|Home | About | Journals | Submit | Contact Us | Français|
Selective gene transfer to tumor is critical in cancer gene therapy. We previously used ionizing radiation to improve adenovirus uptake in intrahepatic tumors but liver cytotoxicity associated with the viral administration still occurred. Here, we explore the potential of radiation for improving gene delivery by a virus-mimicking nanoparticle, transferrin (Tf)-cationic liposome-DNA complex (Tf-lipoplex). Transduction by Tf-lipoplex was highly efficient in various cell lines and further increased by radiation in a dose- and time-dependent manner. This radiation induction, which was associated with an increase in Tf-lipoplex uptake (3- to 4-folds in hepatocytes WB and lung cancer cells, LLC1), was absent when a Tf-deficient complex was used or abolished by the presence of free Tf, suggesting that Tf receptor (TfR) interaction is required for radiation induction. Radiation (10–20 Gy) markedly induced transgene (LacZ) expression in LLC1 xenografts (3.5- to 7.4-folds), correlating with increased plasmid content and TfR expression in irradiated tumors. Moreover, Tf-lipoplex-mediated gene expression was not observed in the liver or other normal tissues regardless of radiation treatment. We conclude that radiation improves Tf-lipoplex gene delivery selectively to tumor cells both in vitro and in vivo. Our findings may provide insight in developing ligand-specific lipoplex for molecularly targeted cancer gene therapy.
Advances in developing viral vectors have greatly raised the hope of gene therapy for cancer during the past decade, but these vectors still have limitations. In general, adenovirus is highly efficient in transducing targeted cells but lacks gene integration capability, thus requiring multiple administrations that raises concerns about toxicity resulting from an immune response to viral antigens.1,2 Adeno-associated virus is less antigenic than adenovirus, but is lower in viral production and packaging capacity. Lentivirus can efficiently and stably transduce cells but is less selective for in-vivo delivery and carries a risk of producing cancer due to its potential for self-replication. Efforts have been made to improve vector targeting and selective transgene expression in tumors. In the treatment of colon cancer liver metastases, a disease that continues to kill approximately 15 000 patients a year, viral vector-mediated gene therapy has advantages in treating diffuse or unresectable liver tumors due to its high efficiency, but is limited by nonselective gene delivery to normal liver that causes toxicity. Previously, we have developed a strategy that uses an adenovirus vector containing the carcinoembryonic antigen (CEA) promoter for selective expression of yeast cytosine deaminase (yCD) in intrahepatic colon cancer metastases. Although we observed improved radiosensitization when a selectively expressed yCD converted the prodrug 5FC into 5FU in tumors, there was also significant toxicity in normal liver due to the nonspecific targeting of adenovirus in liver.3 The need for multiple viral administrations also raises concerns about hormonal immune response to viral antigens that could be potentially lethal.1,2 Development of a nonimmunogenic vector with improved tumor targeting remains a challenge in this field.
These limitations of viral methods have stimulated interest on nonviral vectors. One such approach uses liposome- and polymer-based particles, which possess safety features compared to viral approaches and have, therefore, become preferred in a variety of clinical trials. For instance, a specially formulated cationic liposome has demonstrated some effectiveness for p53 delivery in vivo; however, it may suffer from low gene transfer efficiency, large particle size, poor stability and the lack of targeting capability.4 The recent development of a liposome vector that incorporates human transferrin (Tf) ligand into the complex (Tf-lipoplex) has markedly improved its efficiency and tumor-targeting ability due to the high expression of Tf receptor (TfR) in prostate, breast, pancreas, lung and colon cancers.5–10 Furthermore, Tf-lipoplex has been shown to achieve long-term therapeutic efficacy in systemic p53 gene therapy for human head and neck cancer and prostate cancer, without compromising safety.11–13
Radiation-combined gene therapy has become an increasingly promising approach in cancer treatment. One of such advance is TNFerade, which uses radiation to induce tumor necrosis factor-α expression spatially and temporally in tumors by replication-deficient adenovirus vector containing a radiation-inducible promoter from the egr-1 gene.14,15 Our own study has demonstrated that radiation induces adenovirus uptake in tumors, contributing to enhanced transgene expression.16,17 For liposome-mediated delivery, radiation has been shown to increase drug delivery to tumor blood vessels18 and to sustain an enhanced transgene expression in human breast cancer cells.19 In addition, liposome-mediated delivery of the cytosine deaminase gene has resulted in radiosensitization of rectal cancer.20 However, the mechanism of radiation induction in liposome-mediated gene transfer is not well understood.
