack” introduced by the Mitragotri lab demonstrates the efficiency of simple adsorption to improve the delivery of therapeutics. By optimizing the geometry of NPs and adding a cell-adhesive layer, NPs were able to hitchhike on macrophages mediating their delivery to the tumor site [ 58 ]. Thus, adsorption presents a reversible method of attachment often requiring minimal cell preparation and particle requirements. However, this simple technique can be limited by the low affinity and rapid/unpredictable dissociation of cargo from the cells [ 4 ]. Adsorption of therapeutic retroviral vectors, achieved by incubating antigen-specific T-cells with the retroviral stock, resulted in > 90% non-specific adhesion of the virus on the surface of these cells. Adoptive transfer of these T-cells bearing virus encoding either IL-12 or Herpes simplex virus thymidine kinase (HSVvtk) resulted in site-specific accumulation of these therapeutic viruses which cured metastatic disease better than the antigen-specific T-cells alone[ 78 ]. A consideration in this simple step is the hand-off of the loaded therapeutic specifically to the target site. Cole et al., [ 78 ] reported a receptor or envelope-dependent interaction which governs the hand-off of therapeutic virus and infection of the tumor cells. Despite being a simple technique, factors such as the incubation time and the presence of cell surface proteins can influence the loading and release of therapeutic cargo. Cole et al., [ 78 ] reported that cell-surface heparan sulfate glycosaminoglycans (HSGs) could hinder the access of the virus to cells, and this was overcome by pre-treating cell-loaded virus with heparanase to achieve an efficient hand-off. Seeing that metastatic cells have a high expression of heparanase, this also enables tumor-specific targeting of the vector. Zhao and coworkers [ 108 ] developed an erythrocyte-leveraged chemotherapy platform (ELeCt) to treat early and late-stage lung metastasis in which doxorubicin-loaded NP was adsorbed to the surface of RBCs. Adsorption of NP was attributed to equal contribution by NP-induced membrane stretching which generates a surface tension force and non-covalent interaction between the NP and RBCs. In this study, ELeCt was reported to achieve an improved drug circulation up to 16.6-fold after 20 min administration compared to the free NPs. The RBCs acting as the primary delivery system led to the accumulation of NPs in the lung based on lung physiology, while the NPs being the secondary delivery system resulted in a controlled drug release at the target site. This design was able to substantially inhibit tumor growth and improve survival in both early and late-stage melanoma lung metastasis [ 108 ]. Adsorption can be facilitated via surface coating of the NPs. One such strategy is the use of a biocompatible ionic liquid coating such as choline-2-hexanoate to aid NP attachment to RBC. This was reported to improve NP biodistribution and delay protein adsorption and aggregation- a major problem encountered with coated NPs. This adsorption is via electrostatic interaction between the negatively charged NPs surface due to the coating and the cationic lipids in the outer layer of RBCs [ 107 , 119 ]. In another study, positively charged chitosan nanoparticles were found to tightly bind to RBC via electrostatic interaction [ 120 ]. These chitosan NPs dislodged via shear stress but were reported to have a dis-adsorption rate far lower than negatively charged particles. While this finding was interesting, the formation of protein corona around these particles in vivo and damage caused to cells like hemolysis and cell agglutination could hinder the clinical application of these positively charged particles for cellular hitchhiking if not carefully tuned [ 4 , 120 ]. Covalent attachment is a more specific attachment achieved by a reaction between groups present on the cargo and the cell surface. Covalent coupling offers a higher affinity between the cargo and the cell. Wayteck et al., [ 67 ] designed a liposome-bearing T-cell system with the potential to achieve triggered liposome release at the target site. Liposomes were covalently bound to T-cells by the formation of disulfide bonds between thiols present on the plasma membrane of T-cells and thiol-reactive liposomes. This system achieved more efficient delivery as 50–60% of liposomes are cleaved from the cells at high glutathione concentrations which are found in tumor sites via a thiol-disulfide exchange reaction [ 67 ]. To improve the efficacy of SN-38, a potent chemotherapeutic with poor pharmacokinetics and toxicity, Huang et al, [ 66 ] designed an SN-38 nanocapsule that covalently binds to autologous lymphocytes to facilitate the delivery to lymphoid tissues. Here, covalent attachment was achieved by the reaction of thiols on the T-cell surface with maleimide groups on the surface of NPs. These hitchhiked NPs rapidly reduced tumor burden and improved survival compared to systemic drug therapy even at doses that had no effect using free drug or limited effect using nanocapsules only [ 66 ]. Ligand-aided attachment presents another simple technique that requires no cell modification as it utilizes ligands naturally present on the surface of desired cells. It is considered a reliable and reproducible technique with the potential for in vivo translation [ 4 ]. To improve the delivery and selectivity of NPs to tumor sites, ligands can be used to decorate the NP surface facilitating specific attachment to the desired cell carrier [ 24 ]. This isn’t a new strategy as NP decorated with target ligands have been used to facilitate tumor accumulation of therapeutics. However, studies have shown that only a small percentage of these artificial delivery systems reach the tumor site and even a smaller percentage penetrate cancer cells [ 21 ]. Interestingly, decorating NPs with non-specific ligands such as mAbs and scFv can lead to rapid clearance of the particles from the bloodstream via the RES [ 121 ]. However, decorating NPs with ligands specific for circulatory cells prevents rapid clearance as these particles are perceived as part of the cells. Sofias et al. [ 13 ] investigated the fate of lipid NPs decorated with an α v β 3 -integrin targeting ligand after intravenous administration. In this study, the NPs accumulated within the periphery of the tumor cells within 1 hour and at the core after 4 hours. The use of a specific ligand resulted in superior accumulation compared to NPs with non-specific ligands. The binding of these lipid NPs to the immune cells was reported to be crucial in tumor targeting and distribution. Goa et al [ 122 ] proposed a non-covalent method of cargo hitchhiking quite different from covalent or ligand-aided binding as it aims to simplify the attachment process, improve cell viability, and prevent impairment of cellular functions, and further drive the clinical translation of cellular hitchhiking. This strategy termed “facile, bioorthogonal supramolecular conjugation” utilizes the interaction between β-cyclodextrin (β-CD) inserted into macrophage cell membrane and adamantane (ADA) inserted into the liposome bilayer to achieve an inflammatory guided delivery of therapeutic liposomes. This complex formed by non-covalent interaction has high stability in systemic circulation and could specifically target inflammation in vivo [ 122 ]. It is however important to include studies to further understand the kinetics of the coupling strategy (rate of association and rate of dissociation) when choosing any of these techniques as this influences the circulation time and the biodistribution of the therapeutic [ 121 ]. For example, increasing the affinity of the cargo can increase the residence time, but using a strategy with rapid dissociation allows for early drug distribution to tissues. Hence careful selection and tailoring of the coupling method based on the potential application of the system is important.

