1 European Journal of Pharmaceutical Sciences 13 (2001) locate/ ejps Review Intracellular control of gene trafficking using liposomes as drug carriers Hideyoshi Harashima *, Yasuo Shinohara, Hiroshi Kiwada a, b b a Graduate School of Pharmaceutical Sciences, Hokkaido University, Kita 12, Nishi 6, Sapporo , Japan b Faculty of Pharmaceutical Sciences, The University of Tokushima, Shomachi , Tokushima , Japan Abstract The objective of this review is to summarize some of the critical barriers in gene delivery and recent progress in overcoming such barriers using non-viral carrier systems. Receptor-mediated endocytosis is generally considered to be a principal entering pathway. Therefore, endosomal escape is an essential step for achieving efficient transfection. The nuclear membrane is also a critical barrier in gene delivery and the application of the nuclear localization signal is discussed, based on recent strategies. It is essential to optimize the carrier system, in order to enhance the transfection ability equivalent to a viral system. The importance of developing an intracellular pharmacokinetic model of genes is emphasized in the optimization of non-viral carrier systems Elsevier Science B.V. All rights reserved. Keywords: Liposomes; Drug delivery system; Gene delivery; Carrier system; Nuclear localization signal 1. Introduction their pharmacokinetics in the body. Recent progress in the field of viruses have revealed some surprising mechanisms In the past 30 years, the concept of a drug delivery which they have acquired through evolution. Polymer or system (DDS) has been created and a variety of strategies particulate systems are often used to mimic the multihave been developed for improving the pharmacokinetics functional system which is seen in viruses and to develop of drugs. Such strategies include absorption, distribution, an artificial carrier system. In this paper, we critically metabolism and excretion for the enhancement of ef- review the recent progress on the intracellular control of ficiency, as well as for decreasing side effects. Our macromolecules such as DNA for gene therapy using understanding of the pharmacokinetics of drugs has made liposomes as the drug carrier. remarkable progress and the systematic theory has been completed at the tissue level. However, many of the prior studies focused on the fate of small molecules. When new 2. Across plasma membrane strategies of medical treatments such as gene therapy are considered, the importance of studies on the intracellular The plasma membrane is the first barrier for the fate of macromolecules such as DNA is obvious. In intracellular delivery of macromolecules. Two major enparticular, in the case of gene therapy, major factors trance pathways are considered, i.e. direct fusion and controlling the fate of the introduced gene and the ef- endocytosis. Receptor-mediated endocytosis (RME) is a ficiency of its expression would be expected to be intracel- selective internalizing pathway for macromolecules, allular events. though direct fusion is superior in efficiency (Kato and The development of methods for controlling the intracel- Sugiyama, 1997). In the case of hepatic targeting, galaclular trafficking of macromolecules is one of the most tose receptors are used for hepatocytes, while mannose attractive research subjects in the field of DDS for the 21st receptors are used for Kupffer cells (Takakura et al., century (Harashima et al., 1999). Sophisticated controlling 1996). In the case of tumor cell targeting, liposomes need systems are required for the intracellular delivery of to be modified as long circulating liposomes via the macromolecules to the target site by circumventing several incorporation of polyethyleneglycol (PEG), in order to barriers after transport into the cells as well as controlling avoid recognition by the immune system (Maruyama, 1999). The diameters of liposomes are also critical in *Corresponding author. Tel./ fax: improving the delivery via leaky tumor endothelium. In address: (H. Harashima). most cases, the optimum size is 100 nm in diameter for / 01/ $ see front matter 2001 Elsevier Science B.V. All rights reserved. PII: S (00)
