Therapeutic N 17. Therapeutic Nanodevices · Therapeutic N 17. Therapeutic Nanodevices Therapeutic nanotechnology offers minimally invasive therapies with high densities of func-tion

279

Therapeutic N17. Therapeutic Nanodevices

Therapeutic nanotechnology offers minimallyinvasive therapies with high densities of func-tion concentrated in small volumes, featuresthat may reduce patient morbidity and mortal-ity. Unlike other areas of nanotechnology, novelphysical properties associated with nanoscaledimensionality are not the raison d’etre of thera-peutic nanotechnology, whereas the aggregationof multiple biochemical (or comparably pre-cise) functions into controlled nanoarchitecturesis. Multifunctionality is a hallmark of emergingnanotherapeutic devices, and multifunctional-ity can allow nanotherapeutic devices to performmulti-step work processes, with each func-tional component contributing to one or morenanodevice subroutine such that, in aggregate,subroutines sum to a cogent work process. Can-nonical nanotherapeutic subroutines includetethering (targeting) to sites of disease, dis-pensing measured doses of drug (or bioactivecompound), detection of residual disease af-ter therapy and communication with an externalclinician/operator. Emerging nanotherapeuticsthus blur the boundaries between medical de-vices and traditional pharmaceuticals. Assemblyof therapeutic nanodevices generally exploits ei-ther (bio)material self assembly properties orchemoselective bioconjugation techniques, orboth. Given the complexity, composition, and thenecessity for their tight chemical and structuraldefinition inherent in the nature of nanothera-peutics, their cost of goods (COGs) might exceedthat of (already expensive) biologics. Early thera-peutic nanodevices will likely be applied to diseasestates which exhibit significant unmet patientneed (cancer and cardiovascular disease), whileapplication to other disease states well-servedby conventional therapy may await perfec-tion of nanotherapeutic design and assemblyprotocols.

17.1 Definitions and Scope of Discussion........ 28017.1.1 Design Issues ............................... 28117.1.2 Utility and Scope

of Therapeutic Nanodevices ........... 285

17.2 Synthetic Approaches: “top-down”versus “bottom-up” Approachesfor Nanotherapeutic Device Components 28517.2.1 Production of Nanoporous

Membranes by MicrofabricationMethods: A top-down Approach .... 285

17.2.2 Synthesis of Poly(amido) Amine(PAMAM) Dendrimers:A bottom-up Approach ................. 286

17.2.3 The Limits of top-downand bottom-up Distinctionswith Respect to Nanomaterialsand Nanodevices ......................... 287

17.3 Technological and BiologicalOpportunities ....................................... 28817.3.1 Assembly Approaches ................... 28817.3.2 Targeting: Delimiting

Nanotherapeutic Actionin Three-Dimensional Space .......... 296

17.3.3 Triggering: DelimitingNanotherapeutic Actionin Space and Time ........................ 298

17.3.4 Sensing Modalities ....................... 30217.3.5 Imaging Using Nanotherapeutic

Contrast Agents ............................ 304

17.4 Applications for Nanotherapeutic Devices 30717.4.1 Nanotherapeutic Devices

in Oncology ................................. 30717.4.2 Cardiovascular Applications

of Nanotherapeutics ..................... 31017.4.3 Nanotherapeutics and Specific

Host Immune Responses ............... 311

17.5 Concluding Remarks:Barriers to Practice and Prospects .......... 31517.5.1 Complexity in Biology ................... 31517.5.2 Dissemination

of Biological Information .............. 31517.5.3 Cultural Differences

Between Technologistsand Biologists .............................. 316

References .................................................. 317

PartB

17

280 Part B Nanostructures, Micro/Nanofabrication, and Micro/Nanodevices

17.1 Definitions and Scope of Discussion

Nanotechnology is a field in rapid flux and development,as cursory examination of this volume shows, and defini-tion of its meets and bounds, as well as identification ofsub-disciplines embraced by it, can be elusive. The wordmeans many things to many people, and aspects of mul-tiple disciplines, from physics to information technologyto biotechnology, legitimately fall into the intersectionof the Venn diagram of disciplines that defines nano-technology. The breadth of the field allows almost anyinterested party to contribute to it, but the same am-biguity can render the field diffuse and amorphous. Ifnanotechnology embraces everything, what then is it?The ambiguity fuels cognitive dissonance that can re-sult in frustrating interactions between investigators andfunders, authors and editors, and entrepreneurs and in-

N

Antigen binding domains

Commoneffectordomains

NN

LC

HCC

C C

C

NN

NN

LC

HCC

C C

C

NN

C

a) b) c)

Fig. 17.1a–c Antibodies resemble purpose-built devices with dis-tinct functional domains [17.1]. Native antibodies are composed offour polypeptide chains: two heavy chains (HC) and two light chains(LC), joined by interchain disulfide linkages (lines between HC andLC moieties). Amino and carboxy termini of individual polypep-tide chains are indicated (by N and C). Antigen binding domainsare responsible for specific antigen recognition, vary from antibodyto antibody, and are indicated by the thicker lines. Common effec-tor functions (FC receptor binding, complement fixation, etc.) aredelimited to domains of the antibodies that are constant from mol-ecule to molecule. (a) A native IgG antibody is monospecific butbivalent in its antigen binding capacity. (b) An engineered, bispe-cific, bivalent antibody capable of recognizing two distinct antigens.(c) An engineered antibody fragment (single chain Fv or SCFv) thatis monospecific and monovalent can recognize only one antigenicdeterminant and is engineered to lack common effector functions.This construct is translated as a single, continuous polypeptide chain(hence the name SCFv) because a peptide linker (indicated by theconnecting line in the figure) is incorporated to connect the carboxyend of the HC fragment and the amino end of the LC fragment

vestors. Some consideration of the scope of the fieldtherefore is useful.

To frame the discussion, we will define nanotech-nology as the discipline that aims to satisfy desiredobjectives using materials and devices whose valu-able properties are based on a specific nanometer-scaleelement of their structures. The field is unabashedlyapplication-oriented, so its raison d’être is fulfillmentof tasks of interest: technical information is importantprimarily to the extent that it bears on device design,function, or application.

The meaning of “therapeutic” is largely self-explanatory and refers here to intervention in humandisease processes (although many of the approaches dis-cussed are equally applicable to veterinary medicine).Our discussion will be confined primarily to therapeuticsused in vivo, because such applications clearly benefitfrom the low invasiveness that ultra-small, but multipo-tent, nanotherapeutics potentially offer. It is debatablewhether imaging, diagnostic, or sensing devices can beconsidered therapeutic in this context, though, as wewill see, sensing/diagnostic functionalities are often in-extricable elements of therapeutic nanodevices, and itis difficult to consider so-called smart nanotherapeuticswithout discussion of their sensing capabilities.

Our definition of nanotechnology projects severalcorollaries. First, it embraces macroscale structureswhose useful properties derive from their nanoscaleaspects. Second, the modifier “specific” (as in, “spe-cific nanometer-scale elements”) is intended to excludematerials whose utility derives solely from properties in-herent in being finely divided (high surface-to-volumeratios, for instance), or other bulk chemical and phys-ical properties. We made the exclusion based on ourassessment that therapeutic nanodevices are more in-triguing than nanomaterials per se (see below), thoughwe will engage these attributes where they are germaneto specific devices or therapeutic applications. Third,our definition implies that limited nanotechnology hasbeen available since the 1970s in the form of biotech-nology. Based on their nanoscale structures, individualbiological macromolecules (such as proteins) often ex-hibit the coordinated, modular multifunctionality that ischaracteristic of purpose-built devices (Fig. 17.1). Ananalogous, but perhaps less persuasive, argument canbe made that organic chemistry is an early form ofnanotechnology. Compared to organic small molecules,protein functional capabilities and properties are gen-erally more complex and extremely dependent on their

PartB

17.1

Therapeutic Nanodevices 17.1 Definitions and Scope of Discussion 281

conformation in three-dimensional space at nanometerscale. The nanotechnology sobriquet, therefore, maybe more appropriate to biotechnology than organicchemistry.

Biological macromolecules rely on the deploymentof specific chemical functionalities to specific relativedistributions in space with nanometer (and greater) res-olution for their function, so the inclusion of molecularengineering aspects of biotechnology practice under thenanotechnology rubric is legitimate, despite the discom-fort it may cause traditionally trained engineers. Aswe will see, intervention in human disease often re-quires inclusion of biomolecules in therapeutic devices:frequently, no functional synthetic analogue of activeproteins and nucleic acids is available.

As described above, this chapter focuses primarilyon nanoscale therapeutic devices as opposed to thera-peutic nanomaterials. Devices are integrated functionalstructures and not mixtures of materials. Devices exhibitdesirable emergent properties inherent in their design:the properties emerge as the result of the spatial and/ortemporal organization, and coordination and regulationof action of individual components. The organization ofcomponents in devices allows them to perform multi-step, cogent work processes that can’t be mimicked bysimple admixtures of individual components. In fact, ifdevice functions can be mimicked well by simple mix-tures of components, the labor involved in configuringand constructing a nanoscale device is not warranted.Our device definition thus excludes nanomaterials usedas drug formulation excipients (pharmacologically inertmaterials included in formulations that improve phar-macophore uptake, biodistribution, pharmacokinetic,handling, storage, or other properties), but embracesthose same materials as integral components of drugdelivery or other clinical devices.

17.1.1 Design Issues

The biotechnology industry historically has focusedon production of individual soluble protein and nu-cleic acid molecules for pharmaceutical use, with onlylimited attention paid to functional supramolecularstructures [17.2–7]. This bias toward free moleculesflies in the face of the obvious importance of integratedsupramolecular structures in biology and, to the casualobserver, may seem an odd gap in attention and empha-sis on the part of practicing biotechnologists. The biastoward single molecule, protein therapeutics, however,follows from the fact that biotechnology is an indus-trial activity, governed by market considerations. Of the

myriad potential therapeutics that might be realized frombiotechnology, single protein therapeutics are among theeasiest to realize from both technical and regulatory per-spectives and so warrant extensive industrial attention.This is changing, however, and more complex entities(actual supramolecular therapeutic devices) have andwill appear with increasing frequency in the twenty-firstcentury.

