DNA Duplex-Based Photodynamic Molecular Beacon for Targeted Killing of Retinoblastoma Cell

Abstract

Purposes: Retinoblastoma (RB) is the most common primary intraocular malignancy of infancy. An alternative RB treatment protocol is proposed and tested. It is based on a photodynamic therapy (PDT) with a designed molecular beacon that specifically targets the murine double minute x (MDMX) high-expressed RB cells.

Methods: A MDMX mRNA triggered photodynamic molecular beacon is designed by binding a photosensitizer molecule (pyropheophorbide-a, or PPa) and a black hole quencher-3 (BHQ3) through a complementary oligonucleotide sequence. Cells with and without MDMX high-expression are incubated with the beacon and then irradiated with a laser. The fluorescence and reactive oxygen species are detected in solution to verify the specific activation of PPa by the perfectly matched DNA targets. The cell viabilities are evaluated with CCK-8 and flow cytometry assay.

Results: The fluorescence and photo-cytoxicity of PPa is recovered and significantly higher in the MDMX high-expressed Y79 and WERI-Rb1 cells, compared to that with the MDMX low-expressed cells.

Conclusions: The synthesized beacon exhibits high PDT efficiency toward MDMX high-expressed RB cells. The data suggest that the designed beacon may provide a potential alternative for RB therapy and secures the ground for future investigation.

Retinoblastoma is the most common primary intraocular malignancy of infancy and is virtually fatal if left untreated due to metastasis. The current therapies include radical enucleation and conservative treatments such as local thermotherapy, cryotherapy, brachytherapy or chemotherapy.^1–3 The effectiveness of the treatments is significantly limited due to the delicate anatomical structure of the site and potential late effects. Here lies the need of improved therapies that specifically target retinoblastoma cells while preserving the healthy ocular tissue structure. 

There are two major pathways in cell death, necrosis and apoptosis.⁴ Necrosis is a traumatic cell death that involves minimal autoregulation and is often associated with severe immune responses such as inflammation, swelling, and pain, and is to be avoided in retinoblastoma therapies in general. Apoptosis is a process of programmed cell death. It is initiated either intrinsically due to cellular stress or extrinsically when receiving signals from other cells. It is part of a typical lifecycle of organism with no traumatic symptoms related to the process. Given the delicate anatomical structure, apoptosis is a preferred cell killing mechanism for treating retinoblastoma. 

Photodynamic therapy (PDT) with the photosensitizer Visudyne has been approved by the US Food and Drug Administration for treatment of age-related macular degeneration. It has been suggested that the therapeutic mechanism may also be useful for various ocular tumors, such as retinoblastoma, choroidal hemangioma, and retinal astrocytoma.^5–8 With preadministered photosensitizers, PDT can destroy target cells when light activates the drug to generate cytotoxic reactive oxygen species (ROS), mainly singlet oxygen (¹O₂).^9–11 However, despite the many advantages of PDT over conventional treatments, development of selective PDT agents to improve therapeutic efficiency and reduce side effects remains a significant challenge.^12–15 

The extent of PDT cytotoxicity can be controlled by the irradiation light “dosage.” Furthermore, PDT treatment can be repeated as necessary with minimal side effects or multidrug resistance. It is thus possible, by controlling the PDT dose and using target-specific molecular beacon, to selectively induce apoptosis in the retinoblastoma cells for therapeutic purpose. 

Tumors are associated with abnormal gene expression. Laurie et al.¹⁶ showed that retinoblastoma selectively amplifies murine double minute x (MDMX),^16,17 an antagonist of p53. The aberrant expression leads to an inactivation of the p53 pathway via the formation of inactive p53-MDMX complexes, contributing to the development and progression of retinoblastoma formation.¹⁸ Murine double minute x expression level is closely correlated with MDMX mRNA level and MDMX promoter activity in different tumor cell lines.¹⁹ Murine double minute x was identified as a therapeutic target for p53 reactivation in tumors, especially, as a chemotherapeutic target for the treatment of retinoblastoma.¹⁷ Nutlin-3 is the currently available MDMX inhibitor. It can block the MDMX–p53 interaction. However, its affinity to MDMX is low.^16,20,21 Systemic administration of Nutlin-3 to treat the tumors with MDMX high-expression is not feasible due to its pharmacokinetics and toxicity in multiple organs.²² 

Zheng et al.^23–26 introduced a hairpin-based photodynamic molecular beacon. It comprises a tumor-specific linker, a photosensitizer, and a ¹O₂ quencher. The photoactivity of the photosensitizer is silenced until the linker interacts with a target molecule, protease,^23–26 or RNA.^27,28 This strategy minimizes the cytotoxicity and organs toxicity of the drug to the none-target tissue, and thus may be particular advantage for treating ocular tumors. Clo et al.²⁹ designed a DNA duplