Causing damage to angiogenic vessels is a promising approach for cancer chemotherapy. The present study is a codification of a designed liposomal drug delivery system (DDS) for antineovascular therapy (ANET) with 2′-C-cyano-2′-deoxy-1-β-D-arabino-pentofuranosylcytosine (CNDAC). The authors have previously reported that liposomalized 5′-O-dipalmitoylphosphatidyl CNDAC (DPP–CNDAC), a phospholipid derivative of the novel antitumor nucleoside CNDAC, is quite useful for ANET. DPP–CNDAC liposomes modified with APRPG, a peptide having affinity toward angiogenic vessels, efficiently suppressed tumor growth by damaging angiogenic endothelial cells. In the present study, the authors masked the hydrophilic moiety of DPP–CNDAC, namely, CNDAC, on the liposomal surface with APRPG–polyethyleneglycol (PEG) conjugate to improve the availability of DPP–CNDAC liposomes. The use of the APRPG–PEG conjugate attenuated the negative ζ-potential of the DPP–CNDAC liposomes and reduced the agglutinability of them in the presence of serum. These effects improved the blood level of DPP–CNDAC liposomes in colon 26 NL-17 tumor-bearing BALB/c male mice, resulting in enhanced accumulation of them in the tumor. Laser scanning microscopic observations indicated that APRPG–PEG-modified DPP–CNDAC liposomes (LipCNDAC/APRPG–PEG) colocalized with angiogenic vessels and strongly induced apoptosis of tumor cells, whereas PEG-modified DPP–CNDAC liposomes (LipCNDAC/PEG) did not. In fact, LipCNDAC/APRPG–PEG suppressed the tumor growth more strongly compared to LipCNDAC/PEG and increased significantly the life span of the mice. The present study is a good example of an effective liposomal DDS for ANET that is characterized by: (i) phospholipid derivatization of a certain anticancer drug to suit the liposomal formulation; (ii) PEG-shielding for masking undesirable properties of the drug on the liposomal surface; and (iii) active targeting to angiogenic endothelial cells using a specific probe. (Cancer Sci 2008; 99: 1029–1033)
Because the inhibition of angiogenesis suppresses tumor growth and hematogenous metastases, antiangiogenic therapies have been widely investigated.(1–3) These therapies are also expected to be effective toward a broad spectrum of cancers, including drug-resistant cancers. Besides antiangiogenic therapy, ANET, namely, the causing of indirect lethal damage to tumor cells by the complete cut-off of the supply of oxygen and nutrients through damaging neovessels, is being developed.(4,5) For ANET, the authors previously isolated from a phage-displayed peptide library a peptide that specifically binds to the tumor angiogenic vasculature. The epitope sequence of the peptide was determined to be APRPG.(6,7) The authors demonstrated that APRPG is a useful probe for the active targeting of angiogenic vessels, although the target molecule of APRPG is still unknown. In contrast, PEG-coating of liposomes has been used in a liposomal DDS. It is known that PEG-modified liposomes characteristically remain in the circulation longer than non-modified ones through avoidance of RES-trapping of drug carriers.(8–10) PEG modification of the liposomal surface is known to form a fixed aqueous layer around the liposome due to the interaction between the PEG-polymer and water molecules, and thus prevents the binding of certain serum proteins and opsonins that are responsible for RES-trapping.(11) The feature of long circulation causes liposomal accumulation in tumor tissues, because the angiogenic vasculature in tumor tissues is quite leaky and macromolecules easily accumulate in the interstitial tissues of the tumor due to the EPR effect.(12,13) In the case of ANET, this long time in circulation increases the opportunity for specific binding of ligand-modified liposomes to angiogenic vessels. For this purpose, the authors previously designed a compound composed of APRPG, PEG, and DSPE.(14,15) It has been demonstrated that APRPG–PEG modification is superior to just APRPG modification for enhancing the antitumor activity of liposomal doxorubicin.(16)
In the present study, the authors used CNDAC as a chemotherapeutic agent. CNDAC had been originally synthesized as a novel antitumor nucleoside anti-metabolite by Matsuda et al., who showed that CNDAC has a novel anticancer mechanism and induces DNA strand breaks after its incorporation into tumor cell DNA.(17) The results of phase I clinical studies of its N4-palmitoyl derivative (CS-682) in patients with malignant solid tumors were reported recently.(18,19) The authors previously designed DPP–CNDAC for liposomalization,(20) because liposomal drugs show improved biodistribution and bioavailability in tumor-bearing animals. In fact, liposomal DPP–CNDAC showed enhanced activities for reducing tumor growth and increasing the life span of mice than conventional liposomes or soluble CNDAC.(21,22) As the next step, DPP–CNDAC liposomes were modified with APRPG for the purpose of ANET. APRPG-modification of DPP–CNDAC liposomes actually caused effective tumor growth suppression, possibly through damaging angiogenic endothelial cells.(23) These results also indicated that the therapeutic efficacy should reflect the damage to the angiogenic endothelial cells to which the liposomes gain access, because lipophilic drugs should be delivered to the cells in a liposomal form. However, the in vivo behavior of APRPG-modified DPP–CNDAC liposomes was affected by the presence of the cyano group of DPP–CNDAC on the liposomal surface. It induced aggregation of liposomes, resulting in reduced blood circulation of liposomes.
