Fax: (011) 81 54 264 5705
Induction of intensive tumor suppression by antiangiogenic photodynamic therapy using polycation-modified liposomal photosensitizer
Article first published online: 1 APR 2003
Copyright © 2003 American Cancer Society
Volume 97, Issue 8, pages 2027–2034, 15 April 2003
How to Cite
Takeuchi, Y., Kurohane, K., Ichikawa, K., Yonezawa, S., Nango, M. and Oku, N. (2003), Induction of intensive tumor suppression by antiangiogenic photodynamic therapy using polycation-modified liposomal photosensitizer. Cancer, 97: 2027–2034. doi: 10.1002/cncr.11283
- Issue published online: 1 APR 2003
- Article first published online: 1 APR 2003
- Manuscript Accepted: 16 DEC 2002
- Manuscript Revised: 18 NOV 2002
- Manuscript Received: 31 JUL 2002
- photodynamic therapy;
- antiangiogenic cancer therapy;
- drug delivery system;
- polycation liposome;
- benzoporphyrin derivative monoacid ring A
The authors previously observed that antiangiogenic scheduling of photodynamic therapy (PDT) was effective in causing tumor regression through hemostasis. It would thus be expected that photosensitizer entrapped in polycation liposomes (PCLs) would be efficiently taken up in tumor-derived angiogenic vascular endothelial cells due to the strong electrostatic adhesion between the polycation and the plasma membrane, thus resulting in enhanced phototherapeutic efficacy.
Tumors and angiogenesis were induced by subcutaneous injection of Meth-A sarcoma cells into 5-week-old male BALB/c mice. PDT treatment was performed by an intravenous (i.v.) injection of benzoporphyrin derivative monoacid ring A (BPD-MA)-entrapped liposomes or the PCLs (0.25 mg/kg in terms of BPD-MA), followed by exposure to a laser light of 689 nm with 150 J/cm2 of fluence 15 minutes post injection.
As a result of PDT on angiogenesis-model mice prepared by the dorsal air sac technique, neovascular destruction after laser irradiation was observed when BPD-MA entrapped in PCLs was used. Furthermore, strong suppression of tumor growth was identified by the PCL-mediated PDT treatment along with a prolonged life span for the mice. Destruction of angiogenic vessels and subsequent tumor cell apoptosis were observed after PCL-mediated PDT treatment in an immunofluorescence study. Interestingly, the biodistribution of the injected BPD-MA that was delivered by PCLs indicated invariable photosensitization levels in tumor tissues.
The study revealed that antiangiogenic PDT treatment using a low dose of BPD-MA entrapped in PCLs efficiently induced the destruction of angiogenic vessels and subsequent tumor suppression by vessel occlusion. Cancer 2003;97:2027–34. © 2003 American Cancer Society.
Photodynamic therapy (PDT) is a promising cancer treatment that uses a combination of tissue-penetrating laser light, a photosensitizer (such as a porphyrin, chlorin, or phthalocyanine derivatives), and activated oxygen to afford tumor destruction.1–4 Generally, preferential accumulation of the photosensitizer in malignant tissues is strongly dependent on the complexation between the hydrophobic macrocycle and low-density lipoprotein (LDL) in the blood due to the finding that the LDL is selectively incorporated in tumor cells through their highly expressed LDL receptors.5–9 However, some of the photosensitizer–LDL complexes are snatched away by the reticuloendothelial system (RES) trapping and renal clearance. To reduce such loss, researchers have explored the use of liposomal formulations of photosensitizer as a useful technique with some benefits (i.e., liposomes allow not only photochemotherapeutic applications of water-insoluble macrocycles10 but also prolong the half-life of the photosensitizer in the bloodstream11, 12).
