Photo-Induced Antitumor Effect of 3,6-Bis(1-methyl-4-vinylpyridinium) Carbazole Diiodide

We have applied a fluorescent molecule 3,6-bis(1-methyl-4-vinylpyridinium) carbazole diiodide (BMVC) for tumor targeting and treatment. In this study, we investigated the photo-induced antitumor effect of BMVC. In vitro cell line studies showed that BMVC significantly killed TC-1 tumor cells at light dose greater than 40 J/cm2. The fluorescence of BMVC in the tumor peaked at 3 hours and then gradually decreased to reach the control level after 24 hours. In vivo tumor treatment studies showed BMVC plus light irradiation (iPDT) significantly inhibited the tumor growth. At day 24 after tumor implantation, tumor volume was measured to be 225 ± 79 mm3, 2542 ± 181 mm3, 1533 ± 766 mm3, and 1317 ± 108 mm3 in the iPDT, control, light-only, and BMVC-only groups, respectively. Immunohistochemistry studies showed the microvascular density was significantly lower in the iPDT group. Taken together, our results demonstrated that BMVC may be a potent tumor-specific photosensitizer (PS) for PDT.


Introduction
Tumor-targeting therapy has emerged as an effective and attractive treatment for cancer. Among the various cancer-speci�c targets tested, telomerase has gathered much attention in recent years. Telomerase is detected in about 85% to 90% of cancer cells, but in a low level of normal cells [1]. e maintenance of telomere length by telomerase is required for unlimited proliferation of cancer cells. Telomere has been the target for the development of cancer probes, and telomerase inhibitors have been developed to inhibit telomerase activity and limiting cancer cell growth [2].
In the search for tumor-targeting agents, we have recently developed a �uorescent molecule 3,6-bis(1-methyl-4vinylpyridinium) carbazole diiodide (BMVC) for recognizing speci�c quadruplex structures such as the T 2 AG 3 telomeric repeats and inhibiting the telomerase activity [3][4][5]. Intriguingly, the �uorescence of BMVC detected in cancer cells was much stronger than that in normal cells, suggesting it to be a good candidate for a tumor-targeting agent [3]. e maximum absorption of BMVC is shied from 435 to 460 nm and the �uorescence intensity increases signi�cantly when BMVC interacts with DNA [5]. Because of the ability of telomerase inhibition, BMVC induces accelerated senescence of cancer cells [6].
Photodynamic therapy (PDT) is an effective treatment for cancerous and precancerous lesions [7]. e advantages of PDT are that it can be repeated in the same site if necessary, and it is less destructive than traditional surgery. PDT requires PSs that are activated by speci�c wavelengths of light. Illumination of tumor results in the destruction of cells due to a photochemical reaction. Reactive oxygen species, including singlet oxygen and free radicals, are generated by the photochemical reaction [8,9]. is photochemical reaction is capable of inducing cellular apoptosis and necrosis, by evoking oxidative stress [10]. In addition, PDT may cause tumor cell death indirectly by damaging tumor-associated vasculature or activating host immune responses [9,11]. Previously we have investigated the �uorescence resonance energy transfer (FRET) binary system that consists of BMVC conjugating porphyrin [12]. We found that PDT efficiency is greater when excited by 470 nm of light as compared to 510 nm of light. It is surprised that better PDT efficiency is observed upon exciting the FRET donor (BMVC) rather than the acceptor (porphyrin). is extra phototoxic effect could result from a type I photodynamic reaction [13][14][15] because we detected neither the characteristic spectral signal of singlet oxygen (1270 nm) from BMVC in D 2 O solution nor the decrease of 3-diphenylisobenzofuran (DPBF) signal in organic solvent. Here, we have examined the phototoxicity mechanism of BMVC and illustrated its potential to be used as a photosensitizer (PS) for photodynamic therapy (PDT).
Despite the potential advantages in clinical application, PDT has several limitations that hinder its wide clinical acceptance. Among them, sustained skin photosensitivity and low tumor selectivity are two major problems for the PSs [16]. e purpose of this study was to investigate the photochemical effects of BMVC on tumor cells. Cellular cytotoxicity of BMVC was evaluated in TC-1 cell line. e antitumor effect of BMVC combined with a speci�c wavelength of light was investigated in the animal model.

