Introduction

Numerous nonviral vectors have been developed for use in gene therapy for intractable diseases^1,2. However, low transfection efficiency still remains a bottleneck, preventing its use in clinical applications. It is generally considered that transfection activity is rate limited to a great extent, by a variety of intracellular processes such as endosomal escape, nuclear transfer, and intranuclear transcription.

Along with evolution of life during the past hundreds of millions of years, DNA and RNA viruses have also evolved and have developed sophisticated mechanisms for controlling intracellular trafficking for the efficient delivery of their genomes to nuclei in host cells for symbiosis. Although some nonviral vectors have been evolutionally developed since the first proposal of the concept of gene therapy over 30 years ago³, this history is overwhelmingly short. Therefore, the transfection efficiency of a virus vector is, in general, more prominent than that of a nonviral vector⁴. To improve nonviral vectors, quantitative information concerning why and to what extent the nonviral vector is inferior to the viral one is essential.

For the quantification of intracellular trafficking, we and other researchers have developed methodology to quantify the amount of plasmid DNA in the nucleus by nuclear fractionation followed by the polymerase chain reaction (PCR)^5,6,7. This quantification revealed an important lesson showing that it is necessary to optimize not only the nuclear delivery of plasmid DNA, but also intranuclear transcription efficiency, since transgene expression is remarkably saturated against nucleus-delivered pDNA. In contrast to the nucleus, very few reports are available concerning the amount of pDNA in the endosome/lysosome compartment, and therefore, it is very difficult to evaluate the efficiency of endosomal release. Although the subcellular fractionation of endosomes/lysosomes may solve this issue, many problems, such as the complicated protocol, uncertainties associated with the recovery of the endosomal fraction, and mutual contamination may prevent this strategy from becoming a practical application.

We recently developed a novel strategy, the confocal image-assisted three-dimensionally integrated quantification (CIDIQ) method, which enables the distribution of exogenous DNA in endosome/lysosome, cytosol, and nucleus to be quantified simultaneously in individual cells with sequential Z-series images captured by confocal laser scanning microscopy^8,9. Since intracellular trafficking investigated by CIDIQ can readily explain the differences in transgene expression by various nonviral vectors, this method is useful for identifying the rate-limiting barriers to gene expression.

In the present study, we applied it to the systematic and quantitative comparison of the intracellular distribution of exogenous genes transfected by viral vector and nonviral vector in living cells. It enabled us to examine the rate-limiting processes associated with nonviral vectors that limit their transfection efficiency. As model vectors, we used adenovirus (Ad) and Lipofectamine Plus (LFN), highly potent and widely used viral and nonviral vectors (lipoplex), respectively, in the comparative study.

Results

Comparison of transfection activity between Ad and LFN

We initially compared the time-dependent and applied dose-dependent transfection activity between Ad and LFN. The methodology used to determine the applied dose in terms of luciferase gene copies is described under Materials and Methods. As shown in Fig. 1A, transgene expression is increased in a dose-dependent manner for both vectors 6 h after incubation. In subsequent experiments, we fixed the dose of plasmid DNA (pDNA) in the LFN-mediated transfection at 6.7 10⁵ copies/cell (5 pg/cell) based on the manufacturer's recommended protocol, from which approximately 5 10⁷ RLU/mg protein of luciferase activity was exhibited. In the case of Ad, we fixed the dose at 200 copies/cell, since comparable levels of transgene expression can be achieved at this dose. As shown in Fig. 1B, both vectors exhibited quite comparable time courses for transgene expression at these doses, suggesting that LFN is a potent system for the delivery of pDNA to the nucleus with a speed comparable to that of Ad. However, it should be emphasized that LFN requires 3 orders of magnitude more gene copies than the Ad to achieve comparable gene expression (Fig. 1A). Furthermore, the difference in the required dose for achieving a comparable level of transgene expression is dependent on the expression level. In a comparison of the expression of 1 10⁶ RLU/mg protein, approximately 10,000-fold more copies of plasmid DNA were required for LFN, and at a lower transgene expression level, the difference was even greater. This is due to the nonlinear relationship between the dose and the transfection efficiency in LFN and is also sometimes observed in various nonviral vectors¹⁰.

Figure 1.

Dose–response curve and time course of luciferase gene expression in A549 cells transfected by Ad and LFN. (A) Luciferase gene expression in cells transfected by Ad () or LFN () was measured 6 h after incubation at the indicated dose. (B) Transfection activities were measured at indicated times after incubation with a dose of 200 () or 6.7 10⁵ copies/cell (5 pg/cell) (). The vertical axis represents luciferase activity expressed as relative light units (RLU)/mg protein. These data represent the mean values and standard deviation of three experiments.