Our previous studies and new strategies for delivery of radiation led us to hypothesize that radiation could improve liposome uptake. First, radiation induces adenovirus gene transfer in tumor cells and xenografts by increasing viral uptake through dynamin 2-mediated membrane action.17 Second, Tf-lipoplex, a highly compact nanoparticle with a relatively uniform size of 50–90 nm, structurally resembles that of a virus with a dense core enveloped by a membrane coated with Tf molecules spiking the surface.21 Furthermore, this strategy is becoming increasingly attractive with the development of body stereotactic radiation, which permits high doses of radiation to be selectively administered to tumors compared to normal tissues. Therefore, we investigated the potential for radiation to improve liposome uptake, and thus, gene transfer in tumors. When we found that radiation did increase Tf-lipoplex-mediated gene transfer in tumor cells, we carried out additional mechanistic studies and explored the improvement of gene transfer in a tumor xenograft model.
Rat hepatocytes (WB) were provided by Dr James Trosko (Michigan State University). All other cell lines were obtained from ATCC (Manassas, VA) and were maintained in RPMI media containing 10% fetal bovine serum and antibiotics at 37 °C with 5% CO2. These cells are human cancer cell lines from lung (H1299), liver (HepG2), prostate (PC3) and colon (LoVo), and mouse cancer cell lines for lung (LLC1), colon (CT26) and mouse alveolar macrophage (MH-S).
Cationic liposome consisting of dioleoyl trimethylammonium propane and dioleoyl phosphatidylethanolamine (Avanti Polar Lipids, Alabaster, AL) at a 1:1 molar ratio were prepared by sonication as previously described.15 For in-vitro studies, the Tf-lipoplex was prepared in serum-free culture media to achieve the ratio of Tf:liposome: DNA at 12.5:10:1 (μg:nmol:μg), respectively. Tf (5mgml−1, iron-saturated holo-transferrin; Sigma, St Louis, MO) and was first incubated with liposome in 100 μl of media for 5 min followed by mixing with 100 μl of media containing plasmid DNA cytomegalovirus-green fluorescent protein (CMV-GFP). For in-vivo studies, the complex was formed using plasmid DNA for CMV-LacZ according to the formulation described previously.13 Briefly, 25 μl of Tf, 50μl of liposome (2mM total lipids) and 75 μl of water were mixed and incubated at room temperature for 10 min with agitation. DNA (10 μg) were added into 120 μl of 20mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid buffer (pH 7.4) and suspension was mixed thoroughly followed by incubation at room temperature for 20min with agitation. Thirty microliters of 50% dextrose solution was then added to achieve the ratio of DNA:lipid:Tf at 1:10:12.5 (μg:nmol:μg), respectively.
Radiation was administered using Philips RT 250 Orthovoltage unit (Philips, Bothell, WA) producing 250 kV X-rays at ~2Gy min−1. Radiation doses were calibrated using an electrometer system directly traceable to the National Institute of Standards and Technology standard.
