www.advmat.de www.MaterialsViews.com

REVIEW

Peptide Targeted Lipid Nanoparticles for Anticancer Drug Delivery Timothy R. Pearce, Kamlesh Shroff, and Efrosini Kokkoli*

potential, sustained angiogenesis, and evasion of apoptosis.[4,5] These hallmarks describe how acquired and somatic mutations in the DNA of normal cells that provide cancer cells a Darwinian advantage can result in the development and progression of cancer. The cellular mechanisms that result in the devastating abilities of cancer, including the over-expression of growth promoting proteins, the aberrant expression of proliferative regulatory proteins, the necessity for rapid angiogenesis, and the high mitotic rates, also offer researchers an opportunity to design therapies to treat the disease.[5] Indeed, there are already many chemotherapeutics that have remarkable potential - with more in development. However, many of these drugs suffer from poor pharmacokinetics and dynamics, and off-target delivery to healthy cells, which results in low tumor accumulation and dose limiting toxicity to healthy tissue. To alter the pharmacokinetic and pharmacodynamic profiles of these chemotherapeutic agents aimed at solid tumors, drugs were encapsulated within first-generation nanoparticles. Upon loading a drug into nanoparticles, the fate of the drug is no longer dictated by its own intrinsic properties but is instead controlled by the pharmacokinetics of its carriers. As Matsumura and Maeda first described in 1986, large biological molecules and nanoparticles between 20-400 nanometers (nm) in diameter preferentially accumulate in tumor rather than normal tissue, an effect later called enhanced permeability and retention (EPR).[6] This phenomenon is caused by two unique physiologic features of the tumor microenvironment: underdeveloped ‘leaky’ vasculature that allows for the transfer of large molecules out of the blood stream, and poor lymphatic drainage of the interstitial fluid of the tumor that would typically remove these molecules. It is through this mechanism that nanoparticles alter the pharmacokinetics and pharmacodynamics of the drugs they encapsulate, reducing the off-target toxicity and increasing the accumulation of drugs at the tumor sites. The first nanoparticle-based drug for cancer therapy, called Doxil, was granted FDA approval in 1995. Doxil is lipid vesicles (liposomes) about 100 nm in diameter that are largely composed of two types lipids and encapsulate the anthracycline drug doxorubicin (DOX). A small amount of a polyethylene glycol (PEG)-lipid conjugate is included in the lipid formulations to reduce the clearance of the liposomes from the blood and extend the plasma half-life of the doxorubicin by a factor of ∼40.[7] One year later, the FDA approved another liposomal

Encapsulating anticancer drugs in nanoparticles has proven to be an effective mechanism to alter the pharmacokinetic and pharmacodynamic profiles of the drugs, leading to clinically useful cancer therapeutics like Doxil and DaunoXome. Underdeveloped tumor vasculature and lymphatics allow these first-generation nanoparticles to passively accumulate within the tumor, but work to create the next-generation nanoparticles that actively participate in the tumor targeting process is underway. Lipid nanoparticles functionalized with targeting peptides are among the most often studied. The goal of this article is to review the recently published literature of targeted nanoparticles to highlight successful designs that improved in vivo tumor therapy, and to discuss the current challenges of designing these nanoparticles for effective in vivo performance.

1. Introduction Over the past 50 years astounding medical breakthroughs have eradicated smallpox, created implantable artificial hearts, and allowed for in vitro fertilization. During this time significant effort has been devoted to finding a cure for cancer. Despite significant progress in the treatment of certain cancers its overall impact remains startling high, with 1 in 2 men and 1 in 3 women in the US expected to develop cancer in their lifetime, a death rate that is only marginally lower than it was in 1960, and 23% of US deaths a result of cancer.[1–3] The failure to find a single cure for cancer is, in part, a result of cancer not being one disease but rather a disparate collection of unique diseases that share many of the same characteristics. Hanahan and Weinberg laid out six ‘hallmarks of cancer’: self-sufficiency in growth signals, insensitivity to anti-growth signals, tissue invasion and metastasis, limitless replicative

Prof. E. Kokkoli Department of Chemical Engineering and Materials Science University of Minnesota Minneapolis, MN, 55455, USA E-mail: [email protected] T. R. Pearce Department of Biomedical Engineering University of Minnesota Minneapolis, MN, 55455, USA Dr. K. Shroff Department of Chemical Engineering and Materials Science University of Minnesota Minneapolis, MN, 55455, USA

DOI: 10.1002/adma.201200832

Adv. Mater. 2012, 24, 3803–3822

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

wileyonlinelibrary.com

3803

www.advmat.de

REVIEW

www.MaterialsViews.com

formulation of an anthracycline, daunorubicin, trade name DaunoXome. Similar to Doxil, DaunoXome is composed of two lipids. However, DaunoXome does not contain the additional PEG lipid found in Doxil and thus the half-life of the daunorubicin is only increased by ∼4-fold.[7] Both of these first-generation nanoparticles dramatically alter the pharmacokinetics and dynamics of the encapsulated drugs causing them to have higher tumor accumulation and reduced systemic toxicity when compared with the drugs without their liposome chaperones. ‘Second-generation’ nanoparticles are currently being developed to further improve the pharmacokinetic and pharmacodynamic profiles of drugs to improve the treatment of solid tumors. These second-generation nanoparticles face three challenges that must be addressed to provide effective anticancer therapy. They must 1) evade the mononuclear phagocyte system (MPS) and filtration out of the bloodstream so they can passively accumulate within the tumor via the EPR effect, 2) remain in the tumor tissue and cross through the cancer cell membrane, and 3) ensure the encapsulated drugs are delivered to their site of action within the malignant cells. An understanding of these three challenges is beneficial in discussing nanoparticle-mediated delivery and warrants additional description.

Efie Kokkoli received a Diploma from the Aristotle University of Thessaloniki, Greece and her Ph.D. from the University of Illinois, Urbana-Champaign with Chip Zukoski. She completed her postdoctoral work with Matt Tirrell at the University of Minnesota, and the University of California, Santa Barbara. She is an Associate Professor in the Department of Chemical Engineering and Materials Science at the University of Minnesota. Current research interests include peptide-amphiphiles, aptameramphiphiles and biopolymers for targeted drug delivery and biomaterials applications. Tim Pearce received a B.S. in Biomedical Engineering from the University of WisconsinMadison in 2008 and is currently pursuing a Ph.D. in Biomedical Engineering in the lab of Dr. Kokkoli. His current research focuses on the design and characterization of aptamer-amphiphiles for use in self-assembling targeted drug delivery systems.

1.1 Evasion of the MPS and Removal from the Blood Stream It is important to recognize that the second-generation nanoparticles, often referred to as ‘active targeting nanoparticles’, rely on passive targeting of the tumor via the EPR effect before the targeting peptides functionalizing their surfaces become useful. Extravasation of the nanoparticles out of the blood stream and into the tumor microenvironment is a stochastic phenomenon, and requires that the nanoparticles remain in circulation for a sufficiently long period of time if they are to have substantial passive accumulation. To increase the amount of time nanoparticles spend in circulation the nanoparticles must avoid capture by the MPS, renal clearance, and liver and spleen filtration (Figure 1).

Figure 1. Nanoparticle capture by kidneys, the MPS, and accumulation into the tumor. Nanoparticles smaller than ∼10 nm are effectively filtered out of the bloodstream by the kidneys. Nanoparticles covered by opsonins can be captured by phagocytic cells of the liver and spleen. Long circulating nanoparticles can passively accumulate within the tumor after extravasation through the leaky endothelium.

3804

wileyonlinelibrary.com

Kamlesh Shroff is Postdoctoral Research Associate with Prof. Kokkoli in the Department of Chemical Engineering and Materials Science at the University of Minnesota. He received his Ph.D. in Applied Sciences from the group of Prof. Juergen Ruehe and Prof. Markus Biesalski at Albert-Ludwigs University of Freiburg, Germany and a Master’s degree in Biotechnology from University of Pune, India. His research interests include peptide-functionalized materials for coatings and scaffolding, targeted drug delivery, and bio-MEMS.

The MPS is largely comprised of phagocytic cells, primarily monocytes and macrophages, that reside in connective tissue throughout the body, with large populations taking residence in the spleen, liver, and lymph nodes. These cells are responsible for the sequestration and destruction of inert or foreign material found in the body. The recognition of nanoparticles by these cells is greatly enhanced following a process called opsonization, when opsonin molecules circulating within the

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Adv. Mater. 2012, 24, 3803–3822

www.advmat.de www.MaterialsViews.com

macropinocytosis, or cell drinking, is a non-specific process by which cell membrane invaginations and protrusions pinch off into large intracellular vesicles to capture molecules that reside near the cell surface. Clathrin-mediated endocytosis is a process by which clathrin coated pits at the cell surface are pinched off to form vesicles of ∼120 nm diameters.[14] Clathrin-mediated endocytosis can be triggered by nanoparticle binding to receptors found in these coated pits and is thought to be a prominent mechanism for nanoparticle internalization.[14] Caveolae are 60-80 nm flask-shaped structures that can engulf nanoparticles that bind to their surfaces. Following plasma membrane budding, caveolae form intracellular vesicles typically 50–100 nm in diameter.[15] Clathrin- and caveolae-independent endocytosis comprise a variety of other pathways that appear to require specific lipid membrane compositions and are dependent on cholesterol to form the intracellular vesicles. However, vesicles formed through these endocytic pathways do not appear to play a major role in the internalization of nanoparticles.[14]

REVIEW

bloodstream absorb to the nanoparticles’ surfaces. Once the nanoparticles are opsonized they are readily recognized as foreign and efficiently destroyed by macrophages. PEG is a hydrophilic polymer that is often added to the surface of the nanoparticles to protect nanoparticles from plasma clearance by forming an inert shield surrounding the nanoparticle that discourages the binding of opsonins. The kidneys, spleen and liver pose another opportunity for the body to remove nanoparticles from the blood stream. The kidneys act as blood filters that effectively remove nanoparticles with hydrodynamic diameters smaller than ∼10–20 nm from circulation.[8,9] Filtration of blood through interendothelial cell slits in the splenic sinus walls has been shown to capture nanoparticles larger than 200 nm.[10] Liver fenestrations found in the sinusoids also temporarily remove nanoparticles from the circulation, acting as a nanoparticle depot that may eventually allow the nanoparticles to return back into circulation.[11] These blood filters suggest the size of nanoparticles should be no smaller than 20 nm and no larger 200 nm if long circulation within the blood is desired.

1.3 Ensuring the Delivery of Active Drug to the Site of Action within the Malignant Cell

1.2 Enhancing the Retention and Uptake of Drug into the Tumor The defining feature of second-generation nanoparticles is the attachment of ligands to the nanoparticles’ surface that allow them to bind specific molecules expressed on, or around, the tumor. Many targeting molecules have been used including vitamins, glycoproteins, antibodies and antibody fragments, peptides, and aptamers. These targeting ligands serve two purposes. They increase the likelihood that the nanoparticles that find their way into the tumor mass via the EPR effect remain within the tumor, and they often trigger the transport of the nanoparticles across the cellular membranes of the diseased tissue. Nanoparticles typically cross the cell membrane via four endocytic pathways that are broadly categorized as pinocytosis (and macropinocytosis), clathrin-mediated endocytosis, caveolae-mediated endocytosis, and clathrin- and caveolaeindependent endocytosis (Figure 2).[12,13] Pinocytosis and

Figure 2. A diagram of the various endocytic pathways that allow nanoparticles to cross the cell membrane.

Adv. Mater. 2012, 24, 3803–3822

Many anticancer drugs exert their effect at sites within the cancer cell including the cell membrane, mitochondria, cytoplasm, and nucleus and thus must escape their lipid nanocarriers and the intracellular vesicles from which they crossed the cell membrane. There are fundamental differences between the cytoplasmic trafficking of the different endocytic organelles. For example, following clathrin-mediated endocytosis the intracellular vesicles shed their clathrin coats, allowing the vesicles to fuse with early endosomes. These endosomes are then sorted to late endosomes/lysosomes for degradation of their cargoes in the low pH environment. On the other hand, caveolae vesicles transport and fuse to caveosomes, which have a neutral pH, through a pathway that appears to be slower than the clathrin-mediated endocytosis and often avoids trafficking to lysosomes.[14] When designing nanoparticles for drug delivery, the intracellular fate of the vesicles is paramount. For nanoparticles containing pH-sensitive components, like some lipids, the low pH environment of the endosomes/lysosomes can act as a trigger for drug release from the nanoparticles to the cell cytoplasm (Figure 3). In contrast, the delivery of pH sensitive peptides, proteins, and nucleic acids to the lysosomes should be avoided.[16] Over the past 20 years, much effort has been spent designing and testing targeted nanoparticles for cancer therapy with these three design challenges in mind. A wide variety of materials have been used to create the nanoparticles including blockcopolymers, silica, and carbon nanotubes as well as organic/bioinspired materials like viral capsids, carbohydrate dendrimers, and lipids. Likewise, there have been an incredible number of unique targeting molecules used to functionalize the surface of these nanocarriers including transferrin, folate, antibodies, proteins, aptamers and peptides. Perhaps unsurprising due to the clinical success of the firstgeneration lipid nanoparticles, lipids have been a popular material for constructing second-generation targeted nanoparticles.

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

wileyonlinelibrary.com

3805

www.advmat.de

REVIEW

www.MaterialsViews.com

Despite substantial investments in the design and development of peptide-functionalized lipid nanoparticles for cancer therapy, they have yet to make the transition from animal studies into clinical relevance. The following review describes research using peptide-functionalized lipid nanoparticles for drug delivery published in the past three years (Tables 1 and 2). Specific attention is paid to the approaches taken to address the three challenges described earlier, to support the discussion of promising strategies that may lead to meaningful progress in the treatment of cancer. Reviews describing less recent peptide-functionalized lipid nanoparticles as well as targeted nanoparticles using non-lipid materials like polymers, silica, metals, and carbon nanotubes and other non-peptide targeting ligands like antibodies, proteins, and aptamers, can be found elsewhere.[18,19] Figure 3. Destabilization of endosomal membrane and drug release. After cell endocytosis and localization inside the endosomes, pH-sensitive liposomes go through membrane destabilization and initiate the destabilization of the endosomal membrane that allows for drug efflux into the cell cytoplasm.

