Stimulus-responsive macromolecules and nanoparticles for cancer drug delivery

Abstract

Nanoparticles and macromolecular carriers have been widely used to increase the efficacy of chemotherapeutics, largely through passive accumulation provided by the enhanced permeability and retention effect. Stimulus-responsive peptide and polymer vehicles can further enhance the efficacy of antitumor therapeutics compared with the administration of free drug by three mechanisms: increasing the overall accumulation within solid tumors; providing a homogeneous spatial distribution in tumor tissues; and increasing the intracellular localization of anticancer therapeutics. This article highlights recent developments in ‘smart’ – stimulus-responsive – peptide, polymer and lipid drug carriers designed to enhance the localization and efficacy of therapeutic payloads as compared with free drug.

Tumor accumulation of macromolecules & nanoparticles: increasing local concentration of therapeutics

Anticancer therapeutics are plagued by a low therapeutic index – where the efficacious dose approaches the lethal dose – as most cancer drugs typically act on all actively dividing cells [1]. The primary goal of cancer drug delivery is to improve this typically modest therapeutic index by developing drug carriers that enhance the accumulation of drug within the tumor while limiting accumulation in healthy organs. Macromolecules and nanoparticles can help achieve this goal through passive accumulation within many solid tumors due to the enhanced permeability and retention (EPR) effect, which is a consequence of the aberrant features of tumor vasculature and the poorly developed lymphatic system that limits drainage of molecules from tumor tissue [2]. The rapid development of tumor vasculature by abnormal and poorly controlled angiogenesis results in vessel walls with pores ranging in size from 200 nm to 2 μm, with an average pore size reported to be approximately 400 nm [3]. The leaky tumor vasculature, owing to the presence of these pores, allows macromolecules and nanoparticles to preferentially extravasate into the tumor in contrast to healthy blood vessels, where intact tight junctions between endothelial cells prevent transvascular migration of drug carriers and cargo into healthy tissues (Figure 1) [4]. This enhanced permeability of the tumor vasculature is further affected by anatomical features, such as irregular tumor-vessel architecture, and thus blood flow, and by overexpression of permeability-enhancing factors, including VEGF, bradykinin, matrix metalloproteinases and cytokines that are particular to the tumor environment [4]. In addition, the lack of effective lymphatic drainage prolongs the residence of macromolecules and nanoparticles in the tumor as compared with healthy tissue [2]. The key to exploiting the EPR effect is to retain a high concentration of the drug in the systemic circulation as a function of time to maximize diffusion out of the leaky blood vessels into extravascular tumor tissue. One method to achieve this goal with low-molecular weight (MW) chemotherapeutics is by conjugation to a high-MW carrier (typically ranging from 10⁴ to 10⁶ Da), which results in prolonged circulation, increased accumulation in the tumor, and reduced exposure of critical organs to the drug. For example, the EPR effect, in conjunction with drug carriers, typically leads to a tumor-to-blood drug concentration ratio of 10–30, with a few reports above 2000 [4]. This has the clinically relevant effect of increasing the maximum tolerated dose of anticancer drugs, thereby allowing more drug to be administered to greater effect, and thus increasing the therapeutic window. This principle is generally believed to apply to many solid tumors, making macromolecule and nanoparticle carriers widely useful for anticancer drug delivery. However, there is enough evidence to support the notion that the parameters that control the EPR effect are variable between species [5,6], tumor implantation sites [7,8] and tumor types [7], so that the success of passive targeting, similar to all approaches in drug delivery, is also likely to be variable.

Figure 1. Macromolecule and nanoparticle drug carriers exhibit improved tumor accumulation as compared with low-molecular weight drugs owing to the enhanced permeability and retention effect.

Stimulus-responsive drug carriers can further enhance accumulation of the drug, for instance by temperature-triggered coacervation of thermally responsive carriers in heated tumor tissue. Mild hyperthermia of tumors following systemic administration of temperature-responsive biopolymers leads to aggregation of drug vehicles at the vascular wall. When external heating of the tumor is turned off, the aggregated drug carriers redissolve in the tumor vasculature and are driven into the extravascular tumor space owing to the increased local concentration achieved by their aggregation within the tumor vasculature.

EPR: Enhanced permeability and retention.

Improving tumor accumulation with macromolecules & nanoparticles

The discovery of the EPR effect led to a series of studies that investigated the potential of polymer–drug conjugates to improve upon existing chemotherapy. Polymers, including styrene-maleic acid, polyethylene glycol (PEG), poly(l-glutamic acid) and poly(N-[2-hydroxypropyl] methacrylamide), were used to deliver anticancer therapeutics, such as paclitaxel, doxorubicin, camptothecin and protein drugs, to solid tumors [9]. Conjugation of doxorubicin to poly(N-[2-hydroxypropyl] methacrylamide) prolonged circulation and improved tumor accumulation of the drug, leading to enhanced antitumor efficacy compared with free drug [10]. Conjugation of camptothecin to poly(l-glutamic acid) achieved greater antitumor activity than free drug in melanoma, colon and lung cancers, with greater drug loading and a higher polymer MW increasing this activity [11]. Conjugation of dendrimers with doxorubicin and camptothecin also demonstrated improved pharmacokinetics, biodistribution and antitumor efficacy compared with free drug [12,13]. Protein–drug conjugates have also demonstrated advantages over free drug; one example is albumin-bound paclitaxel (Abraxane^®), in clinical use for breast cancer therapy, which demonstrates significantly lower toxic side effects in comparison to free paclitaxel delivered with Cremophor EL^®-based solvents [14].