C57BL/6 mice (The Jackson Laboratory, Bar Harbor, Maine) were implanted subcutaneously in the flanks with 5×105 of LLC1 cells for 3 weeks or until tumors reached ~64mm3 in size. At 24 h after radiations doses of (0, 5, 10 or 20 Gy) in tumor region mice were injected through the tail vein with Tf-lipoplex containing plasmid DNA (500 μg in the volume of 1 ml) for CMV-LacZ. In separate experiments, tumors were irradiated with 10 Gy followed by injection of the complex immediately (0 h), 24 or 48 h after radiation treatment. At 72 h after complex administration, mice were perfused with saline through the portal vein and inferior vena cava to remove remaining complex in circulation. Tumors and normal tissues (liver, lung, spleen and duodenum) were assessed histologically for x-gal staining as described previously,1 and x-gal-positive cells were counted in five randomly-selected areas by two individuals. All tissues were also collected for real-time PCR assay to measure plasmid DNA content.
Cells were plated on 24-well plates at a density of 2×104 cells per well overnight before being irradiated with various doses (0, 1, 2, 4 or 8 Gy). Cells were transfected with liposome complex conjugated with Tf or without Tf (Tf-lipoplex or lipoplex) immediately (0 h) or at 3 or 24 h after radiation followed by continued culture. The amount of lipoplex was optimized from a pilot experiment for optimal counting and was determined by the amount of plasmid DNA (CMV-GFP, ng per well) for each cell line as follows: WB (5.0 of 25.0), LLC1 (5.0 of 25.0), H1299 (10.0 of 50.0), HepG2 (10.0 of 50.0), CT26 (3.0 of 15.0), PC3 (30.0 of 150.0) and LoVo (0.5 of 2.5). GFP-positive cells were counted using fluorescence microscopy at 48 h after transfection. Radiation induction was calculated as the fold increase compared to nonirradiated cells that were set as ‘1’.
For cell binding, WB and LLC1 cells were first incubated with Tf-lipoplex on ice for 2 h prior to washing for removal of free complex in culture media. Cells prebound with the complex (pre-RT) were then irradiated with various doses (0, 2, 4 or 8 Gy) followed by continued culture at 37 °C for 48 h for quantification of GFP-positive cells. Radiation induction was calculated as the fold increase described above and was compared with cells receiving radiation treatment prior to transfection (post-RT).
To determine specific binding of Tf-lipoplex to cells, WB and LLC1 cells were pretreated with excess free Tf (0, 3, 8 and 40 times) for 1 h before washing to remove free Tf. Cells were then irradiated with 4Gy followed by immediate transfection with Tf-lipoplex for an additional 48 h. GFP-positive cells were counted as described above and were compared to nonirradiated cells.
WB and LLC1 cells were plated and grown overnight on cover slips. Cells were irradiated with 4Gy followed by immediate incubation with SP-DiOC18-labeled (Molecular Probes, Eugene, OR) Tf-lipoplex containing 150 ng of plasmid DNA for CMV-LacZ. After a 1-h incubation, the cells were washed to remove free complex, cell-associated particles (green) were visualized by fluorescence microscopy (Olympus microscope and Olympus MicroSuite Imaging Software, Melville, NY). For in-vivo uptake, mice with or without subcutaneous LLC1 tumor xenografts were irradiated (whole body) prior to intravenous administration of the labeled Tf-lipoplex. Mice were killed at 4, 12, 24, 48 and 72 h after injection for normal or tumor tissue collection after systemic perfusion with saline. Cryosections were analyzed under fluorescence microscopy for complex distribution.
LLC1 tumors in mice were irradiated with 10 or 20 Gy 4, 24 and 72 h prior to tissue collection and homogenization in protein sample buffer (Invitrogen, Carlsbad, CA). Total proteins (~15 μg) were fractionated on 10% acrylamide SDS–polyacrylamide gel electrophoresis (PAGE) gel followed by transfer onto Immobilon-P membrane (Milipore, Billerica, MA). TfR was detected by specific antibody at 1:500 dilution (Zymed Laboratories, South San Francisco, CA) and visualized using ECLluminescence reagent. Expression of β-actin detected by rabbit antibody (Sigma) was used as the internal control for equal sample loading. Specific bands for each protein were analyzed for their intensities by densitometry using NIH Image software. Expression level of TfR was normalized by comparison to the expression of β-actin and compared between nonirradiated and irradiated tumors.