The area of cell-based therapy although relatively new is rapidly excelling with a global revenue of about US $1 billion [ 167 ]. The recent approval of CAR-T therapy has opened up the prospects for other cell-based therapies with a number of them enrolled in clinical trials. The knowledge obtained from research on ex vivo cellular hitchhiking has greatly advanced research efforts leading to the new focus on in situ cellular hitchhiking which presents a straightforward approach to improve drug delivery. However, the success of this approach needs to transcend from benchwork into clinical application, and currently, only a few clinical studies on the hitchhiking strategy exist that have been reviewed elsewhere [ 72 , 167 ]. To a large extent, the concerns that govern cell-based therapies apply to cell-based drug delivery systems. Majorly, there is a need for a robust quality control system to reduce batch-to-batch variations, and maintain cell function and quality. Other concerns include scalability issues and production costs. In situ cellular hitchhiking being a single-step approach holds a lot of promise for clinical translation as it bypasses some of the problems in ex vivo cell-based therapies by shifting the focus from cells back to the cargo, which has simpler manipulation requirements compared to cells. In this review, we try to fill in the knowledge gap on the principles that govern the engineering of NPs for in situ hitchhiking. Being a one-step strategy that avoids the need for cell extraction, a huge part of research is based on the rational design of NPs for specific and efficient cellular hitchhiking. The successful design of NPs for hitchhiking starts by answering critical questions such as: 1) What is the disease focus and what cells play a critical role in this disease state? 2) Can hitchhiking therapeutics to these cells offer any benefit? 3) What will be the in vivo fate of these NPs? Answers to these questions can act as a compass through the design process. A great deal of clinical effort has been made towards NPs with several marketed products and more in active clinical development. However, to further drive its application for in situ cellular hitchhiking, there is a need for preclinical research to include steps to address some of the leading problems in NP therapeutics. This includes 1) the need for stringent safety assays that provide in-depth immunotoxicological information, 2) the development of in vitro , ex vivo , and in vivo models to help understand the relationship between NPs and biological systems in the pathophysiological state and the mechanisms driving NP’s hitchhiking to cells and accumulation to target tissues, and 3) designing hitchhiking NP platforms that can be easily scaled to comply with regulatory requirements. While research efforts are geared towards functionalizing NPs to enable cellular attachment, it is equally important to understand the cellular interactions of these NPs such as the attachment and desorption kinetics and the interaction with other cells. For example, endocytosis can be a method for hitchhiking, however, the type of endocytosis undergone determines the fate of the NP within the cell. As mentioned earlier, experiments assessing the binding efficiency of these NPs in situ should be incorporated into such studies. While it may be difficult to establish ex vivo models that fully represent in vivo experiments, this shouldn’t deter us. We can adopt strategies used in other fields like organ on a chip, the use of microfluidic devices, predictive mathematical models, etc. to have a comprehensive knowledge of the in-situ behavior of NPs. Studies should be directed towards understanding the specificity of NP binding . In situ adsorption of NP to RBC is mainly based on the high density of RBC compared to other blood components. Experiments that prove this preference and selection for RBC as well as evaluating the safety of these NPs if adsorbed to other cells can fast-track the clinical translation of in situ hitchhiking. Efforts using ligands or biomimetic strategies to increase NP specificity show that there is so much opportunity for in situ hitchhiking. However, we know from previous reports the tendency of such surface modification to result in protein interactions, insolubility, and clearance of the NPs by immune cells. Thus, NPs should be designed to have a balance between ligand density which allows for specific cell surface interactions without compromising their stability. In addition to the functionalization of NPs, fine-tuning the properties of the NP itself plays a big role in the success of the system. Studies have shown that the optimization of NP size, shape, and surface charge has the potential to increase its targeting potential. Another area of concern is the clearance of the in situ hitchhiked NPs from the body. One advantage of using cell-based delivery is improving the circulatory time of NPs, however, this can also be a double-edged sword if not fully understood. While NPs do a good job of encapsulating drugs, there is also the well-reported problem of drug leakage over time. It is therefore necessary for studies to consider the potential for unwanted drug effects and toxicity resulting from the prolonged circulation of these drug-loaded particles enabled by cellular hitchhiking. Overall, cellular hitchhiking is an interesting strategy that can greatly transform the application of NPs, and adopting methods to attach NPs in situ can greatly enhance the clinical translation of this strategy. However, research efforts should be geared toward filling the gaps and establishing the safety of this strategy to fast-track the translation of in situ hitchhiking from bench to bedside.

