review article design of nanoparticle-based carriers for...

Review ArticleDesign of Nanoparticle-Based Carriers forTargeted Drug Delivery

Xiaojiao Yu,1 Ian Trase,1 Muqing Ren,2 Kayla Duval,1 Xing Guo,1 and Zi Chen1

1Thayer School of Engineering, Dartmouth College, Hanover, NH 03755, USA2Department of Chemical Physics, University of Science and Technology of China, Hefei, Anhui 230026, China

Correspondence should be addressed to Zi Chen; [email protected]

Received 8 December 2015; Revised 29 March 2016; Accepted 3 April 2016

Academic Editor: Martin Kroger

Copyright © 2016 Xiaojiao Yu et al. This is an open access article distributed under the Creative Commons Attribution License,which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Nanoparticles have shown promise as both drug delivery vehicles and direct antitumor systems, but theymust be properly designedin order to maximize efficacy. Computational modeling is often used both to design new nanoparticles and to better understandexisting ones.Modeled processes include the release of drugs at the tumor site and the physical interaction between the nanoparticleand cancer cells. In this paper, we provide an overview of three different targeted drug delivery methods (passive targeting, activetargeting, and physical targeting) and compare methods of action, advantages, limitations, and the current stages of research. Forthemost commonly used nanoparticle carriers, fabricationmethods are also reviewed.This is followed by a review of computationalsimulations and models on nanoparticle-based drug delivery.

1. Introduction

Nanoparticles (NPs) often exhibit different magnetic, ther-mal, optical, and electrical properties due to their high surfacearea and limited quantum mechanical effects [1]. NPs areoften developed and used as drug carriers, as they can deliverchemotherapeutics targeted to the tumor tissue withoutdamaging normal organs (Figure 1). The ideal NP carriersshould be biodegradable, stable, nonimmunogenic, easy tofabricate, cost-effective, and able to release their payloadsonly at the target site [2].

Medical NPs are often manufactured with a guidedbottom-up method, in which engineered macromolecularcomponents are guided by external stimuli to interact witheach other and self-assemble into complex structures thatotherwise would not be possible [3]. Drugs can be eitherencapsulated within the nanoparticle or attached to thesurface.

A typical drug delivery nanoparticle starts with ananoplatform class, which include liposomes, polymericmicelles, drug conjugated polymers, and dendrimers [4–7]. There are three main methods to transport drug-loadednanoparticles to diseased sites: passive targeting, active

targeting, and physical targeting. Passive targeting worksthrough the increased permeability and retention (EPR)effect, which makes tumor cells preferentially absorb NP-sized bodies [8, 9]. In active targeting, NPs are functionalizedwith ligands such as antibodies, proteins, and peptides [10],which interact with receptors overexpressed at the target site[11]. Physical targeting uses external sources or fields to guideNPs to the target site and to control the release process,for example, in photothermal and magnetic hyperthermiatherapy. For all targeting types, drug release can be triggeredby a change in pH, temperature, or a combination of both.

In order to design an effective NP, one needs to under-stand the combinatorial effects of size, shape, surface chem-istry, patient-specific information, and other parameters.Optimizing all of these parameters through experiments isboth time- and resource-intensive, and so computationalmodeling is used to shrink this possibility space. Simulationshave been used to model the continuum of NP transportand the quantummechanical interactions of ligand receptors.Mesoscale modeling and Monte Carlo simulations are alsooften used when certain values are uncertain [12].

The aim of this paper is to review fabrication meth-ods for the most common nanoparticle types (specifically

Hindawi Publishing CorporationJournal of NanomaterialsVolume 2016, Article ID 1087250, 15 pageshttp://dx.doi.org/10.1155/2016/1087250

2 Journal of Nanomaterials

Tumor

Free drug

(a)

Tumor

Drug-loaded NPs

(b)

Figure 1: Schematic contrast of drug biodistribution after injection of free drug (a) and drug-loaded NPs (b).

self-assembly), targeted drug delivery processes, and thecurrent state of NP computational modeling. Directions forfuture research are also discussed.

2. Self-Assembled Nanoparticles asDelivery Vehicles

Medical nanoparticles, despite their name, are far from thesmallest things that scientists and engineers work with.Rather, they occupy a manufacturing blind spot locatedbetween large and small structures. Objects and materials

with features at, and larger than, themicroscale are now read-ily fabricated through “top-down” approaches like lithogra-phy and precision machining. Objects smaller than nanopar-ticles are generally easily synthesized through “bottom-up”methods in which individual chemicals essentially assemblethemselves under the influence of intermolecular forces.Nanoparticles are too large and complex to be made bysimply mixing their molecular components in a test tubebut much too small to be assembled with even the high-est precision lithographic device. The solution is a guidedbottom-up approach, in which macromolecular componentsare engineered to interact with each other and often external

Journal of Nanomaterials 3

Liposome Micelle

Figure 2: Schematic of two types of amphiphilic nanoparticles, liposomes and micelles. Liposomes have a double layer and a hydrophiliccore, while the core of micelles is hydrophobic.

stimuli fields and self-assemble into structures more complexthan would otherwise be possible [3].

While there are many types of self-assembled nanopar-ticles, the most studied can broadly be sorted by structure.Below is an introduction, with examples, to amphiphilic NPs(the most common type of drug carrying NP), followed by abrief explanation of other novel NP structures: dendrimers,polyrotaxanes, functionalized carbon nanotubes, graphene,and metal solid-core nanoparticles. It is important to notethat all of these NPs can be functionalized to actively targetspecific sites in and on tumor cells.

2.1. Amphiphilic Nanoparticles. The earliest drug NPs werebiomimetic in nature, mimicking the micelles and liposomesalready present in the body. These NPs contain macro-molecules with both hydrophobic and hydrophilic regions.In an aqueous environment such as blood or cellular fluid,the hydrophobic regions will cluster together in order tobe shielded from the surrounding polar water, forming amicelle. Synthetic micelles have hydrophobic cores, whilemore complex amphiphiles can form full liposomes with adouble-layered macromolecular wall and a hydrophilic corewith the same properties as or different properties from thesurrounding fluid (Figure 2). Drugs can be captured eitherin the core or on the surface of these particles, dependingon their hydrophobicity. In either case, the drug is shieldedfrom removal by the immune system and from other in vivohazards.

An early example of micelle-forming particles involvedpoly(ethylene oxide)-poly(aspartic acid) (PEO-PAA) blockcopolymers [21]. These polymers were used to create a self-assembling micelle containing the chemotherapy drug dox-orubicin (DOX). DOX is commonly used to test the efficacyof self-assembling carriers and comes in both hydrophobicand hydrophilic HCl formulations—it can be assumed to behydrophobic unless otherwise stated. In this case, DOX was

bound to the hydrophobic PAA polymer in organic solventand then pushed through a membrane into an aqueous solu-tion, a technique known variously as membrane dialysis ordiafiltration.This transition fromhydrophobic to hydrophilicsolution induces the formation of micelles and leaves DOXencapsulated at the PAA center. The resulting nanoparticlewas extremely soluble and stable in water, increasing the half-life of the payload drug.

A similar experiment was carried out with an amphiphilicpullulan acetate polymer [22]. Cancer cells have been shownto overexpress vitamin H, so the polymer was functionalizedwith it in order to actively target these cancer cells. AfterDOXwas loaded into the micelles using membrane dialysis, theauthors found that the amount of vitamin H expressed onthe surface of the nanoparticle correlated with its uptake bycancer cells, indicating successful active targeting.

More complex micelles can be engineered to releasetheir payloads in response to external stimulation. Bae etal. [23] had designed pH-sensitive nanoparticle carrierswhich follows previous reports on poly(ethylene glycol)-poly(aspartate-hydrazone) copolymers. DOX was bound tothe hydrazone group andmicelles were again formed throughmembrane dialysis. Hydrazone bonds are easily cleaved inacidic conditions, so this micelle was designed to exposeDOX in response to low pH. Micelles are taken up bycells through endocytosis and subsequently engulfed bylysosomes, which have a pHof around 5 and thus trigger drugrelease.The authors found thatDOXconcentration decreasedas a function of pH, with around 30% of the drug released ata pH of 5 and the entire payload released at lower pH.

