Current Pharmaceutical Biotechnology, 2007, 8, 000-000 1

1389-2010/07 $50.00+.00 © 2007 Bentham Science Publishers Ltd.

Photochemical Internalization: A New Tool for Drug Delivery

Kristian Berga,*, Marco Folini

b, Lina Prasmickaite

a, Pål Kristian Selbo

a, Anette Bonsted

a, Birgit Ø. Engesæter

c,

Nadia Zaffaronib, Anette Weyergang

a, Andreas Dietze

a,d , Gunhild M. Mælandsmo

c, Ernst Wagner

e,

Ole-Jacob Noruma,f

, Anders Høgsetg

aDepartment of Radiation Biology, Institute for Cancer Research, The Norwegian Radium Hospital, Montebello, N-0310 Oslo, Nor-

way, bDepartment of Experimental Oncology, Istituto Nazionale Tumori, 20133 Milano, Italy,

cDepartment of Tumor Biology, Insti-

tute for Cancer Research, The Norwegian Radium Hospital, Montebello, N-0310 Oslo, Norway, dDepartment of Orthopedic Surgery,

Vestfold County Hospital, N-3103 Tønsberg, Norway, ePharmaceutical Biology-Biotechnology, Department of Pharmacy, Ludwig-

Maximilians-Universitaet, Butenandtstr. 5-13, D-81377 Munich, Germany, fDepartment of Surgical Oncology, The Norwegian Ra-

dium Hospital, Montebello, N-0310 Oslo, Norway and gPCI Biotech AS, Hoffsvn. 48, N-0377 Oslo, Norway

Abstract: The utilisation of macromolecules in the therapy of cancer and other diseases is becoming increasingly important. Recent ad-

vances in molecular biology and biotechnology have made it possible to improve targeting and design of cytotoxic agents, DNA com-plexes and other macromolecules for clinical applications. In many cases the targets of macromolecular therapeutics are intracellular.

However, degradation of macromolecules in endocytic vesicles after uptake by endocytosis is a major intracellular barrier for the thera-peutic application of macromolecules having intracellular targets of action.

Photochemical internalisation (PCI) is a novel technology for the release of endocytosed macromolecules into the cytosol. The technol-

ogy is based on the activation by light of photosensitizers located in endocytic vesicles to induce the release of macromolecules from the endocytic vesicles. Thereby, endocytosed molecules can be released to reach their target of action before being degraded in lysosomes.

PCI has been shown to stimulate intracellular delivery of a large variety of macromolecules and other molecules that do not readily pene-trate the plasma membrane, including type I ribosome-inactivating proteins (RIPs), DNA delivered as gene-encoding plasmids or by

means of adenovirus or adeno-associated virus, peptide nucleic acids (PNAs) and chemotherapeutic agents such as bleomycin and in some cases doxorubicin. PCI of PNA may be of particular importance due to the low therapeutic efficacy of PNA in the absence of an ef-

ficient delivery technology and the 10-100-fold increased efficacy in combination with PCI. The efficacy and specificity of PCI of mac-romolecular therapeutics has been improved by combining the macromolecules with targeting moieties, such as the epidermal growth

factor. In general, PCI can induce efficient light-directed delivery of macromolecules into the cytosol, indicating that it may have a vari-ety of useful applications for site-specific drug delivery as for example in gene therapy, vaccination and cancer treatment.

Key Words: Photodynamic, photochemical internalisation, photosensitizer, macromolecule, peptide nucleic acid, gene therapy, drug deliv-ery, protein toxin.

INTRODUCTION

A large number of macromolecules are today under develop-ment for use as medicines and a few have already been approved, e.g. a CD33-targeted immunoconjugate (Mylotarg), an HER2-tar- geted antibody (Herceptin), an EGFR-targeted antibody (Cetuximab) and recombinant IL-2 truncated diphtheria toxin fusion protein (Ontak), erythropoitin, interferons and colony-stimulating factors [1]. Macromolecules with therapeutic potential include proteins such as ribosome-inactivating protein toxins for treatment of cancer and other indications (se below), antibodies and growth factors for cell surface targeting, peptides and mRNA for vaccination, DNA for gene therapy, and antisense oligonucleotides, ribozymes, pep-tide nucleic acids (PNAs) and siRNA for gene silencing [2]. There are many extracellular and intracellular barriers for these molecules to overcome before they can arrive at the target cells, enter the cells and reach intracellular therapeutic targets. Degradation by serum enzymes and elimination by cells of the reticulo-endothelial system (RES), penetration into the target tissues through the endothelial lining as well as transport limitations within the tissue are important hurdles to obtain sufficient biological effect of these macromole-cules [3]. The delivery system should also overcome the intracellu-lar barriers, such as the plasma membrane, endocytic membranes and, in gene therapy, the nuclear membrane. Many of the delivery methods developed are based on improved endocytic uptake and release of the therapeutic molecule from the endocytic vesicles into the cytosol. However, a major limitation in the use of macro-

*Address correspondence to this author at the Department of Radiation

Biology, Institute for Cancer Research, The Norwegian Radium Hospital,

Montebello, N-0310 Oslo, Norway; E-mail: ________________________

molecular therapy is still the low rate of penetration through the membranes of endocytic vesicles and degradation of the macro-molecules by lysosomal enzymes [4]. The use of photochemical treatment to stimulate translocation of endocytosed macromolecules into the cytosol is a novel technology to improve therapeutic effi-cacy. The technology as described in this review is derived from photodynamic therapy (PDT) and is named photochemical inter-nalization (PCI). The basis for PDT and PCI is reviewed as well as the present status of PCI as a tool for delivery of drugs.

Photodynamic therapy

PDT is a treatment modality that takes advantage of the photo-toxic effects induced by a photosensitizer and light in the presence of oxygen [5]. PDT is approved for several cancer indications, as well as for age-related macular degeneration. Fluorescence diagno-sis (FD) based upon the fluorescing properties of a PDT-type pho-tosensitizer has been recently approved for detection of bladder dysplasia and cancer [6].

The Photosensitizer

A photosensitizer is defined as a chemical entity, which upon absorption of light induces a chemical or physical alteration of an-other chemical entity. Most photosensitizers and all clinically ap-proved photosensitizers (with the exception of methylene blue) used in photodynamic therapy (PDT) are based on or related to the tetrapyrrole macrocycle [7]. Porphyrins comprise of four pyrrole subunits linked together by four methane bridges as shown in Fig. (1). This tetrapyrrole ring structure is named porphine and deriva-tives of porphins are named porphyrins. Tetrapyrroles are naturally occurring pigments, which are used in many biological processes. These tetrapyrroles are not able to induce any photochemical or

2 Current Pharmaceutical Biotechnology, 2007, Vol. 8, No. 6 Berg et al.

400 500 600 700 800

Wavelength (nm)

Soret Q-bands

IVchlorins

bacteriochlorins

phthalocyanines

Naphtalocyanines

Ab

sorb

ance

photobiophysical reactions in other compounds or are rapidly quenched in their normal surroundings, as in chlorophyll. In all these tetrapyrroles a metal ion is coordinated in the middle of the compounds. The presence of a coordinated metal ion and its elec-tronic properties are of importance for the photocytotoxic potential of porphyrins as photosensitisers. By removal of the metal tetrapyr-roles become efficient photosensitizers as well as acquiring fluo-rescing properties, for example by removal of iron from heme. Most efficient porphyrin-based photosensitizers therefore lack co-ordinated metal ions. Coordinated metal ions increase the probabil-ity for non-radiative decay of the triplet state (see below), paramag-netic metals like Fe

3+ being much more efficient than diamagnetic

metals like Al3+

and Mg2+

. Several metallophotosensitizers have however been developed for clinical purposes. Although they in most cases have lower quantum yields for cell inactivation they have other properties like improved solubility and stability, which makes them interesting as therapeutic substances. The metals used include Zn, Pd, Sn, Ru, Pt, Lu, Gd and Al.

