Composition. Historically, the administration of therapeutic agents has been limited by multiple factors, primary among these being low solubility, stability and rapid clearance. The result is a short circulation half-life and low efficacy, making frequent administration necessary. Additionally, there can be significant side effects in non-diseased tissues that adsorb therapeutic agent. These issues have led to the development of various targeting strategies aimed at increasing therapeutic index, including monoclonal antibodies and immunoconjugates. Some of the strategies are currently used in clinical practice and others are in clinical development.(
46) In addition to these strategies, it has been shown that polymer−drug conjugates can substantially improve the blood residence time and weaken side effects. It is anticipated that engineered multifunctional nanoparticles can address issues with targeting as well as carry a more substantial drug payload. Many polymers have been investigated, including HPMA [
N-(2-hydroxypropyl)methacrylamide], dextran, and polyglutamate. HPMA and polyglutamate drug conjugates represent ~35% of all the polymeric drug delivery systems in clinical development. The clinical success of AP5346, PK1, PK2, Xyotax, and CT-2106 seems to be due to their long circulation half-life, passive targeting ability, and, most importantly, lower toxicity allowing higher dosages (see Table ). Thus far, the main impact of polymer conjugates has been to improve the pharmacokinetic parameters of drugs already in clinical use. These polymeric nanocarriers must be non-toxic, non-immunogenic, and carry sufficient amount of drug and release the drug at the optimal dose. The use of poly(ethylene glycol) (PEG) for nanoparticle surface functionalization has had led to very favorable results due to its intrinsic physicochemical properties but has had limited impact as a drug conjugate carrier due to low drug loadings. PEG polymers have low toxicity and no immunogenicity and are approved by the Food and Drug Administration (FDA) for clinical use. PEG−drug conjugates or nanoparticles functionalized with PEG chains have been described as long-circulating drug delivery systems with potential applications for systemic drug administration.
21,47 Poly(ethylene glycol) or poly(ethylene oxide) refers to an oligomer or polymer of ethylene oxide in linear or branched structures.(
48) PEG is a hydrophilic polymer that can be adsorbed or covalently attached to the surface of nanoparticles. Hrkach et al. have demonstrated the formation of PLA−PEG nanoparticles in a core−corona structure with a solid core and anchored PEG chains on the surface. Their results showed for the first time that PEG chains, covalently attached to a particle surface, could exhibit flexibility similar to that of free PEG polymer dissolved in water. Furthermore, PEG has been shown to substantially reduce nonspecific interactions with proteins through its hydrophilicity and steric repulsion effects, reducing opsonization and complement activation.
49–55 The chain length, shape, and density of PEG on the particle surface have been shown to be the main parameters affecting nanoparticle surface hydrophilicity and phagocytosis. The mechanism involved in phagocytosis of opsonized nanoparticles is receptor-mediated by interaction of specific proteins absorbed on the surface of the nanoparticles with phagocytes. Gref et al. were the first to report the advantages of pegylation on PLGA−PEG nanoparticles, resulting in a substantial increase in blood residence time. Apolipoproteins (A-IV and E) were found to be present on the surface of pure PLA nanoparticles in contrast to the PLA−PEG nanoparticles using gel electrophoresis analysis.(
53) Although the assay did not provide a precise quantitative analysis, it was clearly shown that protein adsorption, particularly apolipoprotein J and complement protein C3, was substantially reduced on the surface of the PLA−PEG nanoparticles compared with pure PLA formulations. These results suggest that these specific apolipoproteins might also play a role in the process of opsonization of the particles. Gref et al.(
53) have systematically studied the effect of PEG chain length in preventing protein adsorption on the surface of the nanoparticles. The results showed that an optimal molecular mass (
Mw) range exists (between 2 and 5 kDa) in order to reduce plasma protein adsorption. The amount of protein absorbed on PLA−PEG 5 kDa was substantially reduced (~80%) compared to the amount of nonpegylated PLA nanoparticles. PEG content as low as 0.5 wt % on the surface of the nanoparticles was able to significantly reduce the total amount of protein when compared to the nonpegylated PLA nanoparticles. The most significant reduction of protein absorption was found for pegylated particles (5 wt %). The effect of PEG on the surface of the nanoparticle in preventing protein absorption correlated with the polymorphonuclear leukocyte (PMN) and human monocyte (THP-1) uptake. Interestingly, a threshold of 1−2 nm space between the PEG chains was estimated for minimal protein absorption. Since then, many research groups have investigated the interaction of polymeric nanoparticles with serum opsonins, characterizing both uptake mechanisms and kinetics.
