500

Biol. Pharm. Bull. 28(3) 500505 (2005)

Vol. 28, No. 3

Enhancing Effects of Galactosylated Dendrimer/a -Cyclodextrin Conjugates on Gene Transfer Efciency

Koki WADA, Hidetoshi ARIMA, Toshihito TSUTSUMI, Fumitoshi HIRAYAMA, and Kaneto UEKAMA*
Graduate School of Pharmaceutical Sciences, Kumamoto University; 51 Oe-honmachi, Kumamoto 8620973, Japan. Received August 25, 2004; accepted December 14, 2004 To improve in vitro gene transfer efciency and/or achieve cell-specic gene delivery of polyamidoamine (PAMAM) starburst dendrimer (generation 2, G2) conjugate with a -cyclodextrin (a -CDE conjugate (G2)), we prepared a -CDE conjugate bearing galactose (Gal-a -CDE conjugates) with the various degrees of substitution of the galactose moiety (DSG) as a novel non-viral vector. The agarose gel electrophoretic studies revealed that Gal-a -CDE conjugates formed complexes with plasmid DNA (pDNA) and protected the degradation of pDNA by DNase I, but these effects impaired as the DSG value increased. Dendrimer and a -CDE conjugate exerted pDNA condensation through the complexation, but Gal-a -CDE conjugates did not. Gal-a -CDE conjugate (DSG 4) was found to have much higher gene transfer activity than dendrimer, a -CDE conjugate and Gal-a -CDE conjugates (DSG 8, 15) in HepG2, NIH3T3 and A549 cells, which are independent of the expression of the asialoglycoprotein receptor. Transfection activity of Gal-a -CDE conjugate (DSG 4) was insensitive to the existence of competitors (asialofetuin and galactose) and serum. In addition, no cytotoxicity after transfection of the complex of pDNA with Gal-a -CDE conjugate (DSG 4) was observed. These results suggest the potential use of Gal-a -CDE conjugate (DSG 4) as a non-viral vector in various cells.
Key words non-viral vector; dendrimer conjugate; a -cyclodextrin; galactose; glycofection

The area of non-viral gene therapy has been gaining in interest.1) Non-viral vectors such as cationic lipids and cationic polymers have some advantages for gene transfer, i.e. they are easily prepared plasmid DNA (pDNA) complexes, are not limited by gene size, and can be vested through structural modication with the ability to carry pDNA to the target cells.2,3) In addition, non-viral vectors are believed to be able to overcome some disadvantages of viral vectors, e.g. immunogenicity, oncogenicity and potential virus recombination. However, further improvement in the gene transfer activity of non-viral vectors has been desired. Of non-viral gene therapy, the glycofection method has recently been come to attention.4) Glycosylated polymers are used for transfection and interact with pDNA to give a glycoplex.5) In general, glycoplexes are used to target to the specic cells and/or to increase gene transfer activity. For example, a galactosylated polyethyleneimine (PEI) has high transfection efciency to hepatocyte expressing asialoglycoprotein receptor.6) In addition, a mannosylated PEI has high transfection efciency to macrophages and dendritic cells, which were mediated by the mannose-specic receptor and DEC205, respectively.7,8) Recently, Fajac and co-workers reported that glycosylation of PEI affected intracellular trafcking of its complex with pDNA.9) Furthermore, some promising ndings that glycosyl residues are considered to be very promising candidates of nuclear localization signal have been reported.10,11) Thus, glycosylation of polymer is one of the effective method to deliver gene to target cells and/or to enhance gene transfer. Cyclodextrins (CyDs) have recently been applied to gene transfer and oligonucleotide delivery.12) CyDs are cyclic (a 1,4)-linked oligosaccharides of a -D-glucopyranose containing a hydrophobic central cavity and hydrophilic outer surface, and they are known to function as novel drug carriers.13) The most common CyDs are a -, b - and g -CyDs, which consist of six-, seven- and eight-D-glucopyranose units, respec To whom correspondence should be addressed.