A particularly interesting trigger-based nanoparticlewas created by combining the pH responsive polymerpoly(acrylic acid) (PAA) with the heat sensitive poly(N-isopropylacrylamide) (PNIPAM) [24]. PNIPAM becomeshydrophobic above its lower critical solution temperatureof around 32∘C, while PAA becomes hydrophobic at a pH


below 4.8. The copolymer will thus form a micelle with aPNIPAM core at high temperature and pH but flip to a PAAcore at low temperature and pH. PAA also binds to DOX inaqueous solution, eliminating the need for organic solventsor membrane dialysis. When it binds at low temperatureand high pH, the PAA-DOX complex becomes hydrophobicand forms micelles. This nanoparticle was shown to exhibitdrug release with both an increase in temperature (whichcauses the micelle to flip inside-out and expose DOX tothe environment) and a drop in pH (which causes PAA toprotonate and release the less positive DOX). PNIPAM hasalso been used to synthesize nanoparticles that can shrink involume in response to temperature [25].

Polymer nanoparticles have also been made with hybridpolymer-lipid amphiphiles, which allow for a broader rangeof potential polymers.This system was used in a nanoparticleable to hold both drugs and DNA using a cationic core-shellsystem [26]. The main polymer chain is hydrophilic poly(N-methyldietheneamine sebacate) (PMDS) grafted with thehydrophobic N-(2-bromoethyl) carbamoyl cholesterol lipidto form an amphiphilic copolymer.The antitumor drug pacli-taxel (PTX) was encapsulated through membrane dialysis,and luciferase-coding DNA was bound to the nanoparticlein order to detect fluorescence. It was found that cancercells successfully expressed luciferase, indicating successfulendocytosis. Lipids can also be used as a molecular shield toincrease drug half-life in blood [27].

It is also possible to change the chemistry of certain poly-mers to distort the resulting nanoparticle [28]. Micelles canbe created in a variety of nonspherical shapes using poly[oligo(ethylene glycol) methacrylate]-block-[poly(styrene)-co-poly(vinyl benzaldehyde)] block polymers. The shape ofthe micelles changed from sphere to rod as the degree ofpolymerization for the P(ST-co-VBA) blocks increased. DOXwas able to be loaded into the micelles as normal.

Huang et al. [29] designed a poly(lactide-co-glycolic acid)nanoparticle coated with sgc8 aptamer capable of carryingboth a hydrophobic and hydrophilic cancer drug. After self-assembly, hydrophilic DOX is located in the poly(ethyleneglycol) shell and hydrophobic paclitaxel is located in thePLGA core. The sgc8 aptamer causes the nanoparticle tobe internalized by cancer cells specifically, an example ofactive targeting. The authors show that the multidrug combowas more effective than either drug individually at reducingcell viability, and the addition of a second drug had aninsignificant effect on the viability of normal cells. It is alsopossible to use transferrin on the nanoparticles to increaseuptake by cancer cells [30].

Xia et al. [31] designed silk-elastin protein polymernanoparticles. Silk-elastin-like proteins are synthetic geneti-cally engineered proteins designed tomimic the properties ofboth silk and elastin.Theproteins are temperature-dependentamphiphiles and will form micelles. In this case, it wasshown that DOX actually triggered micelle formation insome cases by increasing the hydrophobicity of the silk-endchains. The authors reason that this polymer and method ofself-assembly, since it is formed biologically in nonextremeenvironments, will generate fewer toxins and be a saferalternative to traditional methods.

2.2. Novel Nanoparticle Structures. An example of a novelnanoparticle structure is a dendrimer (Figure 3(d)), whichis a repeatedly branching nanostructure that mimics thestructure of tree branches and blood vessels [4]. This shapeis shown to both generate large surface areas and disperse asurface more evenly throughout the structure. Dendrimersare commonly made of poly(amidoamine) but can alsobe made from other polymers. It has been shown thatdendrimers can be used to target cancer cells with highaccuracy [32]. Li et al. reported a novel class of micellesmade of linear PEG and dendritic cholic acids (CA) blockcopolymers (called telodendrimers) [15]. By cross-linking theboronate esters at the core-shell interface, the stability ofthese micelles can be improved (Figure 4) [16]. As the cross-linking reactants, boronic acid and catechol were added to thepolymers through stepwise peptide chemistry.

Rotaxanes are molecular linkages in which a cyclicmolecule encircles a dumbbell-shaped one—the cyclicmolecule can rotate but cannot slide off of the dumbbell. Poly-rotaxanes (Figure 3(a)) can be a useful tool for drug delivery[33]. Liu et al. attached cyclic cyclodextrin to poly(ethyleneglycol) chains in order to form a nanoparticle. Hydrophobiccinnamic acid was attached to the ends of the PEG chains inorder to increase the space between the cyclodextrin rings,providing space for DOX to bind [33]. This system can alsobe used to transport the drug methotrexate [13].

Dendrimers and rotaxanes are used primarily for drugdelivery, but the next three structures are also often usedin thermal therapy, where a localized temperature increaseis used to destroy a tumor. Graphene (Figure 3(b)) is asingle atom-thick hexagonal allotrope of carbon with novelelectrical, thermal, and mechanical properties. Because of itshigh surface area, it is useful as a drug carrier, and its structuremakes it efficient at converting infrared light into heat. It is,however, relatively toxic and must be stabilized and shieldedthrough the addition of polymers to its surface [14, 34].Carbon nanotubes (Figure 3(c)) are rolled tubes of graphenethat exhibit many of the same properties. Nanotubes havebeen used as hybrid drug carries, where the antitumor drugis released only in a specific site when bombarded with near-infrared radiation and the NPs are heated up.This can help tolimit drug release to the tumor site and protect healthy tissue[35, 36].

Metal-core nanoparticles (Figure 3(e)) can be usedphotothermally, like graphene and carbon nanotubes, butthey can also be used for magnetic hyperthermia, wherea magnetic NP oscillates and heats up in response to anexternal magnetic field. Most NPs are made of magnetite ormaghemite cores and are relatively biocompatible. Successfulclinical trials have also been done with a magnetic fluidcomposed of dispersed NPs in water [37].

Oligopeptides have been extensively studied as nanocar-riers due to their intrinsic pH sensitivity resulting fromamino acids. Mo et al. developed a liposome system basedon zwitterionic oligopeptide lipids as nanocarriers [38].The amino acid-based lipids, 1,5-dioctadecyl-l-glutamyl 2-histidyl-hexahydrobenzoic acid (HHG2C

18) and 1,5-distearyl

N-(N-𝛼-(4-mPEG2000) butanedione)-histidyl-l-glutamate(PEGHG2C

18), have a multistage pH-response to first the


(a)

(b)

(c) (d) (e)

Figure 3: Structures of novel nanoparticles. (a) Polyrotaxane NPs are assembled from cyclical molecules threaded around a long polymerchain. Hydrophilic ends are added to the chain in order to induce self-assembly. Drugs are then added to the finished NP. Adapted withpermission from [13]. (b) Graphene functionalized with shielding molecules and ligands. Adapted with permission from [14]. (c) Carbonnanotube schematic. (d) Dendrimer schematic. (e) Metal-core photothermal NP schematic.

tumor microenvironmental pH (pH 6.8) and then the endo/lysosomal pH (pH 4.5).

Supramolecular polymers have significant potentialapplication in drug delivery due to their reversiblemonomer-to-polymer transitions. Figure 5 shows the molecular unitdesigned to form a supramolecular architecture [17]. Thepair is a peptide amphiphile monomer composed of threesegments: a biological signal-bearing sequence, an aminoacid-contained domain, and a hydrophobic alkyl tail. Thesemonomers can form a cylindrical aggregate where twisted𝛽 sheets (red) collapse through hydrophobic interactionsamong alkyl chains, resulting in high signal densities. Theblue regions represent water domains present in the assemblyinterior.

2.3. Challenges in Nanoparticle Development. There aremanychallenges that arise in the early and late stage development ofmedical nanoparticles that are either nonexistent or minimalin non-nanoparticle-based therapies [39].The primary driverof these issues is the hierarchical and nonuniform nature ofnanoparticles, which means that a small change in a single

property can have outsized effects on the pharmacokineticsor therapeutic efficacy of the particle. As an example, onecommon issue is maintaining a narrow distribution of par-ticle size. For most applications, a particle size under 200 nmis desirable. With a broad normal distribution of sizes, thismeans that often the average particle must be too small to beuseful in order to limit the number of particles over 200 nm.Amanufacturing process that ensures a narrow size distributionis thus desirable.