Porphyrins and porphyrin-related dyes used in PDT have sub-stituents in the peripheral positions of the pyrrole rings, on the four methine carbons (meso-positions) and/or coordinated metals. These derivatives are synthesized to influence on the water/lipid solubil-ity, amphiphilicity, pKa and stability of the compounds [8]. These parameters determine the biodistribution of the compounds, i.e. the intracellular localization, tissue distribution and pharmacokinetics.

Mechanisms of PDT

The photosensitizer is usually administered systemically by means of intravenous injection. There is substantial evidence for a preferential retention of the photosensitizers in neoplasic lesions, usually with at tumor:(normal surrounding tissue) ratio of 2-3:1 [9]. At some time interval after photosensitizer administration the target tissue is illuminated with light of an appropriate wavelength. The light source is often a laser, e.g. a diode laser, but non-collimated light sources may also be used. The photosensitizer must absorb the wavelengths emitted by the light source in order to induce a treat-ment effect (Fig. (2)). Clinically, 6-800 nm light is mainly used due to the optimal tissue penetration properties at these wavelengths,

while other wavelengths may be used on cells on culture. The pho-tocytotoxic effects, which are utilized in PDT, are initiated by the

Fig. (2). Absorption spectra of photosensitizers. A full spectrum is shown

for a porphyrin-type photosensitizer and additionally the location of the

main peaks for photosensitizers developed for PDT are shown on the figure

(not in scale). See also Fig. (1).

absorption of light by the photosensitizer. A molecule that has ab-sorbed a light quantum is excited to a short-lived excited state (

1P1,

Fig. (3)). The absorbed energy can be released as heat, emitted as fluorescence (utilized for cancer diagnosis) or undergo intersystem crossing (ISC) to a long-lived triplet state (

3P1). The extra energy in

the triplet state may as for the singlet state be released as heat, as light (phosphorescence) or be used in photochemical or photo-physical reactions to form reactive oxygen species (ROS). The triplet state may react in two ways, either by a type I mechanism involving electron or hydrogen atom transfer between the photosen-sitizer and the target molecule or by a type II mechanism involving energy transfer to molecular oxygen (Fig. (3)). In type I reactions free radicals are formed while in the type II reaction singlet oxygen (1O2) is formed. Singlet oxygen is an excited form of O2, which is

Fig. (1). Basic structure of some photosensitizers.

N N

N N

N N

N N

N N

N N

Porphin Chlorin Bacteriochlorin

Phthalocyanine

N

N

N

N

N

N

N

Naphthalocyanine

N

N

N

N

N

N

N

N

N

Photochemical Internalization Current Pharmaceutical Biotechnology, 2007, Vol. 8, No. 6 3

in the triplet form in its ground state, and is not a radical. Both mechanisms may occur during PDT and the relative importance depends on the surroundings and the nature of the substrate mole-cules. A significant amount of indirect evidence, i.e. by means of quenchers of

1O2 and degradation of photosensitizers, and some

direct evidence exists for the importance of 1O2 in the photoinduced

processes of PDT [10-13]. Singlet oxygen causes mainly membrane damage by oxidizing amino acids (tryptophan, cysteine, histidine, methionine and phenylalanine), unsaturated fatty acids and choles-terol. Additionally, guanine may also be oxidized by singlet oxy-gen. These reactive oxygen species are responsible for irreversible damage to the target tissue. The therapeutic effect of PDT may be due to direct cell killing or vascular damage, and there are also indications for the importance of immune responses [5,14]. The relative importance of these pathways to therapeutic effects de-pends on the photosensitizer and its formulation, the way of ad-ministration and the time between administration of the photosensi-tizer and exposure to light.

Photochemical Internalization (PCI)

Photochemical internalization (PCI) is a novel technology, which may improve the therapeutic effect of macromolecules and other molecules that accumulate in endocytic vesicles [15,16]. PCI utilizes photosensitizers with physico-chemical properties that lead to accumulation in endocytic vesicles [17]. Upon exposure to light the contents of vesicles containing such photosensitizers are re-leased into the cytosol in a functionally active form [18,19]. Exam-ples of photosensitizers and tentative localization in endocytic vesi-cles are shown in Fig. (4). The most efficient photosensitizers for use in PCI have an amphiphilic structure with a hydrophilic part inhibiting penetration through cellular membranes as illustrated in Figs. (3) and (4). Upon exposure to light reactive oxygen species of which singlet oxygen is assumed to be dominant, are formed. Sin-glet oxygen has a short range of action (<20 nm, [20,21]) and the chemical reactions induced by

1O2 leads to release of the contents

of the endocytic vesicles (Fig. (5)). Although the exact structure of the damaged vesicles has not been revealed results with dextran and

Fig. (3). Basic mechanisms in PDT. A simplified Jablonski diagram is shown (left side) where the vibrational levels are omitted. P – photosensitizer; ISC –

intersystem crossing.

Fig. (4). Proposed localization of sulfonated photosensitizers in endocytic vesicles. The figure shows a schematic drawing of an endocytic vesicle and

drawings of the main location of a tetrasulfonated (tetra(4-sulfonatophenyl)porphine,TPPS4) and an amphiphilic disulfonated photosensitizer (tetraphenylpor-phine disulfonate (TPPS2a) and aluminium phthalocyanine disulfonate (AlPcS2a)) (not in dimension).

4 Current Pharmaceutical Biotechnology, 2007, Vol. 8, No. 6 Berg et al.

viral particles indicate that relatively large particles can escape the vesicles after PCI. The various macromolecules that have been shown to increase their biological function by PCI are described in Table 1. PCI of therapeutic macromolecules has been shown to increase the necrotic depth of the treatment and to increase the frac-tion of complete remissions in animal tumor models [22,23].

It was initially thought that the photosensitizer and macromole-cule to be released from the endocytic compartments had to be lo-cated in the same compartments at the time of light exposure. How-ever, it has recently been found that the photochemical treatment can be performed up to 6-8 hours prior to the delivery of a plasmid without reducing the synergistic effect of the combined treatment [24]. The assumed mechanistic basis for such an observation is that newly formed vesicles are able to fuse with photochemically dam-aged vesicles and that the macromolecules are released into the cytosol after fusion of such vesicles. Further studies are however required to fully document this hypothesis.

Photochemical Delivery of Macromolecules

Macromolecules for therapeutic purposes may be divided into 3 main groups, i.e. proteins and peptides, nucleic acids and synthetic polymers for drug delivery and protection. For the PCI technology to induce increased translocation of therapeutic macromolecules to the cytosol they need to be located in endocytic vesicles at some stage in the process. The hydrolytic environment of late endosomes and lysosomes may therefore set limits to the therapeutic effect of

PCI. Furthermore, the reactive oxygen species (ROS) formed by the photochemical treatment may oxidize the macromolecular drugs by for example oxidizing the amino acids in proteins and peptides and guanine in the nucleotides. The quantum yield for inhibition of the therapeutic activity of the macromolecules will depend on the 3-D structure of the macromolecule, i.e. the access of ROS to amino acids or guanine of importance for biologic activity. In addition, the intravesicular localization of the photosensitizer, i.e. distance be-tween the photosensitizer and the macromolecule, will have impact on the rate of macromolecule inactivation. The success of PCI in inducing release of functionally active macromolecules may there-fore depend on the hydrolytic activity of the endocytic vesicles, the sensitivity of the macromolecule to the photochemical treatment as well as the endocytic activity of the target cells. Most cell types, except mature erythrocytes, exert endocytic activity, but the rate of endocytosis may vary. The present status on the use of PCI for acti-vation of various macromolecular structures is reviewed below.