53,56–58 In one study, the relationship between the protein adsorption kinetics of the nanoparticles and their uptake in the liver was investigated.(
56) The results showed that the rate of hepatic uptake of polystyrene nanoparticles (50 nm) was significantly higher when nanoparticles were incubated with serum prior to transfusion in the rat liver. It was suggested that the increased rate of hepatic uptake was mainly mediated by opsonization. In vitro uptake studies showed that the increased amount of complement protein C3 and immunoglobulin G (IgG) adsorbed on nanoparticles was directly reflected in the increased rate of uptake of nanoparticles by Kupffer cells. These results indicate that the amount of opsonins associated with the surface of the nanoparticles increased over time and triggered substantial hepatic agglomeration. Similarly, C. Fang et al.(
59) have recently shown the effect of the molecular mass of PEG for passive targeting of stealth poly(cyanoacrylate-
co-
n-hexadecyl) cyanoacrylate (PHDCA) nanoparticles. As in previous studies, PEG surface modification of nanoparticles was able to dramatically reduce protein absorption as detected using quantitative assays (BCA assay). The amount of protein adsorbed was directly dependent on the molecular mass of the PEG. Medium-sized pegylated nanoparticles (100−200 nm) showed 10−40% protein absorption, and PEG 10 kDa was found to be the most efficient size of PEG as compared to PEG 2 kDa and PEG 5 kDa in preventing protein absorption. The results suggest that a dense PEG shielding over a negatively charged surface is important in preventing protein absorption. The trend in shielding properties was confirmed by the analysis of nanoparticle uptake by murine macrophages and blood clearance kinetics. Interestingly, the level of tumor necrosis factor-α (TNF-α) delivered by nanoparticles into tumor-bearing mice was clearly correlated with the shielding density. Another recent example confirmed the advantages of surface pegylation by utilizing thiolated gelatin nanoparticles (~300 nm) to improve passive tumor targeting in an orthotopic human breast adenocarcinoma xenograft mouse model.(
60) Despite the fact that these particles were twice the size of those in the previous study, the results showed a lower rate of uptake of the PEG-modified nanoparticles by the liver, indicative of the stealth properties of pegylated nanoparticles. In general, pegylated nanoparticles were found to have longer circulation time and higher levels of tumor accumulation than nonpegylated nanoparticles. In summary, much has been learned about PEG molecular mass and PEG density on nanoparticles which has led to reduced plasma protein adsorption, opsonization, and nonspecific uptake. In turn, this has resulted in increased nanoparticle circulation half-life and improved therapeutic efficacy of drugs delivered using pegylated nanocarriers.
| Table 1Clinical Status of Polymeric Drug Delivery Nanoparticles under Development |
Effect of Size
On the basis of physiological parameters such as hepatic filtration, tissue extravasation, tissue diffusion, and kidney excretion, it is clear that, along with surface composition, particle size is a key factor in the biodistribution of long-circulating nanoparticles and achieving therapeutic efficacy (Figure ). In one study, in vivo biodistribution results of polystyrene nanoparticles with consistent composition and varying particle sizes of 50 and 500 nm showed higher levels of agglomeration of the larger nanoparticles in the liver.(
56) It was suggested that the mechanism of hepatic uptake was mediated by surface absorption of proteins leading to opsonization. However, the effect of temperature (37 °C vs 4 °C) on hepatic elimination showed unexpectedly faster uptake of the 50 nm polystyrene nanoparticles at the lower temperature. Similarly, the size of the nanoparticle was shown to have a substantial effect on the protein absorption. Small (<100 nm), medium (100−200 nm), and large (>200 nm) pegylated PHDCA nanoparticles incubated with serum protein for 2 h showed a significant correlation between particle size and protein absorption.(
59)Protein absorption on small nanoparticles (80 nm) was quantified (6%) and compared to the same nanoparticle formulation with a larger size (171 and 243 nm, 23 and 34%, respectively). The effect of protein absorption on different sized nanoparticles was also confirmed with the analysis of nanoparticle uptake by murine macrophages and blood clearance kinetics. Blood clearance of the smaller nanoparticles was twice as slow as with the larger nanoparticle formulations. More importantly, the amount of drug encapsulated (TNF-α) in nanoparticles that accumulated in the tumor within 24 h was twice that of the larger nanoparticle formulation. The results suggest that it might be due to a higher surface PEG density (brushlike) on the surface of smaller nanoparticles. Rijcken et al.(
81) have shown that size and polydispersity can substantially affect the biodistribution of micelles. Cross-linked micelles incubated in PBS (pH 7.4) at 37 °C for more than 3 days did not show substantial increases in size and polydispersity in contrast to non-cross-linked micelles (PD ~ 0.5 after 10 h). Stable micelles showed a long circulation half-life of ~8 h due to a low rate of hepatic uptake (liver ~ 10% of the injected dose; spleen ~ 2% of the injected dose). Similar to results with small liposomes (<100 nm), micelles were found to accumulate in the skin.