Full figure and legend (136K)

Quantification of an intracellular distribution of pDNA and Ad DNA

We then quantified the intracellular distribution of pDNA or Ad by a combination of TaqMan PCR and CIDIQ⁸. The procedures used to determine the distribution of exogenous DNA transfected by LFN and Ad are illustrated in Figs. 2A and 2B. We determined the total cellular uptake (S_PCR(tot)) first by TaqMan PCR in terms of gene copies. In the case of LFN, we determined the fraction of plasmid DNA in the endosome/lysosome, cytosol, and nucleus by CIDIQ (Fig. 2B). A schematic diagram illustrating the principal of CIDIQ analysis is shown in Fig. 4A. After the transfection of rhodamine-labeled pDNA, we stained the endosome/lysosome and nucleus with LysoSensor DND-189 and Hoechst 33342, respectively, to discriminate the subcellular localization of the pDNA (typical images are exhibited in Fig. 4C). We transferred each 8-bit TIFF image to Image-Pro Plus version 4.0 (Media Cybernetics, Inc., Silver Spring, MD, USA) to quantify the total brightness and pixel area of each region of interest. Since we have previously shown that the majority of rhodamine-labeled pDNA was present in the form of clusters in the cytosol and nucleus at 1 h (a typical image is shown in Fig. 4C), we used the pixel area of each cluster in endosomes/lysosomes, s_i(end/lys); cytosol, s_i(cyt); and nucleus, s_i(nuc) as an index of the amount of pDNA⁸. The detailed methodology for the calculation of the fraction of pDNA in the endosome/lysosome, cytosol, and nucleus (F(end/lys), F(cyt), and F(nuc), respectively) is described under Materials and Methods. The copy number in each organelle is calculated as S_PCR(tot) multiplied by the fraction in each organelle (Fig. 2B).

Figure 2.

Schematic diagram illustrating the concept for the quantification of intracellular distribution of exogenous genes. Subcellular distributions of exogenous DNA transfected by Ad and LFN were quantified by TaqMan PCR and CIDIQ analysis as illustrated in (A) and (B), respectively. (A) Total cellular uptake (S_PCR(tot) and nuclear-delivered Ad DNA (S_PCR(nuc)) were quantified by TaqMan PCR. Distributions of Ad in endosome/lysosome and cytosol (F'(end/lys) and F'(cyt), respectively) were also determined by CIDIQ. The copy numbers in the endosome/lysosome and cytosol fractions were calculated as S_PCR(sup) multiplied by F'(end/lys) and F'(cyt), respectively. (B) S_PCR(tot) was first determined by TaqMan PCR in terms of gene copies. The fraction of plasmid DNA in the endosome/lysosome, cytosol, and nucleus was determined by CIDIQ.

Full figure and legend (136K)

Figure 4.

Quantification of the intracellular distribution of exogenous genes by CIDIQ analysis. (A) Schematic diagram of the CIDIQ analysis is illustrated. 20 Z-series images were captured by confocal laser scanning microscopy. The pixel areas corresponding to the pDNA in each x–y plane (Z = j), s_j(k), were summed for the respective compartments and are denoted as S'_{Z = j}(k), where k represents each compartment (e.g., endosome/lysosome, cytosol, and nucleus). S'_{Z = j}(k) in each x–y-plane image was further integrated to obtain S(k), which represents the total pixel area localized in each compartment (k) in one cell. All of the S(k) values were combined to calculate the S(tot), reflecting the total pixel area in a cell. F(k), representing the fraction of plasmid DNA in each compartment to the whole cell, was calculated as S(k) divided by S(tot). (B and C) Typical images for the confocal laser scanning microscopy used in the CIDIQ analysis in Ad and LFN are shown. Texas red-labeled Ad or rhodamine-labeled pDNA (red) were transfected into A549 cells. Endosome/lysosome and nucleus compartments were stained with LysoSensor DND-189 (green) and Hoechst 33342 (blue), respectively. (D and E) Fractions of DNA in the endosome/lysosome, cytosol, and nucleus determined by the strategy illustrated in Fig. 2 after transfection with Ad and LFN are shown. These values represent the mean values of 15 individual cells.

Full figure and legend (425K)

We also determined intracellular distribution for the Ad by CIDIQ (a typical image is shown in Fig. 4B), in which the pixel areas of the Texas red-labeled Ad are used as an index of the amount of Ad. However, the nuclear pixels cannot serve as an index for the amount of Ad DNA, since it is dissociated from the fluorescence-labeled capsid proteins when it is internalized via the nuclear pore complex¹¹. Therefore, we determined nucleus-associated Ad DNA (S_PCR(nuc)) beforehand by nuclear isolation, followed by TaqMan PCR (Fig. 2A). In this case, the supernatant from the nuclear isolation procedure (S_PCR(sup)) includes DNA in both the endosomal/lysosome and the cytosol fractions. The copy numbers in these fractions are calculated as S_PCR(sup) multiplied by the fractions calculated by CIDIQ (F'(end/lys) and F'(cyt), respectively).

Comparison of the cellular uptake process between LFN and Ad

We first quantified the cellular uptake of pDNA transfected with LFN and Ad in terms of copy numbers of luciferase genes by TaqMan PCR. Since the nuclear delivery of exogenous DNA is achieved within 1 h for both of vectors^8,12,13,14, we evaluated cellular uptake and the following intracellular distribution at 1 h to compare the initial disposition of DNA. As a result, the number of copies of pDNA taken up by the cell for LFN was approximately 15,000-fold more than that for Ad (Fig. 3). After we normalized the cellular uptake by the applied dose, more than 40% of the pDNA was taken up by the cell, whereas for the Ad DNA, this value was only 10% (Fig. 3).