In addition, many unique challenges can arise during thetrial and production stages of the nanoparticle. Oftentimes,a procedure for nanoparticle formation that works in a labsetting will not work in a factory, and the synthesis stepsmust be completely reworked. In a factory, the variation innanoparticle structure must be smaller, the yield must behigher, and the synthesis must be more sterile than what isacceptable in a lab. All of these things can make a particleunviable to produce even if it works. If a particle is sold,it must be shelf-stable, which means both that it will notdegrade in solution and that it will not clump over time,as nanoparticles often do. Finally, nanoparticles face extra


+ nn

PEG 5000PEG 5000

PEG5k-L2-(boronic acid)4-CA8 PEG5k-L2-catechol4-CA8NHO

BO

O

HNR

O

B B

B B

C C

CC

Hydrophobicdye or drug

Self-assembleIn situ cross-link

Mannitol and/oracidic pH

Figure 4: Schematic representation of the disulfide cross-linked micelles formed by oxidization of self-assembled thiolated telodendrimerPEG5k-Cys

4

-L8

-CA8

[15]. Schematic representation of the telodendrimer pair [PEG5k-(boronic acid or catechol)4

-CA8

] and the resultingboronate cross-linked micelles (BCM) triggered by mannitol and/or acidic pH values [16].

Nanofiber

Supramolecular polymerMonomer

Lipid tail 𝛽-domain Biological signal

Figure 5: Molecular representation of monomer and the corresponding supramolecular polymer formed after their aggregation throughspecific interactions [17].

regulatory challenges as their toxicity is much more difficultto determine than that of a small molecule. This greatlyincreases the time and cost of clinical trials. These challengescombined mean that it is always prudent to consider ques-tions of scalability and reproducibility even at the earlieststages of development to prevent failure at a later stage.

3. Cancer Cell Targeted Drug Delivery

3.1. Mechanisms of Targeted Cancer Cell Drug Delivery. Can-cer cells are otherwise normal cells with unique mutations

in genes regulating growth, which cause them to divideuncontrollably and give them the ability to metastasize [40].Cancer cells successfully compete with normal cells foroxygen, glucose, and amino acids for division and growth,but a tumor can only grow to about 2mm3 without formingblood vessels (angiogenesis) [41–43]. There are more thanone hundred types of cancer, more than 85% of which aresolid [43]. Current treatment includes surgery, radiotherapy,chemotherapy, hormone therapy, and immunotherapy [43].However, the inability of drugs to specifically target cancercells hinders most treatment [2, 44, 45]. It is often quicker


Red blood cell

Normal cell

Endothelial cell

Tumor cell

Drug-loaded NPs

Blood vessel

Figure 6: Schematic illustration of enhanced permeation and retention (EPR) effect.

and cheaper to design a more effective way to better targetan existing drug than to develop an entirely new one. Drugdelivery targeting is classified as passive, active, or physicaland can target organs, cells, or organelles. Organelle targetingis an especially promising field of research, as many cancersspecifically affect a single one, and certain organelles providealternative paths for drug localization.

3.2. Passive Targeting Nanocarrier Systems. Drugs deliveredintravenously tend to evenly disperse throughout the body.However, tumor cells tend to take up particles of a certain sizeto a greater degree than healthy cells due to a combinationof leaky tumor blood vessels and faulty particle screening.This is known as the enhanced permeation and retention(EPR) effect (Figure 6) and is the mechanism behind passivetargeting [46].

The EPR effect is influenced by NP properties includingparticle size, shape, and surface charge, and it in turn influ-ences circulation time, penetration speed, and intracellularinternalization [47, 48]. For example, phagocytic cells facil-itate larger particle uptake, while nonphagocytic cells favor

the uptake of smaller particles [49]. It has been consistentlyshown that PEGylated nanoparticles reduce plasma proteinadsorption on their surface and reduce hepatic filtrationwhen their size is smaller than 100 nm [50]. The nanoparticlesurface properties could play a central role in blood circula-tion and subsequent cellular internalization [48, 51]. NPswitha negative surface charge will circulate longer in the blood,but positively charged NPs are more readily taken up bycancer cells (which have negative surface charge) [8–10, 52–54]. In order to clarify the influence of shape on the cellularuptake of PEGylated NPs, Li et al. performed large-scalemolecular simulations to study different NP geometries withidentical surface area, ligand-receptor interaction strength,and PEG grafting density (Figure 7) [20]. They found thatspheres exhibited the fastest internalization rate, followed bycubes, while rods and disks were the slowest.

Delivery platforms include liposomes [5], polymericmicelles [6, 7], targeted polymer drug conjugates [4], anddendrites. They all consist of macromolecule collections inwhich drugs are dissolved, entrapped, or conjugated to thesurface [55]. Several liposomal drug delivery systems have


Sphere Rod Cube Disk

(a) (b)

Figure 7: Different shapes of NPs: sphere, rod, cube, and disk. (a) shows the transmission electron microscopy images of these NPs [18, 19].(b) shows the PEGylated NPs with grafting density 1.6 chains per nm2 in molecular simulations [20].

received clinical approval, including ones for doxorubicinand daunorubicin. An albumin-bound nanoparticle carryingpaclitaxel, Abraxane, was also approved by the FDA for breastcancer treatment [56].

Despite the EPR effect, more than 95% of passivelytargeted NPs fail to reach the tumor when administeredintravenously [52]. Targeting can be greatly improved bylocally controlling drug release at the site of the tumor. Thiscan be triggered through changes in the microenvironment(pH, temperature, or enzymatic) or through external stimuli(light, electric fields, magnetic fields, or ultrasound) directedat the tumor site [57–59].

An alternative way to improve the uptake of NPs, bothpassive and active, has been to functionalize their surfaceswith cell-penetrating peptides (CPPs). It has been found thatcertain short (∼30-amino-acid) peptide sequences can piercecell membranes and transport drug cargo into a cell. TheseCPPs can be attached to micelles, liposomes, and other typesof NPs [60]. Certain CPPs can also act as drug carriers ontheir own, carrying small molecules and short stretches ofDNA into cells [60]. Most CPPs are amphiphilic, with a netpositive charge. Because they are shaped similar to lysingpeptides, certain CPPs can exert unwanted toxic side effects.

3.3. Active Targeting. Active targeting uses ligands boundto the NP surface to improve their uptake selectivity. Theseligands can react with target cells and will often protectNPs from enzyme destruction. Ligands with a high bindingaffinity to the target cell will strongly increase deliveryefficiency. The most basic form of active targeting involvesfunctionalizing a NP with a ligand that binds to a moleculeoverexpressed on cancer cells. The issue with this, of course,is that healthy cells still express the same molecule, and ashealthy cells greatly outnumber cancer cells most of the NPsmiss their target.This issue can bemitigated by usingmultipleligands or by using ligands of different types.

Approaches to identify potential receptors in and oncancer cells include in vivo phage screening and aptamerscreening [61]. Using in vivo phage screening, F3 was dis-covered to bind well with nucleolin [62], which is presentat tumor cell surfaces and in tumor endothelial cells. Thecytoplasmic proteins annexin-1 [63], plectin-1 [64], and p32protein [65] were also found through in vivo phage screen-ing. By studying the expression of the known cell surfacereceptors in tumor vessels, other molecular markers can bedetected. For example, 𝜕v𝛽3, 𝜕v𝛽5 integrins, and ED-B were

discovered in angiogenic vessels using this principle [66–69]. Gene expression analysis has also been used to discoveroverexpression of collagen in tumor endothelial cells [70].A detailed review on various markers and their discoverymethods was given by Ruoslahti et al. [61].

Many antibodies have been approved for use in clinicaltreatment by the FDA, such as rituximab, ipilimumab, andtrastuzumab [71]. Antibodies are among the most studiedligands because of their high specificity and availability. Anantibody conjugated dendrimer was found to bind exclu-sively to human prostate adenocarcinoma (LNCaP) cellsthat express PSmA (J591) [32]. Although antibodies havemany merits, they are difficult to conjugate to NPs, resultin a short circulation time, and are expensive. Peptides area promising alternative, as they are smaller, simpler, morestable, and easier to produce. Among peptides, RGD is oftenused due to its strong binding with 𝛼v𝛽3 integrin receptors.Nucleic acid base aptamers combine the advantages of bothantibodies and peptides, but they degrade quickly. Othersmall molecules can also be used as ligands, such as folic acidfor folate receptors [72]. Suchmolecules are small, stable, andeasy to produce. Unfortunately, ligand detection for relevantsubstrates is challenging. Even with proper binding ligandsand receptors, binding incompatibility can limit therapeuticefficiency.Multiple ligandswith different charges can increaseoverall the binding affinity, but the limited binding ability andcapacity of receptors will govern the quantity and quality ofthe binding. For instance, overly strong binding can actuallyreduce tumor penetration, hinder selectivity, and lead to anoverdose of carriers [73].