In summary, it has been consistently shown that pegylated nanoparticles smaller than 100 nm have reduced plasma protein adsorption on their surface and also have reduced hepatic filtration. Further, these small pegylated nanoparticles have a long blood residence time and a high rate of extravasation into permeable tissues, demonstrating the importance of tunable particle size and surface composition for achieving effective, targeted delivery.
Effect of the Core
Shenoy et al.(
82) investigated the biodistribution of stealth poly(β-amino ester) nanoparticles (PbAE) and poly(caprolactone) (PCL)-based nanoparticles with a similar size range of 100−200 nm. In addition, both nanoparticle formulations had a very high positive surface charge of approximately +30 mV. The results clearly showed a significantly higher level of accumulation in the heart and lung tissues for the PbAE nanoparticles. This was correlated with a lower level of accumulation in the liver as compared to PCL nanoparticles. It was suggested that a higher degree of aggregation of PbAE nanoparticles in the presence of serum proteins affected the biodistribution leading, to such high levels of accumulation in the lung. Recently, greater degrees of nanoparticle “flexibility” were hypothesized to improve the binding ability of particles on the cell surface.(
83) Shell cross-linked nanoparticles (SCKs) containing partially hydrochlorinated poly(isoprene) cores were shown to undergo temperature-dependent deformation.(
84) SCK nanoparticles with similar physiochemical properties (size, ca. 20 nm; ζ, ca. −25 mV) possessing a low glass transition temperature (
Tg) with a fluidlike poly(methyl acrylate) (PMA) core or a high
Tg with a glassy poly(styrene) (PS) core were synthesized to evaluate the effects of the rigidity of the polymeric core on the in vivo biodistribution.(
84) The results showed that high-
Tg poly(styrene) core nanoparticles exhibited a significantly higher blood residence time compared to the low-
Tg poly(methyl acrylate) nanoparticles. The low-
Tg core is expected to provide greater flexibility and an increased number of surface interactions of the nanoparticles with the tissues and biological environment. However, it was not clear if the relative rigidity or other physicochemical properties of the polymers (hydrophobicity) affected the blood residence time. The results suggested that the composition of the core has an important effect on the blood residence time. In this case, PEG surface modifications did not have a noticeable effect on kidney accumulation and clearance most probably due to the small size of the nanoparticle formulations. These results are consistent with previous work investigating the effect of the core composition on the nature of absorbed proteins using a series of block copolymers with the same PEG thickness and density.(
85) Copolymers of poly(ethylene glycol) (PEG) and polyesters with increased hydrophobicity (PLGA, PLA, and PCL) were synthesized.(
85) Evidence of C3 cleavage was used as a semiquantitative way to characterize the level of complement activation by naked and pegylated nanocapsules (NCs) formed with different core copolymers. The results showed that PLA NCs prepared from biodegradable polyesters act as strong activators of complement. Since PCL showed a lower level of complement activation, the reason for the stronger complement response did not seem to correlate with the hydrophobicity of the copolymer. However, pegylated NCs had significantly reduced levels of complement activation which were shown to depend on PEG chain length and density. In another study, polymeric nanoparticles coated with polysorbate were shown to be able to accumulate in the brain tissues.(
86) The brain blood vessels are characterized by tight junctions between endothelial cells possessing a TEER (trans-endothelial electrical resistance) of approximately 1500−2000 Ω/cm
2 which prevents paracellular transport of molecules and represents a biological barrier for delivery of the drug to the brain. However, poly(butyl cyanoacrylate) nanoparticles coated with polysorbate 80 were able to deliver significant amounts of doxorubicin to the brain. Multiple formulations were studied, and polysorbate 80-coated nanoparticles exhibited a high doxorubicin concentration (6 μg/g) in brain tissues as compared to noncoated nanoparticles (0.1 μg/g). It was found that apolipoprotein E (apoE) binds substantially to polysorbate 80 surfactants, and the results suggest that apoE is involved in the mediated transport of the polysorbate 80-coated nanoparticles to the brain. Further explanation of the high concentration of doxorubicin drug delivered to the brain can be derived from the fact that polysorbate can act as an anchor for apoE protein binding and interact with LDL receptors expressed in the brain blood vessels. Polysorbate 80 also has been shown to have Pgp inhibition properties which could provide an additional rationale for the delivery of the doxorubicin with nanoparticles having a polysorbate 80 coating.(
87) Recently, Shaw and Weissleder (in press in
Proceedings of the National Academy of Sciences of the United States of America) have found interesting results regarding the core composition and relative toxicities of nanoparticles utilizing an in vitro, high-throughput, multidimensional analysis across varying concentrations utilizing multiple cell types and multiple assays of cellular physiology. Their generalizable, systematic approach indicates that biological effects result from the combined effects of many aspects of nanoparticle composition, including the core composition which was found to have a somewhat surprising effect on biological properties. Future work developing such structure−activity relationships may shed even more light on the complex interplay between nanoparticle properties and subsequent biological effects. In summary, it has been found that physicochemical properties of the core are critical parameters of the nanoparticle formulations. Precise modifications of these properties are able to dramatically affect the extent of interactions with blood and biodistribution of nanoparticles.