To measure cell uptake of Tf-lipoplex, WB and LLC1 cells were plated on 24-well plates at a density of 1×105 cells per well overnight and were irradiated with 2 or 4 Gy followed by immediate incubation with Tf-lipoplex containing 30 ng of plasmid DNA (CMV-LacZ) for various times (15, 30, 60 and 120 min). After washing, cells were lysed in 200 μl of 1×Reporter Lysis Buffer (Promega, Madison, WI) for measurement of DNA content by real-time PCR. Radiation induction of cell uptake was calculated as the fold increase of plasmid DNA compared to nonirradiated cells that were set as ‘1’. To quantify plasmid DNA content in irradiated mice, total DNA was extracted from ~10 mg of tumors using DNeasy tissue kit (Qiagen, Santa Clarita, CA) and was subjected to real-time PCR analysis. The amount of plasmid DNA containing CMV-LacZ in irradiated tumors was compared to that in nonirradiated tumors and was represented as the fold increase. Each sample was also measured for α-actin level as an internal control. Real-time PCR was performed on an Opticon system with software (MJ Research, Waltham, MA) using QuantiTect SYBR Green PCR kit (Qiagen) combined with specific primers for: LacZ gene, 5′-CACGGCAGATAC ACTTGCTG and 3′-ATCGCCATTTGACCACTACC; α-actin gene, 5′-CGAGATCCCTCCAAAATCAA and 3′-TGTGGTCATGAGTCCTTCCA.
Data are presented as the means±the standard errors of four cell culture experiments or four mice for each condition. Densitometrical analysis of bands on western blot is performed in four individual experiments and the average density for each condition is used for comparison. Values were compared by analysis of variance and were considered significantly different when P<0.05.
To test radiation effects on lipoplex-mediated gene transfer, various cell lines were irradiated with 4Gy followed by immediate transfection with a Tf-lipoplex containing plasmid DNA for CMV-GFP gene. Radiation enhanced gene expression, as quantified by the fold increase of GFP-positive cells relative to nonirradiated cells. These inductions were greatest in WB (hepatocytes, 4.09±0.06) and LLC1 (Lewis lung carcinoma, 3.46±0.02), and were moderate in other cancer cell lines from lung (H1299, 2.19±0.03), liver (HepG2, 2.15±0.02), colon (CT26, 2.10±0.03) and prostate (PC3, 1.54±0.02) (Table 1). These inductions were abolished when a Tf-deficient complex (lipoplex) was used for transfection regardless of cell type (Table 1), suggesting that Tf-mediated receptor interaction was responsible for the radiation-induced increase.
To further assess radiation effects, the two most responsive cell lines, WB and LLC1, were treated with various radiation doses and at various times relative to radiation. The number of GFP-expressing cells was significantly increased at the dose of 2Gy in both cell lines by 2.2- and 2.7-fold (117±7 and 150±10 compared to 53±5 and 56±4 in nonirradiated WB and LLC1 cells, respectively) and was further enhanced with an increase of radiation dose, reaching plateau at 4Gy by 4.0- and 3.5- fold in the two cell lines, respectively (Figure 1a). These inductions were maximal when WB and LLC1 cells were transfected with Tf-lipoplex immediately after radiation (0 h) and were markedly reduced when transfections were delayed for 3 h (3.1±0.2- and 3.1±0.1-fold, respectively) and were diminished further in WB (1.0±0.3-fold) and LLC1 (1.9±0.2-fold) at 24-h-delayed transfection (Figure 1b). These results demonstrated that radiation induced Tf-lipoplex gene transfer in a dose- and timedependent fashion in cells, suggesting a radiation-related mechanism(s) in this induction.