This concept of improved delivery of drugs by conjugation to soluble macromolecules was then extended to drug-containing polymer nanoparticles. Amphiphilic block copolymers that self-assemble into micelles are one type of polymer nanoparticle that are effective for the encapsulation of hydrophobic drugs in their core. Doxorubicin has been encapsulated within micelles of PEG–poly(aspartic acid) [15], cisplatin and camptothecin have been encapsulated in PEG–poly(glutamic acid) micelles [16], and styrene–maleic acid copolymer has been used to encapsulate doxorubicin and pirarubicin [17,18], all resulting in formulations that have demonstrated enhanced antitumor activity with reduced toxicity compared with free drug. The improved biodistribution and increased accumulation of these carriers, such as paclitaxel-loaded PEG–poly(aspartic acid) micelles, have resulted in their clinical evaluation for the treatment of colon and stomach cancer [19].

We have developed a class of recombinant polypeptide micelles as drug-delivery vehicles. Conjugation of multiple hydrophobic doxorubicin molecules to the C-terminus of a hydrophilic elastin-like polypeptide imparted sufficient amphiphilicity to the conjugate to trigger its self-assembly into nanoparticles, with doxorubicin buried in the nanoparticle core. This nanoparticle drug carrier provided an increase in tumor accumulation and a reduction in peak heart accumulation at 24 h compared with an equivalent dose of free doxorubicin, resulting in a fourfold increase in the maximum tolerated dose [20] and led to near-complete regression of tumors in the syngeneic C26 colon carcinoma model in BALB/c mice. Liposomal formulations have also achieved improved antitumor efficacy, with a PEGylated liposome that encapsulates doxorubicin (Doxil^®) as an example of a liposomal carrier that is now in clinical use [21,22].

Macromolecular and nanoparticle carriers have been functionalized with tumor-specific peptides, antibodies and aptamers in an effort to improve their tumor targeting and cellular uptake. The underlying hypothesis of this approach is that targeted interactions should result in improved tumor accumulation and retention, and hence to an improved therapeutic effect [23]. Actively targeted carriers have demonstrated enhanced antitumor activity in vivo, but mixed results have been reported regarding their effect on tumor accumulation [24,25]. An emerging view is that the overall tumor accumulation of functionalized nanoparticles is principally controlled by the EPR effect, while improvement in the therapeutic effect of actively targeted carriers, in comparison to passively accumulating nanoparticles, is most likely due to their enhanced uptake by tumor cells rather than their increased overall accumulation in the tumor [26]. As active targeting by presentation of ligands that bind to overexpressed tumor receptors or molecular targets seems to have an inconsistent effect on the tumor distribution of nanoparticle drug carriers, there is a clear need to explore the effects of other targeting approaches that do not solely rely upon receptor–ligand interactions. Stimulus-responsive materials represent an alternative and, we believe, a promising avenue for enhancing tumor accumulation beyond that afforded by the passive targeting of the EPR effect and active targeting with ligand-functionalized drug carriers.

Improving tumor accumulation with stimulus-responsive carriers

Stimulus-responsive carriers provide additional opportunities to improve the tumor accumulation of systemically delivered drug carriers by triggering changes in material properties to increase the local concentration of the drug within solid tumors. The stimulus may be extrinsic, such as the local application of heat, ultrasound or light, or intrinsic to the tumor environment, such as its low extracellular pH or upregulated protease expression. Table 1 provides a summary of the drug carriers discussed in this article.

Table 1.

Summary of drug-delivery vehicles and stimulus-responsive drug carriers that have been developed to increase targeted accumulation of a drug, enhance its distribution, and control its intracellular or subcellular localization, with the ultimate goal of increasing the efficacy of anticancer therapeutics.

