Maltotriose Conjugation to a Chlorin Derivative Enhances the Antitumor Effects of Photodynamic Therapy in Peritoneal Dissemination of Pancreatic Cancer

Pancreatic cancer is a highly aggressive disease with a dismal prognosis. The reported 5-year overall survival rate of patients with pancreatic cancer is less than 5% (1), due mainly to the frequently advanced stage at diagnosis, rapid tumor growth, and metastasis to distant organs such as the liver, lung and peritoneum (2). Peritoneal dissemination of pancreatic cancer poses significant difficulties for both patients and clinicians because the associated poor general condition of affected patients and problems in assessing scattered tumors due to ascites, jaundice, and ileus hamper the administration of standard treatment. The treatment of choice for patients with peritoneal dissemination is palliative systemic chemotherapy; however, its impact on the primary pancreatic cancer is limited (3, 4). Thus, new methods of treating peritoneal dissemination are needed.

One such approach, photodynamic therapy (PDT), is a promising, minimally invasive modality for cancer therapy that uses reactive oxygen species (ROS) generated by the photochemical reactions between a photosensitizer and irradiation (5, 6). The anticancer effects of PDT are consequences of a low-to-moderately selective degree of photosensitizer taken up by malignant cells, direct cytotoxicity due to ROS, and severe tumor vascular damage that impairs blood supply to the treated area (7–9). PDT also has several advantages over conventional radiation and chemotherapy with respect to side effects and damage to normal tissue.

PDT using photofrin, a first-generation photosensitizer, is widely used clinically (10, 11). More recently, PDT using a second-generation photosensitizer, talaporfin, has shown several advantages over PDT with photofrin, including deeper penetration into tissue and less prolonged photosensitization (12, 13), and has already been used to treat a wide range of cancers (14–16). Nevertheless, PDT issues related to insufficient efficacy and skin photosensitivity remain unsolved, and the need remains for photosensitizers with better cancer cell selectivity.

We previously reported that PDT using a newly developed photosensitizer, glucose-conjugated chlorin (G-chlorin), exerted high photocytotoxicity compared with PDT with talaporfin (17, 18). G-chlorin with its four glucose molecules (5,10,15,20-tetrakis-(pentafluorophenyl)-2,3-[methano(N-methyl)iminomethano] chlorin) is likely to elicit high photocytotoxicity in cancer cells due to their tendency to accumulate glucose, called the Warburg effect (19). Unfortunately, G-chlorin is insoluble in water, making it less suitable for clinical use. Furthermore, good water solubility ensures rapid clearance from the body, avoiding cutaneous phototoxicity.

We have therefore developed a straightforward strategy to improve the water solubility of chlorin derivatives by using a oligosaccharide, such as maltotriose (Mal₃), called Mal₃-chlorin and have reported its highly hydrophilicity and photocytotoxicity (20). In the present study, we investigated the efficacy of PDT with a novel photosensitizer, Mal₃-chlorin, in culture and in a mouse model of disseminated peritoneal pancreatic cancer using in vivo fluorescence imaging to monitor sequentially.

Talaporfin sodium (mono-l-aspartyl chlorin6, Laserphyrin) was purchased from Meiji Seika (Tokyo, Japan; Fig. 1A). Mal₃-chlorin (5,10,15,20-tetrakis-[4-(β-D-maltotriosylthio)-2,3,5,6-tetrafluorophenyl]-2,3-[methano-(N-methyl)iminomethano]chlorin) was synthesized (Fig. 1B) and provided by Yamagata University (Yamagata, Japan; ref. 20).

Human pancreatic cancer cell lines, AsPC1 and BxPC3, were obtained from the ATCC on December 3, 2015, whereas the luciferase-transfected AsPC1/luc and BxPC3/luc lines were obtained from the JCRB cell bank on December 1, 2015. Our cell lines were authenticated with STR-PCR profiling at the cell banks and revealed to be not misidentified with any other cell lines. These cell lines were certified to be free of fungal, bacterial, and mycoplasmal contaminations by the cell banks, and were distributed to our institution. All cell lines were cultured in RPMI1640 medium (Wako Pure Chemical Industries Co. Ltd.) supplemented with 10% FBS. Cells were cultured at 37°C in 5% CO₂ humidified air, and experiments were carried out within 3 months of resuscitation.