Despite the enormous potential cellular hitchhiking holds in improving the clinical application of NPs, a major consideration that holds back the application of this strategy is the method of loading- ex vivo vs. in situ . Most of the early work in the field focuses on ex vivo engineering/designing of the cells of interest to carry the NP. This involves the extraction of allogeneic (donor) or autologous (host-derived) cells, altering the cells to carry the payload, and reintroducing these cells into patients. This strategy is greatly limited by the high cost, complicated preparation, and unexpected side effects including immune activation and graft vs. host disease [ 21 , 31 , 123 ]. Thus, there is a need for effective techniques to achieve payload identification and attachment to specific cells in vivo - in situ hitchhiking. Recent research efforts have thus gone into designing and optimizing strategies to allow for the precise attachment of cargoes to specific cells and targeted delivery upon administration, some of these are discussed in the next section and others are summarised in Table 2 [ 13 , 65 , 74 , 87 , 101 , 107 , 119 , 124 – 144 ]. Aside from further simplifying the process and improving the safety of cellular therapeutics, in situ hitchhiking holds the potential to improve the clinical application of other cells in cell-mediated drug delivery. For example, neutrophils have a short half-life of about 7 hours and thus ex vivo cell manipulation and hitchhiking of NPs is limited. In addition to this, only low quantities of neutrophils can be harvested compared to other frequently used cell types. However, with the prospect of achieving specific attachment of particles to distinct cell populations in vivo , this can open more applications for neutrophil-based cellular hitchhiking. In fact, neutrophil is one of the major target cell types used in the in situ hitchhiking approach owing to its relatively high abundance and outstanding capability in endocytosing NPs under inflammatory conditions. As the understanding of NP accumulation and distribution expands, various strategies to achieve in situ hitchhiking of NP are now being explored ( Fig. 2 ). As with ex vivo hitchhiking, the methods for attachment of NPs to cells can also be used to achieve in situ hitchhiking. Adsorption remains a simple way to achieve hitchhiking in vivo . Although surface attachment of therapeutics to cells can affect the stability of the cells, however, this is considered to be to a lesser extent than encapsulation and overall maintains the vitality and function of the cells [ 120 ]. While Wang et al., [ 120 ] designed positively charged chitosan nanoparticles that improved vitamin K circulation by adsorbing to RBCs ex vivo , these NPs could not adsorb to RBCs in vivo due to the formation of a protein corona. Ionic liquid (IL) coating presents a way to achieve spontaneous hitchhiking of NPs to RBCs after intravenous administration. This coating protects NPs from protein adsorption by virtue of its anionic properties and IL-NPs have increased retention in the body even outperforming PEGylated NPs [ 107 ]. The ionic liquid coating imparts stealth properties like PEG allowing the NPs to avoid protein adsorption and to hitchhike to RBCs in vivo [ 119 ]. In a recent study, this group extended the application of these ionic liquid-coated NPs to achieve a higher accumulation of abacavir in the brain overcoming the efflux of anti-retroviral therapy (ART) by the BBB. Injection of these particles via the carotid artery, to achieve in situ RBC hitchhiking and maximum deposition at the first capillary bed downstream of injection, resulted in the accumulation of abacavir in the brain up to 48% with only 3% accumulating in the liver ( Fig. 3 ). NPs accumulation in the brain was observed to follow the intracarotid path of microvascular distribution which caused the colocalization of NPs with endothelial cells after which they could traffic to microglia for selective uptake [ 119 ].. Hydrophobic interactions also play a role in NP binding to RBC as hydrophobic and negatively charged NPs have a stronger interaction with RBC than hydrophilic NPs. The prolonged circulation of NPs by hitchhiking to RBC is thought to be directly dependent on the strength of the interaction [ 145 ]. Controlling the physiochemical properties of NP such as size has also been reported to affect the binding affinity to RBC with increasing diameter resulting in greater interaction [ 53 , 145 ]. However, the presence of hydrodynamic forces and other stressors as well as interaction of particles with other non-target cell components could result in low loading and even aggregation of these NPs in situ . For example, the desorption of NPs from RBC hitchhiked ex vivo has been reported to occur in a shear-dependent manner and RBC-RBC collision can result in particle detachment [ 55 , 108 ]. Thus, the interplay of transport, hydrodynamic forces, and the presence of other nonspecific interactions could result in inefficient adsorption/attachment in vivo . The absence of ex vivo models that closely simulate in vivo physiological conditions also limits our understanding of the complex flow and shear factors. To address this problem, various numerical simulations and computational models have been designed to help understand the in vivo behavior/interactions of NPs with cells [ 146 , 147 ]. These go on to show that controlling the adhesion strength and physicochemical properties of the NPs can control the attachment/detachment process. Hydrophobic interaction presents another non-covalent attachment strategy that is widely used in achieving albumin hitchhiking [ 32 , 33 ]. This is a simple strategy that draws inspiration from albumin’s primary function of transporting hydrophobic solutes and has been utilized to achieve in situ attachment of therapeutics to albumin [ 40 , 92 , 94 , 97 ]. Synthetic polymers have been reported to bind to albumin, particularly polymers with broad-spectrum anti-viral activity by either hydrophobic or anionic interactions. Based on this, macro-molecule prodrugs and albumin-polymer conjugates have been designed to prolong blood circulation of small molecules[ 97 , 98 , 148 , 149 ]. Although the non-covalent attachment of polymers to albumin poses some challenges due to resultant protein aggregation leading to rapid clearance, Frich et al., [ 97 ] overcame this by modifying the terminal domains of the polymer to bring about lipid-mediated albumin binding. A notable application of this was the design of a molecular peptide vaccine (CpG-DNA/peptide amph-vaccine) by Liu et al. [ 39 ] which hitchhiked onto albumin after subcutaneous administration to achieve significant accumulation in the lymph nodes. This was possible due to the presence of a lipophilic albumin-binding tail incorporated into the antigen/adjuvant which resulted in a 30-fold increase in T-cell priming and increased antitumor efficacy. Such amphiphilic modification of therapeutics can enhance the therapeutic efficacy of peptide vaccines ( Fig. 4 ). Xu et al., [ 150 ] further designed a fatty acid- fluorene methoxycarbonyl (FMOC)- modified PEG drug conjugate which imparts a 30-fold improvement in the blood circulation of Deferoxamine (DFO) previously requiring 8 to 12-hour administration 5 days a week. In addition to increasing the circulation time, converting DFO into a prodrug activated by