Nor is it surprising that peptides, with their ability to bind to many different cellular targets with strong affinity, to be produced at relatively low cost and with high fidelity, and to attach to the nanoparticles without eliminating their binding ability, have made them the most often used targeting ligands. Many types of lipid-based nanoparticles have been created including liposomes, lipid-polymer hybrid nanoparticles, nanocapsules, nanoemulsions, and solid lipid nanoparticles as illustrated in Figure 4.[17]

Figure 4. Examples of lipid based nanoparticles. A) Conventional liposomes; B) Functionalized liposomes; C) Nanoemulsions; D) Solid lipid nanoparticles; E) Nanostructured lipid carriers; F) Lipid-polymer hybrid nanoparticles; G) Lipid nanocapsules; H) Lipid grafts. Reproduced with permission.[17] Copyright 2010, Begell House Inc.

3806

wileyonlinelibrary.com

2. Lipid Nanoparticles Targeting Tumor Vasculature Recent research about the role of tumor vasculature in the development and progression of solid tumors suggests that antivascular therapy has the potential to inhibit tumor growth and lead to a state of dormancy.[20] Solid tumors rely on rapid angiogenesis to supply the oxygen and nutrients needed to support their accelerated growth, and use this new vasculature to metastasize as the disease progresses. Antivascular therapies attempt to restrict the blood supply to the tumor-starving it of the required nutrients for rapid proliferation and eliminating a route of metastasis. These therapies have enjoyed modest clinical success,[21] inspiring the exploration of vascular targeting nanoparticles as vehicles to enhance the delivery of antivascular agents to the tumor associated endothelial cells. The following reports of vasculature targeting lipid nanoparticles use peptides that bind to unique molecules found on tumor-associated vasculature or are over-expressed on the surfaces of vascular cells. A common approach to identify and isolate peptides that recognize proteins expressed by tumor-associated endothelium is phage display biopanning. This powerful peptide selection technique can identify individual amino acid sequences from a large candidate pool (typically >109 initial peptide sequences) that have high affinity and specificity for cancer cell membrane proteins.[22,23] When performing phage display biopanning against complex systems like cancer cells and tissues, the molecules to which the isolated peptides bind are initially unknown. While potentially possible to identify the binding partners of these peptides, this is often not attempted. Using in vivo phage display biopanning against human oral cancer xenografts in mice, Chang et al. identified the PIVO-8 (SNPFSKPYGLTV) and PIVO-24 (YPHYSLPGSSTL) peptides that had remarkable binding ability not only to the oral cancer tumor mass, but also to tumor specimens taken from human breast, lung, liver, colon, and pancreatic tumors.[24] These peptides were shown to be specific to tumor endothelium, with no affinity for either the cancer cells themselves or for healthy endothelium. Furthermore, in vitro experiments demonstrated that the PIVO-8 and PIVO-24 peptides allowed the nanoparticles to traverse endothelial cell membranes via receptormediated endocytosis and accumulate intracellularly.[24] In vivo

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Adv. Mater. 2012, 24, 3803–3822

www.advmat.de www.MaterialsViews.com

Peptide

Cancer type

Molecular target

Nanoparticle size [nm]

Peptide mol%

PEG mol%

Encapsulant

Ref

Tumor Vasculature PIVO-8 (SNPFSKPYGLTV), PIVO-24 (YPHYSLPGSSTL)

Lung, Breast, Liver,Pancreas, Colon

Unknown

–

0.5

–

DOX

[24]

Colon

BiP/GRP78

120

3

3 (PEG2000)

DOX

[25]

APRPG, NGR (GNGRG)

Colon

Unknown, APN

110–150

7

7

DOX

[30]

CPRECESA, CNGRCGV

Nerve

APA, APN

90–115

–

–

DOX

[33]

Breast, Sarcoma

APN

100

5

5 (PEG2000)

DOX

[36]

Lung, Breast, Pancreas

Somatostatin Receptor

100

0.5–1

5, 23 (PEG2000)

DOX, 111In

[49,54]

Breast

LHRH Receptor

120–150

–

0.4 (PEG2000)

Mitoxantrone

[50]

Skin

Cholecystokinin Receptor

200–230

10

0

DOX

[51]

Prostate

Bombesin Receptor

200

10

10 (PEG3000)

111

[52]

WIFPWIQL

NGR (cKNGRE)

REVIEW

Table 1. Peptide-functionalized lipid nanoparticles used for in vitro studies.

Tumor Cells FCFWKTCT Pyr-HWSYGLRPG CCK8 (GDYMGWMDF) QWAVGHLM RRPYIL

In

Colon, Muscle

Neurotensin Receptor

90

5

0

DOX

[53]

TAT (AYGRKKRRQRRR)

Colon, Liver

None

100

0.5–2

2 (PEG1600), 3(PEG2000), 0-8 (PEG5000)

DiD, Calcein

[60,61]

Pep-1 (KETWWETWWTEWSQPKKRKVC)

Breast, Skin

None

160

-

0

Au

[62]

Poly-R, NGR (CYGGRGNG)

Skin

None, APN

85–103

0–5, 0–15

0-15 (PEG2000)

Rhodamine

[63]

LSQETFSDLWKLLPEN

Bone

None

100–200

100

0

-

[64]

DMPGTVLP

Breast

Unknown

90–100

0–1

3-5 (PEG2000)

DOX, Rhodamine

[65,66]

Prostate

Unknown

–

–

3-5 (PEG2000)

DOX

[67]

SP5-2 (TDSILRSYDWTY)

Lung

Unknown

–

–

–

DOX, Vinorelbine

[68]

CSNIDARAC

Lung

Unknown

200

1.3

1.3 (PEG2000)

DOX, Cy7.5

[69]

HVGGSSV

Lung

Unknown

100

2

2 (PEG2000)

DOX, AlexaFluor750

[72]

Ovary, Breast, Sarcoma

αv Integrins

90–250

0.3–10

0-40 (PEG2000)

Paclitaxel, DOX, CA-4, Qd, Gd

[78–81,107]

Skin, Breast, Colon, Lung

αv Integrins

35–100

0.3

3-5 (PEG2000)

Matrine, Gd, DiD

[82,84,106]

Lung

αv Integrins, Neuropilin-1

65–75

–

5 (PEG2000)

Paclitaxel

[86]

Breast, Colon, Prostate

α5β1 Integrin

80–150

0.7–6.9

2-5 (PEG750), 2-5 (PEG2000)

DOX, 5-FU,TNF-α, Calcein

[95–97,100]

Breast

α5β1 Integrin

100

–

25 (PEG2000)

DOX

[102]

DTDSHVNL, DVVYALSDD

Tumor Vasculature and Cells RGD peptides

Cyclic RGD peptides ARYCRGDCFDATWLPPR PR_b (KSSPHSRN(SG)5RGDSP) PHSCN CDPGYIGSR tbFGF Flt-1 (WHSDMEWWYLLG), ATWLPPR, LyP-1 (CGNKRTRGC) GCRGRRST, GARYCRGDCFDG TSPB (KRFKQDGGWSHW), TSPA(TRIRQDGGWSHW)

Adv. Mater. 2012, 24, 3803–3822

Skin

Laminin Receptor

100

5

5 (PEG2000)

5-FU

[103]

Prostate, Breast

FGF Receptor

100–160

–

0

DOX, Paclitaxel

[111]

Skin

VEGF Receptor, Neuropilin-1, p32Protein

90–360

1

5 (PEG2000)

DOX

[116]

Skin, Blood

PDGF Receptor, αvIntegrins

90–130

1

10 (PEG750), 5 (PEG2000)

DOX

[117]

Colon

TGF-β

120–150

10

0

DOX

[122]

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

wileyonlinelibrary.com

3807

www.advmat.de

REVIEW

www.MaterialsViews.com Table 1. Continued Peptide

Cancer type

Molecular target

Nanoparticle size [nm]

Peptide mol%

PEG mol%

Encapsulant

Ref

LARLLT

Lung

EGF Receptor

110–150

9

3 (PEG2000)

Rhodamine, Cy5.5

[125]

PH1 (TMGFTAPRFPHY)

Lung, Liver

Tie2

100

–

3–5 (PEG2000)

Cisplatin

[128]

LyP-1 (CGNKRTRGC)

Lung, Breast

p32 Protein

90–100

0.5-4

5-8 (PEG2000)

DOX, DiR

[131–133]

C3d (ASKKPKRNIKA)

Sarcoma

NCAM

120

–

–

DOX, Gd

[134]

Abbreviations: BiP (binding immunoglobulin protein), APN (aminopeptidase N), APA (aminopeptidase A), LHRH (luteinizing hormone releasing hormone), FGF (fibroblast growth factor), VEGF (vascular endothelial growth factor), PDGF (platelet-derived growth factor), TGF-β (transforming growth factor-β), EGF (epidermal growth factor), NCAM (neural cell adhesion molecule), DOX (doxorubicin), 111In (indium-111), Au (gold), CA-4 (Combretastatin A-4), Qd (quantum dot), Gd (gadolinium), Cy7.5, Cy5.5, DiD, DiR (fluorescent dyes), 5-FU (5-fluorouracil), TNF-α (tumor necrosis factor- α).

Table 2. Tumor, liver and spleen accumulation of peptide-functionalized nanoparticles, as well as tumor accumulation of targeting vs. non-targeting nanoparticles. Peptide

Peptide mol%

PEG mol%

Nanoparticle size [nm]

Tumor Accumulation of Targeting Nanoparticles

Tumor Accumulation Targeting/NonTargeting Nanoparticles

Liver, Spleen Accumulation of Targeting Nanoparticles

Ref

Tumor Vasculature PIVO-8 (SNPFSKPYGLTV), PIVO-24(YPHYSLPGSSTL) APRPG, NGR (GNGRG)

0.5

–

–

1.2 μg drug/g

1.6

–

[24]

7

7

–

1% ID/100 mg

1

4.5% ID/100 mg, 5% ID/100 mg

[30]

Tumor Cells QWAVGHLM

10

10 (PEG3000)

200

2.4% ID/g

1.5

16% ID/g, 2.5% ID/g

[52]

TOC (FCYWKTCT)

0.5

5 (PEG2000)

100

4.25% ID/g

–

45% ID/g, 7% ID/g

[54]

TAT (AYGRKKRRQRRR)

0.5

3 (PEG2000), 0-8(PEG5000)

100

–

2.6

–

[60]

SP5-2 (TDSILRSYDWTY)

–

–

–

1.8 μg drug/g

1.9

7 μg drug/g

[68]

Tumor Vasculature and Cells RGDF

2.5

0

175–250

-

2.1

–

[81]

DGARYCRGDCFDG

–

5 (PEG2000)

100

16% ID

4

25% ID, 23% ID

[84]

CDPGYIGSR

5

5 (PEG2000)

100

16% ID

1.3

20% ID, 11% ID

[103]

tbFGF

–

0

100-160

–

2.7

15 μg drug/g, 5 μg drug/g

[111]

Flt-1 (WHSDMEWWYLLG), ATWLPPR, LyP-1(CGNKRTRGC)

1

5 (PEG2000)

90-360

5% ID/g

2.5

12.5% ID/g, 12.5% ID/g

[116]

GCRGRRST, GARYCRGDCFDG

1

5 (PEG2000)

90-130

5% ID/g

2.5

13% ID/g, 11% ID/g

[117]

LyP-1 (CGNKRTRGC)

4

8 (PEG2000)

100

10% ID/g

3.3

–

[133]

C3d (ASKKPKRNIKA)

–

–

120

20 μg drug/g

1.7

–

[134]

experiments with the PEGylated peptide-functionalized liposomes showed almost twice as much accumulation within subcutaneous murine lung cancer tumors as the non-targeting PEGylated liposomes showed. Tumor tissue sections revealed that both peptide-functionalized liposome formulations escaped the tumor endothelium and diffused into the tumor interstitial space between 4 and 24 hours after their injections, causing severe damage to both the tumor endothelium and the cancer cells themselves. Non-targeting liposomes were also found within the tumor itself, although at much lower concentrations. When injected into mice bearing lung (H460), liver (Mahlavu), pancreatic (PaCa-2), and colon (HCT116) tumors, both PIVO-8 and PIVO-24 functionalized liposomes significantly reduced

3808

wileyonlinelibrary.com

tumor volumes in all tumor types and decreased tumor angiogenesis, demonstrating the enhanced and versatile therapeutic potential obtained by targeting tumor vasculature rather than directly targeting tumor cells. The Oku group used comparative proteome analysis in combination with subcellular fractionation to identify endothelial cell membrane proteins that are highly expressed in response to pro-angiogenic vascular endothelial growth factor (VEGF) stimulation.[25] These molecules are attractive targets for antivascular drug delivery therapies as they are often the result of tumor-driven VEGF signaling and are in direct contact with the blood stream.[26] One of the proteins identified through their analysis was the binding immunoglobulin protein (BiP), also

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Adv. Mater. 2012, 24, 3803–3822

www.advmat.de www.MaterialsViews.com

Adv. Mater. 2012, 24, 3803–3822

liposomes containing either the APN or APA targeting peptides were used in a consecutive dosing regimen to target the liposomes to both the endothelial and perivascular cells supporting a murine neuroblastoma tumor model. When administered as single injections, greater than 20% of each type of the peptide-functionalized liposomes that contained DOX remained in circulation 24 hours after injection, and both formulations enhanced the delivery of DOX to the tumor when compared to non-targeting liposomes.[33] Consecutive injections of APA and APN liposomes, each consisting of 50% of the dose that was injected during the single injection experiments to ensure equal amounts of injected DOX, resulted in more DOX delivery to the tumor than occurred by a single injection of either type of liposome formulation. Mice that received a consecutive dosing regimen of APA liposomes followed 30 minutes later by APN liposomes survived significantly longer than mice that received either monotherapy liposome regimen, non-targeting liposomes, or free DOX (Figure 5). The combination treatment group also exhibited enhanced destruction of endothelial and pericyte cells, as well as a large amount of necrosis in the tumor itself.[33] Despite the higher affinity of the cyclic NGR peptide (CNGRCGV), the linear version of the peptide is more commonly used to avoid the potential formation of disulfide bonds between adjacent peptides that reside on the liposomal surface and could compromise their binding ability. The NGR peptide c(KNGRE) was designed that had a 3.5-fold increase in affinity compared to the linear peptide and did not require a disulfide bond.[36] The new cyclic peptide and the standard linear version were used to functionalize thermo-sensitive liposomes that were created by combining 10 mol% lysolipids along with DPPC lipids. As seen in Figure 6, the liposomes were found to have a strong and rapid temperature-dependent release of their encapsulated doxorubicin, with greater than 75% of the doxorubicin released after 4 seconds at 41.3 °C and very little released at both 37 °C and 38 °C over a period of 15 minutes. The drug release from the liposomes at modestly elevated temperatures was designed to allow the systemic delivery of liposomes to be combined with thermal ablation or focal hyperthermia to enhance tumor-specific drug targeting.