Figure 3.

Quantification of cellular uptake and cellular binding of Ad and LFN in A549 cells. A549 cells were incubated with adenovirus or LFN at 37 and 4°C for 1 h to evaluate cellular uptake and cellular binding, respectively. Associated genes were quantified in terms of copy number of luciferase genes by TaqMan PCR. Cellular association was normalized by number of cells, which was quantified by the number of copies of the genomic -actin gene.

Full figure and legend (87K)

Total cellular uptake is a hybrid parameter of cell surface binding and the internalization rate constant (k_int). Therefore, we measured the cellular binding of pDNA and Ad by incubation at 4°C for 1 h. In the case of LFN, 8.8% of the applied pDNA was attached to the cell surface, whereas the corresponding value was only 2.4% for the Ad (Fig. 3). When the k_int, denoted as the total cellular uptake (Fig. 3) divided by the cell surface binding (Fig. 3), were compared, LFN was found to exhibit a value comparable to that of Ad (5.1 and 4.2 h^-1, respectively).

Comparison of intracellular trafficking between Ad and LFN

Concerning the intracellular distribution after transfection with LFN, 47.4% of the pDNA was in the endosome/lysosome fraction and a large part of the pDNA had already escaped from this compartment (Figs. 4E and 5) within 1 h. In addition, we observed significant nuclear distribution (13.5%) (Figs. 4E and 5). This is consistent with rapid gene expression within 3 h (Fig. 1B). Multiplying these fractions by the S_PCR(tot), we calculated the copy numbers in each organelle as shown in Fig. 5.

Figure 5.

Summary of the quantitative comparison of intracellular trafficking in A549 cells between Ad and LFN. These values were quantified by TaqMan PCR and CIDIQ analysis.

Full figure and legend (374K)

In the Ad, we determined nuclear fraction by nuclear isolation, followed by TaqMan PCR. After incubation at 37°C at 1 h, 54.2% of the Ad genome was experimentally recovered from the nuclear fraction. To avoid a situation in which the nuclear delivery of the Ad genome was overestimated, we estimated the efficiency of contamination of Ad in the nucleus during the nuclear isolation with cells that were incubated with Ad at 4°C, at which temperature nuclear delivery was largely excluded. After incubation at 4°C for 1 h, we fractionated the nucleus and quantified cell surface-bound and nucleus-associated Ad. As a result, we calculated the percentage of nuclear Ad to cell surface-bound Ad to be 17.6%. Thus, 17.6% of the nuclear fraction, corresponding to contamination during the isolation process, was subtracted from the experimentally determined nuclear fraction after the incubation at 37°C (54.2%). As a result, we calculated that 36.6% of the total cellular uptake (7.3 copies/nucleus) of DNA reached the nucleus (Fig. 5). Other portions (12.7 copies/cell; 63.4% of the total cellular uptake) were recovered from the supernatant fractions (S_PCR(sup)), which include the endosome/lysosome and cytosol fractions. A CIDIQ analysis showed that the fractions in the endosome/lysosome (F'(end/lys)) and the cytosol (F'(cyt)) compared to the supernatant were 47.9 and 52.1%, respectively. Taking these data into consideration, we calculated the numbers of Ad gene copies in the endosome/lysosome and cytosol as 6.1 copies/cell (30.3% of the total cellular uptake) and 6.6 copies/cell (33.1% of the total cellular uptake), respectively (Figs. 4D and 5). The efficiency of the endosomal escape and nuclear translocation calculated from the data shown in the circle graph in Fig. 5 is summarized in Table 1. We calculated the efficiency of endosomal escape as the fraction that escaped from the endosome/lysosome (nuclear fraction plus cytoplasmic fraction) divided by the total cellular uptake. Similarly, we determined the efficiency of nuclear translocation as the nuclear fraction divided by the fraction that escaped from the endosome/lysosome (nuclear fraction plus cytoplasmic fraction). As a result, the efficiency of endosomal escape is only slightly higher for Ad. In contrast, the efficiency of nuclear translocation is considerably higher for Ad.

Table 1 - Comparison of the efficiencies of endosomal escape and nuclear translocation between Ad and LFN.

Full table

Finally, comparing the nuclear delivery of DNA, 5600-fold more gene copies are delivered to the nucleus in the case of LFN. We calculated transcription efficiency as the expression divided by the gene copies in the nucleus as shown in Fig. 5. It was shown that Ad is 8100 times more efficient than LFN in nuclear transcription.

This conclusion is also applicable to HeLa cells (Table 2). To exhibit a comparable transgene expression, LFN also requires 3 orders of magnitude more gene copies (200 vs 1.4 10⁶ copies/cell). Under these conditions, approximately 4100 times more gene copies reach the nucleus, and therefore, the transcription efficiency of Ad was approximately 7000 times higher than that of LFN.

Table 2 - Comparison of nuclear delivery and transcription efficiency between Ad and LFN in HeLa cells.