PCI for Gene Delivery

Gene therapy is a novel therapeutic modality receiving great attention and being generally recognised as having the potential to constitute treatment for a wide range of diseases [1, 25]. However, although there are some encouraging clinical trials, gene therapy have hitherto largely given quite disappointing conclusions [26] An important reason for this is that methods for efficient and specific delivery of therapeutic genes in vivo are lacking. With most gene

Fig. (5). Schematic representation of photochemical internalisation (PCI). The figure illustrates an endocytic vesicle. In photochemical internalisation

(PCI), photosensitizer molecules and macromolecules accumulate in endocytic vesicles as indicated. The most efficient photosensitizers for PCI are mainly

located in the membranes of the vesicles. Upon illumination, reactive oxygen species, mainly singlet oxygen, are formed, constituents of the membranes are

destroyed and passage of macromolecules to the cytosol becomes possible.

Photochemical Internalization Current Pharmaceutical Biotechnology, 2007, Vol. 8, No. 6 5

delivery systems the therapeutic gene is taken into the cell by endo-cytosis, and for many of these systems, especially non-virus-based, the lack of efficient mechanisms for translocating the gene out of the endocytic vesicles constitutes a major hindrance for realization of the therapeutic potential of the gene.

Photochemical internalisation has been studied as a gene deliv-ery technology (reviewed in [27] both with several non-viral [28-30] and with adenoviral vectors [31] as well as adeno-associated virus [32], mainly by using reporter genes such as genes encoding EGFP (enhanced green fluorescent protein) or ß-galactosidase.

Table 1. In Vitro and In Vivo Studies Demonstrating PCI Induced Activation of the Macromolecules Biological Effects**

Macromolecules Photosensitizer References

IN VITRO STUDIES

Proteins

gelonin TPPS4, TPPS2a [15,73]

AlPCS2a [76]

Chlorine e6 Unpublished results

saporin TPPS2a [15,80]

agostin [15]

horseradish peroxidase TPPS2a [15]

Targeted Proteins

MOC-31-gelonin TPPS2a , AlPcS2a [76]

ALA-PpIX [82]

EGF-saporin TPPS2a [80]

Cetuximab-saporin TPPS2a Unpublished results

Genes (Reporters and Therapeutic-Relevant)

Virus Mediated:

AdCMV-EGFP TPPS2a [32,83]

AdCMV-lacZ AlPcS2a , TPPS2a [83-85]

AdRGD-EGFP TPPS2a [47]

AdSV40-EGFP TPPS2a [48]

AdvCMV-TRAIL TPPS2a [86]

AAV5RSV-EGFP TPPS2a [32]

Non-Virus Mediated:

Polylysine (with reporter gene EGFP) TPPS2a, TPPS4, AlPcS2a [28-30,85]

Polyethyleneimine (PEI) (with reporter genes eGFP, luciferase and therapeutic gene HSV-thymidine kinase) TPPS2a,AlPcS2a [29,33,79]

Glucosylated PEI with p53 gene TPPS2a [34,87]

Glucosylated PEI with PTEN gene TPPS2a [35]

NLS-peptide/plasmid complex complexed to a dendrimer-photosensitizer conjugate Dendrimer phthalocyanine [45]

Cationic lipids (with reporter eGFP)* AlPcS2a [88]

Targeted Gene Delivery:

EGF-PEI (with reporter genes eGFP and luciferase TPPS2a [77,79]

EGF-polymer complexed adenovirus [78]

Oligonucleotides

Peptide nucleic acids (PNA) AlPcS2a, TPPS2a [57]

Various cationic peptide-PNA conjugates AlPcS2a, TPPS2a [58,59]

Chemotherapeutics

Bleomycin TPPS2a, AlPcS2a [75]

doxorubicin TPPS2a [89]

IN VIVO EXPERIMENTS

Gelonin AlPcS2a [22]

Glucosylated PEI with p53 gene AlPcS2a [44]

NLS-peptide/plasmid complex complexed to a dendrimer-photosensitizer conjugate Dendrimer phthalocyanine [45]

Bleomycin AlPcS2a [75,75]

* Not all cationic lipids are suited for combination with PCI and PCI of polyplex/DNA complexes is generally a more efficient combination.

** Abbreviations: EGFP – enhanced green fluorescence protein, EGF – epidermal growth factor: TRAIL – tumor necrosis factor-related apoptosis-inducing ligand; PTEN - phospha-

tase and tensin homolog

6 Current Pharmaceutical Biotechnology, 2007, Vol. 8, No. 6 Berg et al.

However, PCI-mediated gene delivery has also been shown to in-duce the delivery of therapeutic genes, such as the genes encoding HSV-tk (Herpes Simplex Virus thymidine kinase) [33], p53 [34], PTEN [35], TRAIL (unpublished results) and IL-12 (interleukin-12) (unpublished results).

Delivery of genes to cells for therapeutic purposes requires in most cases that the DNA is protected from degradation and that the cellular uptake is facilitated (Fig. (6)). The vehicles used for such purposes, named vectors, may be divided into chemical or non-viral and viral vectors. In addition, physical methods such as electropora-tion may be utilized. PCI may be viewed as a physico-chemical vector.

1. Non-viral vectors. The non-viral vectors may be divided into polymers, polyplexes, and lipids, lipoplexes. Both polyplexes and lipoplexes are generally cationic in order to easily form complexes with DNA and are added in sufficient amount to generate positive charge ratios in order to stimulate association with the cell surface.

PCI has been found to enhance the transgene expression of genes incorporated into plasmids and complexed with many non-viral vectors (Fig. (7)). Whereas transfection with polyplexes in general seems to be enhanced by PCI, not all lipoplexes benefit from a combination with PCI [36]. The observed differences in the PCI response between the different transfection agents raise several interesting questions. It seems logical to think that for transfection agents where transfection is not stimulated by PCI, endosomal re-lease is not a limiting factor for “normal” transfection. Polylysine is a cationic polymer, which forms complexes with DNA, enhances cellular uptake of DNA, but is less efficient in inducing release of DNA into the cytosol [37]. This is probably due to the high pK of the side chain amino-group in lysine (pK = 10.0-10.2, [38]) where the overall charge will not change upon entering the acidic endo-cytic vesicles. Endosomal release is therefore expected to be a lim-iting factor in transfection efficacy and PCI is indeed shown to greatly improve transfection efficacy. In contrast, polyethyle-neimine (PEI) based vectors are supposed to enhance endosomal release by acting as a “proton sponge” due to the pK values of the amino groups being close to physiological pH. Thus, for PEI-based vectors PCI-induced increase in transfection efficacy should not be expected as high as that obtained with polylysine complexes [39]. This is also in accordance with experimental observations [40]. However, the observation that also transfection by PEI complexes can be positively affected by PCI [29,41] gives an indication that

even if endosomal escape is induced by PEI, this process probably is not always very effective. This may be e.g. related to the size of the PEI/DNA complexes employed, as pointed out by Ogris et al. [42]. With the proposed mechanism of PEI as a “proton sponge” [43] it would also be necessary with a certain minimum amount of PEI inside each endocytic vesicle in order to induce endosomal swelling and lysis. The observation that PCI shows the greatest enhancement of PEI-mediated transfection at low doses of DNA/PEI complex or at low PEI/DNA ratios [33] is in accordance with this proposed mechanism, since the positive effects of PCI should be expected to increase with the decrease in the ability of the PEI to effect endosomal release. This observation could also have important implications for the use of PEI and similar agents for in vivo gene therapy. With in vivo delivery of DNA/PEI complexes the amount of complex reaching the target cells will often be very lim-ited, and probably often well below the threshold where PEI is able to induce efficient endosomal escape. In these cases PCI could be a very valuable tool for increasing gene delivery specifically within the target area for the therapy. A recent in vivo study shows that treatment of squamous cell carcinomas deficient in active p53 with intratumoral injection of the photosensitizer AlPcS2a and a plasmid expressing p53 complexed to a glycosylated PEI caused dramatic tumor regression in all the transfected animals [44]. In contrast, similar PDT or PEI-p53 treatments did not delay tumor growth. In this study the tumors were relatively large at the time of the first treatment (500 mm

3) and the tumors were treated once a week for

up to 7 weeks. These results thus indicate that repeated local treat-ments might be feasible to control bulky tumors.