2 86 H. Harashima et al. / European Journal of Pharmaceutical Sciences 13 (2001) Fig. 1. Structures of cationic lipids (from Marshall et al. (1999)). both blood circulation time and permeability in tumor seems to result in lower uptake and lower transfection endothelial cells. The addition of selective ligands such as (Mounkes et al., 1998). transferrin or monoclonal antibody fragments at the edge of the PEG is also a useful strategy for internalization via RME. In the case of gene delivery, cationic lipids have 3. Endosomal escape advantages in terms of enhancing the binding to target cells which are negatively charged. Many types of cationic Endosomal escape mechanisms which have been applied lipids have been developed thus far, as shown in Fig. 1 to gene delivery are summarized in Table 1. After inter- (Marshall et al., 1999). nalization via endocytosis, the complexes exist in endo- A large complex of an aggregated complex of cationic somes and then either fuse with lysosomes or recycle back lipid and DNA is internalized with great difficulty via to plasma membrane. However, the route by which they endocytosis, since the average diameter of endosomes is reach the cytosolic compartment is unknown. It is required nm. Polycations such as poly-l-lysine or poly- that they escape from lysosomal degradation and are ethyleneimine (PEI) can effectively reduce the mean delivered into the cytosol, for enhanced transfection actividiameter of the complex around 100 nm, which enhances ty. Viruses such as an influenza or an adenovirus enter into the transfection activity by protecting the DNA from cell via endocytosis and escape from endosomes by DNAase in serum, enhancing uptake via RME, and by membrane fusion or disruption using the acidic ph in enhancing endosomal escape, as described below. It has endosomes ( 5.0). Mechanisms such as these have been been reported that proteoglycans play an important role as applied to enhance gene delivery by several researchers a binding site on the cell surface in adherent cells, while (Curiel et al., 1991; Wu et al., 1994; Wagner et al., 1992; non-adherent cells do not contain such molecules, which Plank et al., 1994). Dioleoylphosphatidyl ethanolamine Table 1 Strategies for the selective gene delivery by receptor-mediated endocytosis and endosomal escape mechanisms Receptor Target cells Endosomal escape Compaction Reference Transferrin CFT1/ KB/ HeLa Adenovirus(dl312) Poly(L-lysine) Curiel et al., 1991 ASOR Huh7 Adenovirus(dl312) Poly(L-lysine) Wu et al., 1994 Transferrin K562/ HeLa/ BNL Influenza HA-2 Poly(L-lysine) Wagner et al., 1992 CL.2 subunit Transferrin BNL CL.2 Influenza HA-2 Poly(L-lysine) Plank et al., 1994 subunit (dimer) Macrophages Listeriolysin O Lee et al., 1996 ErbB-2-receptor COS-1/ SKBR3 Pseudomonas exotoxin A GAL4 Fominaya and Wels, 1996 (s.s. binding) Folate KB DOPE/ CHEMS Poly(L-lysine) Lee and Huang, 1996
3 H. Harashima et al. / European Journal of Pharmaceutical Sciences 13 (2001) (DOPE) forms a stable lipid bilayer at physiological ph 7, endosomal chloride anion, which diffuses into the endowhile at acidic ph 5 6, the hexagonal-ii structure forms, somes with the protons, leads to an increase in osmotic which destabilizes the membrane. Therefore, DOPE is pressure, thus inducing osmotic swelling. Behr et al. found usually used as a fusogenic lipid for ph-sensitive lipo- that PEI is advantageous in that it has a high density of somes. It is known that ph-sensitive liposomes containing positive charges and bifurcated structures which permit the DOPE release encapsulated macromolecules into the cyto- binding of DNA efficiently. Cationic lipids which contain sol (Tachibana et al., 1998). quaternary amines do not enhance transfection activity due A hypothetical model has been proposed by Xu and to a lack of proton sponge ability. Since cationic lipids Szoka to explain the mechanism of release of cationic such as DOTMA, DOTAP, DC-Chol have weak buffering lipid/ DNA complexes from endosomes as shown in Fig. 2 ability due to one cationic molecule per lipid, they require (Xu and Szoka, 1996). excess amounts of DOPE to obtain