Pancreatic cancer cells were incubated with talaporfin (5 μmol/L) or Mal₃-chlorin (5 μmol/L) for 4 hours and washed once with PBS. Cells were then harvested using 0.25% trypsin-EDTA (GIBCO) and analyzed using a FACSCanto II Analyzer (BD Biosciences; excitation; 405 nm, emission; 660 nm). At least 10,000 events were collected for each sample.

Pancreatic cancer cells (AsPC1, BxPC3, AsPC1/luc, and BxPC3/luc) were incubated with photosensitizer in culture medium for 4 hours. Cells were washed once with PBS, covered with PBS, and irradiated at 13.9 J/cm² (intensity: 30.8 mW/cm²) of LED light (Opto Code Corporation), which emits 660 nm wavelength. Following irradiation, the PBS in the wells was exchanged for medium supplemented with 10% FBS, and the cells were incubated for the specified times before analyses.

Cell viability was analyzed by the WST-8 cell proliferation assay, with 5 × 10³ cells per well of AsPC1, BxPC3, AsPC1/luc, and BxPC3/luc incubated with photosensitizer for 4 hours, and then irradiated as described in the preceding section. After incubation for 24 hours, cells were incubated for 2 hours with the Cell Counting Kit-8 (Dojindo) and absorption was measured with a microplate spectrophotometer (SPECTRA MAX340; Molecular Devices). Cell viability was expressed as a percentage of untreated control cells and the half maximal (50%) inhibitory concentration (IC₅₀) was calculated.

Apoptosis in the AsPC1/luc cells was assessed using the Caspase-Glo3/7 Assay Kit (Promega). AsPC1/luc cells were treated with photosensitizers (0.25–1 μmol/L) at 37°C for 4 hours, and irradiated. Analyses were performed at 4 hours after PDT by adding 100 μL of caspase-3/7 reagent to each well, mixing, and then incubating for 1 hour at room temperature. Luminescence was measured using a Lu mat LB 9507 instrument (EG&G BERTHOLD), and the data were normalized against the average value of nontreated cells. Relative luminescence units (RLU) of the caspase-3/7 activity are expressed as mean ± SD.

AsPC1/luc cells were treated with or without Mal₃-chlorin (0.25–1 μmol/L) for 4 hours with or without subsequent irradiation (13.9 J/cm²). At 1 hour after irradiation, cells were incubated with 100 μg/mL 2′,7′-dichlorofluorescein-diacetate (DCFH-DA, Sigma) at 37°C for 20 minutes, in the dark. The cells were washed two times with warm PBS and viewed with a fluorescence microscope (BZ-9000; Keyence). The fluorescence intensity scores of imaged cells were evaluated using the associated software (21). Furthermore, H₂O₂ released into media collected from AsPC1/luc cells with or without PDT was determined using ROS-Glo H₂O₂ assay (Promega). Cells were incubated with each photosensitizer (0.25–2 μmol/L) for 4 hours and irradiated. At 1 hour after PDT, cells were incubated in a H₂O₂ substrate solution at 37°C for 3 hours and then luciferin detection reagent at room temperature for 1 hour. Luminescence was measured using a Lu mat LB 9507 instrument (EG&G BERTHOLD) and the RLU of H₂O₂ concentrations are expressed as mean ± SD, relative to the average value of nontreated cells.