Full table

Discussion

In the present study, intracellular trafficking and the intranuclear transcription of exogenous DNA transfected by viral and nonviral vectors were quantitatively evaluated. Ad and LFN were used as model vectors in the comparison, since they are both highly potent and widely used vectors. As a result, we found that LFN can accomplish transgene expression comparable to that of the Ad vector, when the protocol is optimized. However, the dose of pDNA in terms of luciferase gene copies under these conditions was 3 orders of magnitude more than that of Ad. Therefore, it would be worthwhile to clarify which of the intracellular processes is rate limiting for LFN.

First, the cellular uptake process was compared. Calculated from the applied dose and cellular uptake (Fig. 3), LFN exhibited significantly higher cellular uptake efficiency than Ad (approximately 45% vs 10%). Further comparison of the cellular binding efficiency by incubation at 4°C (Fig. 3) indicated that the efficient cellular uptake in the case of LFN can be attributed to efficient cellular binding, but not to the internalization rate constant (k_int). The superior cellular surface binding in the case of LFN can be explained by differences in the binding mechanism. Binding of the Ad is based on specific ligand–receptor interactions between the fiber protein and the coxsackievirus and Ad receptor and between the RGD motif in the penton base and the integrin receptors¹⁵. Therefore, the maximum binding is dependent on the total number of these receptors. In contrast, the pDNA/LFN complex can bind to the entire cell surface area via electrostatic interactions. Concerning the k_int value, it has recently been reported that Ad enters, not only via clathrin-mediated endocytosis^16,17, but also via macropinocytosis^18,19, which is actively driven by signal transduction after the binding of the RGD motif to the integrin receptor for cellular entry^20,21,22. Therefore, the internalization of Ad appears to be a highly efficient process. The k_int value of LFN was comparable to that of Ad. A recent study indicated that a certain type of nonviral vector was taken up by cells, not only by the classical endocytosis pathway^23,24, but also by another pathway such as macropinocytosis²⁵. Presumably, multiple pathways are also responsible to the cellular uptake of LFN, and this results in an internalization rate constant comparable to that of Ad. Concerning intracellular trafficking, various types of cellular uptake processes must be considered. Since only acidic compartments were stained by LysoSensor DND-189, the involvement of nonacidic vesicular compartments such as caveola was excluded from the analysis. To analyze the contribution of acidic vesicular transport (i.e., clathrin-mediated endocytosis and macropinocytosis²⁶) to total vesicular transport, plasma membranes were nonspecifically labeled with PKH-26 (Sigma). After incubation for 1 h, all of the labels on the plasma membrane were internalized and detected as clustered forms. Dual staining with PKH-26 and LysoSensor DND-189 revealed that approximately 70% of the PKH-26 clusters were colocalized with LysoSensor DND-189, suggesting that a major part (70%) of the vesicular transport system was acidic compartments (data not shown). In addition, it appears that the size of the lipoplex (generally more than 200 nm^27,28) is too large to be taken up via nonacidic vesicular transport such as caveolin-mediated endocytosis (60 nm) and clathrin- and caveolin-independent endocytosis (90 nm)²⁹. Therefore, we conclude that the staining of vesicular compartments by LysoSensor DND-189 is useful for tracing the main intracellular trafficking of vectors, although it is possible that pDNA taken up via a nonacidic compartment may unexpectedly participate in the efficient trafficking pathway. The mechanism for the efficient internalization rate of the LFN remains to be clarified.

Concerning endosomal escape, it was considered that the dismantling of Ad particles in response to the low endosomal pH is closely related to this event^16,30. Wiethoff et al. have recently shown that partial disassembly of the Ad capsid triggers the release of protein VI, which then lyses the endosomal membrane structure³¹. LFN also had an efficiency comparable to that of adenovirus (Table 1). LFN consists of polycationic lipid 2,3-dioleyloxy-N-[2(spermincarboxamido)ethyl]-N,N-dimethyl-1-propanaminium trifluoroacetate (DOSPA) and corn-shape lipid dioleoyl phosphatidylethanolamine (DOPE). A highly efficient endosomal escape of LFN, as high as that of Ad, was then synergistically achieved by the proton sponge effect³² derived from secondary amines in DOSPA, along with the fusogenic effect^33,34 of DOPE, whereas their strategies were different.

In contrast, once it escapes into the cytosol, adenovirus delivers its DNA to the nucleus more efficiently than LFN. Although LFN can rapidly deliver its DNA to the nucleus presumably due to the electric interaction of cationic lipid and negatively charged lipids of the nuclear membrane³⁵, it is likely that the cytoplasmic delivery of Ad is more sophisticated. It has previously been shown that Ad utilizes a microtubule network to pass through the cytoplasm into the nucleus^36,37. Furthermore, Ad binds to the nuclear pore complex receptor CAN/Nup214 and, thereafter, inserts the genomic DNA into the nucleus with the assistance of nuclear histone H1 cells and the importin family proteins of the host cells^11,38. Such types of multiple/active delivery systems in adenovirus may be responsible for its efficient nuclear delivery.