Recently, a highly sophisticated vector design was described by Kataoka and colleagues [45]. Their design was based on a disulfide-extended cationic peptide complexed with the plasmid. This cati-onic-peptide complex was mixed with a negatively charged den-drimer containing 32 carboxyl groups at the surface and a phthalo-cyanine photosensitzer in the core. It was shown that this complex caused efficient transfection at doses of light causing hardly any cytotoxicity. When this vector was evaluated in the conjunctiva of rats, transgene expression was shown only in the irradiated area without any observable damage to the treated tissue. This opens up therapeutic possibilities for indications such as in monogenic dis-eases, where it is important to avoid tissue damage.

2. Adenoviral vectors. Adenovirus vectors are known to be taken into cells by endocytosis and to be released from endosomes by a regulated process. This endosomal release is usually regarded

Fig. (6). The role of PCI in gene therapy. The transgene is delivered as a plasmid and complexed to non-viral vector or delivered as part of the genome of a

viral vector. The vector facilitates cellular uptake of the vector/DNA complexes (e.g. interaction of the positive overall charge of the non-viral vector/DNA

complexes or Coxackie and adenoviral receptor binding for the adenovirus). The vector/DNA complexes enter the endocytic vesicles and are either degraded

by lysosomal enzymes or are released into the cytosol. Alternatively, PCI stimulates the release from the endocytic vesicles before degradation. The transgene

will the reach the nucleus and the transgene expressed.

Photochemical Internalization Current Pharmaceutical Biotechnology, 2007, Vol. 8, No. 6 7

as a very efficient process [46]. It is therefore somewhat surprising that PCI is able to increase the number of adenoviral transduced cells by up to 30-fold. Nevertheless, PCI with adenoviral vectors has been tested in 12 different cell lines, and in all cases a positive effect on transduction has been observed [47]. The adenovirus acti-vated by means of PCI seems to follow the same cellular pathways as for conventional adenovirus infection, i.e. the fraction of trans-duced cells followed a linear relationship with the Coxsackie and Adenoviral Receptor (CAR) level of the cells and is integrin de-pendent. Furthermore, PCI increase the number of nuclearly located viral DNA molecules as measured by real-time PCR and fluores-cence in situ hybridization (FISH) [48]. The results so far indicate that the main cause of the photochemical activation of adenovirus transgene expression is related to enhanced release of the viral par-ticles from the endocytic vesicles into the cytosol. In accordance with what has been found for PCI of plasmids the adenovirus may be delivered after the photochemical treatment (unpublished re-sults). However, adenovirus may be delivered up to 12 hrs after the photochemical treatment, which is longer than what is effective for PCI of plasmids.

Fig. (7). Enhancement of transgene expression by PCI. The effect of PCI

on expression of the reporter enhanced green fluorescence protein (EGFP)

gene delivered by PEI into HCT-116 cells. Fluorescence microscopy (A)

and flow cytometry (B) analysis of EGFP-expressing cells performed 2 days

after treatment. The cells were transfected with pEGFP-N1 (5 μg/ml) and

either kept in the dark (No PCI) or exposed to light (PCI).

3. Adeno-associated viral vector. Adeno-associated viruses (AAV) derived from endemic non-pathogenic parvoviruses have several properties that make them attractive as vectors in gene ther-apy [49]. These properties include the absence of host inflammatory response, low cytotoxicity, a low cell-mediated immune response, the ability to transduce non-dividing cells, the induction of long-term gene expression and a broad tropism. A major limitation is the packing constraint of about 5 kbs. PCI has recently been found to increase the transduction efficacy of a serotype 5 AAV in one of two glioblastoma cell lines tested. The non-responding cell line required a much lower number of viral particles to transduce the same number of cells and it was speculated that PCI might improve the transduction efficacy only in cells difficult to transduce by AAV’s [32].

Overall, PCI has the potential of being a very useful technology for in vitro and in vivo gene delivery, both because it can improve the delivery of genes in general, and because it is a light dependent

process, rendering the therapeutic gene active only in illuminated sites of the body. The possibility to use photochemically activated promoters may further add to the specificity of such treatment [50,51]. Thus, by the employment of PCI adverse effects due to gene expression in non-target areas of for example a cytotoxic gene could largely be avoided, making PCI especially attractive for gene therapy of cancer and localized diseases.

PCI of Peptide Nucleic Acids

Several oligonucleotides are currently under investigation for regulation of protein expression patterns in neoplastic as well as other cell types [52]. Oligonuleotides have been designed in several ways in order to obtain stability towards degradation in biological environments and penetration through membrane barriers without losing biological activity. Peptide nucleic acids (PNAs) belong to the third generation of antisense oligonucleotides in which the phosphate backbone is replaced by a pseudopeptide backbone com-posed of N-(2-aminoethyl) glycine units [53]. Due to the modified internucleotide linkage, PNAs are resistant to nucleases and pepti-dases and therefore are extremely stable in biological systems [54]. Despite the radical difference in the chemical composition of the backbone, PNA not only retains, but also improves, the hybridisa-tion characteristics to DNA and RNA. However, PNA does not elicit the target RNA cleavage by RNase H as seen with unmodified oligonucleotides. PNAs are relatively easy to synthesise and also quite easy to modify (making for instance PNA-peptide conjugates) without impairing their binding characteristics to nucleic acids [54]. PNAs seem to be non-toxic, as they are uncharged molecules with low affinity for proteins that normally bind nucleic acids [54]. Their favourable properties have led to the use of PNAs for different ap-plications in the field of antisense/antigene technology. Specifi-cally, a number of experiments with permeabilised cells, isolated nuclei and also intact cells has demonstrated the potential of PNAs in antigene or antisense applications to down-regulate the transcrip-tion or translation of cancer-related genes. As PNAs are neutral molecules, solubility and cellular uptake are serious problems that have to be overcome for PNAs to become useful tools in clinical settings [1]. Several approaches have therefore been developed for the delivery of such molecules to intact cells based on electropora-tion, microinjection or on the generation of PNA/DNA complexes [54]. Improved intracellular delivery can be also obtained by cou-pling PNAs to lipids or peptides that are efficiently internalized by cells. Several efforts have been made to exploit the potential of “natural-occurring” peptides such as amphiphilic cationic/hydro- phobic peptides, that are reported to have the capacity for trans- porting molecules such as oligonucleotides across biological mem-branes in a receptor independent way [54]. In this context, the most extensively studied peptide is penetratin, a 16 residue peptide de-rived from Drosophila homeodomain transcription factor, Anten-napedia [54]. Other peptides have also been investigated for their ability to promote intracellular delivery of PNA, for instance the nuclear localisation signal peptide (NLS) [54] and the peptide de-rived from Tat protein of HIV-1 virus [55].