Active targeting alters the natural distribution patterns ofa carrier, directing it to a specific organ, cell, or organelle. Incontrast, passive targeting relies on the natural distribution ofthe drug and the EPR effect. Both of these processes dependon blood circulation and the location of initial drug delivery.However, no actively targeted NPs are commercially availablecurrently.

3.4. Physical Targeting. Physical targeting navigates drugsto cancer cells using external stimulation, such as radiationor magnetic fields. For example, photothermal therapy usesNPs that once delivered efficiently convert near-infrared lightenergy to heat, killing cancer cells, and this process has fewside effects. Currently, most research is being carried outusing gold nanoparticles because they can be well controlledand have low toxicity.Their SPR (surface plasmon resonance)


effect can be tuned by shape, size, and thickness to maximizeexcitation and focus on a specific wavelength. According toa previous report, gold nanoparticles are already being testedin animals [74].

Photothermal agents other than gold have also beenexplored, such as carbon nanotubes which show strongabsorbance in near-infrared region [35], and have been usedwith photothermal treatments to inhibit G2-M cell cycles[36]. Graphene is another material with potential applicationas a photothermal agent. Functionalized graphene oxide withpolymer conjugates is pH sensitive and can photothermallycause cell death [34]. In addition, silica-coated graphenenanosheets functionalizedwith hydrophilic polyethylene gly-col have been used to deliver doxorubicin [14]. In the cases ofcarbon nanotubes and graphene oxide, pH and heat are themechanisms that initiate drug release.

One limitation to photothermal therapy is that cancercells are often tolerant to environmental stress, for example,with heat shock proteins that protect the cancer cell againstfurther damage [75]. Nanoparticle-free radiation therapiesare quite common, however. High-energy X-ray or gammaradiation is cytotoxic and can kill cells in specific regions [76].The details of radiation therapy have been reviewed by Stacyet al. [76].

Magnetic hyperthermia uses the heat energy producedby magnetic nanoparticles oscillating in a magnetic field.The distance betweenmagnetic nanoparticles and target cells,sensitivity to magnetic fields, and magnetic field strength allaffect heat energy production and correlate to therapeuticeffects. Magnetic NPs typically consist of four parts: NPs,protective agents, biomolecules, and surface agents. They areusually synthesized based on magnetite (Fe

3O4), maghemite,

cobalt, or nickel. Among them, iron oxides are most used dueto their biocompatibility and shape controllability.

Magnetic NPs are usually coated with functional poly-mers such as carbohydrates and proteins to protect againstcorrosion and potential toxin release. However, some poly-mers’ mechanical strength and selectivity are not easy tocontrol, so organic linkers are often used to create electro-static interaction. Replacing ions and changing the pH ofthe immediate environment can also modify the bindingstrength of magnetic particles. Both photothermal therapyand magnetic hyperthermia can be done in vivo and in vitro,and several clinical trials have been performed [77]. Unlikeactive targeting or physical targeting alone, this combinationcan effectively promote NP internalization by tumor cells[78].

4. Modeling Applications in Drug Delivery andNanoparticle Device Design

4.1.Models andTheir Applications in the Field. Theuse ofNPsfor drug delivery is complex because the pharmacokineticproperties of the drug itself must be fully understood as wellas all parts of the delivery system. NPs can be affected bymany parameters, including platform (liposomes, polymericmicelles, polymer drug conjugates, and dendrites), physicalparameters (size, shape), and surface chemistry. On its way to

the target, the NP must pass through and interact with mul-tiple biological barriers, including those in the bloodstream,those at the site of the tumor, those at the surface of the cell,and several within the cell. In addition, the efficacy of a NP ishighly dependent on individual patient conditions.

Because nanoparticle research is complex, with manyvariables and costly experiments, it is an excellent candidatefor computer simulations. Simulations can both screen drugand carrier candidates and provide design insights towardsentirely new NPs. Theoretical and computational modelingcan be used for any of the drug delivery processes previouslyreviewed to provide solutions for optimized geometry, sur-face chemistry, or other properties. For instance, continuumbased modeling is used to study transport and dissolution,molecular dynamics are used for cell interaction, and stochas-tic approaches such as Monte Carlo simulations can be usedto calculate random variables and take patient uncertaintiesinto account.

4.1.1. Continuum Modeling. Transport modeling of NPsthrough the vascular network is challenging due to vastlydifferent blood vessel diameters, ranging fromcentimeters forlarger vessels to microns for capillaries. Advection-diffusionmodels are used for transport in the larger vessels, whereblood is modeled as a simple Newtonian fluid [79, 80]. Inthe microvasculature, a convection-diffusion-reactionmodelwas developed for nanoparticle concentration studies. Full-stone et al. [81] used flexible large-scale agent-basedmodeling(FLAM) coupled with computational fluid dynamics (CFD)to study NP distribution in capillaries. In the microvascula-ture, red blood cells aid NP dispersion to the vessel walls andmodify the EPR effect. It was reported that larger NPs (sub-micron size) are more likely to be pushed to the vessel wallsthan smaller NPs, which in turn gives them a greater chanceof permeating through diseased vasculatures to reach thetarget cells [81]. NP adhesion with endothelial cells modeledwith IMEFEM suggests that NP shape affects adhesion, withspherical NPs having a lower lateral drift velocity comparedwith ellipsoids. A NP’s binding probability can be simulatedby randomly assigning initial positions of nanoparticles atthe channel inlet, applying a Brownian adhesion dynamicsmodel, conducting a number of independent trials, calcu-lating the average of number of bonded nanoparticles, andthen normalizing the total number of nanoparticles for acertain depletion layer thickness and shear rate. Nanoparticledeposition and distribution patterns inside the blood vesselnetwork can also be simulated using a continuummodel [12].Saltzman and Radomsky [82] developed a diffusion basedkinetics model to study drug release problems in brain tissue,and their prediction was validated with drug distributionprofiles gained through in vitro experiments.

Pressure, velocity, and blood chemistry will all affect thetransport of NPs and can lead to premature deformation orrelease. An accurate prediction of these parameters is thuscritical to design stable NPs. Factors that affect NP defor-mation and release include blood flow velocity, shear rate,bonding energy, and porosity. It has been shown that NPs canbe modeled using a combination of diffusion, swelling, anderosion [80, 83, 84]. Because of possible swelling and erosion


during circulation, onemust assume a dynamic boundary forthe continuum models during the release process [12]. Finiteelement analysis can be used to solve these models.

4.1.2. Molecular Modeling. Molecular modeling can be usedto understand the size, shape, and charge distribution withina system. It is based on interactions between atoms andmolecules for a fixed time period and uses free energy mini-mization to generate a numerical solution for complex systemproperties. Therefore, it can be used to simulate molecularmotion and elucidate both the uptake process and the influ-ence ofNP characteristics. Depending on howpotential ener-gies are calculated, various molecular modeling methods,including empirical methods, ab initio quantum mechanicalmethods, classic molecular dynamics, and coarse-grainedmolecular dynamics, can be used to understand drug-carrierand carrier-medium interactions [85].

Empirical molecular descriptors that rely heavily onexperimental data fitting can be used to predict physicalproperties, such as drug solubility, and the diffusion coef-ficient, which often leads to complicated artificial param-eters. By deriving the theoretical molecular descriptorsfrom compound chemical structures, limitations imposed byexperimental data availability can be resolved. The quantummechanical ab initiomethod is a useful tool for property cal-culations and higher-level model calibrations by consideringthe electronic degrees of freedom; it uses quantum mechan-ical methods to determine information about electronicbehavior.Thismodel can simulate a few hundred atomswith-out any experimental input and generate information that isunavailable from other methods, such as the electronic state.Using the ab initio quantummechanical method, Adhikari etal. [86] studied the electronic structure, partial surface chargedistribution, and dielectric response of the RGD-4C peptide.Thismethod shed light on the peptide-𝜕v𝛽3 integrin receptorinteraction and proved particularly useful for the study ofbond breaking and formation. Unfortunately, this method isvery complex and computationally intensive to simulate largesystems or to simulate over large time periods.