Finally, the intranuclear transcription of DNA, when transfected with Ad, is 8100 times higher than that of LFN. Considering that the CMV promoter, luciferase (GL3), and BGH polyadenylation sequences in pDNA and the Ad genome are identical to one another, two explanations are possible. One is that the nuclear DNA introduced by the LFN is so well condensed that the transcription process is inhibited. It is generally accepted that the release of pDNA from the vectors is the rate-determining process for transgene expression³⁹. In fact, transgene expression after the nuclear microinjection of DNA as a lipoplex was reported to be severely limited^28,40. The other possibility is that the Ad genome structure and/or proteins coded in the Ad genome affect transgene expression. Since almost all of the E1/E3 region was deleted from the Ad genome, other factors such as the inverted terminal repeat sequence⁴¹, terminal protein⁴², and various proteins derived from the E4 region^43,44 may be involved in the improvement in transcription and nuclear stability. Currently, it has been reported that terminal proteins interact with the nuclear matrix, where they play an important role in nuclear transcription⁴². To evaluate the potential of pDNA vis-a-vis adenovirus DNA in the transcription process, adenoviral DNA was purified by treatment with guanidine, followed by sucrose gradient centrifugation⁴⁵. The nuclear microinjection of 10 copies of plasmid DNA and Ad genome encoding the green fluorescent protein revealed that the Ad genome exhibited only a slightly higher transgene expression compared with plasmid DNA (approximately 35% vs 25%), suggesting that these two types of DNA are equivalent in the transcription process (S. Hama et al., unpublished observation). As a result, the difference in decondensation in the nucleus is a more plausible hypothesis for explaining the difference in intranuclear transcription.

The remarkable difference in transcription activity was also found in HeLa cells, indicating that this phenomenon is generally applicable to various cells. Very recently, it was also shown that the intranuclear transcription of plasmid DNA introduced by nonviral vectors such as polyethylenimine was much lower than that of viral vectors. This conclusion is then generally applicable to various types of current nonviral vectors⁴⁶.

Collectively, we were able successfully to quantify the intracellular trafficking and intranuclear transcription of viral and nonviral vectors. LFN delivers pDNA to the nucleus with a comparable speed and exhibits transgene expression comparable to that of the Ad vector. However, LFN requires 3 orders of magnitude more pDNA than the Ad to exert similar transfection activities. Surprisingly, this remarkable difference principally arises from differences in nuclear transcription efficiency. This is the first systematic and quantitative comparison of the intracellular distribution of the exogenous genes introduced by a viral vector and a nonviral vector. Such quantitative comparisons will be useful for developing new generations of nonviral vectors.

Materials and methods

General

To prepare the reporter gene vector for the pDNA (pcDNA3.1-GL3), an insert fragment encoding luciferase (GL3) was obtained by HindIII/XbaI digestion of the pGL3-Basic vector (Promega, Madison, WI, USA) and ligated to the HindIII/XbaI-digested site of pcDNA3.1 (Invitrogen, Carlsbad, CA, USA). LFN was from Invitrogen Corp. Other chemicals used were commercially available and reagent grade products. As to the viral vector, the E₁^-, E₃^-, replication-deficient serotype 5 adenovirus, in which an expression cassette is inserted in the E₁ position, was used⁴⁷. The expression cassette consists of a cytomegalovirus promoter/enhancer, cDNA encoding luciferase (GL3), and BGH polyadenylation sequences, which are also encoded in the pDNA used in the LFN-mediated transfection.

Quantification of luciferase gene copies by TaqMan PCR

To determine the applied particle titer of the Ad in terms of copy number of luciferase gene, Ad was dismantled by vortexing with 0.1% SDS/10 mM Tris–HCl/5 mM EDTA treatment. The number of luciferase gene copies was determined by TaqMan PCR with a 100-fold diluted sample solution. As a reference, a dilution series of pDNA3.1-GL3 was run along with the virus sample. It was confirmed that 0.001% SDS and 0.05 mM EDTA had no effect on the luciferase gene amplification by TaqMan PCR. The luciferase gene region of Ad genomic DNA or pDNA was amplified and quantified by TaqMan PCR (ABI Prism 7700 sequence detection system; Applied Biosystems). The sequence of the probe was 5'-CCGCTGAATTGGAATCCATCTTGCTC-3' with FAM as a fluorescent dye on the 5' end and TAMRA as a fluorescence quencher dye labeled on the 3' end. This probe is designed to anneal to the target between the sense primer (5'-TTGACCGCCTGAAGTCTCTGA-3') and the antisense primer (5'-ACACCTGCGTCGAAGATGTTG-3') in the luciferase sequence of the Ad genome and pDNA.

For the quantification of cellular uptake and nuclear delivery of the luciferase genes, DNA was purified from cell lysates or an isolated nucleus by means of a GenElute Mammalian Genome DNA Miniprep kit (Sigma–Aldrich, St. Louis, MO, USA) and subjected to the TaqMan PCR with ABI Prism 7700 sequence detection system. The isolation of nuclei followed a previous report^6,7. Briefly, cells were suspended in 0.5 ml of lysis buffer (0.5% Nonidet P-40, 10 mM NaCl, 3 mM MgCl₂, 10 mM Tris–HCl buffer; pH 7.4) to dissolve the plasma membrane, and the nuclear fraction was then isolated by centrifugation at 1400g for 5 min. This treatment (washing) was repeated twice, and the final pellet was used as the nuclear fraction. The high recovery of the Ad genome (>90%) by DNA purification with the GenElute Mammalian Genome DNA Miniprep kit was confirmed by comparison with the luciferase gene amplification by TaqMan PCR.