In the present study, the CNDAC on the liposomal surface was masked with APRPG–PEG conjugate to erase this undesirable property of DPP–CNDAC in liposomalization. The authors integrated their previous observations to formulate angiogenic vessel-targeted long-circulating DPP–CNDAC liposomes and applied them to ANET.
Materials and Methods
Materials. Synthesis of CNDAC and DPP–CNDAC was performed as described previously.(17,20) A phosphatidyl group was introduced into CNDAC through transphosphatidylation from 1,2-dipalmitoyl-3-sn-glycerophosphocholine using phospholipase D. Preparation of DSPE–PEG and DSPE–PEG–APRPG was performed as described previously.(14) DSPC and cholesterol were obtained from Nippon Fine Chemical Co., Ltd (Takasago, Hyogo, Japan). Colon 26 NL-17 colon carcinoma cells were established by Dr Yamori (Japanese Foundation for Cancer Research, Tokyo, Japan) and kindly provided by Dr Nakajima (Johnson & Johnson KK, Tokyo, Japan).
Animals. Five-week-old BALB/c male mice were obtained from Japan SLC Inc. (Shizuoka, Japan). The animals were cared for according to the animal facility guidelines of the University of Shizuoka.
Preparation of liposomes. DPP–CNDAC, DSPC, and cholesterol with DSPE–PEG (LipCNDAC/PEG) or DSPE–PEG–APRPG (LipCNDAC/APRPG–PEG) (10:10:5:2 as a molar ratio), or DPP–CNDAC, DSPC, and cholesterol without PEG-conjugate (LipCNDAC, 10:10:5 as a molar ratio) were dissolved in chloroform/methanol, dried under reduced pressure, and stored in vacuo for at least 1 h. Liposomes were produced by hydration of a thin lipid film with 10 mM phosphate-buffered 0.3 M sucrose (pH 6.8), and frozen and thawed for three cycles using liquid nitrogen. Then the liposomes were sized by extrusion thrice through polycarbonate membrane filters with 100-nm-diameter pores (Nucleopore, Maidstone, UK). The liposomal solutions were centrifuged at 180 000g for 20 min (CS120EX, Hitachi, Japan) to remove the untrapped DPP–CNDAC if present. Then the liposomes were resuspended in 10 mM phosphate-buffered 0.3 M sucrose. To determine the efficacy of trapping DPP–CNDAC in the liposomes, an aliquot of the liposomal solution was solubilized by the addition of reduced Triton X-100 (Sigma-Aldrich Co., St Louis, MO, USA), and the amount of DPP–CNDAC was optically determined at 280 nm after the pH of the solution had been adjusted to 1.0. For a biodistribution study, a trace amount of [1α, 2α(n)-3H] cholesterol oleoyl ether (74 kBq/mouse, Amersham Pharmacia, Buckinghamshire, England) was added to the initial organic solution. To examine the intratumoral localization of liposomes in tumor syngrafts, DiIC18 (Molecular Probes Inc., Eugene, OR, USA) was added to the initial organic solution (DPP–CNDAC:DSPC: cholesterol:DiIC18:DSPE–PEG or DSPE–PEG–APRPG = 10:10:5:0.1:2; DPP–CNDAC : DSPC : cholesterol : DiIC18 = 10:10:5:0.1, as a molar ratio). For the therapeutic study, control liposomes composed of DPPC, DSPC, and cholesterol (10:10:5 as a molar ratio) were prepared similarly as for the other liposomes.