Classical molecular mechanical simulations are less time-consuming compared with the ab initio quantummechanicalmethods. The potential energy is relatively easy to calculate,so they can be used for larger systems and a longer periodof time. In addition, coarse-grained molecular dynamics areused for large systems and simulations with a time periodlarger than 1ms. Instead of calculating all the atoms in thesystem, subunits are selected for system energy calculations.Model parameters need to be fitted with experimental data,so the accuracy is limited by the availability of data. Loverdeet al. used coarse-grained molecular dynamics (CGMD) tostudy the shape effect of wormlike PEG-PCLmicelles in drugdelivery; it was found that PEG and the PEG-PCl interfaceplay an important role in drug release [87]. CGMD can alsobe used for cellular uptake process simulationswithmultiwallCNT. Dr. Gao et al. have modeled the uptake process withCGMDby assuming immobile ligands anddiffusive receptors[88]. They predicted a critical NP size for endocytosisconsistent with experimental results [88].The studies of YangandMa [89] and Li et al. [90] both show that shape and initial

orientations affect the endocytosis process, which is governedby the bending energy of the cell membrane and the ligand-receptor binding energy. Li et al. [91] also used large-scalecoarse-grained MD to understand how PEGylation affectsNP endocytosis. They found that the repulsive interactionenergy between grafted chains and cell membranes is largerthan the membrane bending energy through endocytosis.Optimal grafting density is also predicted for ligand-receptorinteractions.

4.1.3. Stochastic Modeling. Variability between patients, andeven within the same patient, is one of the hardest factors toaccount for in models. Differences in blood flow, red bloodcells, vasculature, and tumors can all be present, and thusit is crucial to consider uncertainties in model parameters.Simply using average values can lead to unacceptable errorsin the modeling and design of a NP platform. One methodcommonly used to overcome the challenges with parameteruncertainties is the Monte Carlo method [12].

The Monte Carlo method uses random generators fromcertain probability distributions to artificially produce sam-ples repetitively and calculate mean and variance of thesamples. For example, Monte Carlo simulations have beenused to study the interaction between ligand-bound NPswith both tumor cells and healthy cells. Many parametervalues were considered as inputs in order to understandtheir effects [92], and it was found that highly selectivebinding to target cell can be realized with multiple weakaffinity ligands binding with the receptors. In addition, Liuet al. used Monte Carlo simulations to understand the effectof surface functionalization of NPs on its binding to theendothelium. They found that antibody coverage is the keyparameter for the binding process to occur [93]. Recently, Liet al. quantified the uncertainty dispersion coefficient of NPsduring transport through microvasculature using a Bayesiancalibration method [90].

4.1.4. Linking of Multiple Length Scales in Modeling. NP cir-culation, endocytosis, and drug release can each be modeledwith appropriate computational methods mentioned above,but drug delivery takes place across various time and spatialscales. Simulating the entire drug delivery process from drugto patient remains a challenge. Recently, there have beensome reports on multiscale model frameworks shown inFigure 8. These models utilize Monte Carlo methods andatomic calculations such as molecular dynamics to predictphysical parameters, diffusivity, and solubility.The input intothese systems can be drug attributes, delivery system charac-teristics, patient-specific information, and so forth. Patient-specific information includes MRI images, which providethe geometry of the vascular network, family history, andgenetics information. After providing initial concentrationand location of injected NPs, the concentration profile ofNPs in the vascular system can be simulated with immersedFEM analysis. Coarse-grained MD simulations can use theconcentration information with the drug-carrier architectureto simulate the interaction between NPs and target cellmembranes. However, when information is passed from onemodel to another, uncertainties are also passed along. Li


NPs characteristics Patient-specific characteristics Injection concentration/location

Input

Coarse-grained MD andfree energy calculations

Molecular modeling:ab initio quantum mechanics

Size, shape, surface properties

Stochastic modeling: Monte Carlo

Continuum modeling:IMEFEM, FD, LBM

NPs near-cell concentrations

Delivery results Output

Figure 8: A multiscale modeling flow diagram for drug delivery.

et al. provided a good example of how to deal with errorpropagation when connecting models together. By couplingdifferent scales together, a more accurate simulation can bedeveloped leading to better drug delivery optimization [90].

4.2. Challenges and Limitations in Modeling. Computationalmodeling has not been widely used in the pharmaceuti-cal industry for drug delivery design due to its limitedpredictability and accuracy. Most current approaches areagent-based empirical models because mechanism basedmodels are unable to realistically mimic the biological andchemical interactions and delivery processes. To address this,the fundamental mechanisms of each process need to bebetter understood, which could be accomplished by smarterexperimental designs. Eachmodel has limitations;most focuson a particular portion of the drug delivery and are unable tosimulate the entire drug delivery process across all length andtime scales. While linking models of different temporal andspatial scales is challenging, it will provide the greatest benefitin terms of accuracy.

Mechanistic models require input from high qualityexperimental data for parameter estimation, model calibra-tion, validation, and error analysis. One particular challengeis the tracking of the nanoparticles throughout the cells

and tissue, and in vitro data allows for more physiologicallyrealistic simulations to be produced. In addition, theseexperiments need to be repeated to allow for statistical distri-butions. Data across scales is necessary to update the modelsand reduce the uncertainty as errors from one part of thesimulation canpropagate leading to anunacceptable solution.

5. Conclusion and Future Works

NPs have proven themselves to be good candidates fortargeted drug delivery; they are widely available, easily func-tionalized, biocompatible, and stable. These characteristicsallow for concentrated doses of toxic drugs to be encapsulatedand delivered directly to the tumor site. Currently there arethree delivery strategies: passive targeting, which relies onthe EPR effect; active targeting, which uses ligand-receptorinteractions for more selective drug delivery; and next-generation photothermal and magnetic hybrid NPs, whichare time and location controlled based on the heat generatedby the particles. Physical targeting is still at the clinical trialstage.

NPs can self-assemble into many different structures,such as amphiphilic micelles and liposomes, or more exoticgeometries such as rotaxanes, dendrimers, and metal-core


particles. Each of these particles has its own unique specificstrengths and weaknesses making them suitable for deliveryof different therapies to different areas of the body. Nanopar-ticle choice depends on the application in mind—there is notyet a gold standard. Most of the NPs used for drug deliveryare micelles and liposomes, but there is an increasing amountof research into other particles. For example, rotaxanescan release cargo in response to a tailored stimulus, anddendrimers have the ability to maximize their surface areaand cargo capacity compared to spherical structures. Inaddition, metal-core particles are a good choice for physical,physical-active, or physical-passive targeting.

Many nanoparticle characteristicsmust be controlled andconsidered for effective drug delivery to the desired sites,without side effects. If the size, geometry, and/or surfacecharge are not ideal, then serious side effects can arise.Environmental differences can also impact NP drug deliveryefficacy. Before NPs can be deployed clinically, they mustundergo in vitro and in vivo testing to ensure they do notcause harm to living tissue or animals. Due to the complex-ities of the drug delivery process and the large amount ofuncertainty involved, computational modeling has proveduseful for NP design optimization. Currently, computationalmodels are not used widely in the pharmaceutical industrydue to their poor predictability. Most models only focus onspecific processes in drug delivery, not the entire processacross the different scales, that is, the drug to the patient scale.Multiscale modeling across different temporal and spatialscales is rare.

The design of nanoparticles can be guided by multiscalemodeling, but these models would still require input fromprevious experiments and to be experimentally validated.However, multiscale modeling using personalized inputs,such as vascular network images, genome, and family history,needs to be improved and validated in clinical settings. Out-side of clinical settings, several computational models havebeen validated through experiments to allow for the design ofmore efficient nanocarriers. For example, Shi et al. [94] usedmolecular docking and MD simulations to screen buildingblocks for nanocarrier synthesis, which were later validatedthrough experimental synthesis and evaluation. In addition,to understand the role of building blocks in telodendrimerself-assembly, Jiang et al. [95] combined multiscale modelingwith experiments. The multiscale simulations run by Li et al.investigated the effect of size, geometry, surface functionality,and stiffness on NP transport through tumor microvas-culature [91, 96]. Collaborative work between experimentsand multiscale computation modeling is the direction forfuture drug delivery researches. Microchip development hasallowed for the simulation of tumor microenvironments andsubsequent testing of nanoparticle efficacy for targeted drugdelivery [97, 98]. Computational models facilitate the designof these devices, further advancing NP development.

Competing Interests

The authors declare that there are no competing interestsregarding the publication of this paper.

Authors’ Contributions

Xiaojiao Yu and Ian Trase contributed equally to this work.

Acknowledgments

Zi Chen acknowledges the support from the Society inScience–The BrancoWeiss Fellowship, administered by ETHZurich. The research reported in this paper was in partsupported by the National Cancer Institute of the NationalInstitutes of Health under Award no. U01CA202123. Theauthors thank S. J. Huang for comments.

References

[1] N. A. Ochekpe, P. O. Olorunfemi, and N. C. Ngwuluka,“Nanotechnology and drug delivery—part 1: background andapplications,” Tropical Journal of Pharmaceutical Research, vol.8, no. 3, pp. 265–274, 2009.