tively. Arima et al. previously reported the potential use of polyamidoamine (PAMAM) starburst dendrimer (generation 3, G3) bearing a -CyD with the degree of substitution (DS) of 2.4 because of the highest transfection efciency in vitro and in vivo with low cytotoxicity.1416) In addition, Arima et al. described that mannosylated a -CDE conjugate with the DS values of 3 of the mannose moiety can enhance gene transfer activity of pDNA in various cells with less cytotoxicity.17) The purpose of this study is to evaluate physicochemical properties, cytotoxicity and in vitro gene delivery efciency of a -CDE conjugate (G2) bearing galactose (Gal-a -CDE conjugates, Fig. 1) with the various degrees of substitution of the galactose moiety (DSG) as a novel non-viral vector in HepG2, NIH3T3 and A549 cells. MATERIALS AND METHODS Materials a -CyD was donated by Nihon Shokuhin Kako (Tokyo, Japan) and recrystallized from water. Polyamidoamine (PAMAM) starburst dendrimer (ethylenediamine core, G2, the terminal amino groups 16, molecular weight 3256) and a -D-galactopyranosylphenyl isothiocyanate, asialofetuin, galactose and poly-L-lysine (MW 1500030000) were purchased from Sigma-Aldrich Japan (Tokyo, Japan). p-Toluenesulfonyl chloride was purchased from Nakalai Tesque (Kyoto, Japan). Fetal calf serum (FCS) was obtained from Nichirei (Tokyo, Japan). Dulbeccos modied Eagles medium (DMEM) and MEM were purchased from Nissui Pharmaceuticals (Tokyo, Japan). Plasmid pRLCMV-Luc vector encoding Renilla luciferase (pDNA) was obtained from Promega (Tokyo, Japan). The purication of pDNA amplied in bacteria was carried out using QIAGEN EndoFree plasmid maxi kit ( 0.1 EU/m g endotoxin). Other chemicals and solvents were of analytical reagent grade. Preparation of Gal-a -CDE Conjugates a -CDE conju 2005 Pharmaceutical Society of Japan

e-mail: uekama@gpo.kumamoto-u.ac.jp

March 2005

501

gate (G2, DS 1) was prepared as previously reported.14) Galactose residues were attached to primary amino residues of a -CDE conjugate according to the method of Liang et al.18) Five hundreds m l of sodium chloride solution (0.9%, pH 9.0) containing a -CDE conjugate (45 mg) and dimethyl sulfoxide (DMSO) solution containing 0.5 ml of a -D-galactopyranosylphenyl isothiocyanate (DSG 1; 4 mg; DSG 4; 16 mg; DSG 5; 20 mg; DSG 8, 32 mg; and DSG 15, 60 mg) were mixed at 25 C for 24 h. Gal-a -CDE conjugates were puried by gel-ltration (TOSOH TSKGel HW-40S, Tokyo, Japan). 1 H-NMR spectra of these conjugates were measured, and the molar ratios of a -CyD and galactose residues were calculated from the peak areas of anomeric proton of a -CyD and phenyl protons of the a -D-galactopyranosylphenyl isothiocyanate. The DS values are regulated by adjusting the amounts of a -D-galactopyranosylphenyl isothiocyanate because the values are proportional to the amount of the additive. The yield of Gal-a -CDE conjugates (DSG, 1, 4, 5, 8 and 15) from a -CDE conjugate was approximately 60%. Interaction between pDNA and Carriers Electrophoretic mobility of the complex of pDNA/dendrimer, pDNA/a -CDE conjugate or pDNA/Gal-a -CDE conjugates was measured using a gel electrophoresis system. Various amounts of dendrimer, a -CDE conjugate or Gal-a -CDE conjugates were mixed with 0.2 m g of pDNA in Hanks balanced salt solution (HBSS, pH 7.4). Gel electrophoresis was carried out at room temperature in TBE buffer (45 mM Trisborate, 1 mM EDTA, pH 8.0) in 1% agarose gel (include 0.1 m g/ml of ethidium bromide) using MupidTM system (Cosmo Bio, Tokyo, Japan) at 100 V for 40 min. The pDNA bands were visualized using an UV illuminator. DNase I Protection Assay The complex of pDNA (0.2 m g) with dendrimer, a -CDE conjugate or Gal-a -CDE conjugates was prepared as described above and then incubated for 30 min at 37 C in the reaction buffer (pH 7.4, 10 mM TrisHCl, 10 mM NaCl, 10 mM CaCl2, 6 mM MgCl2, 2 mM DTT and 0.2 m g/ml of bovine serum albumin) containing 0.1 unit/m l of DNase I. After the incubation, DNase I was inactivated by heating the solution at 70 C for 10 min, 1 m l of EDTA (0.5 M) and 10 m l of 10% SDS were added to each sample and incubated at 37 C for 60 min to dissociate pDNA from the complexes. After the incubation, gel electrophoresis was carried out in the same manner as described above. DNA Condensed Assay pDNA (1 m g), ethidium bromide (1 m g) and carriers were added to 1 ml of HBSS (pH 7.4) with the various charge ratios. The solutions were incubated at 25 C for 15 min, and then the uorescence (l ex 510 nm, l em 590 nm) was measured by uorescence spectrometer Hitachi F-4500 (Tokyo, Japan). Samples containing pDNA (1 m g) and ethidium bromide (1 m g) were used to calibrate the apparatus to 100% uorescence against a background of ethidium bromide (1 m g). Cell Culture HepG2 cells, a human hepatocellular carcinoma cell line was obtained from Riken Bioresource Center (Tsukuba, Japan). NIH3T3 cells, a mouse broblast cell line, and A549 cells, a human lung epithelium cell line, were purchased from American Type Culture Collection (Rockville, MD). HepG2, and NIH3T3 and A549 cells were grown in MEM and DMEM, respectively, containing 1 105 mU/ml of penicillin, 0.1 mg/ml of streptomycin supplemented with