Cytotoxicity Assay.
Cells were grown in 96-well plates (2000 cells/well) and then incubated with BMVC for 6 hours and changed to fresh culture medium (without phenol red), and the cytotoxicity was determined by the Alamar Blue assay [20]. 10% Alamar Blue in 200 L of culture media was added to each test well and analyzed spectrophotometrically at the absorbance difference between 570 and 600 nm.

Free Radical
Assay. e light source for measurement of PDT effect was a white light (a 200 W Xenon lamp that passes through a 400-700 nm mirror module) �nally, the light power was 100 mW/cm −2 on the sample surface.

Animal Study.
All animal studies were approved by the Institutional Animal Care Committee. Male C57BL/6 (B6) mice, weighing 20-28 g, were supplied by the University Laboratory Animal Center and were allowed free access to food and water. Around 5 × 10 4 TC-1 cells in 100 L of HBSS were injected s.c. into B6 mice. BMVC at concentration of 5 mg/kg was injected (single dose) i.p. aer the tumor grew to around 50 to 100 mm 3 . e length ( ) and width ( ) of the TC-1 tumor mass and the body weight of mice at 1, 7, 10, and 14 days aer BMVC injection were recorded. Tumors were measured in two orthogonal directions, and the tumor volumes were estimated as ( 2 )/2 [18,21]. All animals challenged with TC-1 cells had developed a palpable tumor at day 10.

Distribution of BMVC in Mice.
Mice were sacri�ced at 1, 2, 3, 5, 12, 24, and 72 hours aer BMVC injection, and the tissues were removed, homogenized, and analyzed by a �uorescence microplate reader (excitation wavelength at 460 nm). Tissues from mice 3 hours aer BMVC injection were collected, washed with PBS, and then analyzed by the �ow cytometry (FACS Calibur and Cell�uest soware, BD Biosciences).

Photodynamic erapy
2.7.1. In Vitro Study. Cells were incubated with 5 M BMVC in darkness for 6 hours at 37 ∘ C. Aer wash with PBS, the cells were immediately exposed to different doses of light at 445 nm (LDM445, Pkebtunax, Germany). Cell toxicity was determined 24 hours aer light treatment with energy doses of 1, 5, 10, 20, 40, 60, and 80 J/cm 2 .

2.7.2.
In Vivo Study. BMVC (5 mg/kg) was injected to the mice when the size of tumor reached 0.8 cm in diameter (about 14 days aer TC-1 injection). ree hours aer BMVC injection, mice were subjected to interstitial light treatment as described before [22,23]. In brief, a needle (21 G) was pushed percutaneously into the tumor, and the �ber was passed through the needle. e tip of the �ber was managed to be at about 1 mm outside of the needle tip to ensure it actually touched the tumor tissue. A single dose of 150 J was delivered to the tumor by a 445 nm diode laser (LDM445, Pkebtunax, Germany).

Immunohistochemistry.
Cryosections of the tumor at 10 m thickness were �xed with 10% formalin, washed, and then immunostained for -smooth muscle actin ( -SMA, 2 g/mL, DAKO, Glostrup, Denmark). e number of -SMA positive cells was counted in three nonoverlapping regions. Unit counts were expressed as the number of -SMA positive unit per mm 2 of tumor tissue.

Statistical
Analysis. Student's t-test was used to evaluate the response to a change in conditions. Data were subjected to statistical analysis using the SPSS for Windows version 10.

Results
3.1. BMVC Fluorescence Spectra. e chemical structure of BMVC was shown in Figure 1(a). When TC-1 cells were cultured in the presence of BMVC, the �uorescence signals were detected mainly in the cell nuclei (Figure 1(b)). Fluorescence emission spectra of BMVC were depicted in Figure 1(c). It clearly demonstrated that the �uorescence intensity of BMVC increased for almost 2 orders of magnitude upon interacting with cells.