Drug carrier	Stimulus response	Drug	Application	Ref.
Abraxane® (albumin–drug conjugate)	–	Paclitaxel	Clinically approved for the treatment of breast cancer	[14]
NK105 (drug-loaded PEG–poly(aspartic acid) micelle)	–	Paclitaxel	In clinical trials for the treatment of colon and stomach cancers	[19]
Doxil® (drug-loaded PEGylated liposome)	–	Doxorubicin	Clinically approved for the treatment of recurrent ovarian cancer	[22]
ELP unimer	Temperature-triggered aggregation	–	Increased accumulation in hyperthermia treated tumors in mice	[34]
Dextran–peptide–drug conjugate	MMP cleavage of peptide linker for release of free drug	Methotrexate	Improved inhibition of tumor growth in subcutaneous murine models of human fibrosarcoma and glioblastoma	[40]
Poly(histidine)-β-PEG, poly(l-lactic acid)-β-PEG mixed micelle	pH-triggered disassembly for release of free drug	Doxorubicin	Increased accumulation and improved efficacy in subcutaneous breast cancer tumors in mice	[47]
Pluronic® micelle	Ultrasound-induced disassembly for release of free drug	Doxorubicin	Improved regression with ultrasound treatment in subcutaneous colon cancer tumors in mice in comparison to treatment with drug-loaded micelles alone	[50]
Oligoarginine–peptide–oligoglutamate conjugate	MMP-2 cleavage of peptide linker for CPP activation	–	Increased accumulation compared with uncleavable controls in a variety of subcutaneously implanted tumor types in mice	[65]
PEG–poly(l-histidine), poly(l-lactic acid)–PEG–poly(l-histidine)–TAT mixed micelle	pH-triggered display of TAT for CPP activation	Doxorubicin	Selective cellular uptake in slightly acidic conditions of pH 7.0 improved cytotoxicity in drug-resistant breast cancer cells	[71]
RGD-functionalized ELP diblock	Temperature-triggered micelle assembly for polyvalent ligand display	–	Increased accumulation with hyperthermia-triggered micelle assembly in leukemia cells overexpressing αvβ3 integrin	[77]
Thiolated heparin nanogel	Destabilization of disulfide bonds in reducing intracellular environment for free drug release	Heparin	Increased cytotoxicity by induction of apoptosis in melanoma cells as compared with free drug	[84]
ELP–drug conjugate micelle	pH-triggered drug release in acidic endosomal compartment	Doxorubicin	Enhanced accumulation, increased MTD and improved regression of colon carcinoma tumors in mice	[20]
Poly(l-histidine)-based micelle	pH-triggered protonation of histidine for endosomal disruption	Doxorubicin	Improved cytotoxicity of intracellularly released drug in doxorubicin-resistant ovarian carcinoma cells	[92]
Amidized poly(l-lysine)–drug conjugate	pH-triggered charge reversal for nuclear targeting	Camptothecin	Nuclear localization of drug-enhanced cytotoxicity, compared with free drug, in adenocarcinoma cells	[93]
TPP-modified liposome	–	Ceramide	Improved inhibition of tumor growth with mitochondrial targeting in a subcutaneous mouse model of mammary carcinoma tumors	[94]

CPP: Cell-penetrating peptide; ELP: Elastin-like polypeptide; MMP: Matrix metalloproteinase; MTD: Maximum tolerated dose; PEG: Polyethylene glycol; RGD: Arginine–glycine–aspartic acid; TAT: Transactivator of transcription; TPP: Triphenylphosphonium.

Thermally responsive peptide and polymer constructs provide an interesting example of this class of responsive materials, as their local accumulation can be triggered by the extrinsic application of heat (Figure 1); the level of heating required to trigger their accumulation is consistent with that used in mild clinical hyperthermia, so that this modality can be applied in conjunction with existing clinical hyperthermia regimens, including focused microwaves, radiofrequency and ultrasound, that have been developed for localized heating in cancer therapy [27–29]. Temperature-responsive macromolecules exhibiting a lower critical solution temperature transition are soluble below their transition temperature (T_t) and undergo aggregation into insoluble polymer coacervates above their T_t. Adjusting the T_t of thermally responsive polymers to occur between body temperature (37°C) and the temperature approved for mild clinical hyperthermia (42°C) [27] allows these polymers to improve tissue accumulation by localizing the aggregation of systemically delivered carriers to the heated tumor volume [30,31].

The lower critical solution temperature transition behavior of elastin-like polypeptides (ELPs), biopolymers that are inspired by human tropoelastin, can be exploited to improve the delivery of attached therapeutics to tumors that are subjected to mild hyperthermia. The distribution and accumulation of systemically delivered fluorescently tagged ELPs have been evaluated by intravital-scanning confocal microscopy in nude mice bearing tumors in dorsal-fold window chambers [30,32], revealing the aggregation of thermally responsive ELP in real time, along the vascular walls of tumors subjected to hyperthermia. The reversibility of temperature-triggered aggregation in vivo allowed the rapid dissolution of ELP aggregates within the tumor vasculature upon cessation of hyperthermia, and markedly increased the local vascular concentration of soluble ELP. The temperature-triggered aggregation of the ELP provided the necessary concentration gradient for enhanced diffusion of soluble ELP after cessation of hyperthermia into the extravascular space of the tumor and increased tumor accumulation compared with a soluble ELP control that does not undergo its phase transition in the applied temperature range [32–34]. ELPs are one example within a broad class of stimulus-responsive materials whose physicochemical properties can be manipulated by extrinsic cues or triggers intrinsic to the tumor environment to improve tumor accumulation.