Fig. 1.

Chemical Structures of Gal-a -CDE Conjugates

10% FCS at 37 C in a humidied 5% CO2 and 95% air atmosphere. In Vitro Gene Transfer In vitro transfection of the complex of pDNA/dendrimer, pDNA/a -CDE conjugate or pDNA/Gal-a -CDE conjugates was performed utilizing the luciferase expression of pDNA in HepG2, NIH3T3 and A549 cells. The pDNA (2.0 m g) was mixed with dendrimer, a -CDE conjugate or Gal-a -CDE conjugates at the charge ratio of 200/1, 50/1 or 10/1 (carrier/pDNA). The pDNA complex with dendrimer, a -CDE conjugate or Gal-a -CDE conjugates was then allowed to stand for 15 min at room temperature. The cells (2 105 cells per 24 well plate) were seeded 6 h before transfection, and then washed twice with serum-free medium. Two hundred m l of serum-free medium containing pDNA or the complexes with various carriers and 200 m l of medium containing 20% FCS (nal concentration of FCS was 10%) in the absence and presence of asialofetuin or galactose as a competitor at the concentration of 1 mg/ml and 20 mM, respectively, were added to each dish, and then incubated at 37 C for 24 h. After transfection, the gene expression was measured as follows: Renilla luciferase content in the cell lysate was quantied using the Promega Renilla luciferase assay reagent (Tokyo, Japan) and a luminometer (Lumat LB9506, EG&G Berthold Japan, Tokyo, Japan). It was conrmed that a -CyD, a -CDE conjugate and Gal-a CDE conjugates have no inuence on the luciferase assay under the experimental conditions. Total protein content of the supernatant was determined by BCA protein assay kit (Pierce, Rockford, IL). Cytotoxicity The effects of pDNA complex with dendrimer, a -CDE conjugate or Gal-a -CDE conjugates on cell viability were measured as reported previously.14) In brief, HepG2 and NIH3T3 cells (2 104 cells per 96 well plate) were incubated for 6 h and then the culture medium (50 m l) supplemented with 10% FCS containing pDNA or the complexes with various carriers was added to each well, and the cells were incubated in a humidied atmosphere containing 5% CO2 and 95% air at 37 C for 24 h. After washing twice with HBSS (pH 7.4) to remove pDNA and/or various carriers, 100 m l of fresh HBSS and 10 m l of WST-1 reagent were added to the plates and incubated at 37 C for 2 h. The absorbance of the solution was measured at 450 nm, with referring absorbance at 655 nm, with a microplate reader (BioRad Model 550, Tokyo, Japan). Data Analysis Data are given as means S.E.M. Statistical signicance of mean coefcients for the studies was performed by analysis of variance followed by Scheffes test. p-