REVIEW

known as GRP78, a well-characterized molecular chaperone that is typically located in the lumen of the endoplasmic reticulum.[27] Recently, BiP was identified to play a role in tumor driven angiogenesis with a 3-4 fold higher expression on tumor endothelium compared to healthy endothelium, and was also found on the surface of some cancers.[28,29] Liposomes functionalized with BiP targeting peptides (WIFPWIQL) accumulated four times more within VEGF stimulated endothelial cells than within DU145 prostate and C26 colon cancer cells, suggesting that the peptide-functionalized liposomes might effectively and specifically target neovascular cells.[25] Indeed, fluorescent imaging of tumor sections from mice with subcutaneous C26 tumors taken 3 and 6 hours after liposomal injections showed that the peptide-functionalized liposomes had a strong association with the tumor vasculature cells whereas the non-targeting liposomes exhibited diffuse accumulation throughout the tumor mass.[25] Injections of peptide-functionalized liposomes containing doxorubicin reduced the tumor volumes of mice by roughly 50% after 28 days and extended their survival time when compared to mice that received the non-targeting liposomes. Improved vasculature targeting was shown by dual-targeting liposomes functionalized with the NRG (GNGRG) and APRPG peptides, each capable of targeting different molecules expressed on tumor-associated endothelial cells.[30] The NGR peptide was found[31] to target the membrane associated enzyme aminopeptidase N(APN), a cell surface protein thought to play a role in chemokine processing and tumor invasion, while the molecular target of the APRPG peptide was not identified.[32] The NGR and APRPG peptides, when used together to create dual-targeting liposomes, acted cooperatively to cause significantly more in vitro HUVEC targeting and a larger antiproliferative effect than either of the single-targeting liposomes, despite having the same overall amount of peptides per liposome. In vivo results showed that although the dual-targeting liposomes were capable of significantly reducing the accumulation of drug within the spleen when compared to the single-targeting and non-targeting liposomes, there was no change in the accumulation in the liver, kidneys, lung, and heart. In addition, the dualtargeting liposomes did not produce enhanced tumor accumulation compared to the single- and non-targeting liposomes. Interestingly, immunofluorescent staining of tumor tissue harvested 3 hours after liposome injections showed the dualtargeting liposomes, and to a lesser degree the single-targeting liposomes, were localized within the angiogenic vessels while non-targeting liposomes were not. Although there was no difference in overall drug accumulation within the tumor between the different liposomal formulations, the dual-targeting liposomes had the strongest anti-tumor effect. Loi et al. also used two different peptides to target liposomes to tumor-associated vasculature,[33] a cyclic version of the NGR peptide (CNGRCGV) targeting APN reported to have higher affinity and 10-fold greater anti-tumor activity than the linear form,[34] and an aminopeptidase A (APA) targeting peptide (CPRECESA) previously identified with in vivo phage display.[35] Like APN, APA is a surface expressed enzyme and is upregulated and active in perivascular cells that support the tumor vasculature.[35] Rather than creating dual-targeting liposomes that contained both peptides, single-targeting PEGylated

Figure 5. Survival curves of mice with neuroblastoma tumors treated with intravenous injections of HEPES-buffered saline (control), free DOX (DXR), non-targeting DOX liposomes (SL[DXR]), APN- or APA-targeting DOX liposomes (CNGRC-SL[DXR] or CPRECES-SL[DXR]) and APN- and APA-targeting DOX liposomes (COMBO). Targeting liposomes lacking DOX (CNGRC-SL[empty] and CPRECES-SL[empty]) were included as additional controls. Reproduced with permission.[33] Copyright 2010, Elsevier.

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

wileyonlinelibrary.com

3809

www.advmat.de

REVIEW

www.MaterialsViews.com

Figure 6. Temperature-dependent release of doxorubicin from PEGylated cKNGRE peptide liposomes. Reproduced with permission.[36] Copyright 2010, Elsevier.

3. Lipid Nanoparticles Targeting Cancer Cells The ideal anticancer therapy would be capable of eliminating all of the malignant cancer cells found in a patient’s body, including any metastatic cells that are no longer associated with the main tumor and therefore less sensitive to vascular-targeted therapies. Many of the common chemotherapy drugs like doxorubicin, 5-fluorouracil (5-FU), and cisplatin impair cell division and cause significant damage to cancer cells that undergo rapid proliferation.[37] However, these drugs can also cause severe dose-limiting toxicities that restrict the total lifetime doses that can be administered.[37,38] A variety of approaches have recently been used to target peptide-functionalized nanoparticles to cancer cells in a manner that increases the delivery of the drugs to the cancer cells while reduces the off-target delivery to healthy tissues. These approaches included targeting overexpressed or uniquely expressed molecules on the cancer cells like G protein-coupled receptors, utilizing cell penetrating peptides capable of efficiently transporting nanoparticles across the cancer cell membrane, or identifying new molecular targets on the cancer cell membrane and developing nanoparticles that can bind to them. 3.1. Peptides that Target G protein-Coupled Receptors G protein-coupled receptors (GPCRs) are a large and diverse family of transmembrane proteins responsible for interacting with extracellular molecules and relaying their presence within the cell via the cAMP or PI3K signaling pathways.[39] With an important role in cell regulation and easily accessible binding domains, this class of molecules has been widely studied and is the target for roughly 40% of all pharmaceutics on the market.[40] A number of GPCRs with known binding peptides were found to be over-expressed on the surface of cancer cells, including the bombesin receptor,[41] neurotensin receptor,[42] luteinizing hormone releasing hormone (LHRH) receptor,[43] somatostatin receptor,[44] and cholecystokinin receptor[45] and were previously targeted by cytotoxic agent-peptide conjugates.[46–48] To

3810

wileyonlinelibrary.com

provide a more robust therapeutic strategy these five peptides were used to create targeting lipid nanoparticles with the goal of increasing the amount of cytotoxic agent delivered to cancerous cells. A variety of strategies were used to display the targeting peptides at the surface of liposomes. Long PEG spacers were used to link the somatostatin (FCFWKTCT)[49] and LHRH (Pyr-HWSYGLRPG)[50] targeting peptides to liposomes, short PEG spacers were used to link the cholecystokinin targeting peptides (GDYMGWMDF),[51] both long and short PEG spacers were used to link the bombesin targeting peptides (QWAVGHLM),[52] and synthetic lipid-like molecules that contained branched polylysine headgroups each with four peptide conjugation sites were used to link the neurotensin targeting peptides (RRPYIL).[53] Each of these strategies resulted in peptide-functionalized liposomes with enhanced targeting of in vitro cell cultures compared to non-targeting liposomes. Liposomes containing the bombesin peptide added to the ends of the long PEG3000 spacers had more non-specific targeting and less total targeting of bombesin receptor expressing PC-3 cells than liposomes created with short PEG spacers with only 5 PEG repeats; a surprising result as the longer PEG spacers should have provided significantly more steric screening of the liposomes and reduced the non-specific binding of liposomes to cells. The in vivo biodistribution of peptide-functionalized liposomes targeting the bombesin and somatostatin G-protein coupled receptors showed that the majority of the injected liposomes accumulated in the liver, spleen, and kidneys of the mice.[49,52] These organs were similarly targeted by non-functionalized liposomes[49] and liposomes containing scrambled versions of the peptides, suggesting the peptides did not play a major role in the distribution of the liposomes.[49,52] Helbok et al. used a slightly modified somatostatin receptor targeting peptide (FCYWKTCT) called tryorsine-3-octreotide (TOC) and metal chelating lipids to create PEGylated liposomes and micelles that provided tumor-specific delivery of the radioactive metal 111InCl3.[54] TOC micelles and liposomes each showed nanomolar somatostatin receptor affinity, but a considerable amount of their cell binding was not due to the peptidereceptor interaction (63% and 22% of total binding for micelles and liposomes, respectively). Additionally, free TOC peptides injected intravenously into tumor-burdened mice had better tumor targeting (4.3% of the total injected dose/gram of tissue (4.3% ID/g)) than TOC liposomes and micelles (∼2.5% ID/g). In vivo biodistribution experiments with both healthy Lewis rats and tumor-bearing mice showed that in both animal models the liver and spleen captured the largest amount of the nanoparticles. However, TOC liposomes injected into the tumor-bearing mice resulted in substantially more accumulation in the liver than TOC liposomes injected into the healthy Lewis rats (46% ID/g vs. 5% ID/g). 3.2. Non-Specific Cell Penetrating Peptides Cell penetrating peptides (CPPs) are a class of peptides defined by their ability to translocate large molecules and small particles across cellular membranes and into the cytoplasm of cells. The first peptide found to have this ability was the transactivator

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Adv. Mater. 2012, 24, 3803–3822

www.advmat.de www.MaterialsViews.com

Adv. Mater. 2012, 24, 3803–3822

REVIEW

of transcription (TAT) peptide, derived from a larger 86 amino acid HIV-1 TAT protein.[55,56] A variety of peptides have since been isolated that share the same ability to efficiently carry large cargoes across the lipophilic cell membrane. These CPPs typically contain a high proportion of positively charged amino acids, like lysine and arginine, or have an alternating pattern of polar, charged amino acids and non polar, hydrophobic amino acids. Current theories about the mechanism of the CPP-aided translocation can be categorized into three groups differentiated by the method of entry: direct penetration into the membrane, endocytosis-mediated entry, and entry via the formation of a transitory structure. CPPs have been widely used as transporters for intracellular delivery of large molecules like siRNA, plasmid DNA, proteins, quantum dots, and iron oxide nanoparticles, molecules that normally exhibit low uptake into cells.[57] CPPs were also found to enhance the cellular uptake of liposomes when used to decorate the liposomal surface,[58] a strategy that has been widely replicated in efforts to generate enhanced intracellular delivery of many different liposomal cargoes to cancer cells.[59] CPP-mediated uptake into cells requires direct contact between the CPPs and the cell membrane, which can be screened by the addition of a steric PEG brush layer to the liposome surface. For effective in vivo targeting to cancer cells a delicate balance must be struck; the PEG brush layer is required to increase circulation time of the liposomes and enhance passive targeting via the EPR effect but must not eliminate the ability of the CPPs on the liposomes’ surface to interact with the cell membranes. The TAT peptide (AYGRKKRRQRRR), rich in lysine and arginine, was used to create cell penetrating liposomes for the treatment of subcutaneous C26 colon[60] and H22 liver[61] tumors. With the goal of designing liposomes that could passively accumulate within the tumor while maintaining their cell penetrating ability once they were within the tumor, Kuai et al. designed TAT peptide liposomes with a long, thiol-cleavable PEG brush layer and a shorter non-cleavable PEG layer to which a small amount of TAT peptide was conjugated,[60] as shown in Figure 7. The long PEG brush layer acted as a strong steric barrier to reduce opsonization and non-specific cell interactions of the TAT liposomes as they passively accumulated. Once this accumulation occurred, L-Cysteine (L-Cys) was administered to cleave the long PEG shield and expose the cell penetrating TAT peptides to the tumor cells to spur liposome uptake. Extensive in vitro testing identified different peptide and cleavable PEG concentrations that effectively reduced non-specific liposome uptake prior to the cleavage of the long PEG and substantially increased cell uptake following cleavage.[60] In vivo experiments with murine tumor models showed that TAT liposomes with long, cleavable PEG layers had similar passive tumor accumulation as non-cleavable PEG liposomes prior to the addition of L-Cys, suggesting the cleavable PEG barrier remained intact and did not cause additional non-specific uptake into cells and capture by the MPS. Following the intravenous injection of L-Cys, liposomes with the cleavable PEG layer were internalized within tumor cells to a much greater extent than liposomes with a non-cleavable PEG layer, demonstrating tumor accumulation that was controlled by L-Cys induced cleavage.[60,61] Gold nanoparticles can act as both diagnostic and therapeutic agents for cancer treatment but there are several limitations to

Figure 7. Cysteine cleavable PEG5000-lipids were added to TAT peptide liposomes to reduce opsonization and non-specific cell interactions caused by the TAT peptide while the liposomes passively accumulate within the tumor. The long PEG5000 shielding can be cleaved from the liposome by adding L-cysteine to expose the TAT peptide. Reproduced with permission.[60] Copyright 2011, American Chemical Society.

the in vivo use of these nanoparticles. These limitations include the loss of their optical properties, which results from their aggregation in physiologic environments, and a lack of accumulation in the cytoplasm or nucleus of the cancer cells. To address these issues Kang et al. encapsulated gold nanoparticles within the aqueous core of liposomes functionalized with the cell penetrating peptide Pep-1 (KETWWETWWTEWSQPKKKRKVC).[62] Cryo-transmission electron microscopy (cryo-TEM) images of the liposomes, like those seen in Figure 8 showed a mixture of unilamellar, multilamellar, and multivesicular liposomes encapsulating, on average, 5 gold nanoparticles per liposome. There was no appearance of gold nanoparticle aggregation within the liposomes observed in the cryo-TEM images and their characteristic optical properties were unchanged, confirming they remained well dispersed when encapsulated. Dark-field microscopy showed the gold nanoparticles were largely found within the cytoplasm of cells only when delivered within liposomes and not when delivered alone. To add specific tumor targeting capabilities to CPP liposomes, Takara et al. added NGR peptides that target APN to the surfaces of PEGylated liposomes containing polyarginine CPPs as shown in Figure 9.[63] The low amounts of NGR peptides on surfaces of the liposomes were alone unable to trigger substantial receptor mediated endocytosis. Similarly, the polyarginine CPPs used to functionalize the liposomes, consisting of only four arginine residues, were too short to effectively extend past the PEG2000 brush layer and resulted in negligible internalization of the liposomes. However, when used simultaneously to functionalize liposomes a synergistic effect was seen, yielding