Characterization of liposomes. Particle size and ζ-potential of liposomes diluted with PBS(–) were measured using a Zetasizer Nano ZS (Malvern, Worcestershire, UK). Aggregation testing was performed as follows: The liposomal solution was incubated in PBS(–) or in the presence of 50% FBS (Sigma-Aldrich) at 37°C for 1 h. The turbidity of the liposomal solution was determined at 450 nm, and relative turbidity compared with that in 0.3 M sucrose was then calculated.
Biodistribution of liposomes. Colon 26 NL-17 cells were cultured in DME/F12 medium (Nissui, Tokyo, Japan) supplemented with 10% FBS (Sigma-Aldrich). After harvesting of the cells, 1.0 × 106 cells were carefully injected subcutaneously into the posterior flank of 5-week-old BALB/c male mice. The biodistribution study was performed when the tumor size had become approximately 10 mm in diameter. Size-matched colon 26 NL-17 carcinoma-bearing mice were injected with the radiolabeled liposomes containing [1α, 2α (n)-3H] cholesterol oleoyl ether via a tail vein. One hour after the injection, the mice were killed under diethyl ether anesthesia for the collection of blood. The plasma was obtained by centrifugation (600g for 5 min). After the mice had been bled from the carotid artery, the heart, lung, liver, spleen, kidney, and tumor were removed, washed with saline, and weighed. The radioactivity in each organ was determined using a liquid scintillation counter (LSC-3100, Aloka, Tokyo, Japan). Distribution data were presented as percentage dose per gram of tissue or per 0.1 mL plasma. The total amount in plasma was calculated based on the average body weight of the mice, where the average plasma volume was assumed to be 4.27% of the body weight based on the data on total blood volume.
Histochemical analysis of liposomal distribution in tumor syngrafts. DiIC18-labeled liposomes were administered via a tail vein of colon 26 NL-17 carcinoma-bearing mice when the tumor sizes had reached approximately 10 mm in diameter. Two hours after the injection of liposomes, the mice were bled from the carotid artery under diethyl ether anesthesia, and the tumors were dissected. Histochemical analysis was performed according to the method described previously.(16) In brief, solid tumors were embedded in optimal cutting temperature compound (Sakura Finetechnochemical Co. Ltd, Tokyo, Japan) and frozen at –80°C. Five-micrometer tumor sections were prepared using a cryostat microtome (HM 505E, Microm, Walldorf, Germany), mounted on MAS-coated slides (Matsunami Glass Ind., Ltd, Japan), and air-dried for 1 h. The tissue sections were then fixed with acetone, and washed twice with PBS(–). After the sections had been blocked with 1% BSA in PBS(–), they were incubated with biotinylated antimouse CD31 rat monoclonal antibody (Becton Dickinson Laboratory, Franklin Lakes, NJ, USA) for 18 h at 4°C and then visualized after incubation with streptavidin–FITC conjugates (Molecular Probes Inc., Eugene, OR, USA) for 30 min at room temperature in a humid chamber. These sections were observed for fluorescence using an LSM microscope system (Carl Zeiss, Co. Ltd, Jena, Germany).
Determination of apoptotic cells in tumors. LipCNDAC/PEG or LipCNDAC/APRPG–PEG was administered intravenously into colon 26 NL-17 tumor-bearing mice when the tumor size had reached approximately 10 mm in diameter. Twelve hours after injection of the liposomes, the tumors were dissected from the mice; and tumor sections were then prepared. Next, immunostaining of endothelial cells was performed as described above, except that streptavidin–Alexa 594 conjugate (Molecular Probes Inc.) was used as fluorescent dye instead of streptavidin–FITC conjugate. For visualizing apoptotic cells, TUNEL staining was performed using an ApopTag Plus Fluorescein In Situ Apoptosis Detection Kit (Intergen Co., Purchase, NY, USA) according to the recommended procedures supplied in the kit. In brief, tumor sections were washed and equilibrated for 15 min in a humid chamber at room temperature, and then reacted with TdT enzyme for 1 h at 37°C. Thereafter, they were stained with antidigoxigenin–fluorescein antibody, after which the sections were observed using the LSM system. Apoptotic signals (green signals) were analyzed using Image J software (NIH).