502

Vol. 28, No. 3

Values for signicance were set at 0.05. RESULTS AND DISCUSSION In the present study, we used PAMAM starburst dendrimer (G2) because of its extremely low cytotoxicity.14,19) To reveal physicochemical properties of pDNA complexes, we examined the complex formation between pDNA/dendrimer, pDNA/a -CDE conjugate or pDNA/Gal-a -CDE conjugates (DSG 1, 4, 5, 8, 15) using agarose electrophoresis. As shown in Fig. 2, the intensity of the band derived from pDNA decreased as the charge ratio of pDNA/dendrimer increased, and at the charge ratio of 1 the band disappeared. Similar patterns were observed in the pDNA/a -CDE conjugates. In the case of Gal-a -CDE conjugates (DSG 1, 4, 5, 8), the bands vanished at the charge ratio of 1, 1, 2 and 5 (carrier/pDNA), respectively, but the band still remained up to the charge ratio of 5 in the Gal-a -CDE conjugate (DSG 15) system. The addition of a -CyD and/or galactose to pDNA solution did not change the electrophoretic band pattern of pDNA (data not shown), suggesting much less interaction of pDNA with a -CyD and galactose. These results suggest that Gal-a -CDE conjugates form the complexes with pDNA, but the complexation ability with pDNA decreased with increasing the DSG values, due to a decrease in the number of the positively charged primary amino groups. Next, we tested the stabilizing effect of dendrimer, a -CDE conjugate and Gal-a -CDE conjugates on pDNA degradation by DNase I (Fig. 3). The concentration of DNase I was set to 0.1 unit/m l, because the activity in human serum is 2 10 4 to 8.2 10 2 unit/m l.20) When pDNA alone was treated with DNase I for 30 min, it degraded into smaller size fractions. The addition of dendrimer and a -CDE conjugate increased the band intensity corresponding to pDNA as the charge ratio (carrier/pDNA) increased, indicating their pDNA stabilizing effect (Fig. 3). Likewise, Gal-a -CDE conjugates (DSG 1, 4, 5) stabilized pDNA, but not Gal-a -CDE conjugates (DSG 8). Thus, the stabilizing effects of Gal-a -CDE conjugates appear to impair as the DSG value increased. Cationic non-viral vectors such as cationic polymers and cationic lipids are believed to exert pDNA compaction through electrostatic interaction, leading to the enhanced gene transfer activity.21,22) To examine the effects of dendrimer, a -CDE conjugate and Gal-a -CDE conjugates (DSG 4, 5) on pDNA condensation, uorescence intensity of ethidium bromide was determined (Fig. 4). The relative uorescence intensity decreased to 78% as the charge ratio of dendrimer and a -CDE conjugate increased up to 10. However, no condensation effect of Gal-a -CDE conjugates (DSG 4 or 5) was observed. On the other hand, when Poly-L-lysine, a positive control, was used, the relative uorescence intensity decreased to approximately 43%, indicating that Gal-a -CDE conjugates do not have the ability of pDNA compaction. Taken together, these results suggest that the galactosylation of a -CDE conjugates, especially with higher DSG values, causes undesirable physicochemical properties of pDNA complexes for gene transfer because it attenuated the interaction with pDNA and the enzymatic stability, and lost the pDNA condensing ability. To investigate whether Gal-a -CDE conjugates (G2, DSG 4, 5, 8, 15) have cell-specic and/or sufcient gene transfer

Fig. 2. Agarose Gel Electrophoretic Analysis of the pDNA Complexes with Carriers
The solutions containing the pDNA complexes with carriers were incubated for 15 min at room temperature after slight agitation. The electrophoresis was performed at 100 V for about 40 min.

Fig. 3. Effects of Various Carriers on Electrophoretic Mobility of pDNA Treated with DNase I
The pDNA complexes with carriers were incubated for 30 min at 37 C in the reaction buffer containing 0.1 unit/m l of DNase I. The electrophoresis was performed at 100 V for about 40 min.

Fig. 4.

pDNA Compaction in Various Complexes with Carriers

pDNA compaction was determined by dye displacement assay. Closed circle, dendrimer (G2); open circle, a -CDE conjugate; closed square, Gal-a -CDE conjugate (DSG 4); open square, Gal-a -CDE conjugate (DSG 5); open inverted triangle, poly-Llysine. Each point represents the mean S.E.M. of 36 experiments.

activity, Renilla luciferase activity after transfection of pDNA complexes at the charge ratio of 200/1 (carrier/ pDNA) in various cells was determined. Here, we conrmed that the cell surface asialoglycoprotein receptor was expressed in HepG2, but neither NIH3T3 nor A549 cells, by a

March 2005

503

Fig. 5.