Benzoporphyrin derivative monoacid ring A (BPD-MA), also known as Verteporfin™ or Visudyne™ (Novartis, Basel, Switzerland), is one of the second-generation photosensitizers in PDT that is targeted to the choroidal neovascularization of age-related macular degeneration. Photoactivation of BPD-MA induces local damage to the neovascular endothelium, resulting in vessel occlusion.13 In previous studies, blood flow stasis in the tumor-derived angiogenic vessel was shown to play a significant role in BPD-MA-mediated tumor destruction in angiogenesis-model mice prepared by the dorsal air sac technique, and the antiangiogenic PDT treatment for damaging angiogenic vascular endothelial cells (laser irradiation at 5–15 min post BPD-MA injection) was more efficient for tumor suppression than damaging tumor cells directly (laser irradiation at 3 hours post BPD-MA injection).14–18
It would be expected that polycations would adhere to the anionic plasma membrane by electrostatic interaction due to the concentrated positive charge resulting from the polymerization of amino groups. In this study, to increase the efficiency of photosensitizers, we used partially cetylated polyethylenimine (PEI) as a polycation to coat the polycation on the liposomal membrane stably by van der Waals and hydrophobic interactions with the hydrophobic region of the liposome.
As described herein, we developed BPD-MA-entrapped polycation liposome (BPD-MA PCL) as a novel liposomal drug for PDT. We consider that PCL, constructed with partially cetylated PEI (cetyl-PEI), which is located on the surface of the liposome, combines the above-mentioned advantages of both liposome and polycation.19, 20 In the present study, in vivo angiogenic vessel-targeted PDT treatment using BPD-MA PCL was systematically investigated. We clarified that such treatment caused the destruction of angiogenic vessels and subsequent vessel occlusion, resulting in tumor cell apoptosis. Consequently, strong suppression of tumor growth was induced.
MATERIALS AND METHODS
PEI, consisting of 25%, 50%, and 25% for primary, secondary, and tertiary amino groups, respectively, with an average molecular weight of 1800, was kindly provided by Nippon Shokubai Co., Ltd. (Osaka, Japan). Polymer purification was performed by an ultrafiltration technique with an apparatus equipped with a YM-1 ultrafiltration membrane (Amicon Corp., Danvers, MA). 1H NMR spectra were taken with a Varian Gemini 300 instrument (Varian, Inc., Palo Alto, CA) with tetramethylsilane as an internal standard for CDCl3. Visible absorption spectra were recorded on a Beckman DU-70 spectrophotometer (Beckman Instruments, Fullerton, CA). The purified polycation was lyophilized and stored in ethanol before the preparation of cetyl-PEI. DC-Plastikfolien Kieselgel 60 F254 (Merck, Darmstadt, Germany) was used for analytical thin layer chromatography (TLC). Dipalmitoylphosphatidylcholine (DPPC) and dipalmitoylphosphatidylglycerol (DPPG) were kindly provided by Nippon Fine Chemical Co., Ltd. (Hyogo, Japan), and cholesterol was purchased from Sigma Chemical Co. (St. Louis, MO). BPD-MA and radiolabeled [14C]BPD-MA, were generous gifts from QLT PhotoTherapeutics, Inc. (Vancouver, Canada).
Preparation of Cetylated Polyethylenimine (Cetyl-PEI)
Cetyl-PEI was prepared by a procedure similar to that described previously.19, 20 In brief, PEI (2.40 g, 1.33 × 10−3 mol) was stirred at 65 °C in ethanol (10.0 mL) for 30 minutes under reflux conditions and N2 bubbling, and 1-bromohexadecane (3.87 mL, 1.33 × 10−2 mol) and triethylamine (223 μL, 1.60 × 10−3 mol) were added to the solution. After a 7-hour reaction, the solution was evaporated. The resulting solid was dissolved in 20% ethanol aqueous solution (100 mL), and the solution was ultrafiltered for 2 days with substituting 20% EtOH aqueous solution (over 1000 mL). Finally, the solution was lyophilized. The stoichiometric conjugation percentage of the cetyl group, determined by 1H NMR spectra, was 24.1% per ethylenine unit. 1H NMR (CDCl3): δ 0.82–0.95 (10H, t, cetyl –CH3), 1.13–1.44 (101H, br, cetyl –CH2–), 2.36–3.19 (167H, br, PEI–CH2–), 4.84–5.26 (32H, br, PEI amine H).