In Vitro Cytotoxicity and Phototoxicity.
Cytotoxic effects of BMVC to TC-1 cells were evaluated. Low dose BMVC treatment did not have apparent toxicity to the cells. As shown in the Figure 2, ∼80% of cells were still viable when the BMVC concentrations were below 5 M. When the BMVC concentrations were above this level, cytotoxicity effects became obvious. e cell viability decreased to ∼60% when cells were treated with 20 M of BMVC. Since 5 M of BMVC treatment did not show signi�cant cytotoxicity to the cells, we use this concentration for the subsequent PDT treatment. Photo-induced cytotoxicity of BMVC was evaluated as a function of light dose. Figure 3 showed the photo-induced cytotoxicity of BMVC upon the treatment of 0, 1, 5, 10, 20, 40, 60, and 80 J/cm 2 light doses in TC-1 cells. It was found that BMVC coupled with low light doses (below 20 J/cm 2 ) did not show apparent cytotoxicity. However, cell survival of the BMVC coupled with higher light doses (40, 60, and 80 J/cm 2 ) appreciably decreased to 66.9%, 57.5%, and 52.1%, respectively.

Distribution and Kinetics of BMVC in the Tumor and
Normal Mice Tissues. Liver and kidney are the main organs responsible for drug metabolism and excretion. We further measured the distribution of BMVC in the tumor, liver, and kidney tissues by �ow cytometry and �uorescence microscopy at 3 hours aer BMVC injection ( Figure 5). e results showed that the majority (90.8%) of tumor cells were positive for BMVC �uorescence. On the other hand, only 47.2% of the liver cells and 30.9% of the kidney cells were positive for BMVC �uorescence, although the �uorescence intensity was stronger in liver and kidney than in the tumor. Kinetics of BMVC in the tumor and normal mice tissues were also investigated. Aer injection, the �uorescence of BMVC gradually increased in the tumor tissues, peaked at 3 hours, and then gradually decreased to reach the control level aer 24 hours (Figure 6(a)). Similarly, the �uorescence of BMVC in the liver and kidney increased aer drug injection, peaked at 24 and 12 hours, respectively, and then gradually decreased (Figures 6(b) and 6(c)). At 72 hours aer injection, the BMVC �uorescence levels in the liver and kidney were still above the control level. �owever, the BMVC �uorescence in the brain, lung, and muscles showed no appreciable difference from the control tissues (data not shown).  of acute toxicity in this BMVC dose. Figure 7 showed the growth curves of tumors under different treatments. It clearly showed that BMVC plus light (iPDT) signi�cantly inhibited the tumor growth in vivo ( ). At day 24 aer tumor implantation, tumor volume was measured to be 22 ± 79 mm 3 , 2 42 ± 181 mm 3 , 1 33 ± 766 mm 3 , and 1317 ± 1 8 mm 3 , respectively, in the iPDT, control, light-only, and BMVC-only groups. ere is only 6.76% evolution of tumor volume in iPDT group between day 14 and 24. Although the tumors in the light-only and the BMVC-only groups were also smaller than the control group, the difference did not reach statistically signi�cant level.

Immunohistochemistry.
PDT is known to cause microvascular destruction in tumor tissues. e direct cytotoxic activity and microvascular damage contribute to the destruction of tumor [9,28]. To investigate the microvascular density in the tumors, we measured -smooth muscle actin ( -SMA) positive cells in different groups of tumors. Figure 8 showed the �uanti�cation of -SMA positive cells under different conditions. It is found that the microvascular density was signi�cantly lower aer both the light-only and the iPDT treatments. On the other hand, BMVC treatment showed minor effect on the microvascular density of tumors.