To test whether radiation-induced gene transfer was due to an increased presence of complex, fluorescence-labeled Tf-lipoplex particles were incubated with irradiated WB and LLC1 cells followed by microscopic visualization of uptake and PCR quantification of plasmid DNA associated with cells. Radiation with 4Gy greatly enhanced Tf-lipoplex binding and uptake in cells evidenced microscopically by a higher density of particles in irradiated WB and LLC1 cells compared to nonirradiated (−RT) cells (Figure 2a). DNA content in cells was significantly increased by 2 and 4Gy of radiation at 60 min post-treatment inWB (3.9±0.2- and 3.0±0.3-fold, respectively) compared to nonirradiated control cells (1.1±0.1-fold), although these inductions appeared to be narrowed at 120 min due to possible saturation occurring at the basal level in nonirradiated cells (Figure 2b). Similarly, induction of DNA contents in LLC1 cells irradiated by 2 and 4Gy were most prominent at 60 min after radiation treatment (3.1±0.3- and 2.6±0.3-fold, respectively) (Figure 2c), suggesting that radiation could potentially enhance cellular uptake of Tf-lipoplex by improving complex binding to target cells.
To determine whether interaction between Tf and TfR may be regulated by radiation, cells were first incubated with Tf-lipoplex on ice to establish equal binding prior to internalization at 37 °C. We found that radiation inductions were absent in cells preincubated with complex containing CMV-GFP (pre-RT) regardless of radiation dose (Figure 3a). In contrast, transfection with complex immediately after radiation (post-RT) in WB cells induced GFP expression at 2Gy (2.2±0.3-fold), reaching a plateau at 4Gy (4.1±0.4-fold) (Figure 3a). Similar inductions were observed in LLC1 cells when Tf-lipoplex was given after radiation (post-RT) compared to cells preincubated with complex (pre-RT), maximizing at 4Gy (3.5±0.3- vs 1.3±0.2-fold) (Figure 3b). The fact that radiation failed to induce gene transfer from preexistant membrane-bound Tf-lipoplex may suggest that complex binding to cells is regulated by radiation.
To further confirm the role of Tf-TfR interaction during radiation, free Tf was used to compete with Tf-lipoplex binding to cells. As expected, the number of GFP-expressing cells was decreased in the presence of excess free Tf (3-, 8- and 40-folds relative to Tf from lipoplex) in nonirradiated WB (from 53±5 in control to 19±4, 13±4 and 7±1, respectively) and irradiated WB cells (from 217±15 in control to 108±10, 46±9 and 10±1, respectively) (Figure 4a). GFP expression in nonirradiated LLC1 cells was blocked by ~50% (from 56±6 to 28±4) with a threefold excess of free Tf and was maximally inhibited with eightfold excess (20±4) (Figure 4b). Similar to WB cells, GFP expression in irradiated LLC1 cells was significantly decreased in the presence of free Tf at 3- and 8-fold (from 194±16 to 81±10 and 57±8, respectively), and was maximally inhibited at 40-fold excess free Tf (33±9) (Figure 4b). These results suggest that Tf-TfR interaction mediates efficient lipoplex gene transfer and is required for radiation induction of Tf-lipoplex gene transfer in cells.
To investigate the potential use of radiation in improving gene delivery to tumors, mice bearing subcutaneous LLC1 tumors were irradiated locally prior to systemic administration of Tf-lipoplex containing CMV-LacZ gene. Radiation with 10 Gy given 24 h prior to liposome injection substantially induced β-gal gene transfer to targeted tumor cells as represented by the increasing number of x-gal stained cells in tumors (Figure 5a). Although the overall transduction level is low in nonirradiated control tumors (~0.5%) (Figure 5b), radiation treatment greatly improved gene expression up to ~1.6% (3.5±0.2-fold increase) and ~3.5% (7.4±0.3- fold increase) in tumors irradiated with 10 and 20 Gy (Figure 5b). Moreover, induction of gene transfer by 10 Gy of radiation was most effective when administration of Tf-lipoplex was delayed for 24 h after radiation (~1.6%), and had no or moderate effect when Tf-lipoplex was given immediately or at 48 h after radiation (~0.6 or 0.8%, respectively) (Figure 5c). Compared to the use of adenovirus vector in our previous study,3 comparable transgene expression can be achieved by the use of Tf-lipoplex nanoparticle and can be further enhanced by pretreatment with radiation.