The number of -actin DNAs was also determined by the ABI Prism 7700 sequence detection system. PCR was performed according to the manufacturer's instructions with 0.5 M each forward, 5'-TGCGTGACATTAAGGAGAAGCTGTG-3', and reverse, 5'-CAGCGGAACCGCTCATTGCCAATGG-3', primers and the QuantiTect SYBR Green PCR Master Mix (Qiagen, Hilden, Germany). A linear relationship between the number of cells and the threshold cycle for the -actin gene amplification was confirmed (data not shown).

CIDIQ analysis

In the case of LFN-mediated transfection, the pDNA fraction in the endosome/lysosome, cytosol, and nucleus compared to the total cellular association was assessed by CIDIQ, as described in a recent report⁸. After the transfection of rhodamine-labeled pDNA, the endosome/lysosome and nucleus were stained with LysoSensor DND-189 and Hoechst 33342, respectively, to discriminate the subcellular localization of pDNA. Each 8-bit TIFF image was transferred to Image-Pro Plus version 4.0 (Media Cybernetics, Inc.) to quantify the total brightness and pixel area of each region of interest. For data analysis, the pixel areas of each cluster in endosomes/lysosomes, s_i(end/lys); cytosol, s_i(cyt); and nucleus, s_i(nuc) were separately summed for each x–y plane and are denoted as S'_{Z = j}(end/lys), S'_{Z = j}(cyt), and S'_{Z = j}(nuc), respectively. The values of S'_{Z = j}(end/lys), S'_{Z = j}(cyt), and S'_{Z = j}(nuc) in each x–y plane were further summed and are denoted as S(end/lys), S(cyt), and S(nuc), respectively. These parameters represent the total amount of pDNA in each compartment in an individual cell. Furthermore, the total area of the pDNA, denoted as S(tot), was calculated by integrating the S(end/lys), S(cyt), and S(nuc). This value represents the total cellular uptake of pDNA. The fractions of pDNA present in endosomes/lysosomes, cytosol, and nucleus compared to the whole cell are denoted as F(end/lys), F(cyt), and F(nuc), which were calculated as S(end/lys), S(cyt), and S(nuc) divided by S(tot), respectively.

For quantification of the intracellular distribution of Ad, Texas red-labeled Ad was used in the CIDIQ analysis. The nuclear delivery of Ad DNA (S_PCR(nuc)) was quantified beforehand by nuclear fractionation, followed by the TaqMan PCR⁶. The supernatant fraction S_PCR(sup) contains Ad DNA in the endosome/lysosome and cytosol. Therefore, the fractions in these organelles were quantified by CIDIQ. In this case, S(sup), which represents the Ad DNA in the supernatant fraction in the nucleus isolation experiment, was calculated by integrating the S'(end/lys) and S'(cyt). The fractions of Ad DNA present in endosomes/lysosomes and cytosol compared to the supernatant are denoted as F'(end/lys) and F'(cyt), which were calculated as S'(end/lys) and S'(cyt) divided by S(sup), respectively. Numbers of gene copies in these organelles were determined by multiplying each fraction by S_PCR(sup).