hydrolysis of the conjugate results in reduced toxicity of DFO as the nonspecific interaction with other targets is also limited. Overall compared with the free drug, this drug conjugate by hitchhiking on serum albumin exhibited a better iron chelating capacity and has potential in the treatment of iron-overload-related diseases. A major concern for non-covalent hitchhiking to albumin would be the non-specific delivery of the cargo to other organs. The kidneys are known to have a high expression of albumin-binding receptors and are the location for the reabsorption of albumin. However, not a lot of studies have been done to address this concern, and tumor targeting is largely based on passive tumor accumulation. Covalent hitchhiking offers a way to overcome non-specific detachment of payload observed with adsorption [ 108 , 149 ] and this can also be applied to achieve in situ hitchhiking. This often involves decorating the payload with a binding domain specific to the target cell. Covalent hitchhiking unto albumin in situ appears to be a common strategy as it uses the cysteine-34 amino acid residue which happens to be the most abundant and reactive free thiol in the blood. This also guides the specificity of the reaction to albumin as other serum proteins lack this residue[ 36 ]. For example, covalent binding of monomethylauristatin E (MMAE) to albumin in situ was achieved via the reaction of the maleimide group on a glucuronide MMAE prodrug conjugate with the Cys 34 thiol of albumin via a Michael addition. This design overcomes the problem of non-specific desorption as the detachment is driven by the enzymatic catalysis of glucuronide by B-glucuronidase overexpressed in tumors ( Fig. 5 ). Hitchhiking of MMAE onto albumin achieved increased tumor accumulation and was well tolerated compared to the cytotoxic-free drug. Using this strategy in triple negative breast cancer (TNBC) and pancreatic cancer mouse models, 50% of mice exhibited complete remission at day 50 compared to moderate responses recorded for MMAE, and 33% of mice were tumor-free at day 70 post-implantation respectively [ 151 ]. A similar site-specific enzyme-based approach was utilized to achieve a 2.6 times higher accumulation of gemcitabine (GEM). This design was based on the overexpression of a lysosomal cysteine proteinase- Cathepsin B in tumors and involved conversion of GEM to a maleimide prodrug via decoration with a cathepsin B sensitive linker [ 152 ]. Although the work cited applies to macromolecule prodrugs, such a strategy can be applied to NPs for hitchhiking to improve their biodistribution in vivo . Xu et al [ 95 ] also employed covalent binding of a fluorescent dye to albumin to aid imaging of micrometastases. This strategy takes advantage of the overexpression of secreted protein and rich in cysteine (SPARC)- a high-affinity albumin-binding protein in micrometastases. In this design, upon intravenous administration, the meso-chlorine group of IR1080, an organic near-infrared window II (NIR-II) probe, covalently binds to the thiol-group of albumin and this association turns on the fluorescence. This probe proved to be precise achieving a detection rate of 100% for micrometastasis larger than 2mm and 30% for micrometastases with size <500 um which are difficult to detect. This design opens an avenue for in vivo NIR-II imaging which could guide diagnosis and intraoperative detection of micrometastases. In-depth information on chemistry-based approaches for the attachment of cargo to different cells has been discussed in recently published reviews [ 153 – 156 ]. In a study to better understand the in vivo fate of NPs, cRGD decorated liposomes, previously thought to undergo majorly ligand-mediated accumulation in αvβ3-integrin expressing cancer cells and tumor vascular endothelium, were observed to predominantly accumulate in liver and spleen with short blood circulation half-life. Although cRGD-NPs showed more specific tumor accumulation than control non-specific NPs, a real-time study of cellular interaction revealed the uptake of these ligand-decorated NPs by immune cells minutes after injection. These NPs were observed to hitchhike with myeloid cells particularly neutrophils due to the expression of αv and β3-integrin subunits on these cells and the consequent “ligand-mediated” accumulation in tumor largely contributed to by immune cell hitchhiking. This interesting finding when explored could be an avenue to improve in situ hitchhiking in different disease states as it occurred independent of cancer [ 13 ]. Since NPs and other therapeutics when administered interact non-specifically with different cells and proteins, ligand-aided attachment of NPs to carrier cells seems to be important in directing NPs to achieve in situ hitchhiking. Qi et al. [ 24 ] in a strategy to improve tumor accumulation of doxorubicin and A2AR antagonists succeeded in directing NPs to leukocytes in vivo by decorating these micelles with E-selectin. In this case, the similar migration patterns of leukocytes to tumor cells guided the design of the system. Leukocyte recruitment to endothelial cell walls is mediated by the presence of selectins expressed on the surface of the endothelium resulting in a rolling adhesion [ 79 ]. Thus, decorating NPs with a similar ligand facilitates attachment to leukocytes. E-selectin micelles adhered to the surface of leukocytes and these leukocytes guided higher tumor accumulation and delivery of the loaded therapeutics with lower liver accumulation when compared to non-specific micelles ( Fig. 6 ). This design could have an increased application in targeting and preventing metastases as leukocytes interact with cancer cells in the bloodstream: via the colocalization of leukocytes and cancer cells at the endothelium due to their similar size [ 79 ]. To improve the clinical application of RBC hitchhiking of tissue-type plasminogen activator, Zaitsev et al [ 110 ] designed an approach to achieve in situ hitchhiking as well as higher drug loading using a ligand-aided attachment. In this study, a monovalent single-chain fragment antibody (scFv) with specificity for mouse RBC glycophorin A (GPA) was utilized to achieve rapid and stable coupling of high doses of tPA to RBC. This in situ coupling with recombinant RBC targeting tPA achieved higher fibrinolytic activity and intravascular circulation with 4-fold higher concentration than the non-targeted plasminogen activator with no appreciable accumulation in major organs as observed with free plasminogen activators. This strategy has the potential to further extend the application of fibrinolytic for post-surgery prophylaxis of surgery- arresting clot propagation and promoting clot lysis [ 110 ]. Utilizing a similar technique, Glassman et al., [ 121 ] in a process termed “affinity-mediated coupling” achieved the hitchhiking of liposomes to RBC membrane proteins. These liposomes were decorated with Ter 119 mAB or scFV which recognize the GPA on RBC plasma membrane and experienced a shift in circulation from traditionally observed plasma for PEGylated liposomes to blood cellular elements resulting in a 2-fold increase in systemic exposure ( Fig. 7 ). Comparing this system to PEGylated liposomes which are the gold standard for long-circulating NPs, these specific binding liposomes sufficiently prolonged the elimination half-life by 1 to 2 days. Aside from the ligand-aided attachment strategy, the