Intratumoral distribution of macromolecule & nanoparticle payloads: improving tumor penetration of cancer therapeutics

The spatial distribution of drugs within tumors is an important issue that has received inadequate attention in the field of drug delivery. The physiological properties of tumors, such as their cellular packing and dense extracellular matrix, result in limited diffusion of many small-molecule drugs in tumor tissue. This low diffusivity problem is compounded by the high interstitial fluid pressure in the tumor core created by the lack of lymphatic drainage, leading to an outward convective gradient [35]. Intravital confocal microscopy of subcutaneously implanted solid tumors in murine dorsal skin-fold window chambers demonstrated the dependence of permeability on carrier size, as an increase in the hydrodynamic radius of dextran, a macromolecule, from 2 to 25 nm decreased the apparent tumor permeability by twofold [36]. Given that drugs attached to or encapsulated within nanoscale vehicles are unlikely to exhibit the desired diffusion through macroscopic tumors to homogeneously distribute within the tumor, macromolecular and nanoparticle drug carriers must release their low-MW cargo at their target to improve drug diffusion throughout the tumor mass. While many nanoparticles are designed to exhibit degradation-dependent drug release (e.g., by hydrolysis), these mechanisms are not responsive to the tumor microenvironment to provide localized release only within the target tissue. Tumor-responsive release of the drug would be very useful in this regard. Drug release within a tumor can be achieved with stimulus-responsive polymers and nanoparticles that release their low-MW drug cargo in response to a tumor-specific stimulus following accumulation within the tumor mass, via the EPR effect (Figure 2). We next discuss tumor-localized drug release triggered by intrinsic cues specific to the pathological microenvironment of solid tumors or extrinsic triggers locally applied at the disease site.

Figure 2. Improved intratumoral distribution can be achieved with release of small-molecular weight drugs in response to tumor-specific stimuli.

Attachment of a drug to carrier proteins (A) or polymers (B) through protease-cleavable linkers allows drug release in the presence of overexpressed matrix metalloproteinases in the tumor tissue. Disassembly of drug-loaded micelles in response to low pH (C) improves the diffusion of drug into acidic tumor tissues.

Many proteases, especially matrix metalloproteinases (MMPs) such as MMP-2 and MMP-9, are known to be upregulated in the tumor microenvironment [37]. The first attempts to exploit a tumor-specific enzymatic stimulus for localized drug release involved attachment of chemotherapeutics to macromolecules via MMP-cleavable linkers. Conjugation of doxorubicin to albumin, via an MMP-sensitive linker, led to far greater cytotoxicity against renal carcinoma cells compared with an MMP-insensitive control. In addition, this doxorubicin–albumin conjugate also exhibited a greater maximum tolerated dose and better antitumor efficacy than free doxorubicin [38,39]. Similarly, conjugation of methotrexate to dextran or doxorubicin to PEG by an MMP-cleavable linker led to greater antitumor activity compared with an MMP-insensitive control [40,41]. Liposomes with MMP sensitivity have also been formulated by integrating an MMP-cleavable lipopeptide into the liposome formulation. This lipopeptide consisted of a lipid component, distinct from other lipids in the liposome bilayer, while the peptide component formed a triple helical conformation that could be unwound and cleaved by MMP-9. Cleavage of the peptide led to instability of the resulting free-lipid component, causing discontinuities in the lipid bilayer that resulted in rapid drug release [42,43].

Other investigators have synthesized nanoparticles that are capable of responding to the low extracellular pH of the tumor site caused by hypoxia-induced lactic acid build-up [44,45]. Mixed micelles, consisting of poly(histidine)-β-PEG and poly(l-lactic acid)-β-PEG, were shown to disassemble in response to the slightly acidic intratumoral pH of 6.5–6.8 owing to the ionization of histidine residues in the interior of the micelle [46]. Doxorubicin was encapsulated in these pH-sensitive micelles, which were shown to effectively treat MCF-7 [47] and A2780 [48] tumors that were implanted in mice, and achieved greater tumor accumulation and antitumor efficacy in comparison to pH-insensitive micelles and free drug [47].

An extrinsic trigger can also be employed to enhance the accumulation of drug within a tumor and provide a homogeneous spatial distribution by disassembly of a drug carrier and release of its payload at the tumor site. Release of drugs by the selective disruption of poly(ethylene oxide)-co-poly-(propylene oxide)-co-poly(ethylene oxide) triblock copolymer (Pluronic^®) micelles has been achieved with the external application of ultrasound to tumor tissue. Fluorescently labeled micelles exhibited a more uniform distribution in subcutaneous ovarian carcinoma tissue in mice treated with local transdermal ultrasound as compared with contralateral control tumors exposed to micelles but not receiving ultrasound [49]. Ultrasound-mediated micelle disruption enhanced the movement of the carrier within the tumor tissue, making this a potentially attractive strategy for providing a homogeneous distribution of the drug throughout the tumor. Pluronic micelles encapsulating doxorubicin significantly decreased tumor volume in rats bearing subcutaneous colon cancer tumors treated in conjunction with ultrasound as compared with tumors not receiving ultrasound treatment [50]. The increased drug release and enhanced tumor distribution both probably contributed to this improved therapeutic effect.

A similarly triggered release of contents from liposomal carriers has also been envisioned to occur by destabilization of the liposome in response to a localized extrinsic light stimulus. Liposomes composed of amphiphiles, whose hydrophilic head groups and hydrophobic tails are connected by a photocleavable linker can be destabilized when exposed to a UV stimulus that dissociates the amphiphile constituents and induces leakiness in the liposome, permitting release of its contents [51,52]. Further development of this approach and coupling of liposomal release to infrared triggers may provide a useful strategy to target payload release from liposomes, even deep within the tumor tissue.