Female nude mice (BALB/c Slc-nu/nu) aged 4 weeks and weighing 15 to 20 g were obtained from Japan SLC and acclimated to the laboratory for 1 week. They were housed at five animals per cage on pulp-chip bedding in an air-conditioned animal room at 22 ± 2°C and 55 ± 5% humidity. All mice were maintained under specific pathogen-free conditions with a 12-hours light/dark cycle. To prepare the in vivo peritoneal dissemination model, AsPC1/luc cells were injected intraperitoneally with 1 × 10⁶ cells in 200 μL PBS. All animal experiments were performed under protocols approved by the Institutional Animal Care and Use Committee of Nagoya City University Graduate School of Medical Sciences.

The accumulation of photosensitizer in the in vivo models was examined by optical-filter imaging (Longpass Filter VL0470, cut-on; 470 nm, Asahi Spectra Inc.) and using a semiconductor laser with a VLD-M1 spectrometer (M&M Co., Ltd.) that emitted a laser light with a peak wavelength of 405 ± 1 nm and a light output of 140 mW. The spectrometer and its accessory software (BW-Spec V3.24; B&W TEK, Inc.), used to analyze the spectrum waveform, revealed an amplitude peak (relative fluorescent intensity) at 505 nm for autofluorescence and 655 nm for photosensitizers. The relative intensities of the photosensitizers were also measured spectrophotometrically. To reduce measurement error, we divided the relative fluorescence intensity by the autofluorescence intensity to compare relative fluorescence intensity ratios of the tested photosensitizers in the target tissue.

One week after injection of the cells, bioluminescence was measured using a multifunctional in vivo imaging system (IVIS; MIIS, Molecular Devices Japan) to confirm the cell implantations. The mice were randomly allocated to three groups. Then, the mice were given an intraperitoneal injection with photosensitizer (talaporfin and Mal₃-chlorin) at 1.25 μmol/kg in saline. Four hours later, the mice were irradiated with 660-nm LED light (Opto Code Corporation) at a dose of 13.9 J/cm². This light device was placed just above the abdomen so that the irradiation spot on the abdominal skin was 1.77 cm² (diameter 1.5 cm), thus covering the lower abdomen and avoiding the liver. Treatment was performed on day 0 and 7, as shown in Fig. 4A. To visualize the peritoneal tumors in mice, IVIS was performed as follows. The mice were anesthetized and subsequently intraperitoneally injected with D-luciferin (150 μg/g body weight, Wako, Ltd.) in D-PBS. Two minutes later, a monochrome photograph was acquired followed immediately by bioluminescence acquisition for 20 seconds and in vivo fluorescence imaging. From this, total luminescence flux from the region of tumor on days 0, 7, 14 was quantified using MetaMorph-MIIS software (Molecular Devices Japan) to monitor tumor growth. The bioluminescence images were exported as false-colored images using matched visualization scales. At the end of the experiment on day 21, mice were sacrificed and ascites volumes were measured. Another in vivo experiment was also conducted on the mice peritoneal tumor tissue to analyze apoptosis, whereby PDT with each photosensitizer was administered only on day 14 followed by excision of the tumor tissue on day 15 and fixation in 10% phosphate-buffered formalin.

For immunohistochemistry, the sections were immunostained as previously described (22). In brief, apoptotic cells were detected by terminal deoxy nucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) assay as well as cleaved caspase-3 IHC. TUNEL assays were performed using an In Situ Apoptosis Detection Kit from Takara (Otsu, Japan). For detection of cleaved caspase-3, deparaffinized sections of tumor from mice peritoneum were incubated with 1:100 diluted cleaved caspase-3 antibody (Cell Signaling Technology). Antibody binding was visualized by a conventional immunostaining method using an autoimmunostaining apparatus (HX System, Ventana).

The statistical significance of differences was determined using the Dunnett's test, Student t test, or the χ² test as appropriate. Differences were considered statistically significant at P < 0.05. Data are expressed as mean ± SD.