We then tested whether the induction of gene transfer was related to Tf-lipoplex uptake. We found that tumor uptake of Tf-lipoplex was significantly increased as evidenced by the increase of LacZ-containing DNA content in tumors irradiated with 10 and 20Gy (2.1±0.5- and 10.1±1.5-fold, respectively) compared to nonirradiated control tumors (set as ‘1’) (Figure 6a). Radiation with 10 Gy appeared to facilitate the delivery of DNA in a time-dependent manner. The presence of plasmid DNA was increased by 1.6±0.2-, 2.1±0.4- and 3.7±0.8-fold when Tf-lipoplex was administered immediately after radiation (0 h), at 24 and 48 h after radiation, respectively (Figure 6b). It is not clear why gene expression did not continue to go up when Tf-lipoplex was administered 48 h after radiation and uptake was greater at that time point (Figure 5c). However, a general correlation between an increased presence of transgene (LacZ) DNA and an enhanced transgene expression suggests that radiation induced Tf-lipoplex uptake and thus gene expression by tumors.
To determine whether the induction of Tf-lipoplex uptake by radiation is mediated by receptor interaction, irradiated tumors were further assessed for TfR expression by western blot. We found that the protein level for TfR was increased in tumors at 24 h after 10 Gy (Figures 7a and b) or 20 Gy (Figure 7a). It should be noted that the induction of TfR was approximately twofold compared to nonirradiated control tumor (Ctrl) based on densitometric analysis of the corresponding bands on western blot (Figure 7c) and was less than the induction observed for LacZ expression (as many as 7.4±0.3-fold, Figure 5b) or Tf-lipoplex uptake (as many as 10.1±1.5-fold, Figure 6a). However, the induction of TfR in irradiated tumors was consistent with the observation in tumor cell culture that TfR interaction was required for the increase in transgene expression by radiation (Figure 4). These results suggested that interaction between Tf and TfR at least partially contributed to the radiation induction of Tf-lipoplex uptake and gene transfer in tumors.
Tf-lipoplex-mediated gene expression was further assessed in various normal tissues in order to evaluate the nonspecific expression related to potential tissue toxicity. Fluorescence-labeled Tf-lipoplex (Tf-lipoplex-Di) was administered through the tail vein to visualize its distribution. We found that Tf-lipoplex particles were maximally localized in many tested organs within 4 h after injection as visualized by the labeled particles (green) in tissue sections (data not shown). The amount of particles reached a plateau between 12 and 24 h, as shown for liver, lung, kidney and spleen (Figure 8a, top), followed by a gradual decrease. The particles appeared to distribute into specific anatomic structures in these organs. They were highly condensed in the central vein (CV) of the liver with spreading into the sinusoids. Similarly, particles were concentrated in the glomerulus unit of the kidney but sparse in the collecting tube. In the spleen, these particles were observed mostly in the germinal center where macrophages and phagocytes are abundant. In the lung, they were evenly distributed in the interstitial tissue surrounding alveolar spaces. Despite the abundance of Tf-lipoplex, the expression of transgene (LacZ) in these tissues was undetectable by x-gal staining 72 h after injection (Figure 8a, bottom). Moreover, pretreatment with radiation 24 h before injection had no noticeable effect on LacZ expression (data not shown). In contrast, the uptake of Tf-lipoplex in nonirradiated LLC1 tumor (−RT), which was localized mostly within vasculature, was relatively low 24 h after injection (Figure 8b) compared to normal tissues described above. Pretreatment with 10 Gy 24 h prior to injection (+RT) significantly increased the presence of Tf-lipoplex nanoparticles outside vasculature where a closer contact with tumor cells is expected (Figure 8b). This induction was consistent with the increased LacZ expression (Figure 5) and Tf-lipoplex uptake (Figure 6) that occurred at a later time point (72 h), suggesting that radiation selectively induced Tf-lipoplex uptake and gene transfer in tumors, thus may provide a safety benefit for the use of Tf-lipoplex as gene transfer vector for cancer treatment.