We believe that in the effort to develop and optimize new drug-delivery vehicles, it is critical to quantify not only the total accumulation of a drug within a solid tumor, but also its spatial distribution, as both parameters are likely to dictate the ultimate success or failure of any new delivery system. The intratumoral spatial distribution can be observed most directly using the dorsal-fold window chamber model [36], but intratumoral transport of drugs has also been measured using 3D culture [53] spheroids [54], and microdialysis [55]. Future experiments in this area should focus on determining the direct impact of spatial distribution of chemotherapeutics within tumors on their antitumor effect. Careful analysis is required to investigate the independent contributions of passive targeting via the EPR effect, active targeting to tumors via ligand–receptor binding, and stimuli-responsive targeting via extrinsic (e.g., thermal targeting) or intrinsic (e.g., pH) triggers on the overall accumulation and spatial distribution of the drug within tumors, and to reveal the impact of these parameters on the therapeutic efficacy of ‘smart’ – stimulus-responsive – macromolecular and nanoparticle drug-delivery vehicles.

Improving cellular uptake & controlling intracellular fate

Once macromolecules or nanoparticles have accumulated in the tumor, their stimulus-responsive properties can be further exploited to enhance cellular uptake and to control the fate of intracellularly delivered drugs. Delivering therapeutic cargo into target cells is critical for many drugs that act intracellularly to produce their therapeutic effects; these include many conventional chemotherapeutics, siRNA and gene therapies [56–58]. In addition, the response of drug carriers to unique environmental stimuli presented by intracellular compartments can be exploited to ensure that the drug reaches its site of action. Next, we discuss stimulus-responsive vehicles that can provide increased tumor-localized cellular uptake as well as intracellular drug release and subcellular accumulation.

Localized presentation of bioactive molecules in response to the target tissue environment

One means of controlling cellular uptake is to make the presentation of targeted ligands on drug carriers exclusive to the tumor site (Figure 3). A conventional and long-standing approach to this problem is to exploit ligands that bind to receptors or other molecular targets that are exclusively expressed by tumors. Unfortunately, despite decades of effort devoted to the discovery of unique, tumor-specific targets and ligands that bind to them, there appear to be few, if any, magic bullets that exclusively home to tumors. A somewhat different approach seeks to design ‘smart’ carriers that can expose, unshield or activate a targeting moiety in a drug carrier within the tumor in response to an extrinsic stimulus or an intrinsic trigger that is specific to the tumor environment.

Figure 3. Targeted cellular uptake can be achieved with stimulus-responsive drug carriers.

Intrinsic stimuli provide triggers to induce cellular uptake such as cleavage of stealth polymers by tumor proteases (A) and conformational changes in membrane-penetrating peptides in response to the low pH of tumor tissue (B). Extrinsic triggers, such as heat, can be used to trigger multivalent ligand display (C), enhancing the intracellular delivery of anticancer drug vehicles.

Intrinsic triggers have been utilized in conjunction with cell-penetrating peptides (CPPs) to enhance the uptake of drugs by tumor cells. CPPs are cationic, often arginine-rich, peptide sequences that promote receptor-independent intracellular uptake of a variety of conjugated payloads. The uptake of CPPs by cells is initiated by the interaction of the positively charged guanidine groups of a CPP's arginine residues and the negatively charged proteoglycans that are ubiquitous to the surface of most cells [59]. Since the uptake of CPPs by cells appears to occur by a generic, receptor-independent pathway, CPPs have attracted a great deal of attention as a potential universal methodology to enhance the uptake of a variety of drugs and delivery vehicles by a diverse range of cell lines [60–62]. However, the nonspecific mechanism of CPP uptake creates a confounding problem, as systemically injected CPP-modified drug carriers will lead to enhanced cellular uptake in all tissues as the carrier circulates throughout the body. Increased accumulation within a tumor is, hence, also likely to be accompanied by greatly increased nonspecific uptake by healthy tissues, thereby increasing the toxicity of the carrier.

To address this problem, strategies have been developed to unveil or unmask CPPs only within tumor tissue as a means to endow specificity to CPPs, such that they preferentially enhance the intracellular uptake of a drug or delivery vehicle by tumor cells and minimize off-target uptake by healthy tissue. One approach of creating ‘protected’ or ‘masked’ CPPs was accomplished with stimulus-responsive ionic shielding by conjugation of an aspartic acid-containing anionic peptide to a cationic oligoarginine CPP with an intervening peptide linker that is specifically cleaved by prostate-specific antigen – a cell-surface protease overexpressed in prostate cancer [63]. Intramolecular ionic interactions in the peptide led to the formation of hairpin structures that prevented oligoarginine-induced cellular uptake of intact constructs. Tumor-localized cleavage of the linker by prostate-specific antigen liberated the anionic peptide from the construct, allowing its dissociation from the CPP and consequent unmasking of the oligoarginine CPP. The unmasked CPPs were rapidly internalized by PC3M prostate cancer cells and demonstrated significantly greater cellular uptake than that achieved in the absence of prostate-specific antigen. Similar activatable CPPs, composed of oligoarginine and oligoglutamate components that are cleaved by MMP-2, a protease that is upregulated in many tumors [64], led to a two- to six-fold increase in accumulation in subcutaneous fibrosarcoma, melanoma, and cervical, prostate, colon and breast cancer tumors in mice as compared with noncleavable controls [65]. In an alternative strategy, CPPs were sterically shielded by ‘stealth’ PEG polymers that were selectively cleaved by overexpressed proteases in the tumor environment [66–68]. Conjugation of PEG to CPP-functionalized dextran-coated iron oxide nanoparticles via MMP-2-cleavable peptide linkers allowed protease-triggered shedding of the protective PEG coating [69]. Exposure of CPPs on the surface of the nanoparticles after MMP cleavage of PEG resulted in significantly increased uptake of unveiled particles by human fibrosarcoma cells as compared with nanoparticles whose PEG coating remained intact.