Therapeutic experiment. LipCNDAC/PEG, LipCNDAC/APRPG–PEG or control liposomes were administered intravenously into colon 26 NL-17 tumor-bearing mice. The injected dose for each administration was 15 mg/kg as CNDAC moiety. The treatment was started when the tumor volume became approx. 0.1 cm3. The size of the tumor and the body weight of each mouse were monitored daily thereafter. Two bisecting diameters of each tumor were measured with slide calipers to determine the tumor volume. Calculation of the tumor volume was performed using the formula 0.4 (a × b2), where a is the largest and b is the smallest diameter. The calculated tumor volume correlated well with the actual tumor weight (r = 0.980).(24) The life spans of tumor-bearing mice were also monitored.
Statistical analysis. Variance in a group was evaluated using the F-test; and differences in mean tumor volume using Student's t-test.
Characterization of DPP–CNDAC liposomes. The efficiency of entrapment of DPP–CNDAC into liposomes was approximately 100% in all experiments (data not shown). Because DPP–CNDAC was easily incorporated into the lipid bilayer of liposomes as a lipid component, the CNDAC moieties of DPP–CNDAC were speculated to be exposed on the liposomal surface. In fact, the ζ-potential of LipCNDAC was negative due to the presence of the cyano group in CNDAC (Table 1). In contrast, PEG- or APRPG–PEG-modification reduced the negativity of the ζ-potential of DPP–CNDAC liposomes, suggesting that PEG shielded CNDAC moieties on the liposomal surface by forming a fixed aqueous layer (Table 1). As shown in Fig. 1, the agglutinability of both LipCNDAC/PEG and LipCNDAC/APRPG–PEG in the presence of serum was considerably low compared with that of LipCNDAC.
Table 1. Size and ζ-potential of DPP–CNDAC liposomes
Size ± SD (nm)
Particle size and ζ-potential of DPP–CNDAC liposomes diluted with phosphate-buffered saline(–) were measured using a Zetasizer Nano ZS. APRPG, Ala-Pro-Arg-Pro-Gly; CNDAC, 2′-C-cyano-2′-deoxy-1-β-D-arabino-pentofuranosylcytosine; DPP, 5′-O-dipalmitoylphosphatidyl; LipCNDAC, DPP–CNDAC liposomes; LipCNDAC/APRPG–PEG, APRPG–PEG- modified DPP–CNDAC liposomes; LipCNDAC/PEG, PEG-modified DPP–CNDAC liposomes; PEG, polyethyleneglycol.
120.8 ± 3.5
121.5 ± 5.7
102.4 ± 2.2
Biodistribution study. The biodistribution of these three types of liposomes was determined in colon 26 NL-17 carcinoma-bearing mice. At 1 h after administration of these liposomes, the plasma concentrations of LipCNDAC/PEG and LipCNDAC/APRPG–PEG were significantly higher than that of LipCNDAC (Fig. 2). These data suggest that the use of PEG or APRPG–PEG reduced the aggregation of these liposomes in the blood circulation, which prevented recognition of them by RES and endowed them with long circulation. In addition, APRPG–PEG modification significantly improved the blood circulation of DPP–CNDAC liposomes at 3 or 24 h administration of these liposomes (data not shown). Therefore, LipCNDAC/PEG and LipCNDAC/APRPG–PEG showed high accumulation in the tumors compared with LipCNDAC, possibly through the EPR effect. Particularly in LipCNDAC/APRPG–PEG, this characteristic of long circulation would increase the opportunity for specific binding of these liposomes to angiogenic vessels.
Histochemical analysis of the tumor. Colon 26 NL-17-bearing mice were given a single i.v. dose of LipCNDAC/PEG or LipCNDAC/APRPG–PEG. As shown in Fig. 3a–c, when the fluorescently labeled LipCNDAC/PEG was injected, the liposomal distribution (red fluorescence) was observed to be separate from endothelial cells (green fluorescence). In contrast, LipCNDAC/APRPG–PEG was colocalized with endothelial cells (Fig. 3d–f). These data suggest that LipCNDAC/APRPG–PEG became selectively localized on angiogenic endothelial cells. Cellular apoptosis in the tumor tissues was evaluated 12 h after administration of the liposomes (Fig. 3g–l). The signals of apoptotic cells were approximately 4.6-fold greater for LipCNDAC/APRPG–PEG than for LipCNDAC/PEG. CD31-staining did not show any vessel-like structure in the tumor of either liposome-treated group, suggesting that LipCNDAC/PEG also damaged angiogenic endothelial cells to some degree. These results indicate that LipCNDAC/APRPG–PEG had preferentially damaged angiogenic endothelial cells that induced effective apoptosis of tumor cells surrounding the damaged vessels.