Preparation of BPD-MA-Entrapped Liposome and PCL
BPD-MA-entrapped control liposomes that were not modified cetyl-PEI (BPD-MA liposome) were prepared as described previously.12 This formulation consisted of DPPC, cholesterol, DPPG, and BPD-MA (molar ratio, 20:10:5:0.3, respectively). In contrast, the BPD-MA-entrapped PCL (BPD-MA PCL) also contained cetyl-PEI at a molar ratio of 1.75. Lipids, BPD-MA, and cetyl-PEI that were dissolved in CHCl3 were evaporated to construct the thin lipid film and hydrated with saline (final concentration of BPD-MA, 300 nM). The liposomal solution was freeze-thawed by using liquid N2 and sonicated for 15 minutes at 60 °C. Finally, these liposomes were sized at a 100-nm diameter by extrusion through a polycarbonate membrane filter. Particle sizes and ζ-potential of the 2 kinds of liposomes with entrapped BPD-MA were recorded on an ELS-800 electrophoretic light-scattering spectrophotometer (Otsuka Electronics Co., Ltd., Osaka, Japan). The results supported the assigned liposomal sizes, and the ζ-potential was +5.06 mV for BPD-MA liposome and +32.53 mV for BPD-MA PCL.
Animal and Tumor Model
Male BALB/c mice were purchased from Charles River Japan Inc. (Tokyo, Japan) and were housed in cages with a 12-hour light–12-hour dark cycle. Food and water were available ad libitum. The animals were cared for according to the animal facility guidelines of the University of Shizuoka.
Preparation of Dorsal Air Sac-Model Mice
All instruments for preparation of the dorsal air sac were obtained from Millipore Corporation (Bedford, MA). Meth-A sarcoma cells (6.617 × 107 cells/mL, 0.15 mL) were loaded into a Millipore chamber ring sandwiched by Millipore filters having a 0.45-μm pore size. The chamber ring was then implanted subcutaneously (s.c.) into the dorsum of 5-week-old male BALB/c mice ballooned by supplying 8–10 mL of air under pentobarbital anesthesia. On Day 4 after implantation of the chamber ring, PDT treatment was performed by an intravenous (i.v.) injection of BPD-MA liposomes or BPD-MA PCLs (0.25 mg/kg, 0.25 mL), followed 15 minutes post injection under pentobarbital anesthesia by exposure to a laser light of 689 nm with 150 J/cm2 of fluence (0.25 W, 470.8 sec). The mice were sacrificed by an overdose of diethylether after an additional 24-hour carrying, and the dorsal skin that had osculated the chamber ring was detached.
Antitumor Activity In Vivo
Five-week-old male BALB/c mice were injected s.c. in the left posterior flank with 5 × 106 cells/mL of Meth-A sarcoma cells (0.2 ml), which resulted in an identifying cutaneous bulge at Day 7 after tumor implantation, and then they were randomly divided into 3 groups (1 control and 2 treatment groups, each n = 5). The control group consisted of mice that received an i.v. injection of saline (0.25 mL) without laser exposure. PDT treatment was performed as described above. Tumor volume, body weight, and survival of the tumor-bearing mice were regularly monitored after PDT treatment. Tumor volume was determined as described previously.21
Immunofluorescence Double Staining for CD31/Platelet–Endothelial Cell Adhesion Molecule-1 (PECAM-1) and Terminal Deoxynucleotidyl Transferase-mediated Deoxyuridine Triphosphate Nick End Labeling (TUNEL)
Tumor implantation, photosensitizer injection, and PDT treatment were followed as described above. At 24 hours after PDT treatment, tumor tissues were resected. The tissues were embedded in Tissue Tek O.C.T. Compound (Sakura Finetek USA, Inc., Torrance, CA) and frozen in −80 °C acetone. The frozen tissues were sectioned at a 5-μm thickness and mounted on silicon-oxide coated slide glasses (Matsunami Glass Ind., Ltd., Osaka, Japan). The specimens were fixed in 4% paraformaldehyde for 15 minutes at room temperature after tissue air-drying for 1 hour and washed 3 times with phosphate-buffered saline (PBS). Protein-blocking was performed for 10 minutes at room temperature by using 1% bovine serum albumin (BSA)-containing PBS, and the specimens were then washed 3 times with PBS.