References

1. Lai, C. M., Lai, Y. K. and Rakoczy, P. E. (2002). Adenovirus and adeno-associated virus vectors. DNA Cell Biol. 21: 895–913. | Article | PubMed | ChemPort |
2. Yoshida, J., et al. (2004). Human gene therapy for malignant gliomas (glioblastoma multiforme and anaplastic astrocytoma) by in vivo transduction with human interferon beta gene using cationic liposomes. Hum. Gene Ther. 15: 77–86. | Article | PubMed | ISI | ChemPort |
3. Friedmann, T. and Roblin, R. (1972). Gene therapy for human genetic disease?. Science. 175: 949–955. | Article | PubMed | ISI | ChemPort |
4. Dinser, R., et al. (2001). Comparison of long-term transgene expression after non-viral and adenoviral gene transfer into primary articular chondrocytes. Histochem. Cell Biol. 116: 69–77. | PubMed | ChemPort |
5. Oh, Y. K., et al. (2002). Polyethylenimine-mediated cellular uptake, nucleus trafficking and expression of cytokine plasmid DNA. Gene Ther. 9: 1627–1632. | Article | PubMed | ChemPort |
6. Tachibana, R., et al. (2002). Quantitative analysis of correlation between number of nuclear plasmids and gene expression activity after transfection with cationic liposomes. Pharm. Res. 19: 377–381. | Article | PubMed | ChemPort |
7. Tachibana, R., Harashima, H., Shinohara, Y. and Kiwada, H. (2001). Quantitative studies on the nuclear transport of plasmid DNA and gene expression employing nonviral vectors. Adv. Drug Delivery Rev. 52: 219–226. | Article | ChemPort |
8. Akita, H., Ito, R., Khalil, I. A., Futaki, S. and Harashima, H. (2004). Quantitative three-dimensional analysis of the intracellular trafficking of plasmid DNA transfected by a nonviral gene delivery system using confocal laser scanning microscopy. Mol. Ther. 9: 443–451. | Article | PubMed | ChemPort |
9. Masuda, T., Akita, H. and Harashima, H. (2005). Evaluation of nuclear transfer and transcription of plasmid DNA condensed with protamine by microinjection: the use of a nuclear transfer score. FEBS Lett. 579: 2143–2148. | Article | PubMed | ChemPort |
10. Moriguchi, R., et al. (in press). Non-linear pharmacodynamics in a non-viral gene delivery system: positive non-linear relationship between dose and transfection efficiency. J. Controlled Release.
11. Trotman, L. C., Mosberger, N., Fornerod, M., Stidwill, R. P. and Greber, U. F. (2001). Import of adenovirus DNA involves the nuclear pore complex receptor CAN/Nup214 and histone H1. Nat. Cell Biol. 3: 1092–1100. | Article | PubMed | ISI | ChemPort |
12. Leopold, P. L., et al. (1998). Fluorescent virions: dynamic tracking of the pathway of adenoviral gene transfer vectors in living cells. Hum. Gene Ther. 9: 367–378. | PubMed | ISI | ChemPort |
13. Miyazawa, N., et al. (1999). Fiber swap between adenovirus subgroups B and C alters intracellular trafficking of adenovirus gene transfer vectors. J. Virol. 73: 6056–6065. | PubMed | ISI | ChemPort |
14. Nakano, M. Y. and Greber, U. F. (2000). Quantitative microscopy of fluorescent adenovirus entry. J. Struct. Biol. 129: 57–68. | Article | PubMed | ISI | ChemPort |
15. Meier, O. and Greber, U. F. (2003). Adenovirus endocytosis. J. Gene Med. 5: 451–462. | Article | PubMed | ChemPort |
16. Greber, U. F., Willetts, M., Webster, P. and Helenius, A. (1993). Stepwise dismantling of adenovirus 2 during entry into cells. Cell. 75: 477–486. | Article | PubMed | ISI | ChemPort |
17. Varga, M. J., Weibull, C. and Everitt, E. (1991). Infectious entry pathway of adenovirus type 2. J. Virol. 65: 6061–6070. | PubMed | ISI | ChemPort |
18. Imelli, N., Meier, O., Boucke, K., Hemmi, S. and Greber, U. F. (2004). Cholesterol is required for endocytosis and endosomal escape of adenovirus type 2. J. Virol. 78: 3089–3098. | Article | PubMed | ChemPort |
19. Meier, O., et al. (2002). Adenovirus triggers macropinocytosis and endosomal leakage together with its clathrin-mediated uptake. J. Cell Biol. 158: 1119–1131. | Article | PubMed | ISI | ChemPort |
20. Amyere, M., et al. (2000). Constitutive macropinocytosis in oncogene-transformed fibroblasts depends on sequential permanent activation of phosphoinositide 3-kinase and phospholipase C. Mol. Biol. Cell. 11: 3453–3467. | PubMed | ISI | ChemPort |
21. Araki, N., Johnson, M. T. and Swanson, J. A. (1996). A role for phosphoinositide 3-kinase in the completion of macropinocytosis and phagocytosis by macrophages. J. Cell Biol. 135: 1249–1260. | Article | PubMed | ISI | ChemPort |
22. Li, E., Stupack, D., Klemke, R., Cheresh, D. A. and Nemerow, G. R. (1998). Adenovirus endocytosis via alpha(v) integrins requires phosphoinositide-3-OH kinase. J. Virol. 72: 2055–2061. | PubMed | ISI | ChemPort |
23. Zhou, X. and Huang, L. (1994). DNA transfection mediated by cationic liposomes containing lipopolylysine: characterization and mechanism of action. Biochim. Biophys. Acta. 1189: 195–203. | PubMed | ISI | ChemPort |
24. Zuhorn, I. S., Kalicharan, R. and Hoekstra, D. (2002). Lipoplex-mediated transfection of mammalian cells occurs through the cholesterol-dependent clathrin-mediated pathway of endocytosis. J. Biol. Chem. 277: 18021–18028. | Article | PubMed | ISI | ChemPort |