transfection. In this model, the complex of cationic lipid/dna first destabilizes the endosome membrane. Negatively charged lipids in the cytosolic phase move to the endosomal phase 4. Dissociation of DNA from cationic lipids via a flip-flop mechanism. The cationic lipids then diffuse via lateral diffusion to form neutral ion pair with cationic There are a few possibilities as to where the cationic lipids. As a result, DNA which was bound to cationic lipids and DNA dissociate, (1) in endosomes, (2) around lipids electrostatically is displaced and released into the the endosomal membrane when they escape from endocytosol. somes, (3) in the cytosol and (4) in the nucleus. Cationic Behr et al. postulated the proton sponge mechanism to lipids are not generally thought to dissociate from DNA in explain enhanced transfection activity using PEI as an the nucleus, but, rather, prior to entering nucleus, because endosomal escape enhancer (Boussif et al., 1995). PEI has complexes which were microinjected into the nucleus a nitrogen atom which can be protonated and this can directly did not show transfection (Pollard et al., 1998). consume endosomal protons. As a result, an increase in Polyanions such as RNA in the cytosol may substitute for DNA. Negatively charged phospholipids such as phosphatidylserin are also candidates for substitution for DNA (Xu and Szoka, 1996). On the other hand, a PEI/DNA complex has equivalent ability in transfection after microinjection into nucleus, which suggests that PEI may dissociate from DNA in the nucleus (Pollard et al., 1998). 5. Delivery to nucleus The nuclear membrane is thought to be the major barrier in nuclear delivery of macromolecules such as genes (Zabner et al., 1995). Pollard et al. (1998) compared dose response curves between number of microinjected copies of plasmid DNA and transfection activity measured with b-galactosidase after microinjection into cytosol or nucleus. The curve shifted to the lower dose region by several hundred-fold after microinjection into nucleus, compared to cytosolic injection. This suggests that the nuclear membrane seems to be a critical barrier in the nuclear delivery of plasmid DNA (Pollard et al., 1998). Diffusion is the principal transport mechanism in the nuclear delivery of small molecular weight compounds (,20,000 40,000) however, active transport mechanisms are required for such molecules to be transported through the nuclear pore complex (NPC), which has a diameter of 9 nm (Yoneda, 1996). Nuclear proteins require a nuclear localization signal (NLS) which contains basic amino acids such as lysine to be recognized by cytosolic factors called karyopherin alpha and beta (Jans and Hubner, 1996). In the Fig. 2. Hypothetical model for the endosomal release of cationic lipid/ case of active transport using NLS, the diameter of NPC is gene complexes (cited from Xu and Szoka (1996)). known to be expanded to 26 nm (Dworetzky et al., 1988).
4 88 H. Harashima et al. / European Journal of Pharmaceutical Sciences 13 (2001) Fig. 3. Molecular design of plasmid DNA with a single NLS (from Zanta et al. (1999)). will be very important for the successful application of liposomes as gene carriers. In the majority of studies on gene delivery, only the final output, transfection activity, has been measured and all intracellular events are relegated to a black box. However, to enhance the efficiency of gene expression, it is very important to understand which process is the rate limiting step. To answer this question, quantitative evaluation of the introduced gene in a particular subcellular region will be necessary. The intracellular pharmacokinetic model of genes is shown in Fig. 4. This model should be linked with the physiological pharmacokinetic model which describes the disposition of genes in a carrier system in the light of the overall phenomenon after intravenous administration to a target, such as tumor tissues. Little quantitative information about intracellular pharmacokinetics of genes is available, principally because of the lack of a suitable assay method to quantify the intracellular genes in each subcellular compartment. Re- We recently