TUNEL assays were performed with an ApopTag Fluorescein In Situ Apoptosis Detection Kit (Intergen, Purchase, NY) according to the directions supplied with the kit. After tailing of the nicked DNA with digoxigenin-conjugated dNTP by TdT enzyme, the specimens were incubated with biotin-conjugated rat anti-mouse CD31/(PECAM-1) monoclonal antibody (BD Biosciences, Pharmingen, San Diego, CA), diluted at 1:100, for 1 hour at room temperature and subsequently washed 3 times with PBS. Immunofluorescent double staining was performed by using streptavidin-Alexa Fluor 594 (Molecular Probes, Inc., Eugene, OR) and anti-digoxigenin-fluorescein (from ApopTag kit) in the order given. The specimens were finally washed with PBS, and coverslips were mounted on the glass slides with Prolong Antifade (Molecular Probes, Inc., Eugene, OR).
Immunofluorescent CD31 and TUNEL were detected under a Carl Zeiss (Thornwood, NY) LSM510 erect confocal laser-scanning microscope. Alexa Fluor 594 and fluorescein were excited by using a helium–neon laser at 588 nm and an argon ion laser at 458 nm respectively. All lines were passed through a white transmission filter. The LSM510 allows simultaneous recording of 3 channels: an Alexa Fluor 593–derived red fluorescent image, a fluorescein-derived green fluorescent image, and a merged photomicrograph. The objective lens (20× magnification) through the eyepiece (10× magnification) imaged an area of 225 × 225 μm onto 1024 × 1024 pixels, giving a resolution of 72 dpi along both the x- and y-axes. A pinhole size of 100 μm was used in all experiments.
[14C]BPD-MA liposome and [14C]BPD-MA PCL were prepared in a manner similar to that described above except [14C]BPD-MA was used instead of BPD-MA. On Day 7 after Meth-A sarcoma implantation, the tumor-bearing mice were i.v. injected with 0.25 mg/kg of [14C]BPD-MA liposomes and [14C]BPD-MA PCLs. After 15 minutes, the mice were sacrificed by exsanguination under pentobarbital anesthesia, and the plasma and the tissues were collected. These samples were weighed and solubilized in 1 mL of Solvable (Packard Japan, Tokyo, Japan) for 2 days at 50 °C. After bleaching of the solution with a mixture of H2O2 and isopropanol (1 mL; volume:volume ratio, 1:1) overnight at 50 °C, Hionic-Fluor (Packard Japan) was added. Radioactivity of the 14C-labeled macrocycles was measured with an Aloka LSC-3500 liquid scintillation counter (Aloka Co., Ltd., Tokyo, Japan).
Difference between two means of tumor volume and radioactivity was evaluated by Student t test.
Specific Neovascular Damage in Dorsal Air Sac-Model Mice by PDT Treatment Using BPD-MA PCL
PDT using dorsal air sac-model mice provides important information regarding the development of novel photosensitizers and their delivery vehicles and also allows direct evaluation of tumor-induced neovascular damage in antiangiogenic cancer phototherapy. Initially, using angiogenesis-model mice, we examined the neovascular photodamage caused by PDT treatment with BPD-MA liposomes or BPD-MA PCLs. Figure 1 shows photomicrographs of the angiogenic vessels of the implanted dorsal position after PDT treatment. As is apparent from Figure 1 (A and B), successful neovascularization was obtained by implanting the Meth-A sarcoma-loaded chamber. Although PDT-induced neovascular damage was not observed when BPD-MA liposomes were used (Fig. 1C), remarkable damage by the photosensitizer entrapped in PCLs was visually affirmed by the expanded blood smudges indicating blood clots (Fig. 1D).
Antitumor Activity of Antiangiogenic PDT Treatment for Tumor-Bearing Mice
Antitumor activity of tumor-bearing mice after PDT treatment was determined by regularly timed measurement of tumor volume. Figure 2 shows the tumor volume after PDT treatment with 0.25 mg/kg of BPD-MA liposomes or BPD-MA PCLs. The results indicate that a significant suppression of tumor growth was afforded by the BPD-MA PCL; in contrast, the BPD-MA liposome hardly affected the tumor growth compared with the growth observed in the saline-injected control mice. Neither injection of BPD-MA PCLs without laser irradiation nor laser irradiation alone affected the tumor growth (data not shown). Life span was elongated by 24% for the BPD-MA PCL-treated group compared with the control. The toxicity of BPD-MA PCL was investigated by examining the change in body weight, with the results indicating a slight decrease in body weight on the day after the photosensitizer injection (data not shown). Thus, the toxicity of BPD-MA PCL was negligible.