New top-down and bottom-up materials derivedfrom micro/nanotechnology provide the opportunityto complement the traditional limits of biotechnol-ogy by providing scaffolds that can support higherlevel organization of multiple biomolecules to performwork activities they could not perform as free, solublemolecules. Such supramolecular structures have beencalled nanobiotechnological devices [17.8], nanobiolog-ical devices [17.2–7], or semi-synthetic nanodevices andfigure prominently in therapeutic nanotechnology.

Incorporation of / Interactionwith Biomolecules

In general, design of nanodevices is similar to design ofother engineered structures, providing that the specialproperties of the materials (relating to their nanoscale as-pects such as quantum, electrical, mechanical, biologicalproperties, etc.), as well as their impact in therapy, areconsidered. Therapeutics can interact with patients onmultiple levels, ranging from organismal to molecular,but it is reasonable to expect that most nanotherapeu-tics will interface with patients at the nanoscale at leastto some extent [17.2–5, 9–14]. Typically, this meansinteraction between therapeutics and biological macro-molecules, supramolecular structures and organelles,which, in turn, often dictates the incorporation of bio-logical macromolecules (and other biostructures) intonanodevices [17.2, 5, 13–15]. Incorporating biologicalstructures into (nanobiological) devices presents spe-cial challenges that do not occur in other aspects ofengineering practice.

Unlike fully synthetic devices, semi-biologicalnanodevices must incorporate pre-fabricated biolog-

ical components (or derivatives thereof), and thereforethe intact nanodevices are seldom made entirely de novo.As a corollary, knowledge of properties of biologicaldevice components is often incomplete (as they werenot made by human design), and therefore the rangeof activities inherent in any nanobiological device de-sign may be much less obvious and less well-definedthan it is for fully synthetic devices. Further compli-cating the issue, the activities of biological moleculesare often multifaceted (many genes and proteins exhibit

PartB

17.1


plieotropic activities), and the full range of function-ality of individual biological molecules in interactionswith other biological systems (as in nanotherapeutics)is often not known. This makes design and prototyp-ing of biological nanodevices an empirically intensive,iterative process [17.3–5, 14].

Biological macromolecules have properties, partic-ularly those relating to their stability, that can limit theiruse in device contexts. In general, proteins, nucleic acids,lipids, and other biomolecules are more labile to physic-al insult than are synthetic materials. With the possibleexceptions of topical agents or oral delivery and endoso-mal uptake of nanotherapeutics (both involving exposureto low pH), patients can tolerate conditions encoun-tered by nanobiological therapeutics in vivo, and devicelability in the face of physical insult is generally a ma-jor consideration only in ex vivo settings (relating tostorage, sterilization, ex vivo cell culture, etc.). Livingorganisms remodel themselves constantly in responseto stress, development, pathology, and external stimuli.For instance, epithelial tissues and blood componentsare constantly eliminated and regenerated, and bone andvasculature are continuously remodeled. The metabolicfacilities responsible (circulating and tissue-bound pro-teases and other enzymes, various clearance organs, theimmune system, etc.) can potentially process biologic-al components of nanobiological therapeutic devices aswell as endogenous materials, leading to partial or com-plete degradation of nanotherapeutic structure, function,or both. Furthermore, the host immune and wound re-

Table 17.1 Some ideal characteristics of nanodevices. (A) Characteristics of all nanobiological devices [17.2, 5, 13–15].(B) Desirable characteristics of therapeutic platforms [17.2, 5, 13–15]

A

Biological molecules must retain function.Device function is the result of the summed activities of device components.The relative organization of device components drives device function.Device functions can be unprecedented in the biological world.

B

Therapeutics should be minimally invasive.Therapeutics should have the capacity to target sites of disease.Therapeutics should be able to sense disease states in order to:

– report conditions at the disease site to clinicians.– administer metered therapeutic interventions.

Therapeutic functions should be segregated into standardized modules.Modules should be interchangeable to tune therapeutic function.

sponses protect the host against pathogenic organismincursions by mechanisms that involve sequestering anddegrading the pathogens. Nanobiological therapeuticsare subject to the actions of these host defense systemsand to normal remodeling processes. As we will dis-cuss, various strategies to stabilize biomolecules andstructures in heterologous in vivo environments are ap-plicable to nanobiological therapeutics [17.3, 16–20].Conversely, instability of active biocomponents can of-fer a valuable and simple way to delimit the activity ofnanotherapeutics containing biomolecules.

Nanotherapeutic Design ParadigmsSeveral early attempts to codify the canonical propertiesof ideal nanobiological devices, and therapeutic nano-devices in particular, have been made [17.5, 13, 15, 21]and are summarized in Table 17.1. In general, nano-biological devices contain biological components thatretain their function in new (device) contexts. In otherwords, one must abstract enough of a functional bio-logical unit from its native context to allow it to performthe function for which it was selected. If one wishes,for example, to appropriate the specific antigen recogni-tion property of an antibody for a device function (say,in targeting, discussed later), it is not necessary to in-corporate the entire 150,000 atomic mass unit (AMU)antibody, the bulk of which is devoted to functions otherthan antigen recognition (see Fig. 17.1 [17.1]), but it iscritical to incorporate the approximately 20,000 AMUof the antibody essential for specific antibody–antigen

PartB

17.1


nent nanodevices can be even more so. In analogyto biogenic materials, synthetic polymers are bottom-up materials per se, but fabrication of raw polymericmaterials into final device architectures often involves

top-down steps [17.33]. As we will see, therapeuticnanodevices frequently contain both synthetic and bi-ologic components, and strict top-down and bottom-upcategorization of these devices is often not applicable.

17.3 Technological and Biological Opportunities

This section considers selected enablers for therapeuticnanodevices. Some are purely technological: nanomate-rial self-assembly properties, bioconjugation methods,engineered polymers for conditional release of therapeu-tics, external triggering strategies, and so forth. Othersrelate to disease-state tissue or cell-specific biologythat can be exploited by nanotherapeutics, such as theemerging vascular address system and intrinsic trigger-ing approaches. In association with those applications,we will consider additional biological opportunitiesfor nanoscale approaches specific to particular diseasestates.

17.3.1 Assembly Approaches

Assembly of components into devices is amenable tomultiple approaches. In the case of devices comprisinga single molecule or processed from a single crys-tal (some microfabricated structures, single polymers,or grafted polymeric structures) assembly may not bean issue. Integration of multiple, separately microfabri-cated components may sometimes be necessary (as inthe immunoisolation capsule discussed below) and maysometimes drive the need for assembly, even for silicondevices. Furthermore, many therapeutic nanodevicescontain multiple, chemically diverse components thatmust be assembled precisely to support their harmoniouscontribution to device function.

“One off” Nanostructuresand Low Throughput Construction Methods

Direct-write technologies can obtain high (nanome-ter scale) resolution. For instance, electron-beam(e-beam) lithography is a technique requiring nomask, and that can yield resolutions on the order oftens of nanometers, depending on the resist mater-ials used [17.34]. Resolution in e-beam lithographyultimately is limited by electron scattering in the re-sist and electron optics, and like most direct-writeapproaches, e-beam lithography is limited in its through-put. Parallel approaches involving simultaneous writingwith up to 1,000 shaped e-beams are under de-

velopment [17.34] and may mitigate limitations inmanufacturing rate.

Force microscopy approaches utilize an ultrafinecantilever tip (typically with point diameters of 50 nm orless, Fig. 17.6) in contact with, or tapping, a surface ora stage. The technique can be used to image molecules, toanalyze molecular biochemical properties (like ligand-receptor affinity [17.35]), or to manipulate materials atnanoscale. In the latter mode, force microscopy has beenused to manipulate atoms to build individual nanostruc-tures since the mid-1980s. This has led to constructionof structures that are precise to atomic levels of reso-lution (Fig. 17.6), though the manufacturing throughputof “manual” placement of atoms by force microscopy islimited.

Dip-pen nanolithography (DPN) is a force micro-scopy methodology that can achieve high resolutionfeatures (features of 100 nm or less) in a single step.In DPN, the AFM tip is coated with molecules to be de-ployed on a surface, and the molecules are transferredfrom the AFM tip to the surface as the coated tip con-tacts it. DPN also can be used to functionalize surfaceswith two or more constituents and is well suited fordeployment of functional biomolecules on synthetic sur-faces with nanoscale precision [17.36, 37]. DPN suffersthe limitations of synthetic throughput typical of AFMconstruction strategies.

Much as multibeam strategies might improvethroughput in e-beam lithography [17.34], multipletandem probes may increase assembly throughput forconstruction methods that depend on force microscopysignificantly, but probably not sufficiently to allowmanufacture of bulk quantities of nanostructures, aswill likely be needed for consumer nanotherapeu-tic devices. As standard of care evolves increasinglytoward tailored courses of therapy [17.38] and individ-ual therapeutics become increasingly multicapable andpowerful, however, relatively low throughput synthe-sis/assembly methods may become more desirable. Forthe moment, though, ideal manufacturing approaches fornanotherapeutic devices resemble either industrial poly-mer chemistry, occurring in bulk, in convenient buffer

PartB

17.3

Therapeutic Nanodevices 17.3 Technological and Biological Opportunities 289

a)

c)

b)

Fig. 17.6 (a) A schematic depiction of an atomic force mi-croscope cantilever and tip interacting with materials ona surface. Tips typically have points of 50 nm or less indiameter [17.31, 39]. (b) Schematic of multiplexed AFMtips performing multiple operations in parallel [17.31, 39].(c) AFM image of a quantum corral, a structure built us-ing AFM manipulation of individual atoms (from the IBMImage Gallery)

systems, or in massively parallel industrial microfab-rication approaches. In any case, therapy for a singlepatient may involve billions of billions of individual nan-otherapeutic units, so each individual nanotherapeuticstructure must require only minimal input from a humansynthesis/manufacturing technician.