[2] G. Poste and R. Kirsh, “Site-specific (targeted) drug delivery incancer therapy,” Bio/Technology, vol. 1, no. 10, pp. 869–878, 1983.

[3] G.M.Whitesides, J. K. Kriebel, and B. T.Mayers, “Self-assemblyand nanostructured materials,” in Nanoscale Assembly: Chemi-cal Techniques, Nanostructure Science and Technology, pp. 217–239, Springer, Berlin, Germany, 2005.

[4] S. Svenson and D. A. Tomalia, “Dendrimers in biomedicalapplications—reflections on the field,” Advanced Drug DeliveryReviews, vol. 57, no. 15, pp. 2106–2129, 2005.

[5] E. Igarashi, “Factors affecting toxicity and efficacy of polymericnanomedicines,”Toxicology andApplied Pharmacology, vol. 229,no. 1, pp. 121–134, 2008.

[6] A. Mahmud, X.-B. Xiong, H. M. Aliabadi, and A. Lavasanifar,“Polymeric micelles for drug targeting,” Journal of Drug Target-ing, vol. 15, no. 9, pp. 553–584, 2007.

[7] V. P. Torchilin, “Targeted polymeric micelles for delivery ofpoorly soluble drugs,” Cellular and Molecular Life Sciences, vol.61, no. 19-20, pp. 2549–2559, 2004.

[8] H. Maeda, T. Sawa, and T. Konno, “Mechanism of tumor-targeted delivery of macromolecular drugs, including the EPReffect in solid tumor and clinical overview of the prototypepolymeric drug SMANCS,” Journal of Controlled Release, vol.74, no. 1–3, pp. 47–61, 2001.

[9] F. Yuan, M. Dellian, D. Fukumura et al., “Vascular permeabilityin a human tumor xenograft: molecular size dependence andcutoff size,” Cancer Research, vol. 55, no. 17, pp. 3752–3756, 1995.

[10] N. Bertrand, J. Wu, X. Xu, N. Kamaly, and O. C. Farokhzad,“Cancer nanotechnology: the impact of passive and activetargeting in the era of modern cancer biology,” Advanced DrugDelivery Reviews, vol. 66, pp. 2–25, 2014.

[11] E. Ruoslahti, “Specialization of tumour vasculature,” NatureReviews Cancer, vol. 2, no. 2, pp. 83–90, 2002.

[12] N. Haddish-Berhane, J. L. Rickus, and K. Haghighi, “The roleof multiscale computational approaches for rational designof conventional and nanoparticle oral drug delivery systems,”International Journal of Nanomedicine, vol. 2, no. 3, pp. 315–331,2007.

[13] L. Zhang, T. Su, B. He, and Z. Gu, “Self-assembly polyrotaxanesnanoparticles as carriers for anticancer drug methotrexatedelivery,” Nano-Micro Letters, vol. 6, no. 2, pp. 108–115, 2014.


[14] Y. Wang, K. Wang, J. Zhao et al., “Multifunctional meso-porous silica-coated graphene nanosheet used for chemo-photothermal synergistic targeted therapy of glioma,” Journal ofthe American Chemical Society, vol. 135, no. 12, pp. 4799–4804,2013.

[15] Y. Li, K. Xiao, J. Luo et al., “Well-defined, reversible disulfidecross-linked micelles for on-demand paclitaxel delivery,” Bio-materials, vol. 32, no. 27, pp. 6633–6645, 2011.

[16] Y. Li, W. Xiao, K. Xiao et al., “Well-defined, reversible boronatecrosslinked nanocarriers for targeted drug delivery in responseto acidic pH values and cis-diols,” Angewandte Chemie—International Edition, vol. 124, no. 12, pp. 2918–2923, 2012.

[17] T. Aida, E. W. Meijer, and S. I. Stupp, “Functional supramolec-ular polymers,” Science, vol. 335, no. 6070, pp. 813–817, 2012.

[18] X. Xia, M. Yang, Y. Wang et al., “Quantifying the coveragedensity of poly(ethylene glycol) chains on the surface of goldnanostructures,” ACS Nano, vol. 6, no. 1, pp. 512–522, 2012.

[19] B. Tang, S. Xu, J. An, B. Zhao, W. Xu, and J. R. Lombardi,“Kinetic effects of halide ions on the morphological evolutionof silver nanoplates,” Physical Chemistry Chemical Physics, vol.11, no. 44, pp. 10286–10292, 2009.

[20] Y. Li, M. Kroger, and W. K. Liu, “Shape effect in cellular uptakeof PEGylated nanoparticles: comparison between sphere, rod,cube and disk,” Nanoscale, vol. 7, no. 40, pp. 16631–16646, 2015.

[21] G. S. Kwon,M. Yokoyama, T.Okano, Y. Sakurai, andK. Kataoka,“Biodistribution of micelle-forming polymer-drug conjugates,”Pharmaceutical Research, vol. 10, no. 7, pp. 970–974, 1993.

[22] K. Na, T. B. Lee, K.-H. Park, E.-K. Shin, Y.-B. Lee, and H.-K. Choi, “Self-assembled nanoparticles of hydrophobically-modified polysaccharide bearing vitamin H as a targeted anti-cancer drug delivery system,” European Journal of Pharmaceu-tical Sciences, vol. 18, no. 2, pp. 165–173, 2003.

[23] Y. Bae, S. Fukushima, A. Harada, and K. Kataoka, “Design ofenvironment-sensitive supramolecular assemblies for intracel-lular drug delivery: polymeric micelles that are responsive tointracellular pH change,” Angewandte Chemie—InternationalEdition, vol. 42, no. 38, pp. 4640–4643, 2003.

[24] G. Li, S. Song, L. Guo, and S. Ma, “Self-assembly ofthermo- and pH-responsive poly(acrylic acid)-b-poly(N-isopropylacrylamide) micelles for drug delivery,” Journal ofPolymer Science, Part A: Polymer Chemistry, vol. 46, no. 15, pp.5028–5035, 2008.

[25] A. Rahikkala, V. Aseyev, H. Tenhu, E. I. Kauppinen, andJ. Raula, “Thermoresponsive nanoparticles of self-assembledblock copolymers as potential carriers for drug delivery anddiagnostics,” Biomacromolecules, vol. 16, no. 9, pp. 2750–2756,2015.

[26] Y. Wang, S. Gao, W.-H. Ye, H. S. Yoon, and Y.-Y. Yang, “Co-delivery of drugs and DNA from cationic core-shell nanopar-ticles self-assembled from a biodegradable copolymer,” NatureMaterials, vol. 5, no. 10, pp. 791–796, 2006.

[27] L. Zhang, J. M. Chan, F. X. Gu et al., “Self-assembled lipid-polymer hybrid nanoparticles: a robust drug delivery platform,”ACS Nano, vol. 2, no. 8, pp. 1696–1702, 2008.

[28] B. Karagoz, L. Esser, H. T. Duong, J. S. Basuki, C. Boyer, andT. P. Davis, “Polymerization-Induced Self-Assembly (PISA)—control over the morphology of nanoparticles for drug deliveryapplications,”PolymerChemistry, vol. 5, no. 2, pp. 350–355, 2014.

[29] F. Huang, M. You, T. Chen, G. Zhu, H. Liang, andW. Tan, “Self-assembled hybrid nanoparticles for targeted co-delivery of twodrugs into cancer cells,” Chemical Communications, vol. 50, no.23, pp. 3103–3105, 2014.

[30] J. Li, X. Qu, G. F. Payne et al., “Biospecific self-assembly of ananoparticle coating for targeted and stimuli-responsive drugdelivery,” Advanced Functional Materials, vol. 25, no. 9, pp.1404–1417, 2015.

[31] X.-X. Xia,M.Wang, Y. Lin,Q.Xu, andD. L.Kaplan, “Hydropho-bic drug-triggered self-assembly of nanoparticles from silk-elastin-like protein polymers for drug delivery,” Biomacro-molecules, vol. 15, no. 3, pp. 908–914, 2014.

[32] A. K. Patri, A. Myc, J. Beals, T. P. Thomas, N. H. Bander, andJ. R. Baker Jr., “Synthesis and in vitro testing of J591 antibody-dendrimer conjugates for targeted prostate cancer therapy,”Bioconjugate Chemistry, vol. 15, no. 6, pp. 1174–1181, 2004.

[33] R. Liu, Y. Lai, B. He et al., “Supramolecular nanoparticlesgenerated by the self-assembly of polyrotaxanes for antitumordrug delivery,” International Journal of Nanomedicine, vol. 7, pp.5249–5258, 2012.