succeeded in delivering bovine serum albumin as a marker molecule to target to the nucleus by adding NLS (Tachibana et al., 1998). Sebestyen et al. (1998) attached the NLS of SV40 large T antigen to double strand DNA via cyclopropapyrroloindole as a bifunctional cross-linker and examined the accumulation of the NLS modified DNA into the nucleus. As a result, the DNA molecule was not taken up, even when 24 NLS molecules were introduced into 1.0 kbp DNA. For efficient uptake by nuclei, the introduction of 100 NLS peptides/ kbp DNA, i.e. 1 NLS per 10 bp of cently, we have developed quantitative measurement of DNA was necessary. Considering the observation that the plasmid DNA in the nuclear compartment which is useful addition of NLS to poly-l-lysine did not enhance the in determining the number of copies of intact plasmid transfection activity of plasmid DNA, NLS should be delivered per nucleus. This method should be expanded to bound to DNA but not to polycation and a certain amount each subcellular compartment such as cytosol, endosome/ of NLS is also required for enhanced nuclear delivery. In lysosome and plasma membrane, in order to complete the the case of nuclear proteins, single NLS is sufficient for intracellular pharmacokinetics of genes in the carrier nuclear translocation, however, additional amounts are system. The kinetic analysis of intracellular genes will required for large gold particles (Dworetzky et al., 1988). provide us with the principal factors governing each step in The nuclear transport of IgM (970 Da) cannot be enhanced intracellular trafficking of genes, which also provides by ordinary NLS, but nucleoplasmin, containing five NLS important information in optimization of the non-viral per molecule or long NLS can enhance its nuclear trans- carrier system. port (Yoneda et al., 1992). Zanta et al. first designed a hairpin-shaped oligonucleotide containing amino modified thymidine. NLS was then conjugated via bifunctional crosslinker to this oligonucleotide. Thus, the prepared hairpin-shaped oligonucleotide containing NLS was ligated to the end of double stranded DNA as shown in Fig. 3. Using this procedure, they succeeded in introducing NLS to the end of an intact DNA molecule. Compared to the method of Sebestyen et al. (1998), this procedure has two advantages. Firstly, the DNA itself is intact, and secondly, the NLS can be attached to the edge of double strand DNA. In this case, even a single molecule of NLS is capable of enhancing the transfection activities by ten- to hundreds-fold, depending on cell lines. This is the first result which shows enhanced transfection activity by introducing NLS to plasmid DNA using non-viral vectors. Fig. 4. Intracellular pharmacokinetic model of genes. The intracellular fate of genes is schematically shown. A complex of cationic lipid/ gene 6. Perspective binds to the plasma membrane electrostatically and is internalized via endocytosis (k int). Endosomes fuse with lysosomes and endosomal In this article, the status of the control of intracellular contents are degraded enzymatically (k lys). A certain fraction of endo- cytosed genes are released into the cytosol (k rel), where most genes are trafficking of genes using non-viral carriers has been degraded by DNase (k deg,c). Thus, a small fraction of genes may be reviewed. However, these studies are in their early stages. delivered to the nucleus (k nuc) as the intact form, which is transcripted in Further study of each process in intracellular gene delivery the nucleus.
5 H. Harashima et al. / European Journal of Pharmaceutical Sciences 13 (2001) Acknowledgements Proteoglycans mediated cationic liposome-dna complex-based gene delivery in vitro and in vivo. J. Biol. Chem. 273, Plank, C., Oberhauser, B., Mechtler, K., Koch, C., Wagner, E., The This work was supported in part by a Grant-in-Aid for influence of endosome-disruptive peptides on gene transfer using Scientific Research: Priority Area Research Program- synthetic virus-like gene transfer systems. J. Biol. Chem. 269, Biotargeting-(No ) and from the Ministry of Education, Science, Sports and Culture, in Japan. The Pollard, H., Remy, J.S., Loussouarn, G., Demolombe, S., Behr, J.P., authors would also like to thank Dr. M.S. Feather for his Escande, D., Polyethylenimine but not cationic lipids promote transgene delivery to the nucleus in mammalian cells. J. Biol. Chem. helpful advice in writing the English manuscript. 