Immunofluorescence Analysis of Neovascular Destruction-Induced Apoptosis of Tumor Cells
Figure 3 illustrates the results of CD31/TUNEL immunofluorescence double staining of tumor-tissue sections after PDT treatment. No TUNEL-positive cells were observed in the saline-injected control mice (Fig. 3A). In contrast, some degree of apoptotic tumor cell death was detected in the mice treated with BPD-MA liposomes, although the CD31-positive endothelial cells were not decreased as much by the PDT treatment (Fig. 3B). Interestingly, the marked disappearance of CD31-derived red fluorescence and induction of remarkable apoptosis for tumor cells were simultaneously noted in the mice injected with BPD-MA PCLs for the PDT treatment (Fig. 3C).
Biodistribution of Liposomal BPD-MA in Tumor-Bearing Mice
The biodistribution pattern of the photosensitizer clarified the potent factor for the high success of photosensitizer entrapped in PCLs for phototherapeutic treatment of tumor-bearing mice. The results are summarized in Figure 4. Remarkable accumulation of photosensitizer using [14C]BPD-MA liposomes was observed in the spleen; on the contrary, accumulation was observed in the lungs with [14C]BPD-MA PCLs. However, the efficiency of tumor localization of the photosensitizer was not significantly different between the BPD-MA liposomes and BPD-MA PCLs despite the strong suppression of tumor growth by the latter.
Our interest in photosensitizers entrapped in PCLs is motivated by the finding that this drug design has the potential to achieve effective phototherapeutics at a low dose of photosensitizer with minimum side effects, such as skin photosensitivity, etc. From our in vivo investigations, we concluded that PCL promotes the effective suppression of tumor growth and long-term survival in tumor-model mice (Fig. 2) compared with PEI-nonmodified liposomes. Our results demonstrated the feasibility of photosensitizer entrapped in PCLs for antiangiogenic cancer phototherapy.
In general, PDT treatment for the eradication of cancer has been performed post-LDL-mediated photosensitizer accumulation in the tumor tissue. Recently, the schedule for antiangiogenic PDT treatment has been established in tumor-bearing mice, i.e., laser exposure at an early stage, 5–15 minutes, after photosensitizer injection i.v.14–16 Interestingly, angiogenic vessel-targeted PDT has produced more potent phototherapeutics for tumor-bearing mice despite the photosensitization level in tumor tissue being low.15 Furthermore, we observed earlier that angiogenic vessel-targeted PDT effectively induced hemostasis.15 Our previous results were recently confirmed by Dolmans et al.,17, 18 who reported that the 15-minute scheduling of PDT treatment (antiangiogenic PDT) suppressed tumor growth more effectively than the therapy given at 4 hours after injection of the photosensitizer (direct PDT for tumor cells). In the present study using angiogenesis-model mice, tumor-induced neovascular damage was observed in only the PDT treatment group in which BPD-MA PCLs were used (Fig. 1). PCL-mediated PDT induced strong suppression of tumor growth with long-term effects and negligible side effects (Fig. 2). This report on the notable success of PDT-induced phototherapeutics with a low dose of BPD-MA (0.25 mg/kg) entrapped in PCLs also is the first report to our knowledge on the injection of liposomalized BPD-MA.
From the results shown in Figures 1 and 2, we suspect that the strong tumor suppression was mainly due to the destruction of angiogenic vessels in the tumor tissues and thus the consequent loss of oxygen and nutrient supplements needed for tumor growth. To give a few examples from recent reports on antiangiogenic cancer therapy, tranilast was found to reduce the tumor vascularity and increase apoptosis of Lewis lung carcinoma cells in vivo, as judged from an immunohistochemical study;22 enhancement of endothelial cell apoptosis that preceded an increase in tumor cell apoptosis was identified in liver metastasis-model mice that received vascular endothelial growth factor receptor-2 (VEGFR-2), as observed by use of an immunofluorescence double-staining technique;23 and topotecan inhibited the growth of Wilms tumors by the preceding reduction of vascularity.24 Based on a similar immunofluorescence double-staining study using PDT, Engbrecht et al. reported that Photofrin-mediated PDT treatment induced initial apoptosis in tumor endothelial cells and delayed tumor cell apoptosis in a human sarcoma xenograft model.25 The results of our immunofluorescence assay support this sequence of events, as we found that PCL-mediated PDT treatment caused the destruction of angiogenic vessels and subsequent apoptosis of the tumor cells, although we could not detect the apoptosis of endothelial cells before fast degeneration of the cells (Fig. 3).