Despite the low uptake of PNAs in living cells, it has been re-cently demonstrated that naked or chimeric PNAs enter some cells through a liquid-phase endocytosis and tend to accumulate in the endocytic compartment [56]. The development of endosome-disruptive strategies is thus of great importance to allow PNAs to co-localize with their specific targets. For this purpose, PCI repre-sents a new approach to achieve the photochemically inducible release in the cytoplasm of endocytosed PNA. Recently a 15-mer naked PNA (3’-TCCCGACCGCCGACC-5’) directed against bases 157-171 in the hTERT mRNA (hTERT-PNA) was synthesized [57]. The possibility to efficiently deliver the hTERT-PNA into DU145 human prostate carcinoma cells by means of the PCI ap-proach was evaluated. Rhodamine B-labelling of the PNA was utilized to indicate endocytic uptake and PCI-induced release of hTERT-PNA. Furthermore, PCI of hTERT-PNA was found to sig-

8 Current Pharmaceutical Biotechnology, 2007, Vol. 8, No. 6 Berg et al.

nificantly reduce the catalytic activity of telomerase as well as the cell growth (Table 2), which reflected the higher bioavailability of PNA at the level of the target. In an additional step of the study, the antitelomerase activity of a chimeric PNA molecule, made by con-jugating the hTERT-PNA to the HIV-Tat internalizing peptide (Tat-hTERT-PNA) was evaluated [57]. The results of the TRAP (te-lomeric repeat amplification protocol)-assay carried out at different intervals after an overnight exposure of cells to 2 μM of Tat-hTERT-PNA indicated a very modest effect on telomerase activity and significantly lower than that obtained with the hTERT-PNA and light exposure at the same time point (Table 2) [57]. Such re-sults demonstrate that the PCI approach represents a more efficient system for the delivery of PNAs than simple conjugation with in-ternalization peptides.

It has been recently found that cellular uptake of several cell penetrating peptide (CPP)-conjugates occurs predominantly via an endocytotic pathway and that endosomal escape is therefore one of the major rate-limiting steps of CPP-mediated cellular delivery due to trapping of the CPP conjugates in the endosomal/lysosomal compartment [54,56]. In this context, the antisense activity of the Tat-hTERT-PNA was assessed in DU145 prostate cancer cells in combination to PCI treatment. Interestingly, preliminary results showed that an overnight exposure of DU145 cells to 2 μM of pho-tochemically-internalised Tat-hTERT-PNA resulted in an almost complete abrogation of telomerase activity and a marked impair-ment of cell proliferation (Table 2), as compared to cells exposed to the same concentration of Tat-hTERT-PNA alone or naked hTERT-PNA in combination with photochemical internalization (Table 2). These preliminary data are in accordance with recent findings dem-onstrating that the antisense effects of different PNA-conjugates (Tat, Arg7, KLA) were significantly enhanced, both at cytosolic and nuclear level, in HeLa pLuc 705 and p53R cell systems as a conse-quence of the PCI approach [58]. Shiraishi and Nielsen [58] found PCI to enhance the antisense activity of all the cationic conjugates more than 10-fold while PCI of naked PNA was improved only 2.5-fold. The most efficient conjugate was with Tat where PCI en-hanced the antisense effect by two orders of magnitude. The PCI-induced antisense effect was comparable to that of the lysosomo-tropic agent chloroquine, which is known to stimulate release from endocytic vesicles. Similar results were reported by Boe and Hovig, who found PCI of two PNAs towards S100A4 linked to cationic peptides (5+) to inhibit S100A4 expression in a dose- and time-dependent manner [59]. Based on fluorescence microscopy of fluo-rescence labeled PNAs charged as well as neutral PNAs were relo-cated to the nucleus after PCI although a higher extracellular con-centration was needed for the neutral PNAs. PCI of PNA is clearly improved with respect to PNA targeting effects by conjugation to a cationic peptide, but the optimal peptide sequence cannot be de-duced from these reports. The cellular uptake seems to be increased by increasing the peptide charge up to 5+, but the antisense effect

of the PNA-peptide alone and in combination with PCI does not make a simple correlation with the charge of the peptide, e.g. PCI of PNA linked to the Tat-peptide (8+) and a KLA peptide (5+) in-duced similar antisense effects while various arg/lys sequences with charge in the same range were less efficient [58].

Overall, this evidence demonstrated that that the PCI approach represents a new and more efficient system for the internalization of PNAs and emphasise the importance of the endosomal release to enhance the bioavailability, and thus the antisense activity, of naked or chimeric PNAs. As a consequence, PCI makes it possible to reduce the amount of PNA required to induce the specific biologi-cal effect compared to other delivery systems. Such a new delivery approach could be exploited for in vivo applications of PNAs. In fact, the actual efficiency of PNAs as antisense molecules has been mainly determined in in vitro cell systems and needs to be validated also in animal experimental models before these molecules can be added to the clinical armamentarium.

Photochemical Delivery of Other Oligonucleotides

As pointed out above some proteins, so-called CPPs such as HIV-1 Tat, Drosophila Antennapedia homeoprotein and HSV-1 VP22, have been suggested to traverse the cell membrane and have attracted much attention for their ability to deliver other bio-molecules into the cytosol [60]. VP22 has been non-covalently linked to fluorescein or Bodipy 630/650 oligonucleotides (ODNs) in a ratio forming neutral complexes that might form larger aggre-gates [61-63]. These complexes, named Vectosomes, are taken up in a temperature dependent manner (no uptake at 4

0C) and locate

intracellularly in distinct spots. Upon exposure to light these com-plexes are relocated in the cells and the ODNs can exert their bio-logical functions, not being observed in the absence of light. This has been documented with anti-c-raf1 vectosomes both in vitro and in vivo [61]. Alternatively, VP22 may be fused with a therapeuti-cally active peptide, such as the proapoptotic BH3 domain of the Bcl-2 family of proteins, mixed with a photoactively labelled ODN and treated with light after it has entered the cells [62]. The exact mechanisms involved have not been revealed, but it is suggested that ROS (e.g. singlet oxygen) are formed by activation of the fluorophore and induce disaggregation of the complexes. Alumi-num phthalocyanines may also be used for disaggregation of the complexes [64]. It is claimed that the particles do not co-localize with markers of endocytic vesicles [63] and the ability of VP22 to traverse the cell membrane could suggest the complexes to be lo-cated in the cytosol already before exposure to light. However, recent data on the cellular uptake of the Tat peptide indicate that the ability of the CPPs to actually penetrate the cell membranes is over-estimated and the documentation at least partly due to artifacts in the fixation procedures before microscopy [56,65,66]. Independent of the mechanisms these results show in vivo documentation of photochemical activation of therapeutically active ODNs [67].

Table 2. Effects of hTERT-PNA and Tat-hTERT-PNA on Telomerase Activity and Cancer Cell Growth in the Absence (-) or Presence (+) of Pho-

tochemical Treatment (PCI)

PCIa % Telomerase Activity Inhibition

b % of Cell Growth Inhibition

c

“naked” hTERT-PNA _ no inhibition no inhibition

+ 35.0 ± 5.0 %* 68.0 ± 12.9 %*

Tat-hTERT-PNA _ 26.7 ± 3.0 % no inhibition

+ 95.0± 3.4 %** 85.0 ± 6.3 %**

a0.7 μg/ml TPPS2a + light exposure. bResults of TRAP assays carried out 48 h after an overnight exposure to 2 μM hTERT-PNA or Tat-hTERT-PNA, ± PCI and expressed as the percentage of enzyme activity inhibition

in PNA-treated cells compared with controls. Data represent mean values ± s.d. of at least three independent experiments. (*) P<0.01; (**) P<0.001, Student’s t test. cResults of cell growth assay carried out 48 h after an overnight exposure to 2 μM hTERT-PNA or Tat-hTERT-PNA, ± PCI and expressed as the percentage of cell growth inhibition

in PNA-treated cells compared with controls. Data represent mean value ± s.d. of at least three independent experiments. (*) P<0.01; (**) P<0.001, Student’s t test.