Self-Assembly of NanostructuresSelf-assembly has been long recognized as a poten-tially critical labor-saving approach to constructionof nanostructures [17.40], and many organic and in-organic materials have self-assembly properties thatcan be exploited to build structures with controlledconfigurations. Self-assembly processes are driven bythermodynamic forces and generally result in struc-tures that are not covalently linked. Intra/intermolecularforces driving assembly can be electrostatic or hy-drophobic interactions, hydrogen bonds, and van derWaals interactions between and within subunits of theself-assembling structures and the assembly environ-ment. Thus, final configurations are limited by the abilityto “tune” the properties of the subunits and control theassembly environment to generate particular structures.

Self-Assembly of Carbon NanostructuresCarbon nanotubes (Fig. 17.7) spontaneously assembleinto higher order [17.31] structures (nanoropes) asthe result of hydrophobic interactions between indi-vidual tubes. Multi-wall carbon nanotubes (MWCNT)are well-known structures that can be viewed as self-assembled, nested structures of nanotubes with tubediameters decreasing serially from the outermost to in-nermost tubes. The striking resemblance that MWCNThave to macroscale bearings has been noted and ex-ploited [17.41]. MWCNT linear bearings can be actuatedby application of mechanical force to the inner nano-tubes of the MWCNT assembly. Actuation causes theassembly to undergo a reversible telescoping motion.Interestingly, these linear MWCNT bearings exhibit es-sentially no wear as the result of friction between bearingcomponents.

C60 fullerenes and single wall carbon nanotubes(SWCNT) also spontaneously assemble (Fig. 17.7) intohigher order nanostructures called “peapods” [17.42] inwhich fullerene molecules are encapsulated in nano-tubes. The fullerenes of peapods modulate the localelectronic properties of the SWCNT in which they areencapsulated and may allow tuning of carbon nano-tube electrical properties. The potentially fine-levelcontrol of nanotube properties may prove useful innanotube-containing electrical devices, particularly incases wherein nanotubes are serving as molecular wires.In this capacity, carbon nanotubes have been incorpo-rated into FETs (field effect transistors, discussed belowunder sensing architectures [17.43]), and other molecu-lar electronic structures. Ultimately, these architecturesmay result in powerful, ultra-small computers to providethe intelligence of “smart,” indwelling nanotherapeutic

PartB

17.3


Self-Assembly of MaterialsMade by Traditional Polymer Chemistry

In the realms of drug delivery and biomedical microand nanodevices, the most familiar self-assembled struc-tures are micelles [17.44–46]. These structures areformed from the association of block co-polymer sub-units (Fig. 17.8), each individual subunit containinghydrophobic and hydrophilic domains. Micelles spon-taneously form when the concentration of their subunitsexceeds the critical micelle concentration (cmc) ina solvent in which one of the polymeric domains isimmiscible (Fig. 17.8). The cmc is determined by theimmiscible polymeric domain and can be adjusted bycontrol of the chemistry and length of the immiscibledomain, as well as by control of solvent conditions.Micelles formed at low concentrations from low-cmcpolymers are stable at high dilution. Micelles formedfrom polymer monomers with high cmcs can dissociateupon dilution, a phenomenon that might be exploited tocontrol release of therapeutic cargos. If desired, micellescan be stabilized by covalent cross-linking to generateshell-stabilized structures [17.44–47].

The size dispersity and other properties of micellescan be manipulated by control of solvent conditions,incorporation of excipients (to modulate polymer pack-ing properties), temperature, and agitation. From thestandpoint of size, reasonably monodisperse prepara-tions (polydispersity of 1–5%) of nanoscale micellarstructures can be prepared [17.44–47]. The immenseversatility of industrial polymer chemistry allows mi-cellar structures to be tuned chemically to suit the taskat hand. They can be modified for targeting or to sup-port higher order assembly properties. They can bemade to imbibe therapeutic or other molecules for de-livery and caused to dissociate or disgorge themselvesof payloads at desired times or bodily sites under theinfluence of local physical/chemical conditions. Thetunability of these and other properties at the level ofmonomeric polymer subunits as well as the level ofassembled higher order structures make micelles po-tentially powerful nanoscale drug delivery and imagingvehicles.

Fractal materials, such as dendritic polymers,whose synthesis and structure were discussed previ-ously, exhibit packing properties that can be exploitedto assemble higher order aggregate structures called“tecto(dendrimers)” [17.48]. In fact, these self-assemblyproperties are being exploited in oncological nanoth-erapeutics as Fig. 17.3 shows [17.15, 49]. In principle,these self-assembling therapeutic complexes need notbe pre-formed prior to administration. Individual func-

tional modules of the therapeutic assembly mightbe administered sequentially, potentially to tailortherapies more precisely to individual patient re-sponses.

Stoichiometric Control and Self-AssemblyAs the preceding examples demonstrate, self-assemblyapproaches sometimes do not feature precise controlof subunit identity and stoichiometry in the assembledcomplexes. This can be a limitation when the stoichiom-etry and relative arrangement of differentiable individualsubunits is critical to device function. Stoichiometry isless an issue when the self-assembling components areidentical and functionally fungible, as in the synthetic,peptidyl anti-infective illustrated in Fig. 17.9 [17.50,51].

In the anti-infective architecture, individual peptidecomponents are flat, circular molecules. The planar char-acter of the toroidal subunits is a consequence of thealternating chirality of alternating D-L amino acids (aas)in the primary sequence of the peptide rings. Alter-nating D and L aas is not possible in proteins madeby ribosomal synthesis [17.32]. Ribosomes recognizeand incorporate into nascent polypeptides only L aminoacids, and so, as the result of aa chirality and bondstrain, peptides made by ribosomes cannot be made flat,closed toroids like those of the peptidyl anti-infectives.Much as in α-helical domains of ribosomally synthe-sized proteins, however, the aa R-groups (which are ofvarying hydrophobic or hydrophilic chemical specifici-ties [17.32]) are arranged in the plane of the closedD, L rings extending out from the center of the rings.Hydrogen bonds between individual rings govern self-assembly of the toroids into rod-like stacks, while theR-groups dominate interactions between multiple stacksof toroids and other macromolecules and structures (seebelow).

The planar toroidal subunits can be administered asmonomers and self-assemble into multi-toroid rods atthe desired site of action (in biological membranes).But the peptide toroids R-groups are chemically tunedso that the rod structures into which they sponta-neously assemble intercalate preferentially in specificlipid bilayers (i. e., in pathogen vs. host membranes).Moreover, the assembled rods may undergo an addi-tional level of self-assembly into multi-rod structures,spanning pathogen membranes [17.50, 51]. Whether assingle rod or multiple rod assemblies, membrane in-tercalation by stacked toroids reduces the integrity ofpathogen membranes selectively, and therefore partic-ular toroid species exhibit selective toxicity to specificpathogens.

PartB

17.3


Hydrophobic block

Log concentration–10 –9 –8 –7 – 6 –5 – 4 –3

25

20

15

10

5

0

H2O

H2O

H2OH2O

a) b)

Polymer and drug in organic solvent

organicsolvent

Evaporation

Polymer in water

c)

Dialysisagainstwater

Block copolymer DrugHydrophilic block

Incorporation in micelles (arb. units)d)

Fig. 17.8a–d Micellar drug delivery vehicles and theirself-assembly from block copolymers [17.44–46]. (a) Mor-phology of a micelle in aqueous buffer. Hydrophobic andhydrophilic polymer blocks, copolymers containing theblocks, micelles generated from the block copolymers,and (hydrophobic) drugs for encapsulation in the micellesare indicated. (b) Micelle self-assembly and charging withdrug occuring simultaneously when the drug-polymer for-mulation is transitioned from organic to aqueous solventby dialysis. (c) Pre-formed micelles can be passively im-bibed with drugs in organic solvent. Organic solvent isthen removed by evaporation, resulting in compression ofthe (now) drug-bearing hydrophobic core of the micelle.(d) An illustration of concentration-driven micelle forma-tion. At and above the critical micelle concentration (cmc),block copolymer monomers assemble into micelles, ratherthan exist as free block copolymer molecules. The arrowindicates the cmc for this system

PartB

17.3


N

H

N• • • • • • • • • • • • • • • • • • • • •

• • • • • • • • • • • • • • • • • • • • •

• • • • • • • • • • • • • • • • • • • • •

• • • • • • • • • • • • • • • • • • • • •

• • • • • • • • • • • • • • • • • • • • •

O

O

H

H

N

H

N

H

O

O

O

H

N

H O

H

O

O

RNH

NHO

O

R

O

R

HNO

R

HN

O

RHN

O R HN O

R

NH

O

R

NH

D

L

D

L

D

L

D

L

Fig. 17.9 A self-assembling peptide antibiotic nanostructure [17.50, 51]. Peptide linkages and the α-carbons and theirpendant R groups are indicated. The synthetic peptide rings are planar as the result of the alternating chirality (D or L)of their amino acid (aa) constituents. R groups of aas radiate out from the center of the toroid structure. Individ-ual toroids self-assemble (stack) as the result of hydrogen bonding interaction between amine and carboxy groups ofthe peptide backbones of adjacent toroids. The surface chemistry of multi-toroid stacks is tuned at the level of theaa sequence and, therefore, R group content of the synthetic peptide rings. The chemical properties of the stackedtoroid surfaces allow them to intercalate into the membranes of pathogenic organisms with lethal consequences.The specific membrane preferences for intercalation of the compound are tuned by control of R group content oftorroids

These toroidal, synthetic antibiotics, and othernanoscale antimicrobials represent critically needed,novel antibacterial agents. Resistance to traditional,microbially derived antibiotics often is tied to detox-ifying functions associated with secondary metabolitesynthesis; these detoxifying functions are essential forthe viability of many antibiotic producing organisms(for instance, see [17.52]). The genes encoding suchdetoxifying functions are rapidly disseminated to othermicroorganisms, accounting for the rapid evolution ofdrug resistant organisms that has bedeviled antimi-crobial chemotherapy for the last 25 years. Syntheticnanoscale antibiotics, like the peptide toroids [17.50,51]and the N8N antimicrobial nanoemulsion [17.53], actby mechanisms entirely distinct from those of tradi-tional secondary metabolite antibiotics, and no nativedetoxifying gene exists. Therefore, novel nanoscale an-timicrobials may not be subject to the unfortunatelyrapid rise in resistant organisms associated with mostsecondary metabolite antibiotics, though this remains tobe seen. As bacterial infection continues to re-emergeas a major cause of morbidity and mortality in thedeveloped world, consequence of increasing antibiotic-resistant pathogens, novel nanoscale antibiotics willbecome more important.