273, Sebestyen, M.G., Ludtke, J.J., Bassik, M.C., Zhang, G., Budker, V., Lukhtanov, E.A., Hagstrom, J.E., Wolff, J.A., DNA vector References chemistry: the covalent attachment of signal peptides to plasmid DNA. Nat. Biotechnol. 16, Tachibana, R., Harashima, H., Shono, M., Azumano, M., Niwa, M., Boussif, O., Lezoualc h, F., Zanta, M.A., Mergny, M.D., Scherman, D., Futaki, S., Kiwada, H., Intracellular regulation of macromole- Demeneix, B., Behr, J.P., A versatile vector for gene and cules using ph-sensitive liposomes and nuclear localization signal: oligonucleotide transfer into cells in culture and in vivo: polyquantitative and qualitative evaluation of intracellular trafficking. ethylenimine. Proc. Natl. Acad. Sci. USA 92, Biochem. Biophys. Res. Commun. 251, Curiel, D.T., Agarwal, S., Wagner, E., Gotten, M., Adenovirus Takakura, Y., Mahato, R.I., Nishikawa, M., Hashida, M., Control of enhancement of transferrin-polylysine-mediated gene delivery. Proc. pharmacokinetic profiles of drug macromolecule conjugates. Adv. Natl. Acad. Sci. USA 88, Drug Del. Rev. 19, Dworetzky, S.I., Lanford, R.E., Feldherr, C.M., The effects of Wagner, E., Plank, C., Zatlouka, K., Cotten, M., Birnstiel, M.L., variations in the number and sequence of targeting signals on nuclear Influenza virus hemagglutinin HA-2 N-terminal fusogenic peptides uptake. J. Cell Biol. 107, augment gene transfer by transferrinpolylysine DNA complexes: Fominaya, J., Wels, W., Target cell-specific DNA transfer mediated toward a synthetic virus-like gene-transfer vehicle. Proc. Natl. Acad. by a chimeric multidomain protein. Novel non-viral delivery system. J. Sci. USA 89, Biol. Chem. 271, Wu, G.Y., Zhan, P., Sze, L.L., Rosenberg, A.R., Wu, C.H., Harashima, H., Shinohara, Y., Kiwada, H., Perspective on the new Incorporation of adenovirus into a ligand-based DNA carrier system strategy of intracellular delivery of macromolecules with carrier. Drug results in retention of original receptor specificity and enhances Del. Syst. 14, targeted gene expression. J. Biol. Chem. 269, Jans, D.A., Hubner, S., Regulation of protein transport to the Xu, Y., Szoka, F.C., Mechanism of DNA release from cationic nucleus: central role of phosphorylation. Physiol. Rev. 76, liposome/ DNA complexes used in cell transfection. Biochemistry 35, Kato, Y., Sugiyama, Y., Crit. Rev. Ther. Drug Carrier Syst. 14, Yoneda, Y., Semba, T., Kaneda, Y., Noble, R.L., Matsuoka, Y., Kurihara, Lee, K.D., Oh, Y.K., Portnoy, D.A., Swanson, J.A., Delivery of T., Okada, Y., Imamoto, N., A long synthetic peptide containing macromolecules into cytosol using liposomes containing hemolysin a nuclear localization signal and its flanking sequences of 5V40 from Listeria monocytogenes. J. Biol. Chem. 271, T-antigen directs the transport of IgM into the nucleus efficiently. Exp. Lee, R.J., Huang, L., Folate-targeted, anionic liposome-entrapped Cell Res. 201, polylysine-condensed DNA for tumor cell-specific gene transfer. J. Yoneda, Y., Nuclear pore-targeting complex and its role on nuclear Biol. Chem. 271, protein transport. Arch. Histol. Cytol. 59, Marshall, J., Yew, N.S., Eastman, S.J., Jiang, C., Scheule, R.K., Cheng, Zabner, J., Fasbender, A.J., Moninger, T., Poellinger, K.A., Welsh, M.J., S.H., Cationic lipid-mediated gene delivery to the airways. In: Cellular and molecular barriers to gene transfer by a cationic Huang, L., Hung, M.-C., Wagner, E. (Eds.), Nonviral Vectors for Gene lipid. J. Biol. Chem. 270, Therapy. Academic Press, San Diego, CA, USA, pp Zanta, M.A., Belguise-Valladier, P., Behr, J.P., Gene delivery: a Maruyama, K., Possibility of active targeting to tumor tissues with single nuclear localization signal peptide is sufficient to carry DNA to liposomes. Adv. Drug Del. Rev., in press. the cell nucleus. Proc. Natl. Acad. Sci. USA 96, Mounkes, L.C., Zhong, W., Gipres-Palacin, G., Heath, T.D., Debs, R.T.,