Photochemical Internalization Current Pharmaceutical Biotechnology, 2007, Vol. 8, No. 6 9

Photochemical Delivery of Proteins

Protein toxins have been extensively evaluated, especially as part of immunotoxins, for use in cancer therapy as well as in graft versus host disease, bone marrow purging, autoimmune diseases and HIV infection [68]. These protein toxins usually kill cells by inhibiting protein synthesis, either by inactivating the elongation factor 2 (EF-2) or the EF-2 binding site on the 28 S ribosome subunit. The 28S ribosome inactivating proteins (RIPs) are plant toxins that constitute either only one chain (A-chain, Type I toxins) with enzymatic activity or in addition another targeting chain (B-chain, Type II toxins, e.g. ricin) linked to the A-chain by disulfide bonds [69]. Both types of toxin are endocytosed, but while type I RIPs enters the cells mainly through pinocytosis, type II toxins enter the cells by the more efficient receptor-mediated endocytosis due to the B-chain binding to the cell surface. Type I RIPs enter mainly lysosomes where they are degraded while type II RIPs are partly transported retrogradely from endosomes to the trans-Golgi network, endoplasmic reticulum and by a not fully understood mechanism enter the cytosol where they can exert their enzymatic activity [70]. Type I RIPs therefore exert a relatively low cytotoxic effect on intact cells, while type II RIPs are highly cytotoxic com-pounds despite the fact that both types of RIPs have the same abil-ity to inactivate ribosomes. The literature indicates that only 1-10 RIP molecules located in the cytosol are sufficient to kill the cell [71]. Type I RIPs should therefore be good candidates for a highly specific PCI-based activation of cytotoxic proteins.

PCI has been shown to release the type I RIP gelonin from en-docytic vesicles in vitro and to activate its cytotoxic effect both in vitro and in vivo [22,72,73]. Type I RIPs, such as gelonin and saporin, which have been frequently used in PCI are relatively small proteins with a molecular mass of about 30 kDa, and cannot easily be used unmodified for systemic administration due to leak-age through the kidneys and lack of tissue specificity [69]. PCI of gelonin has therefore been evaluated by injecting gelonin intratu-morally into 6-8 mm tumors. With this treatment 60-80% complete response was obtained, while similar photochemical treatment with intratumoral injection of saline induced only 0-10% complete re-sponse [22,23]. A preferential retention of relevant photosensitizers in neoplastic tissues (tumor-to-normal surrounding tissue ratio of about 2-3:1) and the site-directed delivery of the light provide specificity to such a treatment scenario [9].

PCI of Low Molecular Weight Substances

A few chemotherapeutic agents accumulate in endocytic vesi-cles probably due to their size and/or charge. Bleomycin (MW ~ 1400) is widely used chemotherapeutic agent for several types of cancer, but is also known to accumulate in endocytic vesicles [74]. PCI has been shown to improve its cytotoxic effect both in vitro and in vivo [75], indicating that not only large macromolecules like proteins and DNA may profit from a PCI-based treatment, but also smaller molecules unable to penetrate the biological membranes efficiently. This may make it possible to reduce the administered dose of bleomycin, so reducing the side effects without losing clini-cal efficacy. In earlier in vivo studies subcutaneously growing xeno- grafts have been utilized to document the PCI principle. More re-cently, PCI of bleomycin has been found to significantly delay growth of human fibrosarcoma cells injected intramuscularly in athymic mice (Norum et al., unpublished results). PCI may also reverse resistance to weak base chemotherapeutics such as doxoru-bicin when resistance is caused by lowered pH in endocytic vesicles [88].

Targeting and Specificity

In cancer therapy as well in the treatment of many other dis-eases there is a need for improved specificity of the treatment. PCI relies on the use of photosensitizers, which accumulate preferen-tially in neoplastic lesions and induce a cytotoxic reaction only in

the light-exposed areas. A further improved treatment specificity may however be established by cell specific targeting of the photo-sensitizer or the therapeutic macromolecule. Several targeting methods are under development, such as the use of antibodies and derivatives thereof, receptor ligands, peptides, tissue-specific pro-moters and replication. Some targeting principles have been evalu-ated in combinations with PCI, such as immunotoxins [76], trans-ferrin- and epidermal growth factor-conjugated vectors [29,77-79] and protein affinity-toxins [80], transferrin-polylysin vector [29] and RGD sequences for integrin-targeting [47]. These results indi-cate that PCI can be used in combination with other targeting prin-ciples to further improve therapeutic specificity.

A major limitation for adenoviral gene therapy is the ubiquitous expression of the adenovirus primary receptor CAR on most epithe-lial cells, causing a lack of specificity of gene transfer. Moreover, the expression of CAR in cancer cells is in many cases lower than in its normal counterparts, resulting in resistance to Ad5 infection of cancer cells. We have shown that PCI is efficient for enhancing gene transfer from adenoviral vectors, which infect cells through alternate receptors than CAR [78]. Such retargeting of adenovirus to cell-specific receptors combined with a light-specific treatment to enhance gene transfer and expression presents an opportunity to obtain high and site-specific cell-infectivity levels with reduced viral doses.

A potential limitation in the utilization of light for therapeutic purposes is the penetration of light through tissues. The penetration of light through tissues is limited mainly by heme, especially in haemoglobin, and melanin. Efficient penetration of light therefore requires wavelengths for photoactivation above about 600 nm where absorption of light by these chromophores is low [81]. An upper maximum wavelength for photoactivation is usually set by the energy required for excitation of O2 to form singlet oxygen, i.e. 23 kcal/mol which is equivalent to the excess energy of the long lived form of singlet oxygen. This requires that the photosenstizer must be excited at wavelengths below about 850 nm. The therapeu-tic wavelength window is therefore usually defined as 6-800 nm, but shorter wavelengths may be used when thin lesions are to be treated. The treatment of internal organs is generally of no limita-tions any more due to the development of light sources and light applicators in PDT, which are used to treat hollow organs, other internal organs by interstitial fiber optics and surface exposure after surgical procedures.

CONCLUSIONS

Photochemical internalization of macromolecules is an attrac-tive alternative or addition to other delivery methods. PCI has been shown to increase the treatment effects of targeted macromolecules as well as macromolecules with little or no therapeutic effects due to localization and degradation in endocytic vesicles. This treatment modality has been documented to enhance the treatment effects of structurally highly different macromolecules such as proteins, nu-cleic acids delivered as by non-viral and viral (adenoviral and adeno-associated viral) vectors and PNA, as well as some che-motherapeutic agents. Substantial site-direction of the treatment can be obtained by the need for light activation, and can be further en-hanced by combination with targeting delivery technologies for the therapeutic macromolecule. Photochemical delivery of therapeutic compounds may therefore increase the therapeutic window and thereby reduce the adverse reactions of the compounds. Photo-chemical drug and gene activation and delivery could be considered for a large variety of clinical indications, including cancer, cardio-vascular diseases, autoimmune diseases, rheumatoid arthritis, and as a delivery system for DNA vaccines.

ACKNOWLEDGEMENTS

The authors have been financially supported by the Norwegian Research Council, The Norwegian Cancer Society, the Gene Ther-

10 Current Pharmaceutical Biotechnology, 2007, Vol. 8, No. 6 Berg et al.

apy Funding from the Norwegian Ministry of Health and the Nor-wegian Radium Hospital Research Foundation.

REFERENCES

[1] Kreitman, R.J. (1999) Curr. Opin. Immunol., 11(5), 570-578.

[2] Lu, Y. and Low, P.S. (2002) Adv. Drug Deliv. Rev., 54(5), 675-

693.

[3] Jain, R.K. (2001) Adv. Drug Deliv. Rev., 46(1-3), 149-168.

[4] Cho, Y.W.; Kim, J.D. and Park, K. (2003) J. Pharm. Pharmacol.,

55(6), 721-734.

[5] Dolmans, D.E.; Fukumura, D. and Jain, R.K. (2003) Nat. Rev.

Cancer, 3(5), 380-387.

[6] Schmidbauer, J.; Witjes, F.; Schmeller, N.; Donat, R.; Susani, M.

and Marberger, M. (2004) J. Urol., 171(1), 135-138.

[7] Brown, S.B.; Brown, E.A. and Walker, I. (2004) Lancet Oncol.,

5(8), 497-508.