Transfection Efciency of pDNA Complexes with Various Carriers in HepG2, NIH3T3 and A549 Cells

The effects of the degree of substitution of galactose of Gal-a -CDE conjugates on transfection efciency in HepG2 cells (A), NIH3T3 cells (B), and A549 cells (C). The luciferase activity in cell lysates was determined 24 h after transfection. The charge ratio was 200/1 (carrier/pDNA). Each value represents the mean S.E.M. of 39 experiments. * p 0.05, compared with a -CDE conjugate (A), (B) and (C). The effects of asialofetutin and galactose on transfection efciency of the complexes with various carriers in HepG2 cells (D) and NIH3T3 cells (E). The charge ratio was 200/1 (carrier/pDNA). The concentrations of asialofetuin and galactose were 1.0 mg/ml and 20 mM, respectively. Each value represents the mean S.E.M. of 34 experiments. The effects of asialofetuin on transfection efciency of the complexes with various carriers 24 h after transfection at the low charge ratios in HepG2 cells. The charge ratios (carrier/pDNA) were 50/1 (F) and 10/1 (G). The concentration of asialofetuin was 1.0 mg/ml. Each value represents the mean S.E.M. of 34 experiments.

reverse transcription-polymerase chain reaction method and ow cytometric method (data not shown), which is consistent with previous papers.2325) When pDNA alone in the absence and presence of a -CyD was transfected to cells, no luciferase activity was observed (data not shown). In these cell lines used here, the conjugation of a -CyD to PAMAM dendrimer (G2) augmented gene transfer activity (Figs. 5AC), possibly due to the enhancement of endosomal release of pDNA, as described previously,1416) although these data show slight differences due to the different experimental conditions, i.e. pDNA type and the presence of FCS. An additional attachment of the galactose residue to a -CDE conjugate with DSG value of 4, but not that of 8 or 15, elicited much more gene transfer activity in HepG2 and NIH3T3 cells (Figs. 5A, B). Likewise, Gal-a -CDE conjugate (DSG 4) showed higher gene transfer activity than dendrimer and a -CDE conjugate in A549 cells (Fig. 5C). Therefore, these results suggest that

of all of the Gal-a -CDE conjugates, Gal-a -CDE conjugate (DSG 4) augments gene transfer activity in a cell surface asialoglycoprotein receptor-independent manner. To conrm next whether gene transfer activity of Gal-a CDE conjugate (DSG 4) is associated with the galactosebinding lectin recognition, competitive studies using asialofetuin and galactose were performed (Figs. 5D, E). Here we used the pDNA complexes at the low charge ratio of 200/1 (carrier/pDNA). The gene transfer activity of Gal-a CDE conjugate (DSG 4) as well as dendrimer or a -CDE conjugate was decreased by co-incubation with asialofetuin (1.0 mg) and galactose (20 mM) only very slightly in HepG2 and NIH3T3 cells (Figs. 5D, E). It is presumable that the free Gal-a -CDE conjugate is predominant in the suspension containing the pDNA complexes as described before.14) To reduce the extent of a free form of Gal-a -CDE conjugate, we examined gene transfer activity at the charge ratios of 50/1