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

wileyonlinelibrary.com

3811

www.advmat.de

REVIEW

www.MaterialsViews.com

targeting. Missirlis et al. covalently attached the p53 inhibitor peptides to lipid-like molecules to form peptide amphiphiles that assembled into wormlike micelles and were able to anchor within cell membranes and subsequently internalize.[64] These micelles were dynamically stable in aqueous environment and had a slow dissociation rate in serum supplemented media, suggesting they could passively accumulate in tumors via the EPR effect prior to disassembling. The peptide amphiphile monomers underwent endocytosis in a variety of different cell lines but were found to be trapped in the endosomal membranes. A hypertonic solution of sucrose was used to disrupt the endosomes, which released the amphiphiles into the cytoplasm and enhanced the killing of cells that contained the wild-type p53 protein. 3.3. Peptides that Target Cancer Cells with Unknown Cell Binding Locations Figure 8. Cryo-TEM images of a) unencapsulated gold nanoparticles and b-d) gold nanoparticles encapsulated within lipid vesicles. The arrows point to gold nanoparticles. Reproduced with permission.[62] Copyright 2011, The Pharmaceutical Society of Japan.

liposomes capable of specifically binding and internalizing within APN expressing cells. These liposomes specifically localized to cancer cells but internalized via a non-specific CPP mechanism, demonstrating a drug delivery strategy that could avoid a receptor-mediated endocytotic pathway. Inhibiting the binding of the tumor suppressor protein p53 to the MDM2 protein is a promising anticancer therapy for solid tumors that retain their wild-type p53 and have amplified MDM2 expression. Peptide inhibitors like p5314-29 could prove to be effective therapeutic molecules if they could reach the cytoplasm and nucleus of diseased cells, but they are cell impermeable, rapidly degraded, and are not intrinsically tumor

Figure 9. A diagram of the dual-ligand liposomes. NGR peptides that are specific for aminopeptidase N were added to PEGylated liposomes containing cell penetrating peptides. Reproduced with permission.[63] Copyright 2010, Elsevier.

3812

wileyonlinelibrary.com

The Torchilin group used phage display biopanning to isolate a peptide (DMPGTVLP) with strong affinity for MCF-7 human breast cancer cells.[65] To immobilize the peptide at the surface of the clinically approved Doxil (PEGylated liposomes) without relying on chemical conjugation, a 55 amino acid fusion protein containing the targeting peptide and coat protein pVIII was constructed. As shown in Figure 10, the hydrophobic pVIII coat protein spontaneously inserted into the bilayers of pre-formed Doxil nanoparticles and displayed the targeting peptides at the interface. Importantly, the peptides retained their affinity and specificity for MCF-7 cells and did not alter the liposomes’ structures when added to the lipid membranes.[65] In addition to providing a convenient method for peptide functionalization, the phage coat protein was also found to help the liposomes escape from endosomes after the liposomes were internalized. The low pH environment of the endosomes was found to trigger the fusion of phage proteins with the endosomal membranes, which resulted in the endosomal escape of the liposomes and more cell death than liposomes that could not escape the endocytotic vesicles.[66] The phage coat protein was also used by Jayanna et al. to insert two different peptides generated by phage display into liposomes to target PC-3 prostate carcinoma cells.[67] In vitro experiments using PC-3 cells showed

Figure 10. A chimeric fusion protein consisting of a targeting peptide identified via biopanning against MCF-7 breast cancer cells and the major phage coat protein pVIII was directly incorporated within the liposomal bilayer without additional conjugation to lipophilic moieties. The hydrophobic C-terminus of the fusion protein spontaneously inserted into liposomal membranes, immobilizing the targeting peptide at the outer surface of the liposome. Reproduced with permission.[65] Copyright 2010, Future Medicine Ltd.

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Adv. Mater. 2012, 24, 3803–3822

www.advmat.de www.MaterialsViews.com

4. Lipid Nanoparticles Targeting Both Cancer Cells and Tumor Vasculature There are a substantial number of molecular targets that are either uniquely expressed or over-expressed on both cancer

Adv. Mater. 2012, 24, 3803–3822

vasculature and cancer cells, many of which can be classified into three broad categories: integrins, growth factor receptors (GFRs), and G-protein coupled receptors (GPCRs). Molecules from these categories offer attractive targets for anticancer therapeutics as they are often implicated in tumor growth and progression. Some of the best-selling anticancer drugs including Herceptin (targeting the HER2 receptor), Gleevec (Abl, c-Kit, and PDGFR), and Erbitux (EGFR), belong to these categories, and many more are currently FDA approved or in clinical trials.[73–75] Liposomes that are capable of destroying tumor vasculature to deprive the tumor of nutrients while simultaneously killing the cancerous cells may prove to be a more effective strategy for cancer treatment than liposomes that target only a single location. To explore this possibility, lipid nanoparticles targeting integrins, GFRs, GPCRs and a handful of other molecules found on both tumor vasculature and cancer cells have been designed and evaluated.

REVIEW

that both peptides led to greater liposomal uptake and enhanced cytotoxicity compared to liposomes with a non-binding peptide sequence or non-targeting PEGylated liposomes. Phage display biopanning performed by the Wu group and Lee group generated peptides capable of specific binding to non-small cell lung cancer (NSCLC) cells. Biopanning against the CL1-5 cell line generated the SP5-2 peptide (TDSILRSYDWTY) that showed specific in vitro binding to a variety of NSCLC lines including CL1-5, H460, A549, PC13 and H23 and had no binding to a host of other cancer cells.[68] 15 times more peptide was found within the tumor mass than in the brain, heart, and lungs after injecting the SP5-2 peptides into mice bearing subcutaneous H460 tumors, confirming the peptide’s tumor targeting ability. Liposomes functionalized with the SP5-2 peptides did not show the same ability as the peptides themselves to preferentially accumulate at the tumor rather than healthy organs. Biodistribution experiments with SP5-2 functionalized liposomes, non-binding control peptide-functionalized liposomes and non-targeting liposomes containing DOX revealed that the liver received the highest proportion of DOX and the NSCLC tumor substantially less. However, there was five-fold more DOX delivered to the tumor by the SP5-2 functionalized liposomes than by injections of free DOX, and two-fold more accumulation compared to the non-targeting and control peptide-functionalized liposomes, which was accompanied by dramatic improvements in the therapeutic effects of the SP5-2 liposomes. Similarly, He et al. isolated a peptide (CSNIDARAC) capable of effectively binding the H460, A549 and H226 NSCLC cell lines.[69] The tumor targeting ability of this CSNIDARAC peptide was found to be much higher than that of a non-binding control peptide and had negligible accumulation in the liver, lung, spleen, and heart. The distribution of PEGylated CSNICARAC functionalized liposomes in mice with subcutaneous NSCLC tumors was observed using liposomes labeled with a near-infrared Cy7.5 dye. The fluorescent images taken 2 hours after liposomes were injected showed an increased accumulation of CSNICARAC functionalized liposomes at the tumor site compared to liposomes lacking the peptide, but the distribution to other organs was not determined. CSNICARAC functionalized liposomes loaded with DOX slowed the rate of tumor growth when compared to non-targeting DOX liposomes and free DOX injections, an effect that was correlated with increased amounts of apoptosis observed in the tumor tissue by postmortem microscopy. Irradiation of tumors with high-intensity X-rays fundamentally alters the protein expression of the damaged cells and improves their accessibility to molecular targeting agents.[70] The HVGGSSV peptide, identified by in vivo biopanning of irradiated tumors, and HVGGSSV functionalized liposomes were found to distinguish irradiated tumor cells from untreated tumors and normal tissue in tumor models.[71,72]

4.1. Integrin-Targeting Lipid Nanoparticles Integrins are a family of membrane spanning receptors that mediate the attachment of a cell to surrounding extracellular matrix (ECM) proteins including fibronectin, vitronectin, collagen, and laminin, and play a critical role in the cell signaling to control cell shape, motility, and cell division.[76] Integrin expression is upregulated in many solid tumors and tumor vasculature, and has been implicated in tumor progression and metastasis.[77] The tripeptide sequence RGD found in many ECM proteins is a promiscuous binder of integrins and has been widely used as a targeting molecule for anticancer therapies. Liposomes functionalized with RGD peptides were tested for the treatment of various cancer types such as SKOV-3 human ovarian cancer,[78] B16F10 murine melanoma,[79] drug resistant MCF7/A human breast carcinoma,[80] and S-180 murine sarcoma.[81] RGD functionalized liposomes and non-targeting liposomes that contained the hydrophobic drug paclitaxel within their lipid membranes were more effective in inhibiting SKOV-3 tumor growth in mice compared to free paclitaxel injections,[78] while liposomes that contained the lipophilic vascular targeting drug Combretastatin A-4 (CA-4) in the lipid bilayer and DOX in the aqueous core showed a modest increase in targeting of B16F10 cells and HUVECs and inhibited tumor growth more effectively than non-targeting liposomes (Figure 11).[79] Interestingly, peptide-functionalized liposomes containing only DOX or CA-4 were less effective than non-targeting liposomes containing both drugs, demonstrating potency of this drug combination.[79] A cyclic RGD peptide (RGDFC) was used to functionalize liposomes encapsulating matrine for in vitro targeting of integrin over-expressing human Bcap-37 breast cancer, HT-29 colon cancer, and A375 melanoma cells.[82] Cyclic RGD peptides retain their integrin affinity, are more stable at neutral pH, and resist proteolysis, making them an attractive alternative to the linear form of the peptide.[83] The antiproliferative effect of matrine was found to be roughly twice as potent to each of the three cell lines when added to cells within the targeting liposomes rather than as free drug. A second cyclic RGD peptide

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

wileyonlinelibrary.com

3813

www.advmat.de

REVIEW

www.MaterialsViews.com

Figure 12. The four repeats of the fibronectin (FN) fragment III7–10 are shown: far left repeat for III7 to far right for III10. The synergy site PHSRN is in the III9 repeat. The GRGDS is in the III10. The schematic drawing of the PR_b peptide-amphiphile shows the four building blocks of the peptide headgroup: a KSS spacer, the PHSRN synergy site, a (SG)5 linker, and the RGDSP binding site. When the PR_b peptide-amphiphile is used for the preparation of functionalized liposomes, the hydrophobic tail is part of the membrane, and the peptide headgroup is exposed at the interface.

Figure 11. In vivo therapy of C57BL/6 mice with subcutaneous B16F10 melanoma tumors. Mice were injected with five doses of RGD peptide liposomes containing both CA-4 and DOX (RGD-L[CD]), CA-4 (RGDL[C]), DOX (RGD-L[D]), or five doses of non-targeting liposomes with CA-4 and DOX (RGD-L[CD]) at drug doses of 25 mg/kg CA-4 and 0.8 mg/ kg DOX. Saline was used as a control. a) Tumor volumes of animals from each treatment group were measured every other day beginning on day 8. Results are expressed as means ± SD. b) Excised tumors after 18 days of treatment. Reproduced with permission.[79] Copyright 2010, Elsevier.

(DGARYCRGDCFDG) was used to functionalize liposomes and target the paramagnetic MRI contrast agent Gd-DTPA to subcutaneous A549 lung tumors.[84] The liposomes were found to efficiently internalize within both A549 lung cancer cells and HUVECs through an integrin-dependant process and showed a significantly increased MRI signal within A549 tumors that corresponded to 16% of the injected dose of nanoparticles (16% ID) after 4 hours, while non-targeting liposomes and free GdDTPA showed no significant internalization. The cyclic RGD was also used as part of a dual-targeting approach where the liposomes were functionalized with the ARYCRGDCFDATWLPPR peptide that was comprised of the ARYCRGDCFD cyclic RGD peptide and the ATWLPPR sequence[85] targeting neuropilin-1 (NRP-1).[86] The dual-targeting liposomes, encapsulating paclitaxel, had lower IC50 values to A549 cells and HUVECs than paclitaxel administered

3814

wileyonlinelibrary.com

as free drug or within PEGylated non-targeting and RGD functionalized liposomes, supporting the hypothesis that the dualtargeting liposomes were better at targeting cancer cells. Additional support for this hypothesis came from in vivo studies that showed that mice that received the dual-targeting liposomal formulations had a stronger anti-tumor response when compared to mice the received non-targeting and single-targeting liposomes. Although biodistribution experiments were not reported, the strategy of using a single bifunctional peptide to create dual-targeting liposomes is intriguing and warrants additional study. The PR_b peptide (KSSPHSRNSGSGSGSGSGRGDSP) designed by the Kokkoli group[87] combines the RGD motif with a synergistic PHSRN sequence to create a fibronectin mimetic peptide[88] that has high affinity[89] and specificity for the α5β1 integrin (Figure 12).[87,90] While the RGD motif alone non-specifically targets a variety of integrins that have various levels of expression in healthy as well as cancerous tissue, the PR_b peptide specifically targets the α5β1 integrin, which is over-expressed in a variety of cancer cells[91,92] and on tumor vasculature[93] and plays an important role in angiogenesis and metastasis.[94] PR_b functionalized liposomes and polymersomes have shown enhanced binding, intracellular uptake, and delivery of their encapsulated loads compared to non-targeting and GRGDSP functionalized particles targeted to colon cancer cells,[90,95,96] prostate cancer cells,[97,98] and porcine islets of Langerhans,[99] as well as the capability to deliver a wide variety of therapeutic cargoes including 5-FU, DOX, tumor necrosis factor-α, siRNA, and plasmid DNA. PEGylated PR_b liposomes internalized via a receptor-mediated endocytosis and