[8] MacDonald; I.J. and Dougherty, T.J. (2001) J. Porphyrins Phthalo-

cyanines, 5105-129.

[9] Peng, Q. and Moan, J. (1995) Br. J. Cancer, 72(3), 565-574.

[10] Moan, J.; Pettersen, E.O. and Christensen, T. (1979) Br. J. Cancer,

39(4), 398-407.

[11] Sonoda, M.; Krishna, C.M. and Riesz, P. (1987) Photochem. Pho-

tobiol., 46(5), 625-631.

[12] Niedre, M.J.; Secord, A.J.; Patterson, M.S. and Wilson, B.C.

(2003) Cancer Res., 63(22), 7986-7994.

[13] Niedre, M.; Patterson, M.S. and Wilson, B.C. (2002) Photochem.

Photobiol., 75(4), 382-391.

[14] Castano, A.P.; Mroz, P. and Hamblin, M.R. (2006) Nat. Rev. Can-

cer, 6(7), 535-545.

[15] Berg, K.; Selbo, P.K.; Prasmickaite, L.; Tjelle, T.E.; Sandvig, K.;

Moan, J.; Gaudernack, G.; Fodstad, O.; Kjølsrud, S.; Anholt, H.;

Rodal, G.H.; Rodal, S.K. and Høgset, A. (1999) Cancer Res.,

59(6), 1180-1183.

[16] Høgset, A.; Prasmickaite, L.; Selbo, P.K.; Hellum, M.; Engesæter,

B.O.; Bonsted, A. and Berg, K. (2004) Adv. Drug Deliv. Rev.,

56(1), 95-115.

[17] Berg, K.; Western, A.; Bommer, J.C. and Moan, J. (1990) Photo-

chem. Photobiol., 52(3), 481-487.

[18] Berg, K.; Madslien, K.; Bommer, J.C.; Oftebro, R.; Winkelman,

J.W. and Moan, J. (1991) Photochem. Photobiol., 53(2), 203-210.

[19] Berg, K. and Moan, J. (1994) Int. J. Cancer, 59(6), 814-822.

[20] Moan, J. and Berg, K. (1991) Photochem. Photobiol., 53(4), 549-

553.

[21] Niedre, M.J.; Patterson, M.S.; Giles, A. and Wilson, B.C. (2005)

Photochem. Photobiol., 81(4), 941-943.

[22] Selbo, P.K.; Sivam, G.; Fodstad, O.; Sandvig, K. and Berg, K.

(2001) Int. J. Cancer, 92(5), 761-766.

[23] Dietze, A.; Peng, Q.; Selbo, P.K.; Kaalhus, O.; Muller, C.; Bown,

S. and Berg, K. (2005) Br. J. Cancer, 92(11), 2004-2009.

[24] Prasmickaite, L.; Høgset, A.; Selbo, P.K.; Engesæter, B.O.; Hel-

lum, M. and Berg, K. (2002) Br. J. Cancer, 86(4), 652-657.

[25] Somia, N. and Verma, I.M. (2000) Nat. Rev. Genet., 1(2), 91-99.

[26] Hacein-Bey-Abina, S.; von Kalle, C.; Schmidt, M.; Le Deist, F.;

Wulffraat, N.; McIntyre, E.; Radford, I.; Villeval, J.L.; Fraser,

C.C.; Cavazzana-Calvo, M. and Fischer, A. (2003) N. Engl. J.

Med., 348(3), 255-256.

[27] Høgset, A.; Prasmickaite, L.; Engesæter, B.O.; Hellum, M.; Selbo,

P.K.; Olsen, V.M.; Mælandsmo, G.M. and Berg, K. (2003) Curr.

Gene Ther., 3(2), 89-112.

[28] Prasmickaite, L.; Høgset, A. and Berg, K. (2001) Photochem. Pho-

tobiol., 73(4), 388-395.

[29] Prasmickaite, L.; Høgset, A.; Tjelle, T.E.; Olsen, V.M. and Berg,

K. (2000) J. Gene Med., 2(6), 477-488.

[30] Høgset, A.; Prasmickaite, L.; Tjelle, T.E. and Berg, K. (2000) Hum.

Gene Ther., 11(6), 869-880.

[31] Høgset, A.; Engesæter, B.Ø.; Prasmickaite, L.; Berg, K.; Fodstad,

O. and Mælandsmo, G.M. (2002) Cancer Gene Ther., 9(4), 365-

371.

[32] Bonsted, A.; Hogset, A.; Hoover, F. and Berg, K. (2005) Antican-

cer Res., 25(1A), 291-297.

[33] Prasmickaite, L.; Høgset, A.; Olsen, V.M.; Kaalhus, O.; Mikalsen,

S.O. and Berg, K. (2004) Cancer Gene Ther., 11(7), 514-523.

[34] Ndoye, A.; Merlin, J.L.; Leroux, A.; Dolivet, G.; Erbacher, P.;

Behr, J.P.; Berg, K. and Guillemin, F. (2004) J. Gene Med., 6(8),

884-894.

[35] Maurice-Duelli, A.; Ndoye, A.; Bouali, S.; Leroux, A. and Merlin,

J.L. (2004) Technol. Cancer Res. Treat., 3(5), 459-465.

[36] Selbo, P.K.; Høgset, A.; Prasmickaite, L. and Berg, K. (2002)

Tumour. Biol., 23(2), 103-112.

[37] Mannisto, M.; Vanderkerken, S.; Toncheva, V.; Elomaa, M.;

Ruponen, M.; Schacht, E. and Urtti, A. (2002) J. Control Release,

83(1), 169-182.

[38] General, S. and Thunemann, A.F. (2001) Int. J. Pharm., 230(1-2),

11-24.

[39] Choosakoonkriang, S.; Lobo, B.A.; Koe, G.S.; Koe, J.G. and Mid-

daugh, C.R. (2003) J. Pharm. Sci., 92(8), 1710-1722.

[40] Sonawane, N.D.; Szoka, F.C., Jr. and Verkman, A.S. (2003) J.

Biol. Chem., 278(45), 44826-44831.

[41] Høgset, A.; Prasmickaite, L.; Hellum, M.; Engesæter, B.O.; Olsen,

V.M.; Tjelle, T.E.; Wheeler, C.J. and Berg, K. (2002) Somat. Cell

Mol. Genet., 27(1-6), 97-113.

[42] Ogris, M.; Steinlein, P.; Kursa, M.; Mechtler, K.; Kircheis, R. and

Wagner, E. (1998) Gene Ther., 5(10), 1425-1433.

[43] Boussif, O.; Lezoualc'h, F.; Zanta, M.A.; Mergny, M.D.; Scher-

man, D.; Demeneix, B. and Behr, J.P. (1995) Proc. Natl. Acad. Sci.

USA, 92(16), 7297-7301.

[44] Ndoye, A.; Dolivet, G.; Høgset, A.; Leroux, A.; Fifre, A.; Erba-

cher, P.; Berg, K.; Behr, J.P.; Guillemin, F. and Merlin, J.L. (2006)

Mol. Ther., 13(6), 1156-1162.

[45] Nishiyama, N.; Iriyama, A.; Jang, W.D.; Miyata, K.; Itaka, K.;

Inoue, Y.; Takahashi, H.; Yanagi, Y.; Tamaki, Y.; Koyama, H. and

Kataoka, K. (2005) Nat. Mater., 4(12), 934-941.

[46] Greber, U.F.; Willetts, M.; Webster, P. and Helenius, A. (1993)

Cell, 75(3), 477-486.

[47] Engesæter, B.O.; Bonsted, A.; Berg, K.; Høgset, A.; Engebråten,

O.; Fodstad, O.; Curiel, D.T. and Maelandsmo, G.M. (2005) Can-

cer Gene Ther., 12(5), 439-448.