[34] A. Sahu, W. I. Choi, J. H. Lee, and G. Tae, “Graphene oxidemediated delivery of methylene blue for combined photody-namic and photothermal therapy,” Biomaterials, vol. 34, no. 26,pp. 6239–6248, 2013.

[35] A. C. Estrada, A. L. Daniel-da-Silva, and T. Trindade, “Pho-tothermally enhanced drug release by 𝜅-carrageenan hydrogelsreinforcedwithmulti-walled carbonnanotubes,”RSCAdvances,vol. 3, no. 27, pp. 10828–10836, 2013.

[36] H. Zhang, C. Chen, L. Hou et al., “Targeting and hyperthermiaof doxorubicin by the delivery of single-walled carbon nan-otubes to EC-109 cells,” Journal of Drug Targeting, vol. 21, no.3, pp. 312–319, 2013.

[37] B. Thiesen and A. Jordan, “Clinical applications of magneticnanoparticles for hyperthermia,” International Journal of Hyper-thermia, vol. 24, no. 6, pp. 467–474, 2008.

[38] R. Mo, Q. Sun, N. Li, and C. Zhang, “Intracellular deliveryand antitumor effects of pH-sensitive liposomes based onzwitterionic oligopeptide lipids,”Biomaterials, vol. 34, no. 11, pp.2773–2786, 2013.

[39] N. Desai, “Challenges in development of nanoparticle-basedtherapeutics,” AAPS Journal, vol. 14, no. 2, pp. 282–295, 2012.

[40] A. G. Tzakos, E. Briasoulis, T. Thalhammer, W. Jager, and V.Apostolopoulos, “Novel oncology therapeutics: targeted drugdelivery for cancer,” Journal of Drug Delivery, vol. 2013, ArticleID 918304, 5 pages, 2013.

[41] G. D. Grossfeld, P. R. Carroll, N. Lindeman et al.,“Thrombospondin-1 expression in patients with pathologicstage T3 prostate cancer undergoing radical prostatectomy:association with p53 alterations, tumor angiogenesis, andtumor progression,” Urology, vol. 59, no. 1, pp. 97–102, 2002.

[42] A. Jones and A. L. Harris, “New developments in angiogenesis:a major mechanism for tumor growth and target for therapy,”Cancer Journal from Scientific American, vol. 4, no. 4, pp. 209–217, 1998.

[43] J. I. N. Zhang, C.Q. Lan,M. Post, B. Simard, Y. Deslandes, andT.H. Hsieh, “Design of nanoparticles as drug carriers for cancertherapy,” Cancer Genomics and Proteomics, vol. 3, no. 3-4, pp.147–157, 2006.

[44] G. Poste, “Experimental systems for analysis of the malignantphenotype,” Cancer and Metastasis Review, vol. 1, no. 2, pp. 141–199, 1982.

[45] G. Poste and I. J. Fidler, “The pathogenesis of cancermetastasis,”Nature, vol. 283, no. 5743, pp. 139–146, 1980.

[46] Y. Matsumura and H. Maeda, “A new concept for macro-molecular therapeutics in cancer chemotherapy: mechanism of


tumoritropic accumulation of proteins and the antitumor agentSmancs,” Cancer Research, vol. 46, no. 12, pp. 6387–6392, 1986.

[47] L. Y. T. Chou, K. Ming, and W. C. W. Chan, “Strategies forthe intracellular delivery of nanoparticles,” Chemical SocietyReviews, vol. 40, no. 1, pp. 233–245, 2011.

[48] A. Albanese, P. S. Tang, and W. C. W. Chan, “The effect ofnanoparticle size, shape, and surface chemistry on biologicalsystems,” Annual Review of Biomedical Engineering, vol. 14, pp.1–16, 2012.

[49] C. He, Y. Hu, L. Yin, C. Tang, and C. Yin, “Effects of particlesize and surface charge on cellular uptake and biodistribution ofpolymeric nanoparticles,” Biomaterials, vol. 31, no. 13, pp. 3657–3666, 2010.

[50] F. Alexis, E. Pridgen, L. K. Molnar, and O. C. Farokhzad, “Fac-tors affecting the clearance and biodistribution of polymericnanoparticles,”Molecular Pharmaceutics, vol. 5, no. 4, pp. 505–515, 2008.

[51] Y. Li, M. Kroger, and W. K. Liu, “Endocytosis of PEGy-lated nanoparticles accompanied by structural and free energychanges of the grafted polyethylene glycol,”Biomaterials, vol. 35,no. 30, pp. 8467–8478, 2014.

[52] Y.H. Bae andK. Park, “Targeted drug delivery to tumors:myths,reality and possibility,” Journal of Controlled Release, vol. 153, no.3, pp. 198–205, 2011.

[53] D. Oupicky,M. Ogris, K. A. Howard, P. R. Dash, K. Ulbrich, andL. W. Seymour, “Importance of lateral and steric stabilizationof polyelectrolyte gene delivery vectors for extended systemiccirculation,”MolecularTherapy, vol. 5, no. 4, pp. 463–472, 2002.

[54] X. Guo, C. Shi, G. Yang, J. Wang, Z. Cai, and S. Zhou, “Dual-responsive polymer micelles for target-cell-specific anticancerdrug delivery,” Chemistry of Materials, vol. 26, no. 15, pp. 4405–4418, 2014.

[55] J. Kreuter, “Nanoparticles-a historical perspective,” Interna-tional Journal of Pharmaceutics, vol. 331, no. 1, pp. 1–10, 2007.

[56] R. N. Saha, S. Vasanthakumar, G. Bende, and M. Snehalatha,“Nanoparticulate drug delivery systems for cancer chemother-apy,” Molecular Membrane Biology, vol. 27, no. 7, pp. 215–231,2010.

[57] S. Ganta, H. Devalapally, A. Shahiwala, andM. Amiji, “A reviewof stimuli-responsive nanocarriers for drug and gene delivery,”Journal of Controlled Release, vol. 126, no. 3, pp. 187–204, 2008.

[58] A. A. Kale and V. P. Torchilin, “Environment-responsive multi-functional liposomes,” Methods in Molecular Biology, vol. 605,pp. 213–242, 2010.

[59] K. T. Oh, H. Yin, E. S. Lee, and Y. H. Bae, “Polymeric nanovehi-cles for anticancer drugs with triggering release mechanisms,”Journal of Materials Chemistry, vol. 17, no. 38, pp. 3987–4001,2007.

[60] M. Zorko and U. Langel, “Cell-penetrating peptides: mecha-nism and kinetics of cargo delivery,” Advanced Drug DeliveryReviews, vol. 57, no. 4, pp. 529–545, 2005.

[61] E. Ruoslahti, S. N. Bhatia, and M. J. Sailor, “Targeting of drugsand nanoparticles to tumors,” Journal of Cell Biology, vol. 188,no. 6, pp. 759–768, 2010.

[62] S. Christian, J. Pilch, M. E. Akerman, K. Porkka, P. Laakkonen,and E. Ruoslahti, “Nucleolin expressed at the cell surface is amarker of endothelial cells in angiogenic blood vessels,” TheJournal of Cell Biology, vol. 163, no. 4, pp. 871–878, 2003.

[63] P. Oh, Y. Li, J. Yu et al., “Subtractive proteomic mapping of theendothelial surface in lung and solid tumours for tissue-specifictherapy,” Nature, vol. 429, no. 6992, pp. 629–635, 2004.

[64] K. A. Kelly, N. Bardeesy, R. Anbazhagan et al., “Targetednanoparticles for imaging incipient pancreatic ductal adenocar-cinoma,” PLoS Medicine, vol. 5, no. 4, article e85, 2008.

[65] V. Fogal, L. Zhang, S. Krajewski, and E. Ruoslahti,“Mitochondrial/cell-surface protein p32/gC1qR as a moleculartarget in tumor cells and tumor stroma,” Cancer Research, vol.68, no. 17, pp. 7210–7218, 2008.

[66] P. C. Brooks, R. A. F. Clark, and D. A. Cheresh, “Requirement ofvascular integrin alpha v beta 3 for angiogenesis,” Science, vol.264, no. 5158, pp. 569–571, 1994.

[67] A. Erdreich-Epstein, H. Shimada, S. Groshen et al., “Integrins𝛼(v)𝛽3 and 𝛼(v)𝛽5 are expressed by endothelium of high—riskneuroblastoma and their inhibition is associated with increasedendogenous ceramide,” Cancer Research, vol. 60, no. 3, pp. 712–721, 2000.