natural function of cells can be taken advantage of to achieve in situ cellular hitchhiking. Li et al [ 157 ] developed pathogen-mimicking nano-pathogenoids (NPNs) with the ability to hitchhike to neutrophils in vivo . This design takes advantage of the intrinsic ability of neutrophils to detect and take up pathogens via the recognition of pathogen-associated molecular patterns (PAMPs). In this study, PEG-b-PLGA nanoparticles were coated with outer membrane vesicles (OMVs) secreted by Escherichia coli , and this increased the accumulation of these particles in tumors as seen in Fig. 8 . Further investigation of this mechanism of accumulation revealed that 41 % of neutrophils in blood contained NPNs after 4 hours of administration as opposed to only 10 % observed in regular NPs. This highlights NP hitchhiking on neutrophils mediated by TLR2 and TLR4 as the major mechanism of tumor accumulation with 83.4 % of neutrophils taking up NPs compared to other immune cells. This system, used after photothermal therapy (PTT) which creates an acute inflammatory environment to facilitate the recruitment of these neutrophils to tumor sites ( Fig. 8 ), achieved the elimination of residual tumors. Inflammation is an innate and adaptive immune response marked by the recruitment of polymorphonuclear neutrophils. These neutrophils can adhere to and migrate across the endothelial vessels and thus could be instrumental in delivering NPs across endothelial barriers [ 82 , 86 ]. Thus, this design can easily be extended to other disease conditions marked with acute inflammation such as acute lung injury to achieve more targeted drug delivery and higher drug accumulation. Chu et al [ 86 ] successfully increased NP targeting of lungs by hitchhiking to neutrophils adherent to the vessel wall. Here, as opposed to the use of ligands or other forms of attachment discussed earlier, hitchhiking was achieved via the interaction of denatured albumin with Fcɣ receptors present on neutrophils. Bovine serum albumin was used to synthesize drug-loaded albumin NPs with a size of about 130nm compared to PEG-polystyrene NPs of 116nm. Intravital microscopy revealed that neutrophils internalized the albumin NPs as early as 30 minutes after administration, and migration of these neutrophils into the muscles was observed 1 hour later. This system achieved an inflammation-guided targeting of NPs as control mice without acute lung injury did not show infiltration of neutrophils. Furthermore, activation of neutrophils (by the process of inflammation) was also necessary to achieve NP accumulation in the lung as resting neutrophils did not internalize these albumin NPs further confirming an inflammatory targeted drug delivery. With this strategy, lung inflammation in acute lung injury was resolved in a mice model when treated with TPCA-1 (2-[(aminocarbonyl)-amino]-5-(4-fluorophenyl)-3-thiophene carboxamide) loaded albumin NPs with a lowered lung permeability compared to using the free drug. Also, the application of such a strategy was expanded to achieve a 3-fold decline in bacterial proliferation in Pseudomonas aeruginosa -infected lungs when treated with cefoperazone acid-loaded albumin NPs compared to the free drug [ 86 ]. The same group reported the application of this strategy for improved cancer immunotherapy. TA99 monoclonal antibody specific for melanoma antigen was used to initiate neutrophil recruitment to the tumor site. TA99-mediated neutrophil recruitment also increased NP uptake by neutrophils (by 6.2% compared to 0.7% without TA99), further confirming the previous finding that activation of neutrophils plays a key role in this strategy of NP hitchhiking. This strategy was used to localize pyropheophorbide-a (ppa) in tumors for photodynamic therapy and the treatment achieved significantly smaller tumor sizes with an increased survival rate[ 158 ]. In another study, an inflammatory guided delivery of NP to the tumor was achieved by the same group, this time with the use of a ligand anti-CD11b to enable in situ attachment to activated neutrophils expressing CD11b [ 83 ]. In a follow-up study, the authors investigated 1) how efficient the carrier neutrophils deliver NPs to the targeted tumor tissue and 2) what type of cargos are needed for achieving therapeutic efficacy, which are two important fundamental questions for the in situ neutrophil hitchhiking approach [ 85 ]. It was found that the neutrophil-mediated approach could deliver 5% ID/g of NPs to the tumor and this efficiency is independent of particle size (30−200nm) and their dose (10 8 −10 11 NPs). The authors rationalized that in situ neutrophil hitchhiking of NPs to tumors requires a high drug loading capacity of the NPs because neutrophils only have a limited cell volume to carry drug NPs. To achieve this, the authors designed a carrier-free paclitaxel nanocrystal with nearly 100% drug loading and demonstrated that anti-CD11b antibody-decorated nanocrystals could hitchhike on neutrophils and lead to a high drug accumulation in tumors. A similar design by Zhu et al. [ 159 ] involved the use of these albumin NP loaded with glucose oxidase to achieve an inflammatory-guided drug accumulation in ectopic lesions in endometriosis- and estrogen-dependent chronic inflammatory disease. Endometriosis is characterized by increased systemic neutrophil infiltration and an abundance of neutrophils in the peritoneal fluid resulting in a long-term persistent inflammatory microenvironment. This guided the use of this strategy, and peritoneal injection of these albumin NPs which resulted in an enrichment of the NPs in ectopic lesions 16 hours after injection compared to other organs, providing a “non-hormonal easy-to-achieve, and repeatable approach for the treatment of endometriosis” [ 159 ]. A key consideration for in situ cellular hitchhiking is the preservation of cell function, protection of cargo from degradation, and timely/specific release of cargo at the target site. The success of any cell-based delivery system lies greatly in the preservation of cell function after drug loading[ 7 , 44 ]. Generally, drug delivery systems (DDS) such as liposomes, polymeric NPs, and others by encapsulating the drug can protect the cells from the drug’s cytotoxic effect [ 160 ]. However, the biocompatibility of the cargo (drug encapsulated in DDS) with the cell of interest is important and carefully controlling the properties of the cargo can achieve this. Although surface loading is considered safer than encapsulation, the attachment of particles can interfere with normal cell functions and inhibit the functions of cell membrane proteins[ 44 ]. Lenders et al., [ 49 ] in a study to further understand the factors at play at NP-RBC interface during non-covalent adsorption report the potential of PLGA NPs hitchhiked on RBC to destroy these cells. By measuring the release of hemoglobin from the cells after adsorption, it was found that increasing the amount of NP (starting from the RBC: NP ratio of 1:500) caused greater levels of RBC hemolysis. A similar phenomenon was observed in the design of chitosan nanoparticles to improve the delivery of vitamin K via RBC hitchhiking. With higher concentrations of NPs, phosphatidylserine (PS) on the surface of the RBCs increased with corresponding hemolysis of these cells [ 120 ]. Not only did NPs increase the PS expression, but also the osmotic fragility, which describes the resistance