[48] Engesæter, B.O.; Tveito, S.; Bonsted, A.; Engebråten, O.; Berg, K.

and Mælandsmo, G.M. (2006) J. Gene Med., 8(6), 707-718.

[49] Buning, H.; Ried, M.U.; Perabo, L.; Gerner, F.M.; Huttner, N.A.;

Enssle, J. and Hallek, M. (2003) Gene Ther., 10(14), 1142-1151.

[50] Luna, M.C.; Chen, X.; Wong, S.; Tsui, J.; Rucker, N.; Lee, A.S.

and Gomer, C.J. (2002) Cancer Res., 62(5), 1458-1461.

[51] Prasmickaite, L.; Hellum, M.; Kaalhus, O.; Høgset, A.; Wagner, E.

and Berg, K. (2006) Photochem. Photobiol., 82(3), 809-816.

[52] Wang, L.; Prakash, R.K.; Stein, C.A.; Koehn, R.K. and Ruffner,

D.E. (2003) Antisense Nucleic Acid Drug Dev., 13(3), 169-189.

[53] Kurreck, J. (2003) Eur. J. Biochem., 270(8), 1628-1644.

[54] Koppelhus, U. and Nielsen, P.E. (2003) Adv. Drug Deliv. Rev.,

55(2), 267-280.

[55] Brooks, H.; Lebleu, B. and Vives, E. (2005) Adv. Drug Deliv. Rev.,

57(4), 559-577.

[56] Richard, J.P.; Melikov, K.; Vives, E.; Ramos, C.; Verbeure, B.;

Gait, M.J.; Chernomordik, L.V. and Lebleu, B. (2003) J. Biol.

Chem., 278(1), 585-590.

[57] Folini, M.; Berg, K.; Millo, E.; Villa, R.; Prasmickaite, L.; Dai-

done, M.G.; Benatti, U. and Zaffaroni, N. (2003) Cancer Res.,

63(13), 3490-3494.

[58] Shiraishi, T. and Nielsen, P.E. (2006) FEBS Lett., 580(5), 1451-

1456.

[59] Boe, S. and Hovig, E. (2006) Oligonucleotides, 16(2), 145-157.

[60] Lindgren, M.; Hallbrink, M.; Prochiantz, A. and Langel, U. (2000)

Trends Pharmacol. Sci., 21(3), 99-103.

[61] Zavaglia, D.; Normand, N.; Brewis, N.; O'Hare, P.; Favrot, M.C.

and Coll, J.L. (2003) Mol. Ther., 8(5), 840-845.

[62] Brewis, N.D.; Phelan, A.; Normand, N.; Choolun, E. and O'Hare,

P. (2003) Mol. Ther., 7(2), 262-270.

[63] Normand, N.; van Leeuwen, H. and O'Hare, P. (2001) J. Biol.

Chem., 276(18), 15042-15050.

[64] O'Hare, P.F.J., Brewis, N. D., Normand, N. M., and Sunassee, K.

R. (2002) Delivery of substances to cells. US Patent 20020142960.

[65] Koppelhus, U.; Awasthi, S.K.; Zachar, V.; Holst, H.U.; Ebbesen, P.

and Nielsen, P.E. (2002) Antisense Nucleic Acid Drug Dev., 12(2),

51-63.

[66] Pichon, C.; Monsigny, M. and Roche, A.C. (1999) Antisense Nu-

cleic Acid Drug Dev., 9(1), 89-93.

Photochemical Internalization Current Pharmaceutical Biotechnology, 2007, Vol. 8, No. 6 11

[67] Normand, N.; Valamanesh, F.; Savoldelli, M.; Mascarelli, F.;

BenEzra, D.; Courtois, Y. and Behar-Cohen, F. (2005) Mol. Vis.,

11, 184-191.

[68] Thrush, G.R.; Lark, L.R.; Clinchy, B.C. and Vitetta, E.S. (1996)

Annu. Rev. Immunol., 14, 49-71.

[69] Barbieri, L.; Battelli, M.G. and Stirpe, F. (1993) Biochim. Biophys.

Acta, 1154(3-4), 237-282.

[70] Sandvig, K.; Garred, O.; Prydz, K.; Kozlov, J.V.; Hansen, S.H. and

van Deurs, B. (1992) Nature, 358(6386), 510-512.

[71] Eiklid, K.; Olsnes, S. and Pihl, A. (1980) Exp. Cell Res., 126(2),

321-326.

[72] Berg, K.; Prasmickaite, L.; Selbo, P.K.; Hellum, M.; Bonsted, A.

and Hogset, A. (2003) Oftalmologia, 56(1), 67-71.

[73] Selbo, P.K.; Sandvig, K.; Kirveliene, V. and Berg, K. (2000) Bio-

chim. Biophys. Acta, 1475(3), 307-313.

[74] Pron, G.; Mahrour, N.; Orlowski, S.; Tounekti, O.; Poddevin, B.;

Belehradek, J., Jr. and Mir, L.M. (1999) Biochem. Pharmacol.

57(1), 45-56.

[75] Berg, K.; Dietze, A.; Kaalhus, O. and Høgset, A. (2005) Clin. Can-

cer Res., 11(23), 8476-8485.

[76] Selbo, P.K.; Sivam, G.; Fodstad, O.; Sandvig, K. and Berg, K.

(2000) Int. J. Cancer, 87(6), 853-859.

[77] Kloeckner, J.; Boeckle, S.; Persson, D.; Roedl, W.; Ogris, M.;

Berg, K. and Wagner, E. (2006) J. Control Release, In press.

[78] Bonsted, A.; Engesæter, B.O.; Hogset, A.; Mælandsmo, G.M.;

Prasmickaite, L.; D'Oliveira, C.; Hennink, W.E.; van Steenis, J.H.

and Berg, K. (2006) J. Gene Med., 8(3), 286-297.

[79] Kloeckner, J.; Prasmickaite, L.; Hogset, A.; Berg, K. and Wagner,

E. (2004) J. Drug Target, 12(4), 205-213.

[80] Weyergang, A.; Selbo, P.K. and Berg, K. (2006) J. Control Re-

lease., 111(1-2), 165-173.

[81] Konig, K. (2000) J. Microsc., 200(Pt 2), 83-104.

[82] Selbo, P.K.; Kaalhus, O.; Sivam, G. and Berg, K. (2001) Photo-

chem. Photobiol., 74(2), 303-310.

[83] Bonsted, A.; Engesæter, B.O.; Høgset, A.; Mælandsmo, G.M.;

Prasmickaite, L.; Kaalhus, O. and Berg, K. (2004) Gene Ther.,

11(2), 152-160.

[84] Høgset, A.; Engesæter, B.Ø; Prasmickaite, L.; Berg, K.; Fodstad,

O. and Mælandsmo, G.M. (2002) Cancer Gene Ther., 9(4), 365-

371.

[85] Dietze, A.; Bonsted, A.; Høgset, A. and Berg, K. (2003) Photo-

chem. Photobiol., 78(3), 283-289.

[86] Engesæter, B. Ø. 2006 Enhanced gene transfer to cancer cells using

photochemically mediated delivery of adenoviral vectors. PhD the-

sis. Faculty of Medicine, University of Oslo.

[87] Ndoye, A.; Bouali, S.; Dolivet, G.; Leroux, A.; Erbacher, P.; Behr,

J.P.; Berg, K.; Guillemin, F. and Merlin, J.L. (2004) Int. J. Oncol.,

25(6), 1575-1581.

[88] Hellum, M.; Høgset, A.; Engesæter, B.O.; Prasmickaite, L.; Stokke,

T.; Wheeler, C. and Berg, K. (2003) Photochem. Photobiol. Sci.,

2(4), 407-411.

[89] Lou, P.J.; Lai, P.S.; Shieh, M.J.; MacRobert, A.J.; Berg, K. and

Bown, S.G. (2006) Int. J. Cancer, In press.