Discussion
PDT has emerged as an effective means for cancer therapy. PSs are an essential element of PDT. In general, the photoactivation of the PS leads to an oxygen dependent oxidativereaction resulting in cellular photodamage. e mechanisms of PDT involve direct oxidation of biological targets through hydrogen abstraction or electron transfer to yield radical chain reaction (type I reaction), as well as oxidation mediated by singlet oxygen ( 1 O 2 ) through energy transfer from triplets to molecular oxygen to initiate oxidative damage (type II reaction). In this study, we investigated the possibility of using BMVC as a PS for PDT. Illuminating TC-1 cells with . ese results suggest that BMVC is a useful PS for PDT. Tumor-targeting therapy is the treatment of cancer cells without injuring the normal cells. However, one of the major drawbacks of current PSs is the lack of tumor selectivity. Many studies have been focused on the modi�cation of spe-ci�c porphyrin structures to achieve better tumor selectivity. Furthermore, the �uorescence of BMVC detected in cancer cells was much stronger than that in normal cells, letting it to be a good candidate for the tumor-targeting agent [3]. In this study, we showed that 90.8% of tumor cells were positive for BMVC �uorescence at 3 hours a�er i.v. injection ( Figure 5(a)). On the other hand, only 47.2% of the liver cells and 30.9% of the kidney cells were positive for BMVC �uorescence (Figures 5(b) and 5(c)). Together, these results indicated that BMVC had certain degree of tumor selectivity and could be implemented as a tumor-targeting PS for PDT.
In addition to be used in the treatment of cancer, PSs were also implemented in the diagnosis of cancers. In 1924, red �uorescence of porphyrin was observed in experimental rat sarcomas by Policard [29]. is is the �rst observation of PS �uorescence in tumors. e followed experiments con�rmed that some PSs accumulated in cancer cells and were possibly been used in cancer detection (photodynamic diagnosis, PDD) [30]. In PDD, �uorescence in cancer cells could be observed by either �uorescence spectroscopy or �uorescence imaging. Clinical application of PDD had been shown in premalignant oral tissue screening [31]. In this study, we noted that the �uorescence intensity of BMVC increased signi�cantly upon interacting with tumor cells (Figure 1(c)). e emission spectrum was different between free BMVC molecules and BMVC plus TC-1 cells. ese results were in line with our previous molecular chemical studies that showed strong �uorescence of BMVC that peaked at approximately 550 or 575 nm in the presence of DNA structure [5]. Since the �uorescence of BMVC signi�cantly enhanced upon interacting with tumor cells and that 90.8% of in vivo tumor cells exhibited strong BMVC �uorescence, BMVC could be used as a potential molecule for PDD of cancer cells.
Although BMVC-PDT is effective in vitro, there are some limitations in the in vivo animal studies. e absorption wavelength (445 nm) of BMVC is insufficient to pass through the whole bulk of tumor cells. To overcome this problem, we used �ne-needle interstitial light irradiation to shine the light into tumor tissues. is �ne-needle interstitial PDT (iPDT) which combines with the BMVC treatment signi�cantly inhibited tumor growth (Figure 7). ere were 2 possible mechanisms to explain this signi�cant tumor inhibition. One is the direct effect of BMVC on the tumor telomeres. As reported in our previous studies, BMVC was able to suppress the telomerase activity and induce senescence of cancer cells [3,4]. Tumor formation and progression can be suppressed by BMVC treatment alone. e other is the effect of PDT may destruct the intratumoral vasculature and result in tumor ischemia and death by decreases in perfusion [21]. Our immunohistochemistry results in this study con�rmed the destruction of intratumoral vasculature by BMVC-PDT ( Figure 8). Intriguingly, tumors receiving light irradiation alone also showed signi�cant reduction of intratumoral microvascular density. e thermal effect might decrease the vascular density in light irradiation alone group because we treated the mice by a high energy direct treatment. e mild thermal therapy has the ability to change vascular perfusion and oxygen supplied within the tumor microenvironment [32]. Nevertheless, the iPDT group has the smallest tumor volume certainly. In our study, we showed the photochemical effect occurred by BMVC treated combined with light. e mechanism behind this light induced microvascular destruction remains elusive and needs further investigation.

Conclusions
We described a tumor-targeting therapy for the application of BMVC. In summary, the distinct properties of this �uorescent molecule provide a design of PS for PDT treatment. PDT experiments showed that BMVC-PDT signi�cantly inhibited Our results demonstrated that BMVC may be a potent tumor-speci�c PS for PDT. A number of potential re�nements can be incorporated into future studies based on this effect of BMVC, such as conjugate with other PSs to increase phototoxic effects and pi-conjugation lengths, which induce the bathochromic shi to enhance the penetration depth of the PDT laser.