In this study, we demonstrated that ionizing radiation enhanced Tf-lipoplex gene transfer in tumor cells and improved gene delivery in tumor xenografts in mice in a radiation dose- and time-dependent manner. Radiation-induced gene expression in cells correlated with the increase in complex uptake and could be abolished in the presence of excessive Tf. Moreover, radiation had no effect on cells that were preincubated with Tf-lipoplex at 4 °C, suggesting that a receptor-mediated membrane internalization, but not the binding of complex, may be regulated during radiation treatment. Radiation-induced transgene expression was absent in tested normal organs but observed in tumors, consistent with the increase in DNA content from the complex and elevation of TfR. Our results suggest that radiation improves Tf-lipoplex gene transfer selectively in tumors by increasing tumor uptake of complex through Tf-TfR interaction.
Radiation has been shown to improve liposome-mediated gene delivery, expression and cytotoxicity in cancer cells.18,19 However, the mechanism underlying radiation induction has not been fully elucidated. A previous study demonstrated that radiation causes tumor vessel inflammation and can facilitate selective binding of liposome that is surface conjugated with wheat germ agglutinin, a lectin protein specific for inflamed endothelial cells.13 Although liposome-endothelial cell binding appears to be important in recruitment of liposome into the irradiated site, a direct radiation effect on tumor cell uptake of complex could not be ruled out. In the current study, we found that radiation-induced gene transfer in cells was greatly decreased in the presence of excess free Tf and could be completely abolished in cells preincubated with Tf-lipoplex. These data suggest that Tf-TfR interaction is a prerequisite but not a regulatory step for the radiation induction. Using fluorescence-labeled complex, we observed a significantly induced Tf-lipoplex entry in cells, whereas expression of TfR was not affected by radiation treatment (data not shown). This raises the possibility that radiation may increase complex uptake through other membrane proteins, which may act as either coreceptors for complex binding or mediators for internalization. The evidence that the moderate increase of TfR in LLC1 tumor xenografts (Figure 7) did not correlate with the larger magnitude of radiation induction in transgene expression and liposomal uptake (Figures 5 and and6)6) further support this concept. This is similar to our previous finding that radiation induces adenovirus uptake in tumors by increasing the expression of dynamin 2, a membrane-bound GTPase that is important in a receptor-mediated internalization process, whereas coxsackie and adenovirus receptor is not changed by radiation.17 In addition, it has been demonstrated that dynamin is crucial for liposome endocytosis and intracellular trafficking in neurons.22 The role of dynamin 2 in radiation-induced liposome uptake is certainly worth further investigation.