Biomolecules in Therapeutic Nanodevices:Self-Assembly and Orthogonal Conjugation

Biological macromolecules undergo self-assembly atmultiple levels, and like all instances of such construc-tion, biological self-assembly processes are driven bythermodynamic forces. Some biomolecules undergo in-tramolecular self-assembly (as in protein folding fromlinear peptide sequences, Fig. 17.10). Higher orderstructures are, in turn, built by self-assembly of smallerself-assembled subunits (for instance, structures as-sembled by hybridization of multiple oligonucleotides,enzyme complexes, fluid mosaic membranes, ribo-somes, organelles, cells, tissues, etc.).

Proteins are nonrandom copolymers of 20 chem-ically distinct amino acid (aa) subunits [17.32]. Theprecise order of aas (i. e., via interactions between aaside chains) drives the linear polypeptide chains toform specific secondary structures (the α helices andβ sheet structures seen in Fig. 17.10). The secondarystructures have their own preferences for association,which, in turn, leads to the formation of the tertiaryand quartenary structures that constitute the folded pro-tein structures. In its entirety, this process producesconsistent structures that derive their biological func-tions from strict control of the deployment of chemical

PartB

17.3


cuous chemistries (1-ethyl-3-(3-diamethylaminopropyl)carbodiimide or EDC, conjugation, see [17.60]) usedto conjugate cytokines to nanoparticles tend to in-activate hIL-3 and other cytokines whereas thesame protein/particle bioconjugates retain bioactivityif judiciously chosen orthogonal conjugation strate-gies are used (Lee & Parthasarathy, unpublished).Proteins for which only a small portion of their sur-faces contribute to the interesting portions of theirbioactivities (from the standpoint of the nanodevicedesigner), such as some enzymes or intact antibod-ies (Fig. 17.1) may be somewhat less sensitive topromiscuity of the bioconjugate strategy used [17.20,58, 59], but the benefits of orthogonal conjugationstrategies can also apply to these protein bioconju-gates [17.20, 58, 59]. The potential utility of orthogonalconjugation for incorporation of active biological struc-tures into semi-synthetic nanodevices is becomingmore fully recognized and cannot be overestimated[17.58, 59].

17.3.2 Targeting:Delimiting Nanotherapeutic Actionin Three-Dimensional Space

Delivery of therapeutics to sites of action is a key strategyto enhance clinical benefit, particularly for drugs usefulwithin only narrow windows of concentration becauseof their toxicity (i. e., drugs with narrow therapeuticwindows). Diverse targeting approaches are available,ranging from methods exploiting differential extravasa-tion limits of vasculature of different tissues (see thediscussion of oncology below), sizes, and surface chem-istry preferences for cellular uptake (see discussion ofvaccines below); preferential partition of molecules andparticles into specific tissues by virtue of their charges,sizes, surface chemistry, or extent of opsonization (seebelow); or the affinity of biological molecules decorat-ing the nanodevice for counter-receptors on the cells ortissues of interest.

The Reticuloendothelial Systemand Clearance of Foreign Materials

Physical properties such as surface chemistry and par-ticle size can drive targeting of nanomaterials (andpresumably nanodevices containing them) to sometissues. For instance, the pharmacokinetic (Pk) andbiodistribution (Bd) properties [17.32] of many drugsand nanomaterials are driven by their clearance inurine, which is in turn governed by the filtration pref-erences of the kidney. Most molecules making transit

into urine have masses of less than 25 to 50 kilodal-tons (kDa; 25–50 kDa particles corresponding looselyto effective diameters of about 5 nm or less) and arepreferably positively charged; these parameters areroutinely modulated to control clearance rates of ad-ministered drugs. Clearance of low molecular weight(nano)materials in urine can be suppressed by tun-ing their molecular weights and effective diameters,typically accomplished by chemical conjugation, topolymers such as poly(ethylene glycol) [17.19]. Poly-mer conjugation (pegylation) has been applied to manydifferent materials and may provide some degree ofcharge shielding. Pegylation also increases effective mo-lecular weights of small materials above the kidneyexclusion limit, diverting them from rapid clearance inurine.

Coating foreign particles with serum proteins (op-sonization [17.61]) is the first step in the clearanceof foreign materials. Opsonized particles are recog-nized and taken up by tissue dendritic cells (DCs)and specific clearance organs. These tissues (thymus,liver, and spleen, constituting the organs of the retic-uloendothelial system or RES) extract materials fromcirculation by both passive diffusion and active pro-cesses (receptor-mediated endocytosis). Charge-driven,receptor-mediated uptake of synthetic nanomaterialsoccurs in the RES and can result in partition ofpositively charged nanoparticles into the RES. Forinstance, PAMAM dendritic polymers exhibit high pos-itive charge densities related to the large number ofprimary amines on their surfaces [17.28–30]. In experi-mental animals, biodistribution of unmodified PAMAMdendrimers is limited nearly exclusively to RES or-gans [17.62]. This unfavorable biodistribution can bemodulated by “capping” the dendrimers (i. e., derivatiz-ing the dendrimer to another chemical specificity, suchas carboxy or hydroxyl functionalities [17.15, 62]).

Despite legitimate applications of targeting to thekidney and the RES (for instance in glomerular dis-ease [17.63]), intrinsic targeting to clearance sites isof interest primarily as a technical problem that im-pedes therapeutic delivery to other sites. In such cases,numerous targeting strategies are available, some ofwhich depend on synthetic nanomaterial properties (forinstance, see discussion of the enhanced permeabilityand retention or EPR effect in the context of oncology,below) to minimize uptake of nanotherapeutic devicesby clearance systems and maximize delivery to desiredsites. Targeting via biological affinity reagents decorat-ing the surfaces of therapeutic nanodevices may be themost direct approach.

PartB

17.3


The vascular address system has been character-ized by administering a peptide phage display to libraryanimals, resecting individual organs, and extractingphage from the vasculature of the isolated organs(Fig. 17.11 [17.64–66]; see [17.67] for a discussion ofdisplay technology). Amazingly, phage isolated fromdifferent organs exhibited distinct consensus presentedpeptide sequences, indicating that the vasculature ofindividual organs presented unique cognate receptors,each bound by a different short (ten amino acids orfewer) consensus peptide sequence that had been affin-ity selected from the phage display library. Furthermore,the affinity-selected peptides have the capacity to tethernano- to microscale particles to the site of their cognatereceptors (as illustrated by the binding phage particlespresenting the peptides to specific vascular locations).Site-specific drug delivery using the vascular addresssystem has already been demonstrated [17.68,69]: it hasfurther been used to target an apoptotic (cytocidal) agentto the prostate and to direct destruction of the organ inan animal model [17.68].

Mapping of the vascular address system is cur-rently underway [17.64] and holds the promise ofspecific delivery of therapeutic agents to vasculatureof specific organs. It remains to be seen whether eachorgan has a single molecular marker constituting itsaddress that is amenable to binding a single peptidesequence; organs may instead have unique constella-tions of antigenic markers. If so, specific targeting maybe possible using multiple peptides, each peptide bind-ing its cognate receptor on target organ vasculaturevery weakly. Peptides used in such a multivalent tar-geting strategy would be chosen to reflect the uniqueconstellation of address markers present in the tar-get tissue. Affinities of cognates for linear peptidesare often very low [17.67], though their aggregateaffinity may be substantially higher than that of anypeptide-vascular address cognate alone. Under idealconditions, the affinity of such a multipeptide, mul-ticognate complex should be the equivalent of theproducts of the affinities of each constituent peptidefor its individual constituent cognate. Such multivalentinteraction avidities can be extremely high (and the cor-responding effective affinity constants are also high) butseldom fully realize their theoretical maximums (Lee &Parthasarathy, unpublished; see [17.32] for a review ofreceptor biochemistry).

It should be noted that most of the vascular addressesidentified to date deliver materials to the organ vascula-ture: extravasation and access of organ tissue spaces bynanotherapeutics remains a separate issue.

17.3.3 Triggering:Delimiting Nanotherapeutic Actionin Space and Time

Controlled triggering of therapeutic action is the otherside of the targeting coin. If the site and time of nano-therapeutic delivery cannot be adequately controlled, thesite of therapeutic action can be delimited by spatially-or temporally-specific triggering. The triggering eventmight drive release of active therapeutic from a reservoir,or chemical or physical processing of drug materialsfrom an inert to an active form (inert administrationsthat are converted to active form at a specific time orplace drugs which are activated by a chemical reac-tion occurring their sites of action are called prodrugs,Fig. 17.12). Three major triggering strategies are widelyused: external stimuli, intrinsic triggering, and sec-ondary signaling (multicomponent systems). Triggeringstrategies require nanotherapeutic delivery devices tobe sensitive to a controlled triggering event, or a spa-tially/temporally intrinsic triggering event mediated bythe host. Obviously, the triggering event itself must betolerable to the patient.