Therapeutic efficacy of LipCNDAC/APRPG–PEG in tumor-bearing mice. LipCNDAC/APRPG–PEG suppressed tumor growth more efficiently than LipCNDAC/PEG: Significant differences in the tumor volume of the LipCNDAC/APRPG–PEG-treated group from that of the LipCNDAC/PEG-treated group were observed from day 22–28, although the SD data are shown only for day 28 (Fig. 4a). In addition, the tumor volume of the LipCNDAC/APRPG–PEG-treated group was significantly different from that of the control liposome-treated group from day 20–28. The body-weight change, an indicator of side-effects, was not observed in either the LipCNDAC/PEG- or LipCNDAC/APRPG–PEG-treated groups (data not shown). Corresponding to the tumor growth suppression, treatment with LipCNDAC/APRPG–PEG elongated the survival time of the mice: The mean survival times of the control liposomes, LipCNDAC/PEG-, and LipCNDAC/APRPG–PEG-treated groups were 47.6 ± 3.7, 46.6 ± 6.2, and 60.8 ± 4.8 days, respectively (Fig. 4b). The survival time of the LipCNDAC/APRPG–PEG-treated group was significantly longer than that for the mice treated with the control liposomes (P < 0.001) or LipCNDAC/PEG (P < 0.01).
The use of a DDS for targeting tumors is a promising strategy particularly for drugs with severe side-effects such as those used in cancer chemotherapy. CNDAC was developed as an effective anticancer drug,(17) but has severe side-effects like other anticancer drugs. To design a targeting DDS, the authors previously derivatized CNDAC as a phospholipid mimetic(20) because it was readily incorporated into liposomes, the most widely used drug carrier for a DDS. This mimetic, DPP–CNDAC, was well suited to liposomalization for cancer treatment.(21,22)
For the active targeting strategy for delivery of anticancer drugs, angiogenic vessels were selected as a target organ and a novel type of antiangiogenic therapy, antineovascular therapy (ANET), was examined. Vascular targeting has become an interesting issue in DDS, because anticancer drugs or their carriers first meet angiogenic vessels before extravasation into the tumor tissue. The authors previously applied APRPG-modified liposomes for antineovascular therapy using DPP–CNDAC. Because lipophilic drugs should be delivered to the cells in a liposomal form, the therapeutic efficacy should reflect the damage to the cells to which the liposomes gain access rather than a change in the local concentration of the agent in the tumor tissue. If the therapeutic efficacy of APRPG-modified DPP–CNDAC liposomes is superior to that of non-modified liposomal DPP–CNDAC, such a result would suggest that the destruction of angiogenic endothelial cells is superior to the direct destruction of tumor cells for effective tumor treatment. The authors’ previous results indicate that the delivery of DPP–CNDAC to angiogenic endothelial cells is, in fact, useful for the suppression of tumor growth.(23)
PEG-shielding of the liposomal surface should be useful for designing active targeting DDS as well as passive targeting. In the present study, the significant efficacy of APRPG–PEG-modified DPP–CNDAC liposomes for tumor growth suppression was shown. An important aspect of the present study is that PEGylation served for not only RES-avoidance but also construction of a practical liposomal formulation using a lipid-derivatized drug. When the liposomal surface is modified with anticancer drugs such as DPP–CNDAC, the fixed aqueous layer formed by PEG can mask the undesirable properties of such liposomes for DDS. Thus, the RES avoidance afforded by the use of PEG enhanced the accumulation of the liposomes in the tumor tissue, enabled targeting of angiogenic endothelial cells, and caused efficient damage to tumor cells. Therefore, APRPG–PEG-modified liposomal DPP–CNDAC caused efficient tumor growth suppression without severe side-effects.
The present study is a good example of liposomalization if the property of the objective compound is not suitable to liposomalize, and the technology used is applicable to other agents. The present study indicates the importance of designing drug, carrier, and therapeutic strategy in the development of DDS pharmaceutics.
This work was supported by a Grant-in-Aid for Scientific Research.