504

Vol. 28, No. 3

and 10/1 (carrier/pDNA). As shown in Figs. 5F and G, there were no competitive effects of asialofetuin on gene transfer activity of Gal-a -CDE conjugate in HepG2 cells. Thus, it is evident that the mechanism for the enhancing effect of Gala -CDE conjugate (DSG 4) on gene transfer activity may be due to other factors such as changes in intracellular trafcking and/or stability of pDNA, but not the cell surface galactose-specic receptor. In addition, the only very slight recognition ability of Gal-a -CDE conjugates (DSG 4) to the asialoglycoprotein receptor may be attributed to low DSG values and short length between primary amino groups of dendrimers and the galactose moiety. The serum is generally known to affect transfection efciency of non-viral vectors. Then, we examined the effects of serum on the gene transfer activity of Gal-a -CDE conjugate (DSG 4). Interestingly, gene transfer activity of Gal-a -CDE conjugate (DSG 4, 5) was not suppressed by the addition of 10% FCS, although that of dendrimer tended to be suppressed (Fig. 6). Therefore, gene transfer activity of the Gala -CDE conjugate (DSG 4, 5) may be considerably resistant to serum under the experimental conditions. To reveal the cytotoxicity of Gal-a -CDE conjugates, we examined WST-1 method (Fig. 7). No cytotoxicity of pDNA complexes with dendrimer, a -CDE conjugate and Gal-a CDE conjugates (DSG 4, 5) was observed in HepG2 and NIH3T3 cells up to the charge ratio of at least 200/1 (carrier/pDNA). These results indicate that Gal-a -CDE conjugates have great advantages as nonviral vectors, i.e. superior transfection efciency and lower cytotoxicity. The same result was observed in A549 cells (data not shown). Apparently, cytotoxicity is not involved in the enhancing effect of Gal-a -CDE conjugates on gene transfer activity. However, the detail mechanisms for the enhancing effect of Gal-a -CDE conjugate (DSG 4) on gene transfer activity are still unknown. As demonstrated here, Gal-a -CDE conjugate (DSG 4) improved gene transfer activity in an asialoglycoprotein-independ manner compared with a -CDE conjugate, despite Gal-a -CDE conjugates, especially with higher DSG values, have undesirable physicochemical properties of pDNA complexes for gene transfer. Likewise, a -CDE conjugate (G2) with mannose (Man-a -CDE conjugate), which we recently prepared, improves gene transfer activity in various cells in a cell-surface mannose receptor-independent manner, possibly through suppression of pDNA methylation and improvement in nuclear translocation (data not shown); the mechanism by which Man-a -CDE conjugate suppressed pDNA methylation is still unclear, but possibly the suppressive effects of Man-a -CDE conjugate result from the inhibition of DNA methyltransferases such as DNMT1, DNMT3a and DMMT3b and/or on histon deacetylases in a direct or indirect manner.26) Furthermore, cellular association of the pDNA complexes with Man-a -CDE conjugate and pDNA/a CDE conjugate was equivalent to that of the pDNA/dendrimer complex.14) Therefore, three possibilities for the enhancing effect of Gal-a -CDE conjugate (DSG 4) in cells could be proposed: 1) the feasible intracellular trafcking and nuclear translocation of the pDNA complex with Gal-a CDE conjugate (DSG 4), which may be involved through the interaction with intracellular galactose-binding lectins such as galectins27,28) and CBP35, a nuclear galactose-binding lectin29,30) in a direct or indirect manner, but not the cell sur-

Fig. 6. Effects of Serum on Transfection Efciency of pDNA Complexes with Various Carriers in HepG2 Cells
The luciferase activity in cell lysates was determined 24 h after transfection with or without 10% FCS. The charge ratio was 200/1 (carrier/pDNA). Each value represents the mean S.E.M. of 34 experiments.

Fig. 7.

Cytotoxicity of pDNA Complexes with Various Carriers in Cells

(A) HepG2 cells, (B) NIH3T3 cells. Cells were incubated for 60 min with pDNA complexes with carriers. Cell viability was assayed by the WST-1 method. Closed circle, dendrimer (G2); open circle, a -CDE conjugate; closed square, Gal-a -CDE conjugate (DSG 4); open square, Gal-a -CDE conjugate (DSG 5); Each point represents the mean S.E.M. of 34 experiments.

face asialoglycoprotein receptor, 2) the resistance to methylation of pDNA and/or 3) the feasible release of pDNA from the complexes with Gal-a -CDE conjugate (DSG 4) in cytoplasm and/or nucleus because of the lack of pDNA condensation in the Gal-a -CDE conjugate (DSG 4) system (Fig. 4). Anyhow, further investigation should be required to clarify the enhancing mechanisms of Gal-a -CDE conjugates on in vitro gene transfer efciency. In addition, in vivo application of Gal-a -CDE conjugate (DSG 4) is currently under investigation. In conclusion, Gal-a -CDE conjugate (DSG 4), which is our original conjugate, was shown to be the most preferential carrier among the Gal-a -CDE conjugates prepared in a series of the studies, because it provided good gene transfer activity in vitro with no cytotoxicity. Consequently, the potential use of Gal-a -CDE conjugate (DSG 4) could be expected as a non-viral vector to deliver gene and these data may be useful for design of a -CyD and galactose conjugates with other non-viral vectors. Acknowledgements This work was partially supported by a Grant-in-Aid from Tokyo Biochemical Research Foundation, a Grant-in-Aid from the Research Foundation for Pharmaceutical Sciences, a Grant-in-Aid for Encouragement of Young Scientists from the Ministry of Education, Science and Culture of Japan (12771464) and a Grant-in-Aid for Sci-