Carrier/Trigger

Effector

Carrier/Trigger Effector

Carrier/Trigger

Effector

+

Fig. 17.12 One possible configuration of a prodrug, in whichcarrier/trigger and effector functions are separate functionaldomains of the therapeutic. The key feature of prodrugsis that they are therapeutically inert (as indicated by thecolorless effector domain) until an activation event oc-curs (mediated here through the carrier/trigger domain,with effector activation indicated by its change to a starshape). Activation events can involve cleavage of inhibitorycarrier/trigger domains from effectors. Other activationstrategies involve a chemical change or shift in confor-mation of the effector, mediated through the carrier/triggerdomain and in response to an environmental condition

PartB

17.3


Nanotherapeutic TriggeringUsing External Stimuli

External stimuli are provided by an external nanodeviceoperator/clinician, usually in the form of a site-specificenergy input, typically light, ultrasound, or magnetic orelectrical fields. Organic polymeric structures are veryamenable to interaction with these energy sources. Forinstance, micellar structures can be reversibly dissoci-ated with ultrasound, in which case they disgorge theircontents or expose their internal spaces to the envi-ronment during ultrasound pulses (20 to 90 kilohertzrange). The process has been used to control release ofcytotoxin (doxorubicin) from micelles (Fig. 17.13), andshort ultrasound transients might be used for pulsatile orintermittent exposure of patients to therapeutics [17.70].

Light is another popular external triggering modal-ity. Bioactive materials can be covalently associatedto a nanoscale-delivery vehicle by photo-labile link-ages [17.15, 71, 72], or micelles can be constructedso that their permeability is altered as the result of

0

11

10

9

8

7

6

5

420 40 60 80 100 120 140 160 180

Fluorescence (arb. units)

Time (s)

(a)

(b)

Ultrasoundon

Ultrasoundoff

Fig. 17.13 Externally applied stimuli can trigger drug release: hydrophobic drug molecules are reversibly deployedfrom micelles in response to acoustic stimuli [17.70]. Intact pluoronic micelles maintain the hydrophobic cytotoxindoxorubicin in a hydrophobic environment (the core of the micelle, as in Fig. 17.4). Doxorubicin fluorescence is quenchedin aqueous environments, and hence changes in integrity of pluoronic micelles carrying doxorubicin can be monitoredby doxorubicin fluorescence. Here, an ultrasound pulse is applied to suspensions of such micelles or to doxorubicin insolution. Application of ultrasound triggers exposure of micelle-encapsulated doxorubicin to the aqueous solvent (i. e.,drug release), as demonstrated by the reduction in fluorescence. After cessation of the acoustic pulse, doxorubicin isrepackaged into the micelles, as evidenced by the increase in fluorescence after the ultrasound administration (Line a).Ultrasound has no impact on fluorescence of doxorubicin in solution, as expected (Line b)

exposure to light [17.72]. In the latter case, light in-put can cause photopolymerization resulting in micellecompaction that drives release of therapeutic cargo orphoto-oxidation, which causes loss of micellar integrityto release encapsulated materials. The ability of light topenetrate dense tissues is a clear limitation to this ap-proach. This concern can be accommodated by polymersystems responsive to wavelengths that penetrate tissueefficiently (usually in the red-infrared region of the spec-trum), or by use of systems such as fiber optics to deliverlight to deep tissues [17.15, 71, 72].

Externally applied magnetic fields can also beused to control nanotherapeutic activity. For instance,eddy currents induced by alternating magnetic fieldscan heat nanometallic particles and their immediatevicinity, an approach that has been successfully ap-plied to control the bioactivity of individual biologicalmacromolecules [17.73]. Colloidal gold particles are co-valently conjugated to biomolecules (nucleic acid, orNA, duplexes or proteins), and alternating magnetic

PartB

17.3


Other secondary signaling systems rely on DNAhybridization to assemble therapeutic components intocatalytically active complexes at their sites of ac-tion [17.55]: in this system, the individual componentslack therapeutic activity until they are brought withina few nanometers of each other by hybridization toa target single-stranded nucleic acid molecule (SSNAin Fig. 17.15, potentially a RNA species expressed ina tissue or cell type of interest). At this point, the cat-alytic moiety and the prodrug moieties are sufficientlyclose in space to allow catalytic cleavage of the activecomponent from the prodrug. As depicted, this strategymakes no provision for delivery of drug and catalystbioconjugates to the site of interest, nor for their tran-sit across biological membranes. The system is similarto the DNA hybridization-driven fluorescence transfersystem of Fig. 17.10.

Alternatively, secondary signaling systems can ex-ploit competitive displacement of therapeutics froma carrier structure [17.79]. In this case, a non-covalentcomplex of engineered antibody and plasminogen ac-tivator (PA) is tethered to blood and fibrin clots byantibody affinity for fibrin. PA is released from the boundcomplex (to dissociate the clot) using bolus systemic ad-ministration of a nontoxic binding competitor for PA tothe antibody complex. The strategy establishes a highlocal concentration of PA at clot sites, efficiently dis-solving clots and potentially minimizing systemic sideeffects.

Layering Strategies for Fine Controlof Nanotherapeutic Action

It should be clear that many of these approaches may bebroadly applicable to trigger events other than drug re-lease or to drive assembly or disassembly of therapeuticnanostructures in situ. It should also be clear that theseapproaches are often complementary, and that multipleapproaches can be used in single nanotherapeutic de-vices (Fig. 17.3, Table 17.1 [17.11–13, 15, 71, 78, 80]).Layering targeting and triggering approaches tends tomake devices more complex, but it also allows clini-cian/operators to intervene at multiple points in therapy,potentially leading to finer control of the therapeuticprocess and better clinical outcomes.

17.3.4 Sensing Modalities

The need for “smart” therapy is a key theme of therapeu-tic nanotechnology and pharmacology as a whole. Drugswith narrow therapeutic windows should be deliveredonly to their desired site of action and be pharma-

cologically active only when that activity is needed.These strategies can limit undesired secondary effectsof therapy, some of which can be debilitating or lifethreatening, as discussed elsewhere in this chapter. Onepossible approach to this issue is the incorporation ofsensing capability (specifically, the capacity to recog-nize appropriate contexts for therapeutic activity) intonanotherapeutic devices. Sensing capability may allowself-regulation of a therapeutic device, reporting to anexternal clinician/device operator (though this beginsto touch on imaging applications, see below), or both.In the context of our discussion of therapeutic nano-and microscale devices, we will consider primarily elec-trical and electrochemical sensor systems, particularlymicrofabricated (Field Effect Transistor or FET, andcantilever) and conducting polymer sensors.

Sensor SystemsSensing is predominantly a higher-order device func-tionality, depending on multiple device components,though one could argue that some targeting/triggeringstrategies, particularly targeting by bioaffinity and in-trinsic triggering strategies, must, a priori, incorporate atleast limited sensing capability. But biosensors, as theyare typically considered, are multifunctional, multicom-ponent devices [17.81]. Usually a biosensor system iscomposed of signal transducer, sensor interface, biolog-ical detection (bioaffinity) agent, and an associated assaymethodology, with each system component governed byits inherent operational considerations.

The transducer component determines the physi-cal size and portability of the biosensor system. Signaltransducers are moieties that are sensitive to a physical-chemical change in their environment and that undergosome detectable change in chemistry, structure, or stateas the result of analyte (the thing to be sensed) recogni-tion. Analytes for nanotherapeutic application could bebiomolecules, like proteins, small molecules (organicor inorganic), ions (salts or hydrogen ions), or physicalconditions (such as redox state, temperature). Interfacesare the sensor components that interact directly with theanalyte. For sensor use in nanotherapeutic devices, im-mobilized or otherwise captured biological molecules(proteins, nucleic acids) often constitute the sensor in-terface. Whatever the chemical nature of the interface, itdetermines the selectivity, sensitivity, and stability ofthe sensing system and also is a dominant determi-nant of sensor operational limits. Assay methodologydetermines the need (or lack thereof) for analyte tracers,the number of analytical reagents, and the complexityand rapidity of the sensing process. Nanotherapeutics

PartB

17.3


are of interest at least in part because they can beminimally invasive, low complexity, yet robust and ac-curate; convenient assay methods are therefore highlydesirable.

Cantilever BiosensorsMicromechanical cantilevers (discussed above in con-junction with force microscopy, Fig. 17.6) transducesensed events by mechanical means [17.82]. Bothchanges in the resonant frequency and deflection ofcantilevers resulting from analyte binding or disso-ciation can be conveniently and sensitively detected.These changes in cantilever state can be conve-niently detected by optical, capacitive, interferometry, orpiezoresistive/piezoelectric methods, among others. Mi-crofabricated cantilever dimensions range from micronto sub-micron range, with potential for further dimen-sional optimization (by carbon nanotubes appended tothem, for instance [17.82]). They are operationally ver-satile and can be used in air, vacuum, or liquid, althoughthey suffer some degradation in performance in liquidmedia. Like most micromachined structures, they canbe batch fabricated and conveniently multiplexed.

Cantilevers used in atomic force microscopy (AFM,see also discussion of one-off nanostructures above)approaches can be used to study individual biomolec-ular interactions [17.35,81]. In this approach, cantilevertips are derivatized with biomolecules (effectively,one member of a receptor/counter-receptor pair), andthe tip-bound biomolecule is allowed to bind itscounter-receptor (itself bound on a surface). Under non-equilibrium conditions, (i. e., conditions that result inthermodynamically irreversible changes in analyte mo-lecular structures), the force required to disrupt singlemolecular interactions can be measured and related toclassical biochemical parameters of receptor binding.The method has been applied to interactions betweenhormones and their receptors, sugars and lectins, as wellas hybridizing DNA strands.

Cantilever systems sensitively detect changes inmass at their surfaces: changes as small as a mass densityof 0.67 ng/cm−3 are theoretically detectable. This canallow detection of binding of extremely small objects tothe cantilever and has been applied to detection and enu-meration of prokaryotic and eukaryotic cells, as well assmall numbers of macromolecules [17.82]. The incor-poration of biological receptors or affinity reagents onthe cantilever surface can drive specific binding eventsfor particular sensing tasks. Microcantilevers are alsohighly sensitive to temperature, detecting changes as lowas 10−5 K; they can also detect small changes in pH.