of RBCs in hypotonic solutions, was increased compared to normal RBCs [ 120 ]. Thus, determining the safe loading dose of the cargo is important, and particular attention should be paid when designing in situ platforms as such reactions could result in potential life-threatening adverse effects [ 121 ]. Preserving the trafficking capability of migratory carrier cells (e.g., immune cells) upon NP hitchhiking is essential to the success of NP delivery via cellular hitchhiking. However, systemic investigation of the impact of NP hitchhiking on carrier cell trafficking in vivo remains sparse in the literature. Theoretically, trafficking of the carrier cells can be affected by the material and physicochemical properties, location (intracellular vs on cell membrane), and number of hitchhiked NPs. Additionally, the optimal and/or threshold of these factors that can preserve the cell trafficking capability also depends on the carrier cell type. Understanding the effect of NP hitchhiking on carrier cell migration is often an underappreciated question but is critical to further advancing the in situ cellular hitchhiking approach. Controlling the properties of cargo was found to significantly impact the biocompatibility as NPs with higher hydrophobicity had minimum impact on coupling efficiency but a significant influence on RBC lysis. Thus, increasing the electrostatic influences over the hydrophobic interactions for the NP by using a positively charged PEI PLGA, sufficiently lowered RBC hemolysis [ 49 ]. While Zelepukin et al., [ 145 ] report that cationic NPs cause more severe agglutination than anionic NPs this difference was only seen at high concentrations of NPs. Adsorption of NP was also found to cause aggregation of the RBC but interestingly this was found to differ among different species with rabbit RBC having remarkable aggregation compared to more negatively charged mouse and human RBC. These go on to highlight the need for formulation designs to undergo proper biocompatibility evaluations and the use of animal models that closely simulate clinical reality [ 49 , 121 ]. The success of the cellular hitchhiking strategy is also influenced by the stability of the cargo before, during, and after being associated with the target cells. This is especially important with strategies that involve the encapsulation of the cargo. Internalized NPs could remain trapped within the endosome and not efficiently released, undergo endocytic recycling resulting in premature release, or could be subject to lysosomal degradation [ 21 , 79 ]. Designing the cargo to have a positive surface charge by using cationic excipients is one way to evade lysosomal degradation. Positively charged particles are advantages for encapsulation as the positive charge has been reported to facilitate efficient phagocytosis by both macrophages and dendritic cells. These positively charged cargos have buffer capacity which delays acidification of phagosomes and their subsequent fusion with lysosomes [ 12 , 61 ]. Furthermore, Batrakova et al [ 62 ] showed that protein hitchhiked onto macrophages via encapsulation could be protected from lysosomal degradation by using a copolymer synthesized by conjugating cationic polyethyleneimine (PEI) to poly(ethylene glycol) (PEG). Other ways to overcome this include the use of membrane-fusing lipids which deliver the cargo directly into the cytoplasm by forming a hydrophobic channel on the cell membrane thus avoiding the entrapment in endosomes [ 161 ]. Additionally, optimizing the geometry of the NPs could protect them from phagocytosis by the target cells. Champion and Mitragotri explain how the local shape of a particle at the point of contact can influence the polymerization of actin on the surface of macrophages. Disc-shaped NPs with contact angles greater than 45 ° resulted in frustrated phagocytosis and spreading of cells [ 162 ]. Future work can capitalize on this to shift the loading of cargo from within to the surface of immune cells overcoming the problem of drug degradation often encountered. The success of a cell-based delivery system isn’t the attachment of the cargo to the cell but extends to the specific and timely release of the cargo to elicit its pharmacological action. Release of entrapped cargo within cell-based carriers could occur by inactive transport, via efflux pump, or by exocytosis and these can be tuned to respond to biological signals [ 160 , 163 ]. Batrakova et al (20) showed that nanozymes hitchhiked unto BMM to aid delivery to the brain were released via exocytosis and activation of the BMM by phorbol myristate acetate (PMA) could enhance drug release by 50%. Suggesting that drug release was dependent on cell-activation. Albumin hitchhiking results in intracellular drug delivery guided by caveolae-dependent endocytosis, however, the fate of cargo after endocytosis remains vague. This reflects a possible endosomal/lysosomal degradation of the cargo after uptake, a problem currently encountered by many NPs. Although the mechanism of drug release for albumin hitchhiked cargoes is not clearly understood, Kwak et al [ 131 ] were able to report a possible endosomal escape resulting in cytosolic drug release mediated by multivesicular bodies or formation of early or late endosomes. A study by Szymanowski et al [ 164 ] also confirms this observation as the intracellular transport of polymer lipid complexes was reported to occur via a non-degradative caveolae endocytic pathway. Overall studies to elucidate the intracellular traffick of albumin-bound cargo remain at odds as other groups have also reported the lysosomal accumulation of albumin-binding cargoes in cancer cells [ 165 ] and also accumulation of such albumin-bound doxorubicin in the cytoplasm particularly in the Golgi apparatus and endoplasmic reticulum as opposed to the nucleus as free drug [ 36 , 166 ]. This can be overcome with the use of stimuli-responsive linkers that are cleaved to cause extracellular drug release. When considering surfaced loaded cargo, as earlier mentioned the release of the cargo is greatly determined by the method of attachment. Physical adhesion of cargo to the cell (i.e. adsorption) is reversible and this favors the timely release of the cargo upon reaching the target site. However, this is also a limitation due to the untimely release of the cargo facilitated by capillary shear forces and other cell interactions leading to adverse effects. Successful use of this strategy often requires the administration of ex vivo loaded cells via a route closest to the target site. The release of cargoes covalently attached to cells is mostly governed by cleavage of the linker example enzymatic catalysis of the glucuronide on the maleimide group to release MMAE conjugated to the Cys 34 thiol of albumin as described by Zhang et al (57) ( Figure 4 ). Thus, covalent binding can facilitate site-specific drug release by designing the linker to respond to specific pathological environments. While ligand and receptor-mediated loading facilitates the delivery of the cargo to specific sites, mechanisms surrounding the release are not clearly described but could be thought to be governed by the competitive binding of other components to the ligand or receptor and can be tuned to favor site-specific release. More attention needs to be paid to clearly understand the mechanism behind the accumulation and release of cell-based cargoes.