25. Goncalves, C., et al. (2004). Macropinocytosis of polyplexes and recycling of plasmid via the clathrin-dependent pathway impair the transfection efficiency of human hepatocarcinoma cells. Mol. Ther. 10: 373–385. | Article | PubMed | ISI | ChemPort |
26. Wadia, J. S., Stan, R. V. and Dowdy, S. F. (2004). Transducible TAT-HA fusogenic peptide enhances escape of TAT-fusion proteins after lipid raft macropinocytosis. Nat. Med. 10: 310–315. | Article | PubMed | ISI | ChemPort |
27. Joshee, N., Bastola, D. R. and Cheng, P. W. (2002). Transferrin-facilitated lipofection gene delivery strategy: characterization of the transfection complexes and intracellular trafficking. Hum. Gene Ther. 13: 1991–2004. | Article | PubMed | ChemPort |
28. Zabner, J., Fasbender, A. J., Moninger, T., Poellinger, K. A. and Welsh, M. J. (1995). Cellular and molecular barriers to gene transfer by a cationic lipid. J. Biol. Chem. 270: 18997–19007. | Article | PubMed | ISI | ChemPort |
29. Conner, S. D. and Schmid, S. L. (2003). Regulated portals of entry into the cell. Nature. 422: 37–44. | Article | PubMed | ISI | ChemPort |
30. Medina-Kauwe, L. K. (2003). Endocytosis of adenovirus and adenovirus capsid proteins. Adv. Drug Delivery Rev. 55: 1485–1496. | Article | ChemPort |
31. Wiethoff, C. M., Wodrich, H., Gerace, L. and Nemerow, G. R. (2005). Adenovirus protein VI mediates membrane disruption following capsid disassembly. J. Virol. 79: 1992–2000. | Article | PubMed | ISI | ChemPort |
32. Boussif, O., et al. (1995). A versatile vector for gene and oligonucleotide transfer into cells in culture and in vivo: polyethylenimine. Proc. Natl. Acad. Sci. USA. 92: 7297–7301. | Article | PubMed | ChemPort |
33. Bailey, A. L. and Cullis, P. R. (1997). Membrane fusion with cationic liposomes: effects of target membrane lipid composition. Biochemistry. 36: 1628–1634. | Article | PubMed | ChemPort |
34. Wattiaux, R., Jadot, M., Warnier-Pirotte, M. T. and Wattiaux-De Coninck, S. (1997). Cationic lipids destabilize lysosomal membrane in vitro. FEBS Lett. 417: 199–202. | Article | PubMed | ChemPort |
35. Daum, G. (1985). Lipids of mitochondria. Biochim. Biophys. Acta. 822: 1–42. | PubMed | ChemPort |
36. Suomalainen, M., Nakano, M. Y., Boucke, K., Keller, S. and Greber, U. F. (2001). Adenovirus-activated PKA and p38/MAPK pathways boost microtubule-mediated nuclear targeting of virus. EMBO J. 20: 1310–1319. | Article | PubMed | ISI | ChemPort |
37. Suomalainen, M., et al. (1999). Microtubule-dependent plus- and minus end-directed motilities are competing processes for nuclear targeting of adenovirus. J. Cell Biol. 144: 657–672. | Article | PubMed | ISI | ChemPort |
38. Greber, U. F., et al. (1997). The role of the nuclear pore complex in adenovirus DNA entry. EMBO J. 16: 5998–6007. | Article | PubMed | ISI | ChemPort |
39. Bieber, T., Meissner, W., Kostin, S., Niemann, A. and Elsasser, H. P. (2002). Intracellular route and transcriptional competence of polyethylenimine–DNA complexes. J. Controlled Release. 82: 441–454. | Article | ChemPort |
40. Pollard, H., et al. (1998). Polyethylenimine but not cationic lipids promotes transgene delivery to the nucleus in mammalian cells. J. Biol. Chem. 273: 7507–7511. | Article | PubMed | ISI | ChemPort |
41. Hatfield, L. and Hearing, P. (1991). Redundant enhancements in the adenovirus type 5 inverted terminal repeat promote bidirectional transcription in vitro and are important for virus growth in vivo. Virology. 184: 265–276. | Article | PubMed | ChemPort |
42. Fredman, J. N. and Engler, J. A. (1993). Adenovirus precursor to terminal protein interacts with the nuclear matrix in vivo and in vitro. J. Virol. 67: 3384–3395. | PubMed | ISI | ChemPort |
43. Endter, C. and Dobner, T. (2004). Cell transformation by human adenoviruses. Curr. Top. Microbiol. Immunol. 273: 163–214. | PubMed | ISI | ChemPort |
44. Tauber, B. and Dobner, T. (2001). Adenovirus early E4 genes in viral oncogenesis. Oncogene. 20: 7847–7854. | Article | PubMed | ISI | ChemPort |
45. Sharp, P. A., Moore, C. and Haverty, J. L. (1976). The infectivity of adenovirus 5 DNA–protein complex. Virology. 75: 442–456. | Article | PubMed | ChemPort |
46. Varga, C. M., et al. (2005). Quantitative comparison of polyethylenimine formulations and adenoviral vectors in terms of intracellular gene delivery processes. Gene Ther. 12: 1023–1032. | Article | PubMed | ChemPort |
47. Mizuguchi, H., et al. (2001). A simplified system for constructing recombinant adenoviral vectors containing heterologous peptides in the HI loop of their fiber knob. Gene Ther. 8: 730–735. | Article | PubMed | ChemPort |

Acknowledgements

This work was supported in part by Grants-in-Aid for Scientific Research (B) and Grant-in-Aid for Young Scientists (B) from the Ministry of Education, Culture, Sports, Science, and Technology of Japan and by Grants-in-Aid for Scientific Research on Priority Areas from the Japan Society for the Promotion of Science.