One of the major findings in this study is that radiation induced gene transfer in tumor xenografts most effectively when Tf-lipoplex was administered 24 h after radiation treatment, as compared to in vitro, where the greatest induction occurred when complex was administered immediately after radiation. This could be the result of radiation action that is unique in vivo. One such action is the vascular tone change that is common during radiation. In cancer radiotherapy, this change involves tumor vascular endothelial cells that undergo apoptosis after radiation, contributing to vessel damage. For instance, radiation induces intracellular phosphytidylserine (ps) to translocate from the cytoplasma to the outer membrane surface, which in turn, binds to annexin V and triggers apoptotic signaling pathways.23,24 This transient appearance on the membrane is unique under stress conditions such as radiation and has been used for radiation-guided delivery of therapeutic agents conjugated with annexin V protein or single chain variable fragment antibody specific for ps.25 Early development of vasculature damage after radiation could contribute to enhanced endothelial binding and extravasation of the complex into tumor cells. In this study, we showed that Tf-lipoplex nanoparticles are in proximity to the vasculature in liver (CV), kidney (glomerulus) and tumors regardless of radiation treatment (Figure 8a). We also demonstrated that radiation increased the Tf-lipoplex uptake in tumor in a time-dependent manner, being the most significant when the complex was administered at 48 h after radiation (~3.7-fold) as compared to 24 h (~2.1-fold) or immediate administration (~1.6-fold) (Figure 6b), in contrast to the in-vitro data where radiation had minor effects at 24 (Figure 1b) and 48 h (data not shown). It was also observed that induction of transgene expression was greater in tumor radiated 24 h prior to lipoplex treatment than tumor radiated 48 h before lipoplex treatment (Figure 5c). This slight discrepancy between the time of maximal presence of DNA (48 h) and optimal gene expression (24 h) could be the result of continued blood vessel leakage at 48 h after radiation when tumor cells are undergoing apoptosis and necrosis, and thus less capable of transgene expression. The overall correlation between complex presence and gene expression in tumors suggests that radiation could improve Tf-lipoplex-mediated gene transfer by inducing complex uptake in tumors.
Tf-lipoplex nanoparticle offers promise in cancer gene therapy for its enhanced safety without compromising gene transfer efficacy in tumors. Unlike adenovirus vector, Tf-lipoplex-mediated gene expression in normal tissues including liver, lung, kidney and spleen were negligible in this study. This selectivity could be due to both the specific formulation of this complex that is low in liver cell retention and to the endosome/nucleus fusion event, which occurs during active tumor cell proliferation but less often in the more quiescent normal tissue. Although we found that Tf-lipoplex-mediated gene expression is efficient in a normal hepatocyte cell line (WB), the mechanism for this selectivity would be better addressed in nonproliferative primary hepatocyte culture. Moreover, Tf-receptor interaction appeared to provide high affinity binding between Tf-lipoplex and targeted cells as evidenced by the competition assay where excess free Tf (~8-fold) was required for complete blockage of gene transfer in both nonirradiated and irradiated cells (Figure 4). Although it is not clear whether radiation may further induce Tf-TfR interaction (binding) in cells, our data suggest that radiation markedly increased cellular uptake of Tf-lipoplex, leading to improved transgene expression.
Our previous studies demonstrated that an adenovirus-mediated expression of CEA-yCD can selectively convert 5FC to 5FU in intrahepatic colon cancer metastases and thus eliminate or delay tumor growth. The strategy of using radiation for improving Tf-lipoplex gene delivery should provide additional benefit for radiation-combined gene therapy, especially in tumor-specific targeting. This is especially beneficial for treating local tumors when a high radiation dose can be delivered to the tumors through either a single or multiple fractions by the use of stereotactic radiation technique. For instance, typical fractionation schemes for the treatment of lung cancer are 60Gy in three 20Gy fractions for smaller tumors, and 66Gy in three fractions for large tumors. Likewise, intrahepatic tumors are treated in three fractions of 15–20Gy each.26 We demonstrate here that Tf-lipoplex achieves gene transfer efficacy comparable to adenovirus vector. The low normal tissue toxicity from the use of Tf-lipoplex will allow for multiple administration of this vector without significant host immune response, which is often encountered during adenovirus gene therapy. Therefore, further improvement of therapeutic index is anticipated and currently under investigation.
We thank Mary Davis (Radiation Oncology, University of Michigan) for her review of the paper. This work was supported partly by NIH grants CA80145, CA121830, CA128220, and CA84117, and the Cancer Center Core Grant CA46592 from the University of Michigan.