Cell-based drug delivery is a new generation of biomimetic delivery system that utilizes the physiological properties of cells to enable the in vivo delivery of therapeutics[ 7 , 23 ]. This mimics one of the natural ways for cellular transportation observed within biological systems. Some examples of intracellular hitchhiking include the association of mRNAs with endosomes to aid the transport of mRNAs to the mitochondria, the association of RNA granules on lysosomes via a linker protein and hitchhiking of precursor miRNAs on endolysosomal cargoes in the neurons for transportation to through axons [ 22 ]. However, circulatory cells provide a means to improve the biodistribution and accumulation of NPs in a highly specific manner [ 24 ]. Blood cells and components such as serum albumin, red blood cells (RBCs), and immune cells have attracted attention in recent years for drug delivery purposes. These cells/components are characterized by their natural stealth properties, long circulation time, high mobility and flexibility, and inherent targeting ability [ 4 , 5 ]. Thus, mimicking the functions of these cells and components can provide a way to improve the PK/PD of NPs via cellular hitchhiking- where synthetic carriers are associated with circulatory cells or serum albumin. Albumin is the most abundant serum protein accounting for ~ 60 % of the total proteins in the blood[ 25 , 26 ]. Serum albumin is a small globular water-soluble protein with an average half-life of 19 days[ 25 , 27 ]. It functions primarily in the transportation of hydrophobic molecules like fatty acids, hormones, vitamins, and divalent cations[ 28 ] by reversible binding in a noncovalent manner[ 29 ]. Other functions include maintaining the osmotic pressure of blood, free radical scavenging in inflammation, and wound healing[ 26 ]. Despite being a major serum protein, 60 % of albumin is found in the interstitial space and this ability to cross biological barriers while also avoiding renal clearance makes it attractive for drug delivery [ 25 , 30 , 31 ]. Despite not being a cell, albumin is a versatile carrier for application in targeted drug delivery, due to albumin’s abundance in the blood, the presence of multiple hydrophobic binding pockets on albumin, and its capability of reversible binding enabling the transport of both endogenous and exogenous compounds including drugs, genes, and peptides [ 32 , 33 ]. Because of these reasons, we have included albumin in the discussions in this review. Modification of active pharmaceutical ingredient (API) to express hydrophobic regions to improve its biodistribution via in situ binding to albumin has been utilized in drug products including insulin detemir (Levemir ® ) and Liraglutide (Victoza ® ) [ 33 , 34 ]. Covalent binding of drugs to albumins to improve their biodistribution and bioavailability has also been achieved via the free cysteine (cys 34) residues on the surface of albumin, nucleophilic attack of NH 2 , OH or SH, or reaction with the lysine group on albumin [ 33 ]. In addition to this, the overexpression of albumin-binding receptors in tumors leading to the accumulation of albumin at tumor sites makes it attractive in chemotherapy as it can improve the circulation half-life and tumor accumulation of albumin-hitchhiking therapeutics [ 32 , 35 ]. By associating NPs with albumin, there is a shift in NPs transport across cell membranes from endocytosis to caveola-mediated transcytosis preventing the digestion of the NPs [ 28 , 36 , 37 ]. This is the principle underlying the design and approval of Abraxane ® for the treatment of metastatic breast cancer [ 29 , 38 ]. Moving on from this, Irvine’s research group at MIT by exploiting albumin’s transportation function successfully designed a peptide vaccine that accumulates in lymph nodes by in situ albumin hitchhiking where it primes T-cells against tumors[ 39 ]. Lou et al. [ 40 ] also designed a paclitaxel NP expressing an albumin-binding maleimide group that selectively conjugates to the cys34 residue of albumin via a thiol-maleimide reaction. This prodrug assembly exhibited improved plasma stability, high accumulation at the tumor site, and efficient targeted release via a tumor-redox-responsive drug release mechanism. It should be noted that in this case, the size of the cargo (NP) is larger than that of the carrier albumin and multiple albumin molecules might be associated with one NP. RBCs with a primary function of oxygen transport represent the most abundant circulatory cells[ 41 , 42 ]; research spanning over the past four decades has been geared to utilize the natural transport mechanisms of RBCs in improving therapeutic outcomes[ 43 ]. RBCs possess some unique properties that make them an attractive delivery system and some of these properties include a long circulating half-life of 3 months and the ability to avoid immune clearance until senescent, high surface area, flexibility alongside reversible deformation allowing them to squeeze through small capillaries, and the presence of surface ligands to aid cellular transport[ 43 , 44 ]. Studies using RBC for drug delivery span from the delivery of small molecule drugs like dexamethasone[ 45 ] to larger molecules like antibodies and peptides[ 46 , 47 ]. Substances bound to RBCs can escape renal and RES clearance but are excreted via a pathway for hemoglobin degradation- bile and RES clearance of senescent RBCs. This shows their potential to improve drug bioavailability while ensuring adequate tissue redistribution [ 48 – 50 ]. This is evident by reports on reduced liver and spleen accumulation of NPs hitchhiked on RBCs [ 51 – 53 ]. Controlled release of therapeutics over a long period can also be achieved by the combination of the long circulation of RBCs and the design of the NPs. Thus, RBCs could act as a depot system to maintain constant drug concentrations [ 4 , 48 ]. Aside from its role in the transport of endogenous substances, pathogens such as bacteria and viruses utilize these properties of RBCs to achieve infection and avoid rapid clearance [ 54 ]. Some of these pathogens reside within the RBC such as Plasmodium falciparum while others bind to the surface of the RBC via covalent, noncovalent, or ligand-aided mechanisms to remain in circulation for weeks. Inspired by this observation, Chambers and Mitragotri [ 55 ] designed a polystyrene NP that utilizes non-covalent binding to RBC to achieve prolonged circulation compared to unbound particles which were cleared within 2 minutes of administration. NPs delivered this way do not alter the circulation of RBCs and this technique has the capability of increasing drug payload in vivo as particles up to 450 nm could be delivered as opposed to the traditional 150 nm particles used for in vivo circulation [ 55 ]. In contrast to this, RBC also has an innate immune function that captures pathogens such as bacterial and other immune complexes, neutralizes them via oxycytosis, and presents them to Kupfer cells in the liver and splenic antigen presenting cells (APCs)[ 56 , 57 ]. Exploiting this phenomenon, Ukidve et al., [ 52 ] designed an erythrocyte-driven immune targeting (EDIT) antigenic NP which adsorbs to RBC to facilitate splenic delivery. The particles were reported to be adequately internalized by dendritic cells without the capture of the RBCs, which in turn improved immune response against the antigens. Immune cells (e.g., monocytes/macrophages, neutrophils, lymphocytes) circulate through the vasculature and can overcome endothelial barriers to penetrate deep into tissues. Mostly these cells reside in tissues where they maintain tissue