Field Effect Transistor BiosensorsField effect transistor (FET) architectures are anothersensing architecture that can be conveniently producedby micro-nanofabrication. FETs consist of a currentsource, a current drain, a conductive path (sensing chan-nel) between them, and a sensing gate to which a biascan be applied. Analyte binding to the sensing channelinduces a charge transfer resulting in a dipole betweenthe surface and the underlying depletion region of thesemiconductor: current that passes between the sourceand drain of a semiconductor FET is quite sensitiveto the charge state and potential of the surface in theconnecting channel region. Moving a standard siliconFET from depletion to strong inversion (i. e., shift-ing the surface potential by >∼ 0.5 eV) requires lessthan ∼ 10−7 C/cm2 or ∼ 6 × 1012 charges/cm2, corre-sponding to transfer of 6.25 × 1011 e/cm2. With FETs of2,000 square micrometers, detection of biological ana-lytes in sub-nanomolar concentrations is easily feasible.Specificity for binding of macromolecular analytes ofinterest can be provided by deployment of biologicalaffinity reagents in the FET sensing channel. Submi-cron FETs are routinely manufactured; use of carbonnanotubes in FETS will offer still greater miniaturiza-tion [17.43, 83].

Carbon nanotubes also have excellent mechani-cal properties and chemical stability in addition topotentially tunable electrical properties, making themhighly desirable electrode/nanoelectrical materials forany number of nanoelectrical applications [17.84].Biomolecules can be bound to carbon nanotubes, par-ticularly in FET and nanoelectrode applications. Mostbiomolecules bound to carbon nanotubes are not cova-lently bound (as discussed above) and do not exhibitdirect electrical communication with the nanotube,though redox enzymes bound to nanotubes and otherconductive nanomaterials may [17.84, 85]. Flavin ade-nine dinucleotide (FAD) and flavoenyzme glucoseoxidase (Gox) both display quasi-reversible one electrontransfer when absorbed onto unannealed carbon nano-tubes in glassy carbon electrodes. Gox so immobilizedretains its substrate-specific (glucose) oxidative activity,leading to applications in sensing circulating glucosefor diabetes and, perhaps, to a strategy of harvestingelectrical power from metabolic energy.

Conducting Polymersand Sensor Biocompatibility

Biocompatibility of most metallic structures (as mightbe used in the bioelectrical sensors described) is limitedat best; metal structures rapidly foul with serum pro-

PartB

17.3

Therapeutic Nanodevices 17.4 Applications for Nanotherapeutic Devices 307

17.4 Applications for Nanotherapeutic Devices

As discussed above, nanotherapeutic devices are novel,emerging therapeutics with properties not fully under-stood or predictable. Nanotherapeutics, therefore, mustbe justifiable on at least two levels. As we have seen,the nature of the therapeutic task and the state ofcurrent nanoscale-materials technology make the incor-poration of biological macromolecules unavoidable formany nanotherapeutics. Proteins, for example, typicallyare substantially more expensive than small moleculetherapeutics, and precise nanostructures containing pro-teins will be more costly still. Nanotherapeutics mustjustify their high COGs. Secondly, as new therapeu-tic modalities, nanotherapeutics may carry significantlylarger risks than those associated with more conven-tional therapies. Expensive, novel moieties, such asnanobiotechnological therapeutic devices, are thereforemost likely to be accepted for treatment of condi-tions that not only are accessible to intervention atthe nanoscale but also for which existing therapeuticmodalities have acknowledged shortcomings in patientmorbidity or mortality. We have selected two diseasestates sufficiently grave and sufficiently unserved towarrant nanotherapy: cancer and cardiovascular disease.Modulation of immune responses and vaccination is ourthird application area.

17.4.1 Nanotherapeutic Devices in Oncology

The economic burden imposed by cancer is immense,measuring in the billions of dollars annually in theUnited States alone. Existing therapies such as sur-gical resection, radiotherapy, and chemotherapy haveprofoundly limited efficacy and frequently provideunfavorable outcomes as the result of catastrophic thera-peutic side effects. Additionally, the biology, chemistry,and physics of cancer, in general, and solid tumors inparticular, provide therapeutic avenues accessible onlyby nanoscale therapeutics. Oncology is thus an idealarena for emerging nanotechnological therapies.

Tumor Architecture and PropertiesTumors as tissues are relatively chaotic structures ex-hibiting vast structural heterogeneity as a function ofboth time and space [17.9, 77, 111, 112]. In healthy tis-sues, vasculature resembles a regular mesh in which themean distance of tissue spaces to the nearest vessel istightly controlled and highly uniform. On the other hand,the vasculature of tumors resembles a percolation net-work containing regions experiencing vastly different

levels of perfusion. Tumors of 1 mm3 or larger typicallycontain measurably hypoxic domains, with pO2 valuesas much as two- to threefold lower than in normal tissue.High levels of hypoxia are characteristic of enhancedmetastatic potential and tumor progression. Also dueto insufficiency of perfusion, tumors frequently containnecrotic domains.

The average torturosity of vascular flow paths intumors is also much greater than in healthy tissue, andtransient thrombotic events lead to enhanced resistanceto flow and ongoing vascular remodeling. Aside fromits plasticity, tumor vasculature itself is highly irregular,may be incompletely lined with endothelial cells, andoften exhibits significantly higher extravasation limits(the highest molecular weight of materials that can leavethe vasculature and diffuse into the interstitial spaces)than normal vasculature. Tumor tissue is also poorlydrained by lymphatics, so extravasated materials tend toremain in situ in tumors and are not cleared efficiently.

These biological phenomena all differentiate tumortissue from normal tissue and can be exploited in therapy.For instance, the special vascular integrity and lymphaticdrainage properties of tumors constitute the EnhancedPermeability and Retention effect (EPR) [17.11, 12,80, 113]. EPR presents an obvious opportunity forintervention with nanoscale therapeutics (Fig. 17.3). Ex-travasation limits for normal tissues are variable, butstructures larger than a few nanometers in diameter donot leave circulation efficiently in most tissues, whereastumor vasculature frequently allows egress of materialsin the tens to hundreds of nanometer range. Further,nanomaterials, once extravasated, are not cleared bylymphatic drainage. EPR provides tumor targeting thatdoes not depend on biological affinity reagents: tumortissue provides preferential depot sites for extravasateddrug delivery devices. Targeting to a desired site of ac-tion is highly desirable for cytotoxic therapeutics, andEPR provides the basis for a growing class of polymertherapeutics [17.11,12,80,113]. That said, EPR and tar-geting by biological affinity are not necessarily mutuallyexclusive. A number of antigens more or less specificallyrelated to tumors are known (tumor associated antigensor TAAs), and as discussed above tumor-/organ-specificvascular addresses might be exploited for delivery oftherapeutics [17.65,66]. For instance, various biologicalreagents that recognize TAAs (i. e., antibody to carci-noembryonic antigen, transferrin) have been conjugatedto nanoscale contrast agents to enhance contrast agentlocalization to tumoral sites [17.104].

PartB

17.4


15

10

1.0

0.1

30 90 15 30 90

Tumor

Normaltissue

a)

LabileLinker

Doxorubicin

HPMA backbone

b)

Tumor-localized therapeutic (µg/kg)

Time after administration of 10 mg/kg iv (min)

SMANCS N C S

c)

Fig. 17.16a–c Therapeutic nanodevices can be specificallydelivered to tumors by virtue of their size [17.11, 12, 80,113]. (a) An illustration of the relatively low integrity oftumor vasculature. Therapeutic nanodevices (in this case,micelles carrying cytotoxins) cannot leave (i. e., extravasatefrom) the vasculature of normal tissue, but the vasculatureof tumors is sufficiently leaky to allow specific deliveryto tumoral interstitial spaces. This is the basis of size-dependent targeting to tumors by the enhanced permeabilityand retention, or EPR effect. (b) A polymer nanothera-peutic (PK1, [17.80]) that exploits EPR effect to delivera cytotoxin (doxorubicin) to tumors. The overall mass ofthe complex is tuned by control of the mass of the N-2-(hydroxypropyl) methacrylimide backbone, so as to beabove extravasation limits for normal tissues but belowboth tumor extravasation limits and the size limit for excre-tion through the kidney. A final feature of the device thatfavors delimiting doxorubicin cytotoxicity to tumoral sitesis the fact that the device is a prodrug. The cytotoxicity ofdoxorubicin is severely curtailed when the doxorubicin iscovalently linked to the polymer therapeutic, though it man-ifests its full toxicity when it is released from the complexby cleavage of the labile (peptidic) linker. The labile linkercontains a cleavage site recognized by a protease known tobe over-expressed in the target tumor type. Thus, the de-vice produces a depot of inactive doxorubicin in tumorsthat is activated preferentially to full toxicity by conditionsprevalent in the tumor. (c) EPR can drive localization ofappropriately sized therapeutics to tumoral sites [17.113].The polymer therapeutic SMANCS accumulates rapidly inmurine tumors, whereas the NCS, the parent therapeutic(not engineered for EPR delivery), does not

Tumor (EPR) Properties and NanotherapeuticsPK1 (Fig. 17.16) is an early example of a growing andincreasingly sophisticated class of polymer therapeu-tics [17.11, 12, 80, 113]. It is a nanoscale moleculartherapeutic device wherein the therapeutic moiety (dox-orubicin) is covalently linked to the polymeric backbone.As we will see, each component of the polymer thera-peutic fulfills a discrete function, so the device rubric iswarranted [17.6].