March 2005

505 15) 16) 17) 18) 19) 20) Kihara F., Arima H., Tsutsumi T., Hirayama F., Uekama K., Bioconjug. Chem., 13, 12111219 (2002). Kihara F., Arima H., Tsutsumi T., Hirayama F., Uekama K., Bioconjug. Chem., 14, 342350 (2003). Arima H., Wada K., Kihara F., Tsutsumi T., Hirayama F., Uekama K., J. Incl. Phenom. Macro. Med., 44, 361364 (2004) Liang W., Shi X., Deshpande D., Malanga C., Rojanasakul Y., Biochim. Biophys. Acta, 1279, 227234 (1996). Haensler J., Szoka F. C., Jr., Bioconjug. Chem., 4, 372379 (1993). Avall-Lundqvist E., Economidou-Karaoglou A., Sjovall K., Lans M., Taper H. S., Roberfroid M., Eneroth P., Clin. Chim. Acta, 185, 3543 (1989). Schwartz B., Ivanov M. A., Pitard B., Escriou V Rangara R., Byk G., ., Wils P., Crouzet J., Scherman D., Gene Ther., 6, 282292 (1999). Thomas M., Klibanov A. M., Proc. Natl. Acad. Sci. U.S.A., 99, 1464014645 (2002). Schwartz A. L., Fridovich S. E., Knowles B. B., Lodish H. F., J. Biol. Chem., 256, 88788881 (1981). Yu H. Y., Lin J. H., J. Formos. Med. Assoc., 99, 936941 (2000). Niidome T., Urakawa M., Sato H., Takahara Y., Anai T., Hatakayama T., Wada A., Hirayama T., Aoyagi H., Biomaterials., 21, 18111819 (2000). Szyf M. Drug Resist. Updat., 6, 341353 (2003). Liu F. T., Patterson R. J., Wang J. L., Biochim. Biophys. Acta, 1572, 263273 (2002). Wang J. L., Gray R. M., Haudek K. C., Patterson R. J., Biochim. Biophys. Acta, 1673, 7593 (2004). Wang J. L., Laing J. G., Anderson R. L., Glycobiology, 1, 243252 (1991). Cowles E. A., Moutsatsos I. K., Wang J. L., Anderson R. L., Exp. Gerontol., 24, 577585 (1989).

entic Research (C) from Japan Society for the Promotion of Science (14572158). REFERENCES
1) 2) 3) 4) 5) 6) 7) 8) 9) Thomas M., Klibanov A. M., Appl. Microbiol. Biotechnol., 62, 2734 (2003). Rolland A. P., Crit. Rev. Ther. Drug Carrier Syst., 15, 143198 (1998). Nishikawa M., Huang L., Hum. Gene Ther., 12, 861870 (2001). Roche A. C., Fajac I., Grosse S., Frison N., Rondanino C., Mayer R., Monsigny M., Cell. Mol. Life Sci., 60, 288297 (2003). Fajac I., Briand P., Monsigny M., Glycoconj. J., 18, 723729 (2001). Zanta M. A., Boussif O., Adib A., Behr J. P., Bioconjug. Chem., 8, 839844 (1997). Sato A., Kawakami S., Yamada M., Yamashita F., Hashida M., J. Drug Target., 9, 201207 (2001). Diebold S. S., Plank C., Cotten M., Wagner E., Zenke M., Somat. Cell Mol. Genet., 27, 6574 (2002). Fajac I., Thevenot G., Bedouet L., Danel C., Riquet M., Merten M., Figarella C., DallAva-Santucci J., Monsigny M., Briand P., J. Gene Med., 5, 3848 (2003). Grosse S., Aron Y., Honore I., Thevenot G., Danel C., Roche A. C., Monsigny M., Fajac I., J. Gene Med., 6, 345356 (2004). Duverger E., Carpentier V Roche A. C., Monsigny M., Exp. Cell Res., ., 207, 197201 (1993). Freeman D. J., Niven R. W., Pharm. Res., 13, 202209 (1996). Uekama K., Hirayama F., Irie T., Chem. Rev., 98, 20452076 (1998). Arima H., Kihara F., Hirayama F., Uekama K., Bioconjug. Chem., 12, 476484 (2001).

21) 22) 23) 24) 25)

10) 11) 12) 13) 14)

26) 27) 28) 29) 30)