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Adv. Mater. 2012, 24, 3803–3822

www.advmat.de www.MaterialsViews.com

were initially found in early endosomes but localized within late endosomes and lysosomes at later times.[100] In agreement with work by Garg et al.,[96] the binding efficiency of PEGylated PR_b liposomes increased with increasing PR_b functionalization and PEGylated PR_b liposomes were significantly better at delivering cytotoxic loads to cancer cells than PEGylated GRGDSP functionalized liposomes.[95,97,100] To expedite intracellular release of drugs encapsulated within liposomes after internalization, Garg and Kokkoli designed pH-sensitive liposomes that started to release their encapsulated loads when they were in early and late endosomes.[95] Solution experiments confirmed the pH-responsive release of their encapsulated cargo. Confocal microscopy and plate assays (Figure 13) were used to visualize and quantify the binding, internalization, and subsequent endosomal release from both inert and pH-sensitive PEGylated PR_b liposomes within CT26 cells. The PEGylated PR_b liposomes showed substantial intracellular release of encapsulated dye after 1 hour while inert PEGylated PR_b liposomes that were not pH-sensitive took longer to begin their release.[95] The integrin binding PHSCN peptide–derived from the synergy region of fibronectin–acts as an α5β1 antagonist, exhibits anti-tumor activity in multiple animal models[101] and was used to functionalize PEGylated DOX liposomes.[102] Free DOX was observed to accumulate within the cell nuclei of MDA-MB-231 breast cancer cells and HUVECs, while liposomal DOX was seen not only in the nuclei but also within endosomes and lysosomes. There was much more intracellular DOX delivered by targeting liposomes than by non-targeting liposomes, an effect that was shown to be specific via blocking studies. Another ECM derived peptide sequence, YIGSR, found in laminin and integral for mediating ECM-cell adhesion was

Adv. Mater. 2012, 24, 3803–3822

REVIEW

Figure 13. Binding and intracellular uptake of various PEGylated liposomal formulations with low concentrations of PEG2000 loaded with calcein. Liposomes were incubated with CT26.WT colon cancer cells at 37 °C for 4 hours (light grey) and 16 hours (dark grey). At the conclusion of the experiment cells were washed and lysed, and the total fluorescence of samples was measured. Student’s t-test analysis for # † ¶ &: p < 0.001, # † ¶ & and for ∗ ‡: p < 0.01, indicating significant statistical difference for all groups. Data shown as the mean ± SD. Reproduced with permission.[95] Copyright 2011, Bentham Science Publishers.

used to create 5-FU loaded liposomes to target both endothelial and tumor cells.[103] The YIGSR peptide itself has been found to have an inhibitory effect on tumor growth and anti-metastatic activity in several human cancers.[104] Liposomes functionalized with laminin peptides (CDPGYIGSR) were found to have a 7-fold higher binding to in vitro cultures of HUVECs than nontargeting liposomes. Encapsulating 5-FU in liposomes allowed ∼20% of the injected dose to remain in the circulation of mice bearing subcutaneous melanoma tumors after 24 hours while free 5-FU was completely cleared after 4 hours. Both non-targeting and targeting liposomes caused substantially more drug accumulation in the liver than in the tumor but also allowed for much improved tumor targeting compared to injections of free 5-FU. The benefit of the targeting liposomes was clearly shown by a significant reduction in the occurrence of spontaneous lung metastases, angiogenesis, and tumor volume while extending survival time in comparison to non-targeting liposomes and free 5-FU. Other lipid nanoparticles including solid lipid nanoparticles (SLNs), and micelles have been functionalized with integrin binding peptides to create nanoparticles for lipophilic drug delivery. The lipid cores of micelles and SLNs provide a larger and more suitable hydrophobic environment for the encapsulation of the lipophilic drugs than typically afforded by the lipid bilayers of liposomes.[105] Goutayer et al. created core-shell SLNs about 35 nm in diameter that were functionalized with cyclic RGD peptides, cyclic RAD peptides (non-binding control), or hydroxyl groups and used to target cells over-expressing integrin αvβ3.[106] RGD functionalized nanoparticles were readily bound and internalized by a HEK293 cell line genetically modified to over-express αvβ3 integrins while the RAD and hydroxyl SLNs showed much less binding. All three formulations were tested in a murine tumor model of mammary cancer known to express low levels of αvβ3 integrins. Each formulation, irrespective of their targeting capabilities, showed similar tumor to skin florescence ratios, indicating passive accumulation was the dominate force behind the tumor accumulation and the targeting peptide was less important. Lipid nanocarriers have also been utilized for multimodal imaging of tumor tissue. Mulder et al. designed paramagnetic quantum dot micelles functionalized with an αvβ3-integrin targeting peptide to image tumor angiogenesis.[107,108] These nanoprobes were prepared by encapsulating quantum dots in a monolayer of PEGylated lipids and paramagnetic lipids, which allowed them to be used for in vivo optical and MR imaging. Three complementary in vivo imaging modalities capable of detecting the αvβ3−targeted quantum dot micelles were leveraged simultaneously to visualize the micelle accumulation within the angiogenic regions of tumor-bearing mice. Intravital fluorescence microscopy, whole body fluorescence imaging and high-resolution T1-weighted MR imaging clearly showed signal enhancement at the tumor periphery where the majority of angiogenesis was occurring. 4.2. Growth Factor Receptor Targeting Lipid Nanoparticles In recent years, considerable attention has been paid to the role of growth factors and growth factor receptors in the occurrence

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

wileyonlinelibrary.com

3815

www.advmat.de

REVIEW

www.MaterialsViews.com

3816

and progression of cancer.[109,110] Over-expressed and constitutively activate growth factor receptors can result in hyperproliferation, decreased apoptosis, angiogenesis, invasion and metastasis of cancer cells.[4] Small molecule therapeutics like gefitnib and erlotinib and monoclonal antibody therapeutics like cetuximab and trastuzumab target these rogue growth factor receptors to inhibit their aberrant behavior. Unsurprisingly, these surface-expressed cancer proteins are commonly used as targets for many nanoparticle therapies, including the following peptide-functionalized lipid nanoparticles. Chen et al. created and characterized a liposomal system that used the ionic interactions between cationic liposomes and negatively charged peptides to immobilize truncated basic fibroblast growth factor (tbFGF) targeting peptides, which consist of residues 30–115 of bFGF, at liposomal surfaces.[111] The bFGF receptor is over-expressed on the surface of many tumor cells[112] including prostate, melanoma and breast as well as tumor neovasculature.[113] Despite containing no PEG brush layer to resist opsonization and non-specific cell binding, tbFGF functionalized liposomes generated ∼10 times more paclitaxel accumulation in the tumor compared with free paclitaxel and ∼3 times more than non-targeting liposomes and also delivered significantly less paclitaxel to the liver compared to the nontargeting and free paclitaxel formulations.[111,114] A strong antitumor effect was observed using the tbFGF functionalized liposomes encapsulating DOX or paclitaxel to treat multiple subcutaneous murine tumor models, demonstrating the versatility of these liposomes. A metal chelating lipid[115] that inserts into the bilayer of liposomes provided engraftments sites for five different histidinetagged peptides: Flt-1 peptide (WHSDMEWWYLLG) targeting the VEGF receptor; neuropilin-1 targeting peptide (ATWLPPR); αV integrin subunit targeting peptide (GARYCRGDCFDG); PDGF receptor (GCRGRRST) targeting peptide; and Lyp-1 peptide (CGNKRTRGC) targeting tumor lymphatic cells.[116,117] When compared to liposomes functionalized with non-binding peptides, the PDGF receptor targeting liposomes were seen to dramatically increase cell targeting in vitro but failed to enhance tumor accumulation in vivo. Conversely, the integrin-targeting liposomes had a marginal increase in cell targeting in vitro but a significant increase in tumor accumulation in vivo.[117] Regardless of the engrafted peptide, the major sites of accumulation for the liposomes were the liver, spleen, and lungs. Separate in vivo experiments using the same tumor model showed that neuropilin-1 targeting liposomes modestly improved tumor targeting while the VEGFR targeting liposomes were no better than non-binding control peptide liposomes.[116] Again, the major sites of liposome accumulation were the liver, spleen, and lungs, each with 8–16% ID/g. In an attempt to improve tumor targeting, neuropilin-1 and VEGFR targeting liposomes with shorter PEGylations were used in an identical experiment. These shorter PEG lipids doubled the tumor accumulation of both the neuropilin-1 and VEGFR targeting liposomes without increasing the accumulation of liposomes in the liver or spleen. Thrombospondin proteins (TSPs) bind to multiple surface expressed molecules–including transforming growth factor–β (TGF-β) and integrins[118]−and impact angiogenesis,[119] apoptosis,[120] and activation of TGF-β.[121] The KRFKQDGGWSHW

wileyonlinelibrary.com

peptide (TSPB) isolated from a TSP protein contains a WHSW motif responsible for TGF-β binding and a KRFK motif implicated in TGF-β activation. A second peptide sequence, TRIRQDGGWSHW (TSPA), also contains the WHSW motif but replaces the TRIR sequence with KRFK. Rivera-Fillat et al. found the aspartamide analogs of these peptides were better ligands for tumor cells than the native peptides and therefore used them to create targeting liposomes that bound to endothelial cells and tumor tissue.[122] Liposomes functionalized with the [Abu6]TSPA peptides containing only the TGF-β binding motif significantly reduced murine tumor growth rate while treatment with liposomes functionalized with peptides containing the [Abu6]TSPB peptides containing the TGF-β activation domains were unable to slow tumor growth any more than non-targeting liposomes.[122] Further testing revealed the TGF-β activation domain of the [Abu6]TSPB peptide increased HUVEC proliferation and new blood vessel formation, which suggested it is a pro-angiogenic molecule that may support in vivo tumor progression and a poor choice for use as a targeting molecule. The epidermal growth factor receptor (EGFR) is a cell membrane bound receptor tyrosine kinase that once activated by EGF binding initiates intracellular cell signaling that results in DNA synthesis and cell proliferation.[123] Mutations in the EGF receptor that upregulate its expression or increase its activity often lead to the uncontrolled cell proliferation associated with cancer.[124] An EGFR binding peptide (LARLLT) was identified using a computer assisted design process, yielding a peptide that bound to the receptor in a separate location from the EGF protein and did not result in receptor activation.[125] In vitro experiments suggested LARLLT functionalized liposomes bound and internalized within NSCLC cells via an active receptor-mediated internalization pathway. 4.3. Lipid Nanoparticles Targeting other Surface-Expressed Molecules Tie2 is a receptor tyrosine kinase that plays an essential role in the initiation, progression, and maturation of neovasculature and is dramatically over-expressed in tumor endothelium.[126] Blocking or interfering with the Tie2 signaling pathway inhibits angiogenesis and reduces tumor growth.[127] In vitro phage display biopanning against the recombinant Tie2 protein identified a 12 amino acid peptide PH1 (TMGFTAPRFPHY) with moderate affinity for the Tie2 protein (KD = 15.8 μM).[128] The PH1 peptide was found to bind to tumor endothelium as well as various other cancer cells that over-express the Tie2 receptor. Liposomes functionalized with the PH1 peptides and containing cisplatin internalized within Tie2 expressing cells 2- to 3-fold more than non-targeting liposomes and were more toxic to these cells than free cisplatin and cisplatin contained within non-targeting liposomes. The underdeveloped and rapidly growing vascular system that supplies the tumor microenvironment with oxygen and nutrients also provides a significant aid to the extravasation of nanoparticles from the blood stream. The lymphatic system of tumors, also underdeveloped and undergoing lymphangiongenesis, is responsible for removing and filtering excess fluid

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Adv. Mater. 2012, 24, 3803–3822

www.advmat.de www.MaterialsViews.com

Figure 14. A diagram of a cooperative nanosystem. Irradiation of the photosensitive gold nanorods located in the tumor stimulated changes in the tumor microenvironment that allowed nanoworms and liposomes functionalized with the Lyp-1 peptide to more effectively bind to the tumor associated cells. Transmission electron microscopy images of the gold nanorods (left), nanoworms (top right) and liposomes (bottom right) are shown with scale bars of 50 nanometers. Reproduced with permission.[133] Copyright 2010, National Academy of Science, USA.

and debris from the tumor and acts as a metastatic route away from the primary tumor.[129] A cyclic 9 amino acid peptide, LyP-1 (CGNKRTRGC), was isolated by in vivo phage display and found to bind to stress-related protein p32 that is over-expressed on certain tumor cells as well as tumor lymphatics,[130] offering the potential for multisite targeting to tumor cells as well as lymphatic endothelial cells. Liposomes functionalized with the LyP-1 peptide were found to reduce lymphatic vessel density and lymph node growth in metastatic lung[131] and breast cancer[132] tumor models. The LyP-1 peptide was also used to create a cooperative tumor therapy consisting of injections of two separate types of nanoparticles: passively targeting gold nanorods that heated the tumor and upregulated the expression of p32 stress proteins and DOX liposomes functionalized with Lyp-1 peptides that actively targeted p32 to enhance tumor destruction (Figure 14).[133] Three times more DOX was delivered by LyP-1 functionalized liposomes than non-targeting liposomes to tumors that had been photothermally primed, and substantially fewer Lyp-1 DOX liposomes accumulated within tumors that had not received the near infrared induced heat therapy. Mice that received a single low dose of LyP-1 liposomes following phototherapeutic priming were found to have significant tumor regression or elimination, a result not seen with non-targeting liposomes or LyP-1 functionalized liposomes that were not first thermally activated. Grange et al. demonstrated the ability for liposomes to act as both therapeutic and diagnostic agents by creating liposomes loaded with both DOX and the MRI imaging agent gadolinium (Gd).[134] The liposomes were functionalized with the C3d peptide (ASKKPKRNIKA) that binds with high affinity to the neural cell adhesion molecule (NCAM), which is expressed by several solid tumors, some leukemia, and tumor endothelial cells.[135,136] The C3d functionalized liposomes were internalized by both Kaposi’s sarcoma cells and tumor endothelial cells in vitro, but when injected into mice with Kaposi’s sarcomas the liver and spleen received a substantial amount of the liposomes.