PK1 consists of a polymeric backbone (HPMA, hy-droxyl polymethacrylamide) which, as discussed below,targets desired sites by virtue of its specific nanoscalesize (by EPR effect) and a cytotoxin (doxorubicin),covalently linked together by a peptide. PK1 exhibitsmultiple design features that allow it to preferentiallydeliver doxorubicin to tumors. The size of the complex

PartB

17.4

Therapeutic Nanodevices 17.5 Concluding Remarks: Barriers to Practice and Prospects 315

17.5 Concluding Remarks: Barriers to Practice and Prospects

The effort to produce nanoscale therapeutic devicesis clearly highly interdisciplinary. As we have seen,it touches on numerous established disciplines, en-compassing elements of physiology, biotechnology,bioconjugate chemistry, electrical engineering, and ma-terials science, to name just a few of the fields involved.Obviously, this broad sweep of knowledge is difficult forany one investigator to master fully. The breadth of theeffort constitutes just one of the major barriers to entryin the field. Other challenges include the raw complex-ity of biology, the fashion in which biologists hold anddistribute information, and cultural differences betweenengineers and biological scientists.

17.5.1 Complexity in Biology

Biology is characterized by particularity: nuances ofbiological systems are often unique to the system athand and highly idiosyncratic. This follows from thefact that biological systems are not purpose-built, asare designed devices (although they can give that ap-pearance, as seen in Fig. 17.1), but rather arose as theconsequence of evolutionary processes. Evolution isa highly chaotic business, with the outcome of anygiven evolutionary process highly sensitive to initialconditions (populations of organisms subjected to se-lection, other organisms in the environment, biologyof all organisms involved, resource availability, otherenvironmental factors, etc.). Moreover, many of thevariables affecting natural selection processes are non-static and change, as a function of time or space orboth, while selection is exerted against a populationof organisms. Conditions leading to one evolutionaryadaptation or another are therefore seldom duplicatedexactly, so that individual adaptive features are idiosyn-cratic, with elements relating not only to their biologicalfunctions but also to their evolutionary history. Indi-vidual systems and adaptive features thereof can bealmost baroquely complex because the structures them-selves arose under unique conditions, under uniqueselective pressures, and from unique initial biologicalsystems. The extent of the complexity of biologicalsystems is apparent even in individual macromolec-ular constituents of biological systems and can bemade clear by comparison of synthetic nanostructuresto biological nanostructures with similar dimensionalaspects.

SWCNT (single wall carbon nanotubes) have a min-imal diameter of about 1.3 nm, and many proteins

likewise have diameters of few nms. SWCNTs are reg-ular and homogeneous polymeric structures composedentirely of carbon atoms (Fig. 17.6). Proteins, on theother hand, are not polymers of one repeating subunit,but rather are nonrandom copolymers of 20 chemicallydistinct amino acids (aas). The order of the aas drivesspecific folding events that produce three-dimensionalstructures that, in turn, present specific aas and their sidechains at specific positions in space (Fig. 17.10). Thiscontrol of aa position in space drives protein activity:small changes in aa sequence can perturb structure andfunction profoundly. Therefore, though they fall withinthe same broad size regime, proteins are structurally andfunctionally much more complex than SWCNTs. Theextent of biological complexity can be glimpsed whenone considers that individual living things are orderedaggregates of multiple macromolecules and supramolec-ular structures (potentially, billions and billions of them),belonging to distinct chemical classes (lipids, proteins,nucleic acids, etc.), with each individual macromoleculebeing at least as complex in structure and function as theprotein of Fig. 17.10. The nearly irreducible complexityof biological systems is a central fact of practice in thebiological sciences and drives the way biologists gatherand disseminate information. It is the key informant ofthe culture of biologists.

17.5.2 Disseminationof Biological Information

Biologists typically consider themselves primarily asscientists, as opposed to being engineers or technologydevelopers. The product biologists generate, then, is in-formation rather than devices or structures. Their interestin technological application of the knowledge they pro-duce is typically secondary to their desire to develop theinformation itself. Additionally, as we have just seen,biology is a ferociously complex discipline.

As a practical matter, biological data itself is oftenmuch more rich (and often more ambiguous) than datafrom harder scientific endeavors. The data can be so veryrich that biologists must make choices as to what data isrelevant to a given phenomena and, therefore, what datathey will publish. With absolutely no intent to concealor mislead, biologists are often driven to publish onlya small fraction of the data they gather.

Since they see themselves primarily as scientists,biologists tend to publish data they believe to be ofbroad scientific significance. Typically, the chosen in-

PartB

17.5


formation does not include data that might be criticalto technology development, frequently making the bio-logical literature an inadequate resource for engineers.The omitted information thus becomes lore. That is tosay, the infomation is critical for the practice or develop-ment of a given technology but usually is not accessibleto persons outside the field (as many nanotherapeuticdevice producers may be). This can be illustrated by re-cent experiences around phage display-mediated affinitymaturation of four helical bundle proteins [17.141–143].

Four helical bundle proteins are loosely related smallproteins consisting of four α helices arranged in a spe-cific configuration (see the dynamite stick representationof insulin that occurs in Fig. 17.18). They include in-sulin, growth hormone, most cytokines, and variousother molecules involved in cell to cell signaling. Theyhave a wide range of therapeutically valuable propertiesthat might be made even more desirable if their poten-cies could be enhanced. There is thus significant interestin variant four-helical bundle proteins with improvedbioactivities.

Phage display [17.67] is a method that can be used tosort protein variant libraries for variants with enhancedaffinity for a target receptor. This is usually accom-plished by iterative rounds of affinity selection on thereceptor followed by propagation of selected phage. Asrounds of selection and propagation proceed, the meanaffinity of the selected variants for the receptor increases(so-called affinity maturation). A phage display methodis depicted in Fig. 17.11.

Affinity maturation by phage display can be usedto identify variant proteins from libraries that have en-hanced biological activities, providing that affinity forreceptor is limiting in the overall activity of the parentprotein [17.134]. Phage display affinity maturation hasbeen successfully used to select some enhanced activityvariants of four helical bundle proteins [17.144, 145]but is not applicable to engineering increased activ-ity for all four helical bundle proteins [17.134]. Thereason for the inapplicability of display methods toengineer high-activity human growth hormone (hGH)variants became apparent in a recent publication fromGenentech [17.142].

In the case of hGH, the affinity of the parent mol-ecule for receptor is in excess of that required to drivemaximal biological responses [17.142]. This fact wasinferred after a multiyear effort (beginning in the late1980s and culminating in 1999) to derive more ac-tive hGH proteins, involving literally thousands of hGHvariants. Many of the variants did indeed exhibit en-hanced affinity for the hGH binding protein (receptor),

but none exhibited significantly enhanced biological ac-tivity. Each data point (i. e., the biological activity ofeach individual variant) essentially constituted a fail-ure of a technology to provide a desired result andwas therefore seen, not unreasonably, as negative (andtherefore uninteresting) data by the investigators. Con-sequently, the information was not broadly disseminateduntil a significant conclusion could be presented (i. e.,that biological responses to hGH are not limited byligand-receptor affinity). Thus, between the late 1980sand the 1999 publication, this information was knownprimarily and unambiguously only to the investigatorsinvolved, even though it would have been critical to anytechnologist planning to affinity mature hGH by phagedisplay for enhanced activity [17.134].

17.5.3 Cultural DifferencesBetween Technologistsand Biologists

Biology is the realm of particularity, whereas more phys-ical sciences are realms within which general rules canbe derived and broadly applied. The vast complexityof biology usually drives biologists to be highly spe-cialized, perhaps making a career from the study ofparticular macromolecules (specific enzymes, growthfactors, genetic elements, etc.) or biological systems.Biologists are, therefore, neither trained nor encouragedby funding agencies to see themselves as fungible, andfrequently they are very reluctant to step away fromtheir primary research focuses into new areas. Doing sois a major undertaking, and a substantial risk, for mostbiologists.

Engineers, on the other hand, often see themselvesas operating from first principles and are more (if notentirely) willing to step into new areas. After all, cross-ing disciplines is feasible for engineers working in thephysical sciences, whereas it is much more difficult, notonly for reasons of training but also because of the bio-logical realities discussed above, for biologists. Also byreason of their training and experience, engineers oftenvastly underestimate the complexity of biological sys-tems and sometimes exhibit what biologists perceive asnaïveté in their approach to biological systems. Unfor-tunately, or fortunately, depending on your perspective,therapeutic nanodevices require both biological and en-gineering expertise. An often bruising debate as to howbest to prepare students for practice in this extremelychallenging field is currently playing out in biomedi-cal engineering and other related academic departmentsacross the country.

PartB

17.5

Therapeutic Nanodevices References 317

The traditional approach for training biomedical en-gineers has been to train conventional engineers witha smattering of biological knowledge. This “primarilyengineers” model of biomedical engineering has beenreasonably successful in multiple endeavors (orthope-dics, imaging technologies, etc.), but it seems unlikelyin the extreme that it will suffice for robust practicein nanotherapeutic devices. As this chapter shows, therole of biology is pivotal to success in the field, and so-phisticated nanotherapeutics are virtually impossible todesign without deep knowledge of the biology involved.A reasonable argument can be made that successfulengineering of therapeutic nanostructures will demandintense focus on biology and a smattering of engineer-ing knowledge (a “primarily biologists” approach, as

opposed to the historically prevalent “primarily engi-neers” training model). This view is often strongly andpassionately resisted in colleges of engineering, whichare the usual homes of biomedical engineering and re-lated efforts. The outcome of the debate is not finalized,but it is critically important: it will determine whetheror not a ready cadre of researchers is prepared for a ca-reer in the area. If not, the field will progress muchas it has, very slowly bringing the benefits of the re-search to patients. It seems apparent to us, and it alsoseems that experience to date shows, that a substan-tial biological component of training for the disciplineis simply unavoidable if we are committed to real-izing the vast potential of nanotechnology in humantherapy.