The low extracellular tumor pH [44] has also been exploited as an intrinsic stimulus to trigger localized exposure of bioactive targeting ligands. Nanoparticles capable of revealing targeting ligands in response to the weakly acidic tumor tissue include those formulated from a mixture of PEG–poly(l-histidine) and ligand-functionalized poly(l-lactic acid)–PEG–poly(l-histidine)–biotin [70]. At physiologic pH of 7.4, the polymer mixture formed micelles with a poly(l-lactic acid)- and poly(l-histidine)-containing core, a PEG corona, and biotin ligands buried within the PEG shell at the interface between the hydrophilic corona and the hydrophobic core. A slight decrease in pH preferentially ionized the histidine side chains in the poly(l-lactic acid)–PEG–poly(l-histidine)–biotin segment of the micelle, increasing the hydrophilicity of the biotin-terminated poly(l-histidine) block and allowing its extension into the hydrophilic corona to display biotin beyond the PEG shell. A decrease in pH from 7.2 to 6.8 led to sufficient exposure of biotin to significantly increase the uptake of fluorescently labeled carriers in breast adenocarcinoma cells expressing biotin receptors. Extension of this approach to selectively display the TAT CPP sequence on the micelle surface demonstrated a 30-fold increase in cellular uptake at pH 7.0 in comparison to physiological pH of 7.4 [71]. Encapsulation of doxorubicin in these micelles led to pH-selective cytotoxicity, with decreased cellular viability triggered only upon lowering of pH to 7.0, which led to greater cytotoxicity than free doxorubicin in drug-resistant breast tumor cells.

The acidic environment of tumors can also serve as an intrinsic stimulus to trigger a conformational change in a carrier that activates a bioactive function. One example of this approach is the use of the pH (low) insertion peptide that transforms from an inactive state at physiologic pH to an α-helical conformation in acidic conditions that allows its insertion into cellular membranes, resulting in its cellular uptake [72]. Localized membrane transduction and targeted tumor accumulation of pH (low) insertion peptide conjugates has been demonstrated in low pH environments of breast adenocarcinoma seeded in mouse flanks [72].

Extrinsic stimuli offer an alternative approach to achieve localized cellular uptake. For instance, external triggers can be used to locally trigger polyvalent display of low- affinity ligands, creating a drug carrier with high avidity for upregulated tumor cell receptors only in target tissue. The linear arginine–glycine–aspartic acid (RGD) peptide was used to demonstrate this approach as it is a low-affinity ligand to the α_vβ₃ integrin [73] that is believed to be overexpressed in the vasculature of many tumors [74,75]. Temperature-responsive amphiphilic ELP diblocks that present the linear RGD peptide on the terminus of the hydrophilic block were shown to exhibit thermally triggered self-assembly into micelles above their critical micelle temperature, leading to polyvalent display of the RGD ligand [76]. These diblock ELPs are soluble unimers below their critical micelle temperature, which ensures monovalent presentation of the ligand and hence a low affinity of the ELP for the α_vβ₃ integrin. Heating the ELPs above their critical micelle temperature triggered their self-assembly into micelles that present multiple copies – approximately 50–160 depending on the specific diblock ELP – of the RGD ligand on their corona. We demonstrated that temperature-triggered micelle assembly of RGD-presenting diblock ELPs under hyperthermic conditions led to significant uptake of the ELP by K562 human leukemia cells that overexpress the α_vβ₃ integrin, leading to a high level of integrin-mediated cellular uptake. In comparison, insignificant cellular uptake occurred with the same ELP under normothermic conditions in which the RGD-terminated diblock ELPs exist as unimers [77]. These results suggest the intriguing possibility of using an external – thermal – stimulus to trigger the assembly of high-affinity nanoparticles from low-affinity constituents, although further development of this system is needed to evaluate its potential for in vivo applications under temperature conditions consistent with the clinical application of mild hyperthermia. We believe that the temperature-triggered self-assembly of high-avidity drug carriers with enhanced cellular uptake in the tumor, while maintaining low affinity in healthy tissues at off-target sites, has the potential to be a powerful approach in targeting cancer markers that are overexpressed in tumors, but also exhibit low levels of expression in healthy tissues, such that enhanced tumor accumulation can be achieved while minimizing systemic uptake and decreasing toxicity.