Another potential demonstration of tumor cell apoptosis is that neovascular occlusion and subsequent destruction of the endothelial cells brought about tumor hypoxia. Hypoxia-inducible transcription factor-1α (HIF-1α) was found to undergo ubiquitin-dependent degradation by proteasomes following the binding of p53 with Mdm-2, resulting in cell-cycle arrest and subsequent apoptosis.26 In addition, hypoxia has been shown to induce apoptosis via changes in p53 expression level and caspase activation accompanying cytochrome c release from mitochondria.27–29
To clarify the reason for the strong phototherapeutic effect of BPD-MA PCL, we performed the biodistribution assay using radiolabeled BPD-MA (Fig. 4). The results indicated that PCL-mediated photosensitization circumvented the unfavorable infiltration in RES tissues (especially in the spleen that is found with the regular liposomes. It is possible that the remarkable lung infiltration by the [14C]BPD-MA PCLs was due to strong interaction of PCL with preexisting vessels. However, the toxicity of PCL in the mice was negligible. As a result, we suggest that liposomal photosensitizer is rapidly taken up into the surrounding vascular endothelial cells in the microvasculature due to the proximate distance between the PCL and plasma membrane. We consider that this phenomenon also occurs in the angiogenic vasculature at the periphery of the tumor tissue and that photodestruction of vascular endothelial cells and subsequent vascular occlusion occurs there by PDT treatment. In addition, established microvasculature is not affected due to the finding that the vascular damage is initially induced by laser exposure. The angiogenic vessel-targeted PDT-induced tumor suppression also is supported by the results of PCL-specific neovascular photodamage in angiogenesis-model mice and strong suppression of tumor growth and by the finding that the accumulation level of photosensitizer in the tumor was not significantly different between the mice treated with the photosensitizer entrapped in PCLs and those receiving it in nonmodified control liposomes, although partial contribution of a direct effect of phototherapy on the tumor tissue also is possible. Such evidence that lesser accumulation of an agent in tumor caused more efficient therapeutic effect due to the localization change was observed in antineovascular chemotherapy; angiogenic endothelial cell-targeted chemotherapy was found to cause strong tumor growth suppression compared with the nontargeted therapy, although the bulk accumulation of the agent was lower for the former treatment than for the latter.30
In conclusion, antiangiogenic PDT treatment using BPD-MA entrapped in PCLs resulted in significant tumor destruction with a low dose of the photosensitizer. The mechanism underlying photodestruction of the tumor is considered to be the following: 1) induction of a strong electrostatic interaction between modified PEI, which was on the surface of the liposomal membrane, and the anionic plasma membrane of angiogenic vascular endothelial cells, and subsequent uptake of the photosensitizer via endocytosis; and 2) photodestruction of angiogenic vessels by the damage to the vascular endothelial cells after PDT treatment. The loss of oxygen and nutrient supplements delivered via angiogenic vessels thus resulted in the apoptosis of the tumor cells.
The authors thank Mr. Hidetsugu Ori, Mr. Takayuki Koishi, and Mr. Masaharu Kondo at Nagoya Institute of Technology for assistance in the synthesis of PEI derivatives; Mr. Tomohiro Asai at University of Shizuoka for technical assistance in immunofluorescence analysis; and Mr. Takayuki Uchida and Ms. Masami Kondo at University of Shizuoka for technical support in conducting the biodistribution assay. Thanks also are given to Dr. Jorn R. North at QLT PhotoTherapeutics, Inc., and Dr. Yukihiro Namba at Nippon Fine Chemical Co., Ltd., for generous gifts of BPD-MA derivatives and phospholipids, respectively, and to Mr. Katsuhiko Sato at Suzuki Motor Co., Ltd., for providing the diode-laser apparatus.