References

17.1 C. A. Janeway, P. Travers, M. Walport, J. D. Capra:Immunobiology (Elsevier, London 1999)

17.2 S. C. Lee: Biotechnology for nanotechnology, TrendsBiotechnol. 16 (1998) 239–240

17.3 S. C. Lee: Engineering the protein components ofnanobiological devices. In: Biological Molecules inNanotechnology: The Convergence of Biotechnol-ogy, Polymer Chemistry and Materials Science, ed.by S. C. Lee, L. Savage (IBC, Southborough 1998)pp. 67–74

17.4 S. C. Lee: How a molecular biologist can wind uporganizing nanotechnology meetings. In: Biologi-cal Molecules in Nanotechnology: The Convergenceof Biotechnology, Polymer Chemistry and MaterialsScience, ed. by S. C. Lee, L. Savage (IBC, Southbor-ough 1998)

17.5 S. C. Lee: The nanobiological strategy for construc-tion of nanodevices. In: Biological Molecules inNanotechnology: The Convergence of Biotechnol-ogy, Polymer Chemistry and Materials Science, ed.by S. C. Lee, L. Savage (IBC, Southborough 1998)pp. 3–14

17.6 S. C. Lee: A biological nanodevice for drug delivery.In: National Science and Technology Council. IWGNWorkshop Report: Nanotechnology Research Direc-tions. International Technology Research Institute,World Technology Division (Kluwer, Baltimore 1999)pp. 91–92

17.7 S. C. Lee, R. Parthasarathy, K. Botwin: Protein-polymer conjugates: Synthesis of simple nanobio-technological devices, Polymer Preprints 40 (1999)449–450

17.8 L. Jelinski: Biologically related aspects of nanopar-ticles, nanostructured materials and nanodevices.In: Nanostructure Science and Technology, ed. by

R. W. Siegel, E. Hu, M. C. Roco (Kluwer, Dordrecht1999) pp. 113–130

17.9 J. Baish, Y. Gazit, D. Berk, M. Nozue, L. T. Bax-ter: Role of tumor vascular architecture in nutrientand drug delivery: An invasion percolation-basednetwork model, Microvasc. Res. 51 (1996) 327–346

17.10 J. R. Baker Jr.: Therapeutic nanodevices. In:Biological Molecules in Nanotechnology: The Con-vergence of Biotechnology, Polymer Chemistry andMaterials Science, ed. by S. C. Lee, L. Savage (IBC,Southborough 1998) pp. 173–183

17.11 R. Duncan: Drug targeting: Where are we now andwhere are we heading?, J. Drug Targeting 5 (1997)1–4

17.12 R. Duncan, S. Gac-Breton, R. Keane, Y. N. Sat,R. Satchi, F. Searle: Polymer-drug conjugates,PDEPT and PELT: Basic principles for design andtransfer from the laboratory to clinic, J. Cont. Re-lease 74 (2001) 135–146

17.13 D. S. Goldin, C. A. Dahl, K. L. Olsen, L. H. Ostrach,R. D. Klausner: Biomedicine. The NASA-NCI collab-oration on biomolecular sensors, Science 292 (2001)443–444

17.14 S. C. Lee: Dendrimers in nanobiological devices. In:Dendrimers and other Dendritic Polymers, ed. byD. Tomalia, J. Frechet (Wiley, London 2001) pp. 548–557

17.15 J. R. Baker Jr., A. Quintana, L. Piehler, M. Banazak-Holl, D. Tomalia, E. Racka: The synthesis and testingof anti-cancer therapeutic nanodevices, Biomed.Microdevices 3 (2001) 61–69

17.16 T. Desai, D. Hansford, L. Kulinsky, A. Nashat,G. Rasi, J. Tu, Y. Wang, M. Zhang, M. Ferrari:Nanopore technology for biomedical applications,Biomed. Microdevices 21 (1999) 11–40

PartB

17


17.17 T. A. Desai, W. H. Chu, J. K. Tu, G. M. Beattie,A. Hayek, M. Ferrari: Microfabricated immunoiso-lating biocapsules, Biotechnol. Bioeng. 57 (1998)118–120

17.18 M. Ferrari, W. H. Chu, T. A, Desai, J. Tu: Micro-fabricated silicon biocapsule for immunoisolationof pancreatic islets, Advanced Manufacturing Sys-tems, CISM Courses and Lectures, Vol. 372, ed. byE. Kuljanic (Springer, Wien 1996) pp. 559–568

17.19 J. M. Harris, N. E. Martin, M. Modi: Pegylation:A novel process for modifying pharmacokinetics,Clin. Pharmacokin. 40 (2001) 539–551

17.20 S. B. H. Kent: Building proteins through chemistry:Total chemical synthesis of protein molecules bychemical ligation of unprotected protein segments.In: Biological Molecules in Nanotechnology: TheConvergence of Biotechnology, Polymer Chemistryand Materials Science, ed. by S. C. Lee, L. Savage(IBC, Southborough 1998) pp. 75–92

17.21 S. C. Lee, R. Parthasarathy, T. Duffin, K. Botwin,T. Beck, G. Lange, J. Zobel, D. Jansson, D. Kun-neman, E. Rowold, C. F. Voliva: Antibodies toPAMAM dendrimers: Reagents for immune detec-tion assembly and patterning of dendrimers. In:Dendrimers and other Dendritic Polymers, ed. byD. Tomalia, J. Frechet (Wiley, London 2001) pp. 559–566

17.22 S. C. Lee, M. S. Leusch, V. A. Luckow, P. Olins:Method of producing recombinant viruses in bac-teria, (1993) US Patent 5,348,886

17.23 M. S. Leusch, S. C. Lee, P. O. Olins: A novel host-vector system for direct selection of recombinantbaculoviruses (bacmids) in Escherichia coli, Gene.160 (1995) 191–194

17.24 V. A. Luckow, S. C. Lee, G. F. Barry, P. O. Olins: Effi-cient generation of infectious recombinant bac-uloviruses by site-specific, transposon-mediatedinsertion of foreign DNA into a baculovirus genomepropagated in E. coli, J. Virol. 67 (1993) 4566–4579

17.25 T. Gardner, C. R. Cantor, J. J. Collins: Construction ofa genetic toggle switch in Escherichia coli, Nature403 (2000) 339–342

17.26 J. Hasty, F. Isaacs, M. Dolnik, D. McMillen,J. J. Collins: Designer gene networks: Towardsfundamental cellular control, Chaos 11 (2001) 107–220

17.27 E. Di Fabrisio, A. Nucara, M. Gentili, R. Cingolani:Design of a beamline for soft and deep lithographyon third generation synchrotron radiation source,Rev. Sci. Instrum. 70 (1999) 1605–1613

17.28 G. M. Dykes: Dendrimers: A review of their appealand applications, J. Chem. Tech. Biotech. 76 (2001)903–918

17.29 R. Spindler: PAMAM starburst dendrimers: Designednanoscopic reagents for biological applications. In:Biological Molecules in Nanotechnology: The Con-vergence of Biotechnology, Polymer Chemistry and

Materials Science, ed. by S. C. Lee, L. Savage (IBC,Southborough 1998) pp. 15–32

17.30 D. Tomalia, H. M. Brothers II: Regiospecific conju-gation to dendritic polymers to produce nanode-vices. In: Biological Molecules in Nanotechnology:The Convergence of Biotechnology, Polymer Chem-istry and Materials Science, ed. by S. C. Lee,L. Savage (IBC, Southborough 1998) pp. 107–120

17.31 B. I. Yacobson, R. E. Smalley: Fullerene nanotubes:C1000000 and beyond, Amer. Scient. 85 (1997) 324–337

17.32 D. J. A. Crommelin, R. D. Sindelar: Pharmaceut-ical Biotechnology (Harwood Academic Publishers,Amsterdam 1997)

17.33 L. J. Lee: BioMEMS and micro-/nano-processing ofpolymers – An overview, Chin. J. Chem. Eng. 34(2003) 25–46

17.34 T. R. Groves, D. Pickard, B. Rafferty, N. Crosland,D. Adam, G. Schubert: Maskless electron beamlithography: Propects, progress and challenges,Microelectron. Eng. 61 (2002) 285–293

17.35 M. Guthold, R. Superfine, R. Taylor: The rulesare changing: Force measurements on singlemolecules and how they relate to bulk reac-tion kinetics and energies, Biomed. Microdevices3 (2001) 9–18

17.36 L. M. Demers, D. S. Ginger, S.-J. Park, Z. Li,S.-W. Chung, C. A. Mirkin: Direct patterning ofmodified oligonucleotides on metals and insulatosby dip-pen nanolithography, Science 296 (2002)1836–1838

17.37 K.-B. Lee, S.-J. Park, C. A. Mirkin, J. C. Smith,M. Mrksich: Protein nanoarrays generated by dip-pen nanolithography, Science 295 (2002) 1702–1705

17.38 M. Ferrari, J. Liu: The engineered course of treat-ment, Mech. Eng. 123 (2001) 44–47

17.39 H. W. Rohrs, R. S. Ruoff: The use of carbon nano-tubes in hybrid nanometer-scale devices. In:Biological Molecules in Nanotechnology: The Con-vergence of Biotechnology, Polymer Chemistry andMaterials Science, ed. by S. C. Lee, L. Savage (IBC,Southborough 1998) pp. 33–38

17.40 K. E. Drexler: Engines of Creation: The Coming Eraof Nanotechnology (Anchor Books, New York 1986)

17.41 J. Cumings, A. Zetti: Low-friction nanoscale linearbearing realized from multiwall carbon nanotubes,Science 289 (2000) 602–604

17.42 D. J. Hornbaker, S.-J. Kahng, S. Mirsa, B. W. Smith,A. T. Johnson, E. J. Mele, D. E. Luzzi, A. Yazdoni:Mapping the one-dimensional electronic states ofnanotube peapod structures, Science 295 (2002)828–831

17.43 C. Dekker: Carbon nanotubes as molecular quan-tum wires, Phys. Today 28 (1999) 22–28

17.44 M.-C. Jones, J.-C. Leroux: Polymeric micelles-a new generation of colloidal drug carriers, Eur.J. Pharma. Biopharma. 48 (1999) 101–111

PartB

17

Documents

Therapeutic N 17. Therapeutic Nanodevices · Therapeutic N 17. Therapeutic Nanodevices Therapeutic nanotechnology offers minimally invasive therapies with high densities of func-tion