[68] J. S. Desgrosellier and D. A. Cheresh, “Integrins in cancer:biological implications and therapeutic opportunities,” NatureReviews Cancer, vol. 10, no. 1, pp. 9–22, 2010.

[69] F.Nilsson,H. Kosmehl, L. Zardi, andD.Neri, “Targeted deliveryof tissue factor to the ED-B domain of fibronectin, a marker ofangiogenesis, mediates the infarction of solid tumors in mice,”Cancer Research, vol. 61, no. 2, pp. 711–716, 2001.

[70] E. B. Carson-Walter, D. N. Watkins, A. Nanda, B. Vogelstein,K. W. Kinzler, and B. St Croix, “Cell surface tumor endothelialmarkers are conserved in mice and humans,” Cancer Research,vol. 61, no. 18, pp. 6649–6655, 2001.

[71] A. A. Gabizon, “Pegylated liposomal doxorubicin: metamor-phosis of an old drug into a new form of chemotherapy,” CancerInvestigation, vol. 19, no. 4, pp. 424–436, 2001.

[72] P. S. Low,W.A.Henne, andD.D. Doorneweerd, “Discovery anddevelopment of folic-acid-based receptor targeting for imagingand therapy of cancer and inflammatory diseases,” Accounts ofChemical Research, vol. 41, no. 1, pp. 120–129, 2008.

[73] Z. Cheng, A. Al Zaki, J. Z. Hui, V. R. Muzykantov, and A.Tsourkas, “Multifunctional nanoparticles: cost versus benefit ofadding targeting and imaging capabilities,” Science, vol. 338, no.6109, pp. 903–910, 2012.

[74] X. Huang, P. K. Jain, I. H. El-Sayed, and M. A. El-Sayed, “Plas-monic photothermal therapy (PPTT) using gold nanoparticles,”Lasers in Medical Science, vol. 23, no. 3, pp. 217–228, 2008.

[75] A. S. Sreedhar and P. Csermely, “Heat shock proteins in theregulation of apoptosis: new strategies in tumor therapy: acomprehensive review,” Pharmacology & Therapeutics, vol. 101,no. 3, pp. 227–257, 2004.

[76] D. R. Stacy, B. Lu, and D. E. Hallahan, “Radiation-guided drugdelivery systems,” Expert Review of Anticancer Therapy, vol. 4,no. 2, 2014.

[77] S. C. McBain, H. H. P. Yiu, and J. Dobson, “Magnetic nanopar-ticles for gene and drug delivery,” International Journal ofNanomedicine, vol. 3, no. 2, pp. 169–180, 2008.

[78] N. Schleich, C. Po, D. Jacobs et al., “Comparison of active,passive and magnetic targeting to tumors of multifunctionalpaclitaxel/SPIO-loaded nanoparticles for tumor imaging andtherapy,” Journal of Controlled Release, vol. 194, pp. 82–91, 2014.

[79] S. R. McDougall, A. R. A. Anderson, M. A. J. Chaplain, and J.A. Sherratt, “Mathematical modelling of flow through vascularnetworks: implications for tumour-induced angiogenesis andchemotherapy strategies,” Bulletin of Mathematical Biology, vol.64, no. 4, pp. 673–702, 2002.

[80] J. Siepmann and N. A. Peppas, “Mathematical modeling ofcontrolled drug delivery,” Advanced Drug Delivery Reviews, vol.48, no. 2-3, pp. 137–138, 2001.


[81] G. Fullstone, J. Wood, M. Holcombe, and G. Battaglia, “Mod-elling the transport of nanoparticles under blood flow usingan agent-based approach,” Scientific Reports, vol. 5, Article ID10649, 2015.

[82] W. M. Saltzman and M. L. Radomsky, “Drugs released frompolymers: diffusion and elimination in brain tissue,” ChemicalEngineering Science, vol. 46, no. 10, pp. 2429–2444, 1991.

[83] R. Langer and N. Peppas, “Chemical and physical structure ofpolymers as carriers for controlled release of bioactive agents: areview,” Journal of Macromolecular Science, Part C, vol. 23, pp.61–126, 2006.

[84] J. Siepmann and A. Gopferich, “Mathematical modeling ofbioerodible, polymeric drug delivery systems,” Advanced DrugDelivery Reviews, vol. 48, no. 2-3, pp. 229–247, 2001.

[85] D. Neumann, C.-M. Lehr, H.-P. Lenhof, and O. Kohlbacher,“Computational modeling of the sugar-lectin interaction,”Advanced Drug Delivery Reviews, vol. 56, no. 4, pp. 437–457,2004.

[86] P. Adhikari, A. M. Wen, R. H. French et al., “Electronicstructure, dielectric response, and surface charge distributionof RGD (1FUV) peptide,” Scientific Reports, vol. 4, article 5605,2014.

[87] S. M. Loverde, M. L. Klein, and D. E. Discher, “Nanoparti-cle shape improves delivery: rational coarse grain moleculardynamics (rCG-MD) of taxol in worm-like PEG-PCLmicelles,”Advanced Materials, vol. 24, no. 28, pp. 3823–3830, 2012.

[88] H. Gao, W. Shi, and L. B. Freund, “Mechanics of receptor-mediated endocytosis,” Proceedings of the National Academy ofSciences of the United States of America, vol. 102, no. 27, pp.9469–9474, 2005.

[89] K. Yang and Y.-Q. Ma, “Computer simulation of the transloca-tion of nanoparticles with different shapes across a lipid bilayer,”Nature Nanotechnology, vol. 5, no. 8, pp. 579–583, 2010.

[90] Y. Li, W. Stroberg, T.-R. Lee et al., “Multiscale modeling anduncertainty quantification in nanoparticle-mediated drug/genedelivery,” Computational Mechanics, vol. 53, no. 3, pp. 511–537,2014.

[91] Y. Li, Y. Lian, L. T. Zhang et al., “Cell and nanoparticle transportin tumour microvasculature: the role of size, shape and surfacefunctionality of nanoparticles,” Interface Focus, vol. 6, no. 1,Article ID 20150086, 2016.

[92] G. A. Duncan and M. A. Bevan, “Computational design ofnanoparticle drug delivery systems for selective targeting,”Nanoscale, vol. 7, no. 37, pp. 15332–15340, 2015.

[93] J. Liu, G. E. R. Weller, B. Zern et al., “Computational model fornanocarrier binding to endothelium validated using in vivo, invitro, and atomic force microscopy experiments,” Proceedings ofthe National Academy of Sciences of the United States of America,vol. 107, no. 38, pp. 16530–16535, 2010.

[94] C. Shi, D. Guo, K. Xiao, X. Wang, L. Wang, and J. Luo, “A drug-specific nanocarrier design for efficient anticancer therapy,”Nature Communications, vol. 6, article 7449, 2015.

[95] W. Jiang, J. Luo, and S. Nangia, “Multiscale approach toinvestigate self-assembly of telodendrimer based nanocarriersfor anticancer drug delivery,” Langmuir, vol. 31, no. 14, pp. 4270–4280, 2015.

[96] Z. Shen,M.-P. Nieh, and Y. Li, “Decorating nanoparticle surfacefor targeted drug delivery: opportunities and challenges,” Poly-mers, vol. 8, no. 3, article 83, 2016.

[97] G. Bao, Y. Bazilevs, J. Chung et al., “USNCTAMperspectives onmechanics in medicine,” Journal of The Royal Society Interface,vol. 11, no. 97, Article ID 20140301, 2014.

[98] N. S. Bhise, J. Ribas, V. Manoharan et al., “Organ-on-a-chip platforms for studying drug delivery systems,” Journal ofControlled Release, vol. 190, pp. 82–93, 2014.

Submit your manuscripts athttp://www.hindawi.com

ScientificaHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

CorrosionInternational Journal of

Hindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Polymer ScienceInternational Journal of



CeramicsJournal of


CompositesJournal of

NanoparticlesJournal of



International Journal of

Biomaterials


NanoscienceJournal of

TextilesHindawi Publishing Corporation http://www.hindawi.com Volume 2014

Journal of

NanotechnologyHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Journal of

CrystallographyJournal of


The Scientific World JournalHindawi Publishing Corporation http://www.hindawi.com Volume 2014


CoatingsJournal of

Advances in

Materials Science and EngineeringHindawi Publishing Corporationhttp://www.hindawi.com Volume 2014

Smart Materials Research



MetallurgyJournal of


BioMed Research International

MaterialsJournal of


Nano

materials


Journal ofNanomaterials

review article design of nanoparticle-based carriers for...

Documents