Effect of Surface Functionality and Charge
It has been established that the physicochemical characteristics of a polymeric nanoparticle such as surface charge and functional groups can affect its uptake by the cells of the phagocytic system. It was previously shown that polystyrene microparticles with a primary amine at the surface underwent significantly more phagocytosis as compared to microparticles having sulfate, hydroxyl, and carboxyl groups. Therefore, it is well accepted that positively charged nanoparticles have a higher rate of cell uptake compared to neutral or negatively charged formulations. Nanoparticles carrying a positively charged surface are also expected to have a high nonspecific internalization rate and short blood circulation half-life. Nanoshells having a negative surface charge have shown a marked reduction in the rate of uptake. The ζ potential for nanoshells with BSA absorbed on the surface of the nanoparticle was characterized by a shift to a more negative value. However, the BSA absorption did not promote a higher rate of cell uptake. Interestingly, when the biodistribution of Tyr- and Tyr-Glu-PEG/PDLLA micelles was investigated in mice,(
88) both the Tyr- (neutral) and Tyr-Glu (negatively charged) PEG/PDLLA micelles exhibited no remarkable difference in their blood clearance kinetics. The pharmacokinetic parameters suggest that all the micelle formulations are distributed mainly into the extracellular space volume of the spleen and liver. However, the anionic Tyr-Glu-PEG/PDLLA micelles showed a lower distribution (10 times lower) into the liver and spleen 4 h post-injection. Remarkably, a lower level of accumulation of Tyr-Glu-PEG/PDLLA micelles into the liver and spleen is believed to be due to synergic steric and electrostatic repulsion which decrease the rate of cellular uptake. In addition, urinary excretion was the major excretion route, confirming a low rate of hepatic uptake of the micelle formulation. Recently, thiolated gel nanoparticles (~250 nm; ca. −5 mV)(
60) have had a higher level of tumor accumulation than nonthiolated nanoparticles, and it was correlated with a very short plasma half-life of ~3 h, indicating a possible preferential uptake or accumulation in the tumor tissues. In addition, the thiolated nanoparticles were found to have a higher rate of uptake into the spleen. The reasons for such an impact of the thiolated functional group on the surface of the nanoparticles are not very clear, but it could be due to aggregation of nanoparticles through disulfide bond formation or reaction with thiolated molecules circulating in the blood. In summary, it is established that neutral or negatively charged surface nanoparticles have a reduced plasma protein adsorption and low rate of nonspecific cellular uptake. Thus, surface functionality is another critical parameter in controlling the development of long-circulating nanoparticles.
Effect of Active Targeting
Active targeting of nanoparticles involves the conjugation of targeting ligands to the surface of nanoparticles. These ligands can include antibodies, engineered antibody fragments, proteins, peptides, small molecules, and aptamers. The active targeting mechanism takes advantage of highly specific interactions between the targeting ligand and certain tissues or cells within the body to promote the accumulation of nanoparticles.