We first examined the uptake of talaporfin and Mal₃-chlorin in AsPC1/luc cells using flow cytometric analysis, and then measured intensity of the characteristic red fluorescence at the single-cell level. The cells treated with Mal₃-chlorin showed a stronger signal (mean intensity: 1,545) than both the nontreated cells (mean intensity: 44) and the talaporfin-treated cells (mean intensity: 80), indicated that more Mal₃-chlorin was taken up by cancer cells compared with talaporfin (Fig. 1C).

We next evaluated the cell death induced by PDT with Mal₃-chlorin. As shown in Fig. 2A, PDT with Mal₃-chlorin induced cell death. However, neither Mal₃-chlorin nor light applied alone caused any toxicity in AsPC1/luc cells. Similar results were obtained with the other cell lines (AsPC1, BxPC3, and BxPC3/luc; data not shown). Furthermore, neither Mal₃-chlorin nor talaporfin without irradiation induced any cytotoxicity (Fig. 2B; Supplementary Fig. S1, respectively). We next examined the IC₅₀ of talaporfin and Mal₃-chlorin at 24 hours after PDT. As summarized in Table 1, PDT with Mal₃-chlorin induced cell death with 5 to 27 times more cytotoxicity to all cancer cells than PDT with talaporfin. At the same time, in AsPC1/luc cells, we confirmed that the luciferase luminescence intensities reduced in parallel with cell viability by luciferase assay (Supplementary Fig. S2).

Moreover, the level of apoptosis after PDT with Mal₃-chlorin was increased in a dose-dependent manner compared with after PDT with talaporfin, whereas neither Mal₃-chlorin nor talaporfin alone induced apoptosis (Fig. 2C). To further explore these cytotoxic effects, we then analyzed the levels of intracellular ROS in AsPC1/luc cells using the DCFH-DA assay and found significantly elevated ROS intensities after PDT with Mal₃-chlorin (Fig. 2D and E). Likewise, the H₂O₂ assay also showed dose-dependent increases in ROS generations after PDT with Mal₃-chlorin that were higher than the increases observed after PDT with talaporfin (Fig. 2F).

We examined the ability of Mal₃-chlorin to accumulate in xenograft tumors established by subcutaneously implanting AsPC1/luc pancreatic cancer cells into mice. At 2 weeks after the cell injections, we first confirmed successful cell implantation by bioluminescence imaging (Fig. 3A). Then, the peritoneal cavities were washed with saline and imaged under white light (Fig. 3B), followed by fluorescence imaging of the same view at 4 hours after intraperitoneal injection of 1.25 μmol/kg Mal₃-chlorin. We observed the characteristic red fluorescence of Mal₃-chlorin in the disseminated peritoneal tumors (Fig. 3C). Similarly, following an intraperitoneal injection of 1.25 μmol/kg talaporfin or Mal₃-chlorin, tumor tissue and some organs in each group were excised and observed under white light and fluorescence imaging (Fig. 3D). We then measured the biodistribution of talaporfin and Mal₃-chlorin in xenograft models by spectrophotometry. Notable accumulation of Mal₃-chlorin in the tumor tissue was detected (Fig. 3E), while among the organ tissues, Mal₃-chlorin tended to accumulate in liver and stomach, although at a much lower level than those in tumor. The stomach accumulation was measured at the lower stomach due to the strong autofluorescence of the upper stomach.

To investigate the antitumor effects of PDT with Mal₃-chlorin on intraperitoneal dissemination in vivo, each PDT was performed on a pancreatic cancer model established by intraperitoneal implantation of AsPC1/luc cells into nude mice, as shown in Fig. 4A. All animals remained healthy throughout the experimental period, and there was no significant difference in the mean body weight among the groups at the end of the study (Supplementary Fig. S3). We detected the inhibitory effect of PDT with Mal₃-chlorin on intraperitoneal dissemination with bioluminescence imaging (Fig. 4B), although there was no effect on abdominal skin (Supplementary Fig. S4). Numeric evaluations revealed that PDT with Mal₃-chlorin significantly suppressed tumor growth compared with control treatment (P = 0.036) and tended to suppress it more than PDT with talaporfin (P = 0.074; Fig. 4C). We also observed that PDT with Mal₃-chlorin tended to inhibit the incidence of ascites formation (Supplementary Table S1) and the mean volume of ascites compared with mice in the control and PDT with talaporfin group (P = 0.066 and P = 0.159, respectively; Fig. 4D).