Nanoparticle response to intracellular environments

The endocytosis of nanoscale drug-delivery vehicles often results in accumulation of the drug carrier and its therapeutic cargo within intracellular endosomal vesicles and lysosomal compartments. The environment of these organelles – their reductive potential and acidic pH – is distinct from the extracellular space, and provides an opportunity to exploit these unique properties to trigger drug release from the macromolecular carrier.

Differences in the reducing potential between the extra- and intra-cellular space [78] provides another, complementary trigger for drug delivery. Disulfide bonds, for example, can be used to attach a drug to a carrier, and exposure to the reductive environment of endosomes or the cytosol can lead to cleavage of these disulfide bonds and result in release of the drug from its carrier [79]. Alternatively, a drug can be encapsulated within a nanostructure whose core is stabilized by disulfide bonds, so that disassembly of the nanoparticle in a reductive environment triggers the release of the encapsulated drug from the nanoparticle [80–82]. An example of the former approach includes diblock copolymers of hydrophilic (poly[N-isopropylacrylamide]) and hydrophobic (2-hydroxyethyl methacrylate) connected by a disulfide linker, which were capable of micelle assembly and encapsulation of Nile red dye in their hydrophobic cores [79]. Under the reducing conditions created by the addition of dithiothreitol or glutathione, the cleavage of the diblock copolymers led to micelle disassembly and precipitation of the hydrophobic dye as it was released from the micelle core [83]. The latter approach has been demonstrated with nanogels formulated with PEG and thiolated heparin – a drug that can induce apoptosis of cancer cells [84]. The intermolecular disulfide bonds between heparin molecules created stable nanogels of approximately 250 nm in diameter that demonstrated enhanced cellular uptake when compared with free heparin. In the reducing intracellular environment, the destabilization of disulfide bonds allowed the release of free heparin, inducing apoptosis in B16F10 mouse melanoma cells and resulting in a greater cytotoxic effect than that achieved with free heparin.

The decrease in pH in the progression from endosomes to lysosomes can similarly provide a trigger for intracellular drug release. A simple means of exploiting this local change in pH is the conjugation of drug to a polypeptide or polymer carrier through a pH-labile linker [85,86]. These macromolecular carriers provide the usual advantages conferred by hydrophilic carriers to hydrophobic, insoluble cancer drugs, such as improved solubility, prolonged circulation and delayed clearance, but also allow release of free drug only in acidic lysosomal compartments. In an example of this approach, conjugation of a single doxorubicin molecule to ELP, through a pH-sensitive hydrazone bond, allowed release of drug from the macromolecular carrier and the pH-sensitive release rate of the drug could be tuned by the length of the pH-labile drug linker [87,88]. In an evolution of this approach, multiple doxorubicin molecules were then conjugated to each ELP using the optimal linker identified in the previous study to increase the drug loading to a level suitable for therapy; the conjugation of multiple drug molecules triggered the self-assembly of the conjugate into approximately 40-nm diameter nanoparticles. These ELP–doxorubicin nanoparticles were stable in blood upon systemic injection, exhibited a long – approximately 9.5 h – terminal half-life and good blood exposure, and exhibited significant accumulation in subcutaneous tumors implanted in mice. The pH-labile drug linkers that were used to attach the drug to the ELP enabled facile release of the drug in the acidic endosomal compartment of cells, with eventual localization of doxorubicin to tumor-cell nuclei [20].

The low pH of endosomal and lysosomal compartments has also been exploited to trigger changes in material properties that induce lysis of intracellular vesicles [70,89]. For example, protonation of histidine residues at low pH encourages destabilization of endosomal compartments allowing release of their contents into the cytoplasm [90,91]. The inclusion of histidines in peptide–polymer diblocks capable of micelle assembly provides a histidine-rich core that allows encapsulation of hydrophobic drugs, such as doxorubicin, and promotes the release of drug molecules into the cytosol following micelle disassembly and histidine protonation in early endosomes. The resulting delivery of large quantities of drug to the cytoplasm improved cytotoxicity, even in otherwise doxorubicin-resistant ovarian carcinoma cells [92].

As many chemotherapeutics exert their therapeutic action in the nucleus, the targeting of a drug to the nucleus of tumor cells is an attractive approach to improve its efficacy. Localization of the drug carrier to the nucleus by incorporation of a positively charged component that is attracted to the negative charge of the nucleus is a simple means of promoting nuclear accumulation. A high density of positive charge on a polymer drug carrier, however, has the undesirable consequence of also promoting nonspecific uptake of the carrier by cells in blood and normal tissues that come into contact with the carrier as it transits through the systemic circulation to the tumor. Hence, spatiotemporal control of the charge of a drug carrier is required for effective nuclear localization and to minimize systemic toxicity by restricting the display of positive charge subsequent to its uptake by tumor cells. In one implementation of this approach, the amine groups in poly(l-lysine) were modified with β-carboxylic acid so that the polymer was negatively charged at neutral pH, but was hydrolyzed at low pH to return poly(l-lysine) to its positively charged state [93]. The negative charge of the amidized polymer at physiologic pH should prevent interaction with blood proteins and reduce its toxicity, while the pH-triggered reversal to the cationic state in the acidic environment of the endosome should promote endosomal escape and localization to the cell nucleus. Conjugation of camptothecin, which acts on topoisomerase, a nuclear enzyme, to amidized poly(l-lactic acid) by a linker that is cleaved by glutathione, allowed the release of free drug in the reducing environment of the cytosol once the carrier was taken up by cells, thus enhancing the localization of the drug to the nucleus. Fluorescently labeled constructs exhibited nuclear localization in adenocarcinoma SKOV-3 cells and achieved enhanced cytotoxicity in comparison to free drug, presumably due to the enhanced accumulation of drug at the nuclear site of action.