3,89–92 In the case of weak binding ligands, low affinity can be offset by increased avidity through the surface functionalization of multiple molecules or multivalent designs and has been shown to be a valid approach. There are several examples of FDA-approved antibodies in clinical practice today,
93,94 including Rituxan (target, CD20-positive B-cells for the treatment of non-Hodgkin’s lymphoma and rheumatoid arthritis), Herceptin (target, HER-2-overexpressing breast cancer cells), Erbitux [target, epidermal growth factor receptor (EGFR) for the treatment of colorectal cancer], Iressa (target, EGFR for the treatment of non-small cell lung cancer and metastatic breast cancer), and Avastin [target, vascular epidermal growth factor (VEGF) for the treatment of metastatic colorectal, non-small lung, and breast cancers]. While the progress with monoclocal antibodies has been encouraging, they have not been shown to be curative. This has led to the development of immunoconjugates with the intent of utilizing the targeting specificity of the antibody to deliver a relatively potent drug. The development of immunoconjugates has been slow due to various challenges but has resulted in the FDA approval of three immunoconjugates(
46) such as Mylotarg (conjugate of calicheamicin targeting CD33 for the treatment of acute myeloid leukemia), Zevalin (yttrium-90 radio-immunotherapy conjugates targeting CD20-positive B-cells for the treatment of non-Hodgkin’s lymphoma), and tositumomab (
131I radio-immunotherapy conjugates targeting CD20-positive B-cells for the treatment of non-Hodgkin’s lymphoma). Currently, drug delivery carriers are being functionalized with proteins, including antibodies or antibody fragments and various other targeting ligands, with the goal of both delivering a high therapeutic dose and delivering this high therapeutic dose to specific tissues or cells. Conjugation approaches for controlling the amount of targeting proteins on the surface of the nanoparticles have been developed to increase specificity and binding affinity. Research using proteins for targeting applications has led to a better understanding of the effect of stability and size of the ligand for successful targeting and clinical development.
95,96 Antibody molecules are very large and complex and can be expensive to manufacture relative to small-molecule drugs. In addition, antibodies have a hydrodynamic size of ~20 nm that can increase the size of the NPs. Recently, Davis et al.(
97) have demonstrated the therapeutic impact of differential distribution between intracellular and extracellular tumor domains. Cyclodextran-based nanoparticles containing transferrin targeting ligand (NP size of ~70 nm) showed enhanced intracellular accumulation in a human tumor xenograft mouse model. Interestingly, the model of tumor accumulation and uptake fit the experimental data, suggesting that tumor retention time and internalization into tumor cells are the critical factors affecting accumulation of nanoparticles in targeted tissues. The same group(
98) reported the first toxicology study of a targeted, polymeric nanoparticle platform in non-human primates as reported above. This type of study is crucial for answering the numerous critical questions about the toxicological response to polymeric nanoparticles circulating in the blood. The results clearly show that their intravenously injected polymeric nanoparticles were safely administered to non-human primates even after multiple administrations over 3 weeks. These promising results led to the submission of an investigational new drug application.
While targeting with proteins has been shown to be advantageous in some cases, for various reasons, it may be desirable to achieve targeting with relatively small targeting ligands that are potentially easier to manufacture. The discovery of new peptide targeting domains(
99) provides the advantage of being able to utilize small synthetic molecules for active targeting. One such example is peptides which are relatively stable compared to antibodies and are also less likely to be immunogenic. Similarly, small molecules can be very attractive for use as targeting ligands, and small molecules such as folic acid or sugar molecules have been extensively used. For example, cell surface membrane lectins have been shown to be overexpressed on the surface of numerous cancer cells and are able to specifically internalize sugar molecules (lactose, galactose, and mannose).(
5) Similarly, nucleic acid aptamers are able to fold into unique structures capable of binding to specific targets with high affinity and specificity.(
100) Favorable characteristics of aptamers have resulted in their rapid progress into clinical applications. Our group is interested in developing this class of molecules for targeted delivery of controlled drug-releasing polymer vehicles.(
101) We described the first proof-of-concept drug delivery polymeric nanoparticle utilizing aptamers as targeting ligands in vitro(
102) and subsequently showed the efficacy of similarly designed nanoparticles against prostate cancer tumors in vivo.(
103) In vivo results using a human prostate cancer tumor xenograft mouse model showed a significant therapeutic efficacy of aptamer-targeted nanoparticles loaded with docetaxel compared to nontargeted formulations. Targeted nanoparticles substantially reduced the size of the tumor after a single intratumoral injection, and all of the treated mice survived more than 3 months in contrast to other controls. More recently, we reported a novel strategy for facile synthesis of targeted nanoparticle formulations.(
104) We have shown that targeted nanoparticles specifically accumulated in the prostate tumor xenograft mouse model (~1.8%ID/g of tissue) compared to nontargeted nanoparticles (0.5%ID/g of tissue). In summary, it has now been shown both in vitro and in vivo that targeted nanoparticles are able to agglomerate in specific tissues to improve the therapeutic efficacy, and there is a growing body of literature in this area. Clearly, active targeting provides a powerful approach to increasing the drug delivery load at a specific site for treatment.