Peritoneal tumor tissues from the PDT with the Mal₃-chlorin group showed some chromatin-condensed single cells, therefore TUNEL assays were performed and positivity was found in such dead lesions (Fig. 5A). PDT with Mal₃-chlorin significantly increased the TUNEL apoptotic indices compared with control treatment or PDT with talaporfin (Fig. 5B), and additional immunohistochemistry revealed some TUNEL-positive cells expressing cleaved caspase-3 (Fig. 5A). In contrast, cleaved caspase-3–positive cells were scarcely found in peritoneal tumors of control and PDT with talaporfin-treated tumors.

The poor outcome of pancreatic cancer is due, at least in part, to peritoneal dissemination (3, 4). In pancreatic cancer patients with peritoneal dissemination, the lack of effective chemotherapeutic and targeted agents highlight the urgent need for novel treatment strategies. Against this background, PDT is emerging as a clinically promising locoregional therapy for pancreatic cancer that could also provide aggressive treatment for peritoneal dissemination in partnership with chemotherapy. Such therapy could also be useful to control ascites resulting in improved quality of life during palliative treatments. Supporting this promise, we demonstrated herein a suppressive effect of PDT with Mal₃-chlorin in a mouse model of disseminated peritoneal pancreatic cancer, based on in vivo luminescent imaging and measurement of ascites volume. We could not elucidate the median survival time in the scope of this study within the protocols approved by the Institutional Animal Care and Use Committee. However, we expect that PDT with Mal₃-chlorin could improve survival rates in such experimental models compared to untreated mice due to the reduced volumes of both peritoneal tumor and ascites.

At clinical sites, the tumoritropic characteristics of photosensitizers become crucial in reducing the damage to adjacent normal tissue. There have been many attempts to develop new photosensitizers that show preferential accumulation within the target tumor tissue through combinations with various active targeting approaches, such as conjugation with peptides or antibodies (23–26), incorporation within liposomes (27, 28), and encapsulation within polymeric nanoparticles (29–32). To initiate accumulation in the target tumor, G-chlorin was initially synthesized by linking glucose to the photosensitizer chlorine based on tumors consuming higher levels of glucose than normal cells and indeed, G-chlorin showed improved cancer cell targeting compared to chlorin and other photosensitizers (5, 17, 18). For clinical use, we recently developed Mal₃-chlorin introduced water-solubility. As for the ¹O₂ generation efficiency and photocytotoxicity, Mal₃-chlorin was proved to have a very high performance comparable to that of G-chlorin. In addition, the crucial water-soluble advantage of Mal₃-chlorin allows its injection without the use of toxic organic solvents, with hydrophilicity similar to that of talaporfin (20).

Talaporfin, a second-generation photosensitizer, shows improved efficacy compared with first-generation photosensitizers (12). Now, our in vivo results indicate that Mal₃-chlorin–mediated PDT tends to be more effective than talaporfin-mediated PDT at a low photosensitizer concentration (1.25 μmol/kg). Although we could not find the effects of PDT with talaporfin, we speculate that talaporfin required over five times more dose (>6.25 μmol/kg) than the dose which was used in the present study, according to a previous study (33, 34). This indicated that Mal₃-chlorin was more efficiently taken up by pancreatic cancer cells than talaporfin. In support of this, previous PDT experiments with photosensitizers including talaporfin used irradiation doses of >100 J/cm² in carcinoma xenografts models (33, 35, 36), whereas in this study, PDT with Mal₃-chlorin showed antitumor effects in vitro and in vivo at a low irradiation dose of 13.9 J/cm². The high cancer cell selectivity of Mal₃-chlorin could, therefore, substantially reduce the total energy of light irradiation needed in PDT treatment.