Adv. Mater. 2012, 24, 3803–3822

Targeted nanoparticle drug delivery has been widely hailed as a revolutionary technology, capable of altering the landscape of the pharmaceutical industry. By creating a peptide-functionalized nanoparticle delivery platform that can carry a wide variety of drugs to specific cells in the body by simply changing the targeting molecule, larger amounts of drugs could be delivered to the cells with fewer side effects and lower overall drug doses, leading to significant improvements in the clinical outcomes of patients. Designing such nanoparticle systems has been the focus of intensive research over the past two decades and substantial progress has been made. Lipid-based nanoparticles are simple to construct, non-toxic, biodegradable, and easily engineered to carry the desired targeting molecules and therapeutic drugs, which make them an attractive type of nanoparticle for targeted delivery vehicle construction.[137,138] In recent years lipid-based nanoparticles have been functionalized with a variety of targeting ligands and used to deliver both hydrophilic and hydrophobic small molecule drugs, peptides, proteins, nucleic acids, and imaging molecules to tumors and tumor cells.[79,84,96,137,139–141] While lipid nanoparticles provide a robust platform that can be tailored to accommodate the desired targeting peptides and therapeutic drugs, targeted lipid nanoparticles have not yet produced the therapeutic benefits many predicted and have not yet significantly impacted the treatment of human disease.[142] As discussed in the introduction to this review, rapid clearance from the blood stream that reduces passive accumulation of the nanoparticles within the tumor microenvironment, tumor cell membranes that are typically impenetrable to large molecules, and sequestration and degradation of the nanoparticles within endocytic vesicles that keep the drugs from reaching their sites of action pose three obstacles to successful targetednanoparticle drug delivery. The recently published peptidefunctionalized lipid nanoparticle literature that was reviewed in the previous sections describes a variety of strategies to overcome these barriers. The following sections will summarize these strategies to help identify the successful approaches and recognize the challenges that have not yet been adequately addressed.

REVIEW

5. Discussion

5.1. Extending the Nanoparticle Circulation time while Avoiding MPS The therapeutic benefit of Doxil relies on the ability of liposomes to extend the plasma half-life of the drug they encapsulate, which allows for passive targeting of cancer cells through the EPR effect,[143] and to reduce the cardiotoxicity caused by non-specific delivery of the drug to the heart, which is caused by direct injections of DOX into the blood.[144] A critical component of these liposomes is the PEG brush coating that covers their surfaces and allows them to hide from recognition by the MPS as well as reduces non-specific interactions with non-cancerous cells.[144] Peptides used to create the second-generation lipid nanoparticles must be able to interact with the cell surfaces to generate cell-specific targeting without compromising the nanoparticles’ ability to passively accumulate at the tumor site. There

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

wileyonlinelibrary.com

3817

www.advmat.de

REVIEW

www.MaterialsViews.com

3818

are three common techniques for functionalizing the surfaces of lipid nanoparticles with peptides: conjugating the peptides at the tips of the PEG molecules present on the surface of preformed liposomes,[33,72,103] adding peptide amphiphiles to the lipid mixture during liposome formation,[84,86] and transferring peptide amphiphiles from micelles into the lipid bilayers of liposomes.[24] Conjugation of the peptides to the ends of the PEG molecules places the peptides at the outer edge of the steric PEG barrier, granting them easy access to their surrounding environment but placing them at risk of opsonization and capture by the MPS. Alternately, the peptides can be attached at the nanoparticle surface and within the PEG brush layer. This strategy should more effectively inhibit opsonization and capture by the MPS and reduce the occurrence of non-specific interactions of the nanoparticles and healthy cells that could be triggered by the presence of the peptides.[145] An early step in the development of many peptide-functionalized nanoparticles is in vitro testing, carried out with the goal of reducing the non-specific accumulation of the nanoparticles in cells that do not express the target to which the peptide binds. The effect of PEG and peptide concentration has been tested in vitro and results illustrate the trade-off between PEG and peptides on cell targeting.[96,97] As expected, lower molar amounts of PEG or shorter PEG molecules less effectively screened liposomes from their surrounding environment and led to increased non-specific cell binding.[96,97] Conversely, the addition of peptides onto the liposomal surfaces generated cell targeting that was specific and dependant on the amounts of peptides that were used.[96,97] The recently published literature clearly shows peptide-functionalized nanoparticles can be designed to effectively reduce the non-specific targeting of ‘healthy’ cells while maintaining the targeting ability of the cancer cells.[117,131] However, these in vitro experiments cannot adequately mimic the additional challenge of avoiding opsonization and clearance from the blood making it difficult to predict their in vivo success even if they display ‘ideal’ in vitro performance. In vivo experiments demonstrated that the incorporation of different drugs[103] and imaging agents[84] into peptide-functionalized nanoparticles significantly increased their plasma halflife regardless of the method used to functionalize the nanoparticle.[24,33,84,103] In many cases the plasma half-life of drugs encapsulated within non-targeting PEGylated nanoparticles and peptide-functionalized nanoparticles were very similar, and in some cases greater than 20% of the drug remained in circulation after 24 hours.[33,103] For comparison, the plasma concentration of doxorubicin encapsulated within liposomes without PEG was reduced by >90% just 30 minutes following the intravenous injection of the nanoparticles.[81] Measurements of nanoparticle and drug plasma concentrations do not provide a complete understanding of the distribution and elimination processes that occur after the nanoparticles are injected. To better understand the pharmacokinetics of the targeted nanoparticles, quantitative assessment of the in vivo biodistribution of the nanoparticles and drugs to the healthy tissues and the tumor can be made by directly measuring the drug or nanoparticle concentrations within the organs and tumor. Results from in vivo biodistribution studies often report the delivery of drugs or nanoparticles to any number of organs

wileyonlinelibrary.com

including liver, spleen, kidneys, heart, lungs, brain, and muscles. Many of the recent peptide-functionalized lipid nanoparticle designs reviewed here reported that the liver and spleen received the greatest proportion of the nanoparticles that were injected, while the tumor received substantially less.[52,54,68,143] The liver and spleen not only act as nanoparticle filters capable of sequestering the nanoparticles based on their physical size, but are also home to significant populations of macrophages and monocytes of the MPS, making it tempting to attribute their high accumulation to nanoparticle opsonization.[146] Conventional theory suggests that PEGylating the surfaces of nanoparticles should produce long circulating nanoparticles that avoid capture by the liver and spleen, while the addition of the peptides to these nanoparticles has the potential to increase the opsonization of the nanoparticles and cause them to be more quickly captured by these organs.[143] However, many examples from the recently published literature demonstrated that there was no significant increase in the distribution of the PEGylated peptide-functionalized nanoparticles into the liver and spleen when compared to PEGylated non-targeting nanoparticles.[52,78,116,117] In fact, in some cases the peptide-functionalized nanoparticles were shown to have less accumulation than the non-targeting liposomes.[30,78] These results suggest that the capture of the nanoparticles by the liver and spleen is not solely due to opsonization of the peptides used to functionalize the nanoparticles. Therefore, results up to now suggest the accumulation of anticancer drugs in tumors can be modestly enhanced (Table 2) by peptide-functionalized nanoparticles when compared to nontargeting nanoparticles, but there is typically low overall delivery to the tumor and substantial delivery to other ‘healthy’ tissues like the liver and spleen. PEGylation of nanoparticles has long been considered as an effective strategy for extending the nanoparticles’ circulation times and reducing their accumulation in these organs caused by MPS capture.[147] While, without question, PEGylation is a useful approach, the majority of PEGylated targeting- and non-targeting nanoparticles are still removed from circulation within hours after intravenous injection.[49,54] Thus, there is still room for improvement that will further enhance the long-term circulation of these nanoparticles and their passive targeting and reduce their off-target delivery to healthy tissues. 5.2. Enhancing the Retention of Drugs in the Tumor Microenvironment and Triggering Internalization of Drugs into Tumor Cells While the nanoparticles are circulating in the blood stream, the role of the peptides is minimal. This changes once the nanoparticles extravasate into the tumor tissue and encounter tumor cells. The specific interactions between the peptides of the targeting nanoparticles and the tumor cells increase the likelihood that the nanoparticles will remain in the vicinity of the diseased cells and will not return to the circulation, therefore increasing the concentration of drug within the tumor.[148] Specific interactions between the cancer cells and peptides can also trigger nanoparticle endocytosis,[149] thus overcoming the intracellular drug delivery barrier and providing additional insurance that the drug will remain within the tumor.

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Adv. Mater. 2012, 24, 3803–3822

www.advmat.de www.MaterialsViews.com

Adv. Mater. 2012, 24, 3803–3822

these experiments demonstrate the therapeutic and diagnostic abilities of the nanoparticles, they do not provide a direct way to quantitatively assess the targeting effect and biodistribution of these nanoparticles. As Table 1 and Table 2 show, the second challenge of increasing the nanoparticles’ binding and internalization within tumor cells has been well addressed in the reviewed literature. High affinity peptides and dual-targeting strategies that afford peptide-functionalized nanoparticles strong and specific interactions with cells that express the target molecules have demonstrated an enhanced in vitro tumor cell binding and internalization.[30,86] However, the capability of the peptidefunctionalized nanoparticles to display similarly enhanced cell targeting when injected in vivo is complicated by the heterogeneity of cells within the tumor microenvironment and the variable expression levels of the target proteins.[103,114,153,154] Nonetheless, these issues do not negate the benefits gained by the peptide-functionalized nanoparticles currently used to improve cell binding and internalization, but rather suggest that the nanoparticles face additional challenges when used for in vivo therapy.

REVIEW

Internalization is strongly dependant on the length of time a nanoparticle remains associated with the cancer cell surface and prolonged interactions between peptide-functionalized nanoparticles and surface-expressed molecules of cancer cells are more likely to occur if the interaction is of higher affinity.[149] Not surprisingly, when choosing targeting peptides for drug delivery the peptides with the highest affinity for the specific cancer cell target are often preferred. High affinity peptides can be isolated from the binding domains of large proteins that interact with the receptors that are expressed on cancer cell surfaces,[103,114] as well as by phage display biopanning.[24,69] An alternate strategy employed to generate higher binding affinity and specificity between the peptide nanoparticles and the tumors cells was to create dual-targeting liposomes,[30,86] a strategy that has been shown to increase the targeting efficacy by enhancing the nanoparticles’ uptake into cells and reducing nanoparticle diffusion away from the tumor and back into the blood stream.[148] Dual-targeting nanoparticles offer several potential benefits over single-targeting nanoparticles as they can bind to two targets on tumor cells, not only increasing the affinity of the nanoparticles for the cells of interest but also decreasing the off-target binding to cells that do not express both targets.[30,86] PEGylation of the peptide-functionalized nanoparticles has the potential to mask the peptides that reside on the nanoparticles’ surface and impair the peptides’ abilities to enhance the accumulation of the nanoparticles within the tumor.[116] In vitro studies have shown that by varying the amount and length of PEG and the peptide concentration on the surface of the liposomes an optimal combination can be identified that will allow for the steric repulsion of the PEG molecules without sacrificing the specific targeting conferred by the peptide ligand.[96] Another strategy to overcome this issue is to use PEG molecules that can be cleaved from the nanoparticles’ surfaces once the nanoparticles extravasate into the tumor. Cleavage of PEG molecules from the surfaces of nanoparticles can be induced by low pH,[150] reducing agents,[151] and enzymatic stimuli[152] present in the extracellular microenvironment of the tumor. Kuai et al. recently employed this strategy to create liposomes functionalized with the cell penetrating TAT peptides that were initially shielded by a cysteine-cleavable PEG layer.[60,61] The liposomes showed no difference in their passive accumulation within tumors when compared to non-functionalized liposomes, but enhanced the intracellular uptake of the liposomes following an injection of cysteine.[61] Li et al.[84] who used a cyclic version of the RGD peptide, and Park et al.[133] who used the Lyp-1 peptide, reported significantly higher amounts of delivery to the tumor from peptide-functionalized liposomes than from non-targeting liposomes. These two examples showed clear and quantifiable improvements in the delivery of drugs to tumors. It is important to mention that quantitative evaluation of the different strategies can be aided by how the authors choose to report the drug delivery to the tumors and other organs (i.e.,% ID,% ID/g). Many of the recent examples of peptide-functionalized nanoparticles instead reported the effects of the targeting nanoparticles on tumor size and animal survival time as proxies for nanoparticle drug delivery.[78–80,86,131,134] And in other cases, fluorescence,[25,49] magnetic resonance,[107] or gamma camera imaging[52] were used to qualitatively assess the delivery of nanoparticles. While

5.3. Moving the Therapeutic Payload to its Site of Action to Produce Effective Anticancer Therapy The last obstacle to overcome after delivering the drug to the tumor is the translocation of the drug across the cell membrane and into the intracellular environment where it can begin its therapeutic action. Many peptide-functionalized nanoparticles target molecules that undergo constitutive or binding-triggered internalization via a clathrin-mediated process.[14] Upon binding to these molecules the nanoparticles are transferred across the cell membrane and into lipid vesicles (early/late endosomes and lysosomes). These intracellular compartments sequester the drugs in a location where they are inactive[64] and have low pH and enzymes that are damaging to many biological drugs like peptides, proteins, and nucleic acids.[12] Fast and extensive release of the drugs from these endocytic vesicles has been shown to improve the therapeutic effects of some drugs.[155] On the other hand, the environments of the endosomes and lysosomes can be used to enhance escape of the drugs from their lipid nanoparticles. The Kokkoli[95] and Torchilin[66] groups each designed peptide-functionalized liposomes that can rapidly release their encapsulated payloads once the liposomes have been internalized. Garg and Kokkoli created pH-sensitive peptidefunctionalized liposomes that allowed the encapsulated drug to escape rapidly from the pH-sensitive liposomes after the nanoparticles were internalized.[95] The phage fusion coat protein used by Wang et al. to immobilize the targeting peptide on the liposomal membrane was also found to disrupt the endosomal membranes of cells as the pH within the endosomes dropped.[66] This disruption allowed the encapsulated drugs to escape the endocytic pathway before they reached the lysosomes. Endosomes that were not allowed to acidify were not able to effectively trigger the release the encapsulated drugs from the liposomes. This led to a sequestration of drugs within the endosomes that significantly reduced

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

wileyonlinelibrary.com

3819

www.advmat.de

REVIEW

www.MaterialsViews.com

their therapeutic effects and demonstrated the importance of escaping the endosomes. Nanoparticles functionalized with CPPs that do not target specific cell-expressed molecules are thought to employ multiple endocytic pathways including macropinocytosis, caveolaemediated endocytosis, as well as direct translocation across the cell membrane that can avoid the late endosomes and lysosomes.[14] The He,[60,61] Harashima,[63] and Choi[62] groups successfully used these peptides to move nanoparticles into the cytoplasm of cells but the specific mechanism(s) of internalization mediated by CPPs was not explored. Drugs carried within nanoparticles functionalized with CPPs that are internalized by vesicles that are not trafficked through endosomes to lysosomes do not encounter the enzymes or low pH environments that can be used to trigger their release and must therefore rely on alternate strategies to help them escape from these organelles. Thermo-sensitive liposomes, like those created by Negussie et al., that rapidly release their encapsulated drugs when exposed to elevated temperatures, may provide a potential escape strategy for drugs contained within non-endosomal organelles.[36] Relatively little attention has been paid to the challenge of designing peptide-functionalized lipid nanoparticles that can escape from endocytic organelles, considering the effort put forth in the design of nanoparticles that can avoid capture by the MPS and enhance their tumor accumulation. The complexity and incomplete understanding of the mechanisms that control internalization and sub-cellular trafficking of the peptide-functionalized nanoparticles may explain the lack of focus given to this challenge.[14,156] Nevertheless, several strategies have been shown to enhance the intracellular release of different drugs.[65,95] As we continue to learn more about the complex relationships that exist between the nanoparticles and the endocytic organelles that ultimately determine the fate of the targeted delivery systems, new opportunities to enhance intracellular drug delivery will likely be identified.