homeostasis and primarily function to protect the host from foreign materials via effective phagocytosis and clearance[ 58 , 59 ]. They have the potential for use in drug delivery and could facilitate disease targeting as they have inflammation/injury-responsive migratory capability. In addition to these, they can overcome structural barriers such as the blood-brain barrier (BBB) and can reach hypoxic tumor areas with low blood flow [ 4 , 13 ]. Based on these properties, Dou et al [ 60 ] leveraged bone-marrow-derived macrophages as a trojan horse to facilitate the delivery of indinavir NPs. In this study, adoptive transfer of the macrophage-carried NPs demonstrated robust drug distribution, reduced viral infected cells, and improved CD4 + T-cell population compared to free drug. These immune cells also can achieve targeted drug transport to disease sites due to their intrinsic homing ability [ 61 ]. A study by Batrakova et al. [ 62 ] utilized the ability of mononuclear phagocytes to accumulate in the brain via neuroinflammatory-induced chemokine gradient to deliver a catalase-encapsulated NP for the treatment of Parkisons’ disease (PD) using bone-marrow macrophages as a drug carrier. This delivery system achieved an extended enzyme release over about 5 days in vitro and a 2-fold increase in brain catalase concentrations compared to free nanozyme in a PD model. Monocytes differ from macrophages in that they are circulatory immune cells that migrate into tissues before differentiation. Hence this homing nature can be utilized to facilitate the delivery of NPs to pathological tissues [ 63 ]. Monocytes can achieve deep tumor penetration by utilizing homing signals released by tumor and thus have been investigated in different studies to achieve targeted tumor accumulation [ 64 , 65 ]. However, a well-known barrier to NP application is the clearance by immune cells via non-specific recognition [ 13 ]. Hence, an appropriate design of NPs must be achieved to ensure drug preservation in immune cell-mediated delivery. Other immune cells such as neutrophils [ 13 ] and T-cells [ 66 , 67 ] have also been reported to be useful in cell-mediated drug delivery. Other cells that are not considered blood cells such as stem cells (neural and mesenchymal)[ 68 , 69 ], adipocytes [ 70 ], and engineered cells such as CAR-T cells [ 71 ] have also been investigated as carriers for drug delivery. Recent progress in drug delivery technologies based on different cells can be found in recently published reviews elsewhere [ 41 , 72 – 76 ].

Over the past decades, tremendous advances have been made in nanoparticle (NP) research, especially in the application of nanoparticles (NPs) for the diagnosis and treatment of diseases. NPs are attractive in drug delivery because of their unique properties such as nanoscale size range which favors single-cell interactions, ease in surface functionalization which enables specific cell targeting, and large relative surface area which modulates nano-bio interactions [ 1 ]. NPs have transformed drug delivery by offering effective ways to enhance the efficacy of various therapeutic modalities ranging from small molecules to biologics by improving their stability and modulating their pharmacokinetics (PK) while promoting drug transport across biological barriers [ 2 , 3 ]. Despite exciting advances, the clinical application of NPs in drug delivery is limited by their poor biodistribution, inadequate avoidance of biological barriers, and poor targeting efficiency [ 4 , 5 ]. Physiological barriers play an important role in protecting the body from foreign substances and sadly this includes NPs. Thus, the effectiveness of NPs is largely dependent on their ability to overcome these barriers[ 6 ]. Common physiological barriers that affect the biodistribution of NPs include clearance by the reticuloendothelial system (RES) which is a major barrier for intravenously administered NPs[ 1 , 6 , 7 ], clearance by mucociliary and alveolar macrophages and transcytosis into the bloodstream after inhalation of NPs[ 8 , 9 ], interstitial lymphatic flow rate for subcutaneously administered NPs, and drug leakage into surrounding tissues due to high interstitial pressure for intratumoral administered NPs [ 1 , 10 , 11 ]. Additionally, the interaction of NPs with body fluids forms a protein corona which affects their functionality by causing dissolution or aggregation, increases their clearance, and reduces specificity by masking conjugated ligands [ 6 , 12 ]. Aside from biological clearance, NPs must navigate through endothelial tight junctions which prevent the translocation of NPs to target sites. Majority of the currently approved NPs rely on passive targeting to achieve their efficacy. While research on the active targeting of NPs has progressed, clinical translation remains difficult due to a limited understanding of their in vivo targeting mechanisms [ 13 ]. As the biodistribution of NPs is thought to be dependent on their physiochemical properties [ 6 ], efforts have been focused on manipulating these properties to improve NPs’ PK. Examples include adjusting the size, shape[ 14 , 15 ], and surface charge[ 16 , 17 ] of NPs to improve circulation time, and PEGylation of NPs which provides steric stabilization and prevents protein adsorption [ 2 , 3 , 18 ]. With increasing knowledge of biology and the desire for more personalized medicine, advances have been made to further improve the PK and pharmacodynamics (PD) of NPs by incorporating complex architectures and responsive or targeting moieties [ 2 ]. Such approaches include the use of bio-inspired systems such as lipids, polymeric micelles, peptides, and glycans [ 3 ] and material biointerface engineering where NPs are designed to have features similar to biological structures[ 19 ]. By taking advantage of biological systems, this strategy eliminates the detection of NP as foreign objects and alters their interaction with biological surroundings and vice versa. Thus, the rise of biomimetic systems has given new hope for NP drug delivery and significantly improved the PK/PD of NPs by extending their circulation time, reducing clearance, and aiding target specificity [ 1 , 7 , 20 ]. Cellular hitchhiking is a non-canonical form of cellular transportation observed in fungal, plant, and animal cells that involves the co-transport of a secondary cargo by a primary cargo already in motion. The secondary payload is thus recognized as a part of the cell during circulation and evades clearance [ 21 ]. The interactions between the primary and secondary cargo, their co-migration, and transportation that occurs due to the movement of the primary cargo and not vice versa are criteria that govern cellular hitchhiking [ 22 ]. Cellular hitchhiking, by disguising the cargo as ‘self’, enables them to evade clearance and overcome endothelial barriers while extending their circulation and reducing toxicity by limiting off-target interactions ( Fig. 1 ). By virtue of these properties, cellular hitchhiking of NPs- where synthetic NPs are associated with circulatory cells for targeted delivery, can offer benefits in improving the efficacy of NP therapeutics[ 7 ]. In this review, we discuss advances in cellular hitchhiking of NPs for drug delivery. We first explain the concept of cellular hitchhiking after which we describe the properties of circulatory cells that make them relevant carriers for the therapeutic payloads. We also discuss considerations for cellular hitchhiking comparing ex vivo to in situ methods and highlight current applications and areas for the future application of this strategy.