Owing to their important role in regulating cell death, mitochondria are alternative subcellular targets that are particularly important for the delivery of apoptosis-inducing therapeutic agents. Although mitochondrial targeting has yet to be incorporated into a stimulus-responsive delivery vehicle, progress in mitochondrial targeting has been made with the identification of mitochondriotropic agents that localize to the mitochondria. For example, the incorporation of a mitochondria-targeting triphenylphosphonium cation to liposome surfaces led to colocalization of fluorescently labeled carriers with mitochondria within 2 h of incubation with mammary carcinoma cells [94]. Systemic delivery of ceramide drug-loaded triphenylphosphonium-modified liposomes to subcutaneous tumors in mice revealed greater inhibition of tumor growth with mitochondria-targeted carriers in comparison to nontargeted controls, despite similar levels of accumulation in the tumor tissue. Future incorporation of such mitochondrial targeting components into stimulus-responsive drug vehicles may further improve drug efficacy while ensuring the prevention of off-target effects in healthy tissues.

Conclusion

Macromolecular and nanoparticle carriers are capable of improving the efficacy of chemotherapeutics by their enhanced passive accumulation in tumor tissue as compared with low-MW drugs. Designing drug vehicles with stimulus-responsive constituents allows for dynamic material responses to intrinsic and extrinsic triggers that can be leveraged to further improve upon the favorable tumor accumulation of therapeutic carriers. The changes in material properties of stimulus-responsive polymers triggered by intrinsic characteristics of the pathophysiological tumor environment or tumor-localized extrinsic cues can be used to increase tumor drug concentration, enhance the intratumoral distribution of a drug, and promote cellular uptake and intracellular drug release specifically within the tumor. We believe that the evolution of ‘smart’ – stimulus-responsive – drug carriers will further advance available cancer treatments to improve targeted delivery of drugs to tumors whilst minimizing harmful effects in healthy tissues, thereby providing more effective and better-tolerated cancer therapies.

Future perspective

This article highlights the challenges in designing macromolecule and nanoparticle drug carriers that can achieve localized therapeutic success while minimizing systemic toxicity. Tumor-specific delivery, tumor distribution and cellular uptake must be optimized not only for therapeutic action, but also for detection, diagnostic and imaging purposes. Stimulus-responsive macromolecules and nanoparticles provide a potential solution to each of these challenges, and could lead to approaches applicable to a variety of cancers. In conclusion, the challenges in the design of ‘smart’ nanocarriers include the discovery of new and better tumor-specific targeting moieties, and their integration into ‘smart’ nanocarriers that exhibit self-assembly (for targeting) or disassembly (for drug release) within a tumor in response to tumor-specific intrinsic or extrinsic triggers. Design of such drug-delivery vehicles aims to achieve high uptake and uniform spatial distribution of the drug within the tumor with minimal uptake by normal tissues, thereby minimizing systemic toxicity and maximizing therapeutic efficacy. Finally, the development of stimulus-responsive drug carriers for cancer therapy will require rigorous preclinical, in vivo studies in relevant orthotopic animal models to translate these promising targeted delivery approaches from the laboratory to the clinic.

Executive summary.

Tumor accumulation of macromolecules & nanoparticles

• Leaky vasculature and lack of lymphatics allow increased accumulation of systemically delivered macromolecules and nanoparticles in tumor tissue.

• Macromolecular carriers and nanoparticle vehicles enhance the circulation, accumulation and efficacy of chemotherapeutics.

• Thermally responsive peptide polymers improve tumor accumulation by creating local depots of aggregated carriers in heated tumor vasculature.

Improved tumor distribution of macromolecule & nanoparticle payloads

• High interstitial fluid pressure and low diffusivity inhibit the intratumoral penetration of anticancer therapeutics encapsulated in drug vehicles.

• Release of drug from responsive carriers triggered by localized intrinsic or extrinsic cues can improve the penetration of drug into tumor tissue.

• Further investigation is needed to verify the effect of tumor-localized release of drug from conjugates or carriers on intratumoral drug distribution and to demonstrate its therapeutic benefit.

Enhanced cellular uptake of drug carriers & control of the intracellular fate of drug cargo

• Cellular uptake and intracellular drug release is critical for the therapeutic action of many targeted anticancer drugs.

• Tumor-specific presentation of bioactive molecules in response to enzymatic, pH or temperature stimuli affords improved specificity and efficacy of targeted drug delivery to tumors.

• Reducing conditions and low pH can trigger drug release in intracellular compartments.