6. Conclusion The recently published examples of peptide-functionalized lipid nanoparticles reviewed in this article demonstrated a wide range of approaches for overcoming the barriers that impede effective delivery of drugs to tumors. Encouragingly, many of these approaches were more effective than the first-generation nontargeting nanoparticles at treating a variety of tumor models. To ensure these second generation nanoparticles achieve the clinical success that is widely expected, a continued effort to minimize the accumulation of the nanoparticles in the liver and spleen, to increase the amount of drug within the tumor, and to enhance the specific delivery of the therapeutic load at its site of action, must be made.

Acknowledgements This work was supported by the CAREER award NSF/CBET-0846274. Received: February 28, 2012 Published online: June 5, 2012

3820

wileyonlinelibrary.com

[1] M. D. Stubblefield, M. W. O’Dell, Cancer Rehabilitation Principles and Practice, Demos Medical, 2009. [2] American Cancer Society, Cancer Fact & Figures 2011, Atlanta 2011. [3] J. Xu, in National Vital Statistics Report, US Department of Health and Human Services, Center for Disease Control and Prevention, 2010. [4] D. Hanahan, R. A. Weinberg, Cell 2000, 100, 57. [5] D. Hanahan, R. A. Weinberg, Cell 2011, 144, 646. [6] Y. Matsumura, H. Maeda, Cancer Res. 1986, 46, 6387. [7] H. Pinedo, Drugs Affecting Growth of Tumors, Springer, 2006. [8] C. H. Choi, J. E. Zuckerman, P. Webster, M. E. Davis, Proc. Natl. Acad. Sci. USA 2011, 108, 6656. [9] H. S. Choi, W. Liu, P. Misra, E. Tanaka, J. P. Zimmer, B. Itty Ipe, M. G. Bawendi, J. V. Frangioni, Nat. Biotechnol. 2007, 25, 1165. [10] S. M. Moghimi, C. J. Porter, I. S. Muir, L. Illum, S. S. Davis, Biochem. Biophys. Res. Commun. 1991, 177, 861. [11] T. Schluep, J. Hwang, I. J. Hildebrandt, J. Czernin, C. H. Choi, C. A. Alabi, B. C. Mack, M. E. Davis, Proc. Natl. Acad. Sci. USA 2009, 106, 11394. [12] L. M. Bareford, P. W. Swaan, Adv. Drug Deliv. Rev. 2007, 59, 748. [13] T. G. Iversen, T. Skotland, K. Sandvig, Nano Today 2011, 6, 176. [14] G. Sahay, D. Y. Alakhova, A. V. Kabanov, J. Control. Release 2010, 145, 182. [15] Z. Wang, C. Tiruppathi, J. Cho, R. D. Minshall, A. B. Malik, IUBMB Life 2011, 63, 659. [16] J. Wang, J. D. Byrne, M. E. Napier, J. M. DeSimone, Small 2011, 7, 1919. [17] N. Zhang, D. Liu, Critical Reviews in Therapeutic Drug Carrier Systems 2010, 27, 371. [18] T. O. Pangburn, M. A. Petersen, B. Waybrant, M. M. Adil, E. Kokkoli, J. Biomech. Eng. 2009, 131, 074005. [19] E. Blanco, A. Hsiao, A. P. Mann, M. G. Landry, F. Meric-Bernstam, M. Ferrari, Cancer Science 2011, 102, 1247. [20] G. Bergers, L. E. Benjamin, Nat. Rev. Cancer 2003, 3, 401. [21] N. Ferrara, R. S. Kerbel, Nature 2005, 438, 967. [22] G. P. Smith, V. A. Petrenko, Chem. Rev. 1997, 97, 391. [23] R. Pasqualini, E. Ruoslahti, Nature 1996, 380, 364. [24] D. K. Chang, C. Y. Chiu, S. Y. Kuo, W. C. Lin, A. Lo, Y. P. Wang, P. C. Li, H. C. Wu, J. Biol. Chem. 2009, 284, 12905. [25] Y. Katanasaka, T. Ishii, T. Asai, H. Naitou, N. Maeda, F. Koizumi, S. Miyagawa, N. Ohashi, N. Oku, Int. J. Cancer 2010, 127, 2685. [26] M. A. Arap, J. Lahdenranta, P. J. Mintz, A. Hajitou, A. S. Sarkis, W. Arap, R. Pasqualini, Cancer Cell 2004, 6, 275. [27] S. Luo, C. Mao, B. Lee, A. S. Lee, Mol. Cell. Biol. 2006, 26, 5688. [28] J. J. Virrey, D. Dong, C. Stiles, J. B. Patterson, L. Pen, M. Ni, A. H. Schonthal, T. C. Chen, F. M. Hofman, A. S. Lee, Mol. Cancer Res. 2008, 6, 1268. [29] U. K. Misra, R. Deedwania, S. V. Pizzo, J. Biol. Chem. 2005, 280, 26278. [30] Y. Murase, T. Asai, Y. Katanasaka, T. Sugiyama, K. Shimizu, N. Maeda, N. Oku, Cancer Lett. 2010, 287, 165. [31] R. Pasqualini, E. Koivunen, R. Kain, J. Lahdenranta, M. Sakamoto, A. Stryhn, R. A. Ashmun, L. H. Shapiro, W. Arap, E. Ruoslahti, Cancer Res. 2000, 60, 722. [32] N. Oku, T. Asai, K. Watanabe, K. Kuromi, M. Nagatsuka, K. Kurohane, H. Kikkawa, K. Ogino, M. Tanaka, D. Ishikawa, H. Tsukada, M. Momose, J. Nakayama, T. Taki, Oncogene 2002, 21, 2662. [33] M. Loi, S. Marchio, P. Becherini, D. Di Paolo, M. Soster, F. Curnis, C. Brignole, G. Pagnan, P. Perri, I. Caffa, R. Longhi, B. Nico, F. Bussolino, C. Gambini, D. Ribatti, M. Cilli, W. Arap, R. Pasqualini, T. M. Allen, A. Corti, M. Ponzoni, F. Pastorino, J. Control. Release 2010, 145, 66. [34] G. Colombo, F. Curnis, G. M. De Mori, A. Gasparri, C. Longoni, A. Sacchi, R. Longhi, A. Corti, J. Biol. Chem. 2002, 277, 47891.

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Adv. Mater. 2012, 24, 3803–3822

www.advmat.de www.MaterialsViews.com

Adv. Mater. 2012, 24, 3803–3822

[69] X. He, M. H. Na, J. S. Kim, G. Y. Lee, J. Y. Park, A. S. Hoffman, J. O. Nam, S. E. Han, G. Y. Sim, Y. K. Oh, I. S. Kim, B. H. Lee, Mol. Pharm. 2011, 8, 430. [70] D. Hallahan, L. Geng, S. Qu, C. Scarfone, T. Giorgio, E. Donnelly, X. Gao, J. Clanton, Cancer Cell 2003, 3, 63. [71] Z. Han, A. Fu, H. Wang, R. Diaz, L. Geng, H. Onishko, D. E. Hallahan, Nat. Med. 2008, 14, 343. [72] A. Lowery, H. Onishko, D. E. Hallahan, Z. Han, J. Control. Release 2011, 150, 117. [73] P. M. Lorusso, A. Jimeno, G. Dy, A. Adjei, J. Berlin, L. Leichman, J. A. Low, D. Colburn, I. Chang, S. Cheeti, J. Y. Jin, R. A. Graham, Clin. Cancer Res. 2011, 17, 5774. [74] M. Gross-Goupil, C. Massard, A. Ravaud, Curr. Urol. Rep. 2012, 13, 16. [75] R. L. Yauch, J. Settleman, Curr. Opin. Genet. Dev. 2012, 22, 45. [76] E. A. Clark, J. S. Brugge, Science 1995, 268, 233. [77] J. S. Desgrosellier, D. A. Cheresh, Nat. Rev. Cancer 2010, 10, 9. [78] H. Zhao, J. C. Wang, Q. S. Sun, C. L. Luo, Q. Zhang, J. Drug Target. 2009, 17, 10. [79] Y. F. Zhang, J. C. Wang, D. Y. Bian, X. Zhang, Q. Zhang, Eur. J. Pharm. Biopharm. 2010, 74, 467. [80] J. Jiang, S. J. Yang, J. C. Wang, L. J. Yang, Z. Z. Xu, T. Yang, X. Y. Liu, Q. Zhang, Eur. J. Pharm. Biopharm. 2010, 76, 170. [81] H. Du, C. Cui, L. Wang, H. Liu, G. Cui, Mol. Pharm. 2011, 8, 1224. [82] X. Y. Liu, L. M. Ruan, W. W. Mao, J. Q. Wang, Y. Q. Shen, M. H. Sui, Int. J. Med. Sci. 2010, 7, 197. [83] S. J. Bogdanowich-Knipp, S. Chakrabarti, T. D. Williams, R. K. Dillman, T. J. Siahaan, J. Pept. Res. 1999, 53, 530. [84] W. Li, B. Su, S. Meng, L. Ju, L. Yan, Y. Ding, Y. Song, W. Zhou, H. Li, L. Tang, Y. Zhao, C. Zhou, Eur. J. Radiol. 2011, 80, 598. [85] R. Binetruy-Tournaire, C. Demangel, B. Malavaud, R. Vassy, S. Rouyre, M. Kraemer, J. Plouet, C. Derbin, G. Perret, J. C. Mazie, EMBO J. 2000, 19, 1525. [86] S. Meng, B. Su, W. Li, Y. Ding, L. Tang, W. Zhou, Y. Song, H. Li, C. Zhou, Nanotechnology 2010, 21, 415103. [87] A. Mardilovich, J. A. Craig, M. Q. McCammon, A. Garg, E. Kokkoli, Langmuir 2006, 22, 3259. [88] E. Kokkoli, A. Mardilovich, A. Wedekind, E. L. Rexeisen, A. Garg, J. A. Craig, Soft Matter 2006, 2, 1015. [89] K. Shroff, T. R. Pearce, E. Kokkoli, Langmuir 2012, 28, 1858. [90] T. O. Pangburn, F. S. Bates, E. Kokkoli, Soft Matter 2012, 8, 4449. [91] L. M. Ellis, Am. Surg. 2003, 69, 3. [92] Y. Jia, Z. Z. Zeng, S. M. Markwart, K. F. Rockwood, K. M. Ignatoski, S. P. Ethier, D. L. Livant, Cancer Res. 2004, 64, 8674. [93] S. Kim, K. Bell, S. A. Mousa, J. A. Varner, Am. J. Pathol. 2000, 156, 1345. [94] P. C. Brooks, Cancer Metastasis Rev. 1996, 15, 187. [95] A. Garg, E. Kokkoli, Curr. Pharm. Biotechnol. 2011, 12, 1135. [96] A. Garg, A. W. Tisdale, E. Haidari, E. Kokkoli, Int. J. Pharm. 2009, 366, 201. [97] D. Demirgoz, A. Garg, E. Kokkoli, Langmuir 2008, 24, 13518. [98] D. Demirgoz, T. O. Pangburn, K. P. Davis, S. Lee, F. S. Bates, E. Kokkoli, Soft Matter 2009, 5, 2011. [99] N. A. Atchison, W. Fan, K. K. Papas, B. J. Hering, M. Tsapatsis, E. Kokkoli, Langmuir 2010, 26, 14081. [100] K. Shroff, E. Kokkoli, Langmuir 2012, 28, 4729. [101] P. Khalili, A. Arakelian, G. Chen, M. L. Plunkett, I. Beck, G. C. Parry, F. Donate, D. E. Shaw, A. P. Mazar, S. A. Rabbani, Mol. Cancer Ther. 2006, 5, 2271. [102] W. Dai, T. Yang, X. Wang, J. Wang, X. Zhang, Q. Zhang, J. Drug Target. 2010, 18, 254. [103] P. K. Dubey, D. Singodia, S. P. Vyas, J. Drug Target. 2010, 18, 373. [104] K. Yamamura, M. C. Kibbey, S. H. Jun, H. K. Kleinman, Semin. Cancer Biol. 1993, 4, 259. [105] B. Heurtault, P. Saulnier, B. Pech, J. E. Proust, J. P. Benoit, Biomaterials 2003, 24, 4283.