Glucose transporter 1 (GLUT1) is believed to maintain basal glucose transport in most cell types (37, 38) and is predominant in many types of cancer, including pancreatic cancer (39). We investigated the correlation of GLUT1 to the uptake of Mal₃-chlorin using a pharmacological GLUT1 inhibitor, WZB117 (40), and found that WZB117 inhibited the uptake of Mal₃-chlorin (Supplementary Fig. S5). The experiment also hinted at an involvement for GLUT1 in the mechanism by which Mal₃-chlorin is taken up by cancer cells, although what underlies such a mechanism remains unclear. In addition, several recent reports showed that mutated KRAS caused higher glucose uptake and accumulation possibly by upregulation of GLUT1 (41, 42). KRAS mutations occur in more than 90% of pancreatic cancers, thus G-chlorin and Mal₃-chlorin might produce more effects in pancreatic cancer than in other cancers expressing wild-type KRAS.

Importantly, Mal₃-chlorin might have additional, prospective value as a photodynamic diagnosis (PDD) reagent to detect and discriminate tumor locations due to its high cancer cell selectivity. Our in vitro findings revealed dose- and cell number-dependent increases in fluorescence intensities with Mal₃-chlorin (Supplementary Fig. S6). However, it is currently too difficult to evaluate the correlation between tumor fluorescence and either the dose of Mal₃-chlorin or tumor size in vivo because of mice interindividual differences in microvascular construction and tumor microenvironment. Further studies are therefore needed to establish the usefulness of Mal₃-chlorin in PDD.

It should also be noted that the xenograft peritoneal cancer mouse model used in the present study shows differences in microenvironment compared to tumors from human patients, including in the vasculature, stromal cells, and immune cells (43, 44). We therefore plan to further investigate the effects shown here in a future study using a spontaneous peritoneal disseminated model.

Collectively, our results indicated that PDT with Mal₃-chlorin could effectively suppress the growth of peritoneal disseminated xenograft mouse models of human pancreatic cancer cells through apoptosis without observable damage to the adjacent normal tissue. Furthermore, the newly developed water-soluble Mal₃-chlorin had superior cancer cell selectivity without skin phototoxicity (Supplementary Fig. S4). Thus, Mal₃-chlorin stands as a candidate for the next-generation photosensitizers and our findings may offer new opportunities for the clinical use of PDT in treating peritoneal dissemination of pancreatic cancer.

No potential conflicts of interest were disclosed.

Conception and design: A. Kato, H. Kataoka, M. Natsume

Development of methodology: A. Kato, K. Hayashi, N. Hayashi, M. Yoshida

Acquisition of data (provided animals, acquired and managed patients, provided facilities, etc.): A. Kato, M. Yoshida, S. Takahashi

Analysis and interpretation of data (e.g., statistical analysis, biostatistics, computational analysis): A. Kato, M. Yoshida, Y. Fujita, M. Natsume, A. Naiki-Ito

Writing, review, and/or revision of the manuscript: A. Kato, H. Kataoka, S. Yano, K. Miyabe, H. Kondo, Y. Hori

Administrative, technical, or material support (i.e., reporting or organizing data, constructing databases): A. Kato, S. Yano, M. Tanaka, I. Naitoh, T. Ban, T. Murakami, A. Nomoto

Study supervision: H. Kataoka, T. Joh

Other (synthesis of PDT photosensitizer): A. Narumi

This study was partially supported by JSPS KAKENHI grant numbers 26460947, the Translational Research Network Program of the Japan Agency for Medical Research and Development, AMED, 2015–2016 (to H. Kataoka), JSPS KAKENHI grant numbers 19350031, 25288028, the Japan–German Exchange Program supported by the JSPS and the Deutsche Forschungsgemeinschaft (DFG; to S. Yano), and Pancreas Research Foundation of Japan (to A. Kato).

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.