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

wileyonlinelibrary.com

REVIEW

[35] S. Marchio, J. Lahdenranta, R. O. Schlingemann, D. Valdembri, P. Wesseling, M. A. Arap, A. Hajitou, M. G. Ozawa, M. Trepel, R. J. Giordano, D. M. Nanus, H. B. Dijkman, E. Oosterwijk, R. L. Sidman, M. D. Cooper, F. Bussolino, R. Pasqualini, W. Arap, Cancer Cell 2004, 5, 151. [36] A. H. Negussie, J. L. Miller, G. Reddy, S. K. Drake, B. J. Wood, M. R. Dreher, J. Control. Release 2010, 143, 265. [37] G. N. Hortobagyi, Drugs 1997, 54 Suppl 4, 1. [38] T. Safra, Oncologist 2003, 8 Suppl 2, 17. [39] A. G. Gilman, Annu. Rev. Biochem. 1987, 56, 615. [40] D. Filmore, Modern Drug Discovery 2004, 7, 4. [41] R. Markwalder, J. C. Reubi, Cancer Res. 1999, 59, 1152. [42] I. U. Khan, A. G. Beck-Sickinger, Anticancer Agents Med. Chem. 2008, 8, 186. [43] A. V. Schally, A. M. Comaru-Schally, A. Nagy, M. Kovacs, K. Szepeshazi, A. Plonowski, J. L. Varga, G. Halmos, Front. Neuroendocrinol. 2001, 22, 248. [44] M. Volante, R. Rosas, E. Allia, R. Granata, A. Baragli, G. Muccioli, M. Papotti, Mol. Cell. Endocrinol. 2008, 286, 219. [45] S. A. Wank, Am. J. Physiol. 1995, 269, G628. [46] L. J. Krebs, X. Wang, H. E. Pudavar, E. J. Bergey, A. V. Schally, A. Nagy, P. N. Prasad, C. Liebow, Cancer Res. 2000, 60, 4194. [47] J. O. Olsen, R. V. Pozderac, G. Hinkle, T. Hill, T. M. O’Dorisio, W. J. Schirmer, E. C. Ellison, M. S. O’Dorisio, Semin. Nucl. Med. 1995, 25, 251. [48] C. J. Smith, W. A. Volkert, T. J. Hoffman, Nucl. Med. Biol. 2005, 32, 733. [49] J. Zhang, W. Jin, X. Wang, J. Wang, X. Zhang, Q. Zhang, Mol. Pharm. 2010, 7, 1159. [50] Y. He, L. Zhang, C. Song, Int. J. Nanomedicine 2010, 5, 697. [51] A. Morisco, A. Accardo, D. Tesauro, R. Palumbo, E. Benedetti, G. Morelli, Biopolymers 2011, 96, 88. [52] A. Accardo, R. Mansi, A. Morisco, G. Mangiapia, L. Paduano, D. Tesauro, A. Radulescu, M. Aurilio, L. Aloj, C. Arra, G. Morelli, Mol. Biosyst. 2010, 6, 878. [53] C. Falciani, A. Accardo, J. Brunetti, D. Tesauro, B. Lelli, A. Pini, L. Bracci, G. Morelli, ChemMedChem 2011, 6, 678. [54] A. Helbok, C. Rangger, E. von Guggenberg, M. Saba-Lepek, T. Radolf, G. Thurner, F. Andreae, R. Prassl, C. Decristoforo, Nanomedicine 2012, 8, 112. [55] A. D. Frankel, C. O. Pabo, Cell 1988, 55, 1189. [56] M. Green, P. M. Loewenstein, Cell 1988, 55, 1179. [57] H. Brooks, B. Lebleu, E. Vives, Adv. Drug Deliv. Rev. 2005, 57, 559. [58] V. P. Torchilin, R. Rammohan, V. Weissig, T. S. Levchenko, Proc. Natl. Acad. Sci. USA 2001, 98, 8786. [59] B. Gupta, T. S. Levchenko, V. P. Torchilin, Adv. Drug Deliv. Rev. 2005, 57, 637. [60] R. Kuai, W. Yuan, W. Li, Y. Qin, J. Tang, M. Yuan, L. Fu, R. Ran, Z. Zhang, Q. He, Mol. Pharm. 2011, 8, 2151. [61] R. Kuai, W. Yuan, Y. Qin, H. Chen, J. Tang, M. Yuan, Z. Zhang, Q. He, Mol. Pharm. 2010, 7, 1816. [62] M. J. Kang, S. Lee, B. K. Kim, J. Y. Eum, S. H. Park, M. H. Kang, C. H. Oh, J. Choo, Y. W. Choi, Chem. Pharm. Bull. (Tokyo) 2011, 59, 109. [63] K. Takara, H. Hatakeyama, N. Ohga, K. Hida, H. Harashima, Int. J. Pharm. 2010, 396, 143. [64] D. Missirlis, D. V. Krogstad, M. Tirrell, Mol. Pharm. 2010, 7, 2173. [65] T. Wang, G. G. D’Souza, D. Bedi, O. A. Fagbohun, L. P. Potturi, B. Papahadjopoulos-Sternberg, V. A. Petrenko, V. P. Torchilin, Nanomedicine 2010, 5, 563. [66] T. Wang, S. Yang, V. A. Petrenko, V. P. Torchilin, Mol. Pharm. 2010, 7, 1149. [67] P. K. Jayanna, D. Bedi, J. W. Gillespie, P. DeInnocentes, T. Wang, V. P. Torchilin, R. C. Bird, V. A. Petrenko, Nanomedicine 2010, 6, 538. [68] D. K. Chang, C. T. Lin, C. H. Wu, H. C. Wu, PLoS One 2009, 4, e4171.

3821

www.advmat.de

REVIEW

www.MaterialsViews.com

3822

[106] M. Goutayer, S. Dufort, V. Josserand, A. Royere, E. Heinrich, F. Vinet, J. Bibette, J. L. Coll, I. Texier, Eur. J. Pharm. Biopharm. 2010, 75, 137. [107] W. J. Mulder, K. Castermans, J. R. van Beijnum, M. G. Oude Egbrink, P. T. Chin, Z. A. Fayad, C. W. Lowik, E. L. Kaijzel, I. Que, G. Storm, G. J. Strijkers, A. W. Griffioen, K. Nicolay, Angiogenesis 2009, 12, 17. [108] W. J. Mulder, R. Koole, R. J. Brandwijk, G. Storm, P. T. Chin, G. J. Strijkers, C. de Mello Donega, K. Nicolay, A. W. Griffioen, Nano Lett. 2006, 6, 1. [109] S. V. Sharma, D. W. Bell, J. Settleman, D. A. Haber, Nat. Rev. Cancer 2007, 7, 169. [110] N. Turner, R. Grose, Nat. Rev. Cancer 2010, 10, 116. [111] X. Chen, X. Wang, Y. Wang, L. Yang, J. Hu, W. Xiao, A. Fu, L. Cai, X. Li, X. Ye, Y. Liu, W. Wu, X. Shao, Y. Mao, Y. Wei, L. Chen, J. Control. Release 2010, 145, 17. [112] T. Terada, M. Mizobata, S. Kawakami, Y. Yabe, F. Yamashita, M. Hashida, J. Drug Target. 2006, 14, 536. [113] C. J. Powers, S. W. McLeskey, A. Wellstein, Endocr. Relat. Cancer 2000, 7, 165. [114] X. Wang, L. Deng, X. Chen, H. Pei, L. Cai, X. Zhao, Y. Wei, L. Chen, J. Pharm. Sci. 2011, 100, 1196. [115] T. P. Herringson, J. G. Altin, J. Control. Release 2009, 139, 229. [116] T. P. Herringson, J. G. Altin, Int. J. Pharm. 2011, 411, 206. [117] T. P. Herringson, J. G. Altin, J. Drug Target. 2011, 19, 681. [118] H. Chen, M. E. Herndon, J. Lawler, Matrix Biol. 2000, 19, 597. [119] M. L. Iruela-Arispe, M. Lombardo, H. C. Krutzsch, J. Lawler, D. D. Roberts, Circulation 1999, 100, 1423. [120] N. H. Guo, H. C. Krutzsch, J. K. Inman, C. S. Shannon, D. D. Roberts, J. Pept. Res. 1997, 50, 210. [121] J. E. Murphy-Ullrich, M. Poczatek, Cytokine Growth Factor Rev. 2000, 11, 59. [122] M. P. Rivera-Fillat, F. Reig, E. M. Martinez, M. R. Grau-Oliete, J. Pept. Sci. 2010, 16, 315. [123] R. S. Herbst, Int. J. Radiat. Oncol. Biol. Phys. 2004, 59, 21. [124] H. Zhang, A. Berezov, Q. Wang, G. Zhang, J. Drebin, R. Murali, M. I. Greene, J. Clin. Invest. 2007, 117, 2051. [125] S. Song, D. Liu, J. Peng, H. Deng, Y. Guo, L. X. Xu, A. D. Miller, Y. Xu, FASEB J. 2009, 23, 1396. [126] J. Partanen, D. J. Dumont, Curr. Top Microbiol. Immunol. 1999, 237, 159. [127] R. Tournaire, M. P. Simon, F. le Noble, A. Eichmann, P. England, J. Pouyssegur, EMBO Rep. 2004, 5, 262. [128] J. Mai, S. Song, M. Rui, D. Liu, Q. Ding, J. Peng, Y. Xu, J. Control. Release 2009, 139, 174. [129] K. Alitalo, T. Tammela, T. V. Petrova, Nature 2005, 438, 946. [130] P. Laakkonen, M. E. Akerman, H. Biliran, M. Yang, F. Ferrer, T. Karpanen, R. M. Hoffman, E. Ruoslahti, Proc. Natl. Acad. Sci. USA 2004, 101, 9381.

wileyonlinelibrary.com

[131] Z. Yan, F. Wang, Z. Wen, C. Zhan, L. Feng, Y. Liu, X. Wei, C. Xie, W. Lu, J. Control. Release 2011, 157, 118. [132] Z. Yan, C. Zhan, Z. Wen, L. Feng, F. Wang, Y. Liu, X. Yang, Q. Dong, M. Liu, W. Lu, Nanotechnology 2011, 22, 415103. [133] J. H. Park, G. von Maltzahn, M. J. Xu, V. Fogal, V. R. Kotamraju, E. Ruoslahti, S. N. Bhatia, M. J. Sailor, Proc. Natl. Acad. Sci. USA 2010, 107, 981. [134] C. Grange, S. Geninatti-Crich, G. Esposito, D. Alberti, L. Tei, B. Bussolati, S. Aime, G. Camussi, Cancer Res. 2010, 70, 2180. [135] M. Jensen, F. Berthold, Cancer Lett. 2007, 258, 9. [136] L. C. Ronn, M. Olsen, S. Ostergaard, V. Kiselyov, V. Berezin, M. T. Mortensen, M. H. Lerche, P. H. Jensen, V. Soroka, J. L. Saffell, P. Doherty, F. M. Poulsen, E. Bock, A. Holm, Nat. Biotechnol. 1999, 17, 1000. [137] A. J. Almeida, E. Souto, Adv. Drug Deliv. Rev. 2007, 59, 478. [138] R. H. Muller, C. M. Keck, J. Biotechnol. 2004, 113, 151. [139] A. Schroeder, C. G. Levins, C. Cortez, R. Langer, D. G. Anderson, J. Intern. Med. 2010, 267, 9. [140] J. E. Schroeder, I. Shweky, H. Shmeeda, U. Banin, A. Gabizon, J. Control. Release 2007, 124, 28. [141] S. Sofou, G. Sgouros, Expert. Opin. Drug Deliv. 2008, 5, 189. [142] Y. H. Bae, K. Park, J. Control. Release 2011, 153, 198. [143] D. Papahadjopoulos, T. M. Allen, A. Gabizon, E. Mayhew, K. Matthay, S. K. Huang, K. D. Lee, M. C. Woodle, D. D. Lasic, C. Redemann, et al., Proc. Natl. Acad. Sci USA 1991, 88, 11460. [144] O. Lyass, B. Uziely, R. Ben-Yosef, D. Tzemach, N. I. Heshing, M. Lotem, G. Brufman, A. Gabizon, Cancer 2000, 89, 1037. [145] R. A. Petros, J. M. DeSimone, Nat. Rev. Drug Discov. 2010, 9, 615. [146] E. Ruoslahti, S. N. Bhatia, M. J. Sailor, J. Cell Bio. 2010, 188, 759. [147] S. M. Moghimi, A. C. Hunter, J. C. Murray, Pharmacol. Rev. 2001, 53, 283. [148] D. Peer, J. M. Karp, S. Hong, O. C. Farokhzad, R. Margalit, R. Langer, Nat. Nanotechnol. 2007, 2, 751. [149] W. Jiang, B. Y. Kim, J. T. Rutka, W. C. Chan, Nat. Nanotechnol. 2008, 3, 145. [150] B. Romberg, W. E. Hennink, G. Storm, Pharm. Res. 2008, 25, 55. [151] K. Oumzil, S. Khiati, M. W. Grinstaff, P. Barthelemy, J. Control. Release 2011, 151, 123. [152] H. Hatakeyama, H. Akita, K. Kogure, M. Oishi, Y. Nagasaki, Y. Kihira, M. Ueno, H. Kobayashi, H. Kikuchi, H. Harashima, Gene Ther. 2007, 14, 68. [153] C. P. Leamon, S. R. Cooper, G. E. Hardee, Bioconjug. Chem. 2003, 14, 738. [154] D. Iliopoulos, H. A. Hirsch, G. Wang, K. Struhl, Proc. Natl. Acad. Sci. USA 2011, 108, 1397. [155] E. Mamasheva, C. O’Donnell, A. Bandekar, S. Sofou, Mol. Pharm. 2011, 8, 2224. [156] H. Hillaireau, P. Couvreur, Cell. Mol. Life. Sci. 2009, 66, 2873.

© 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

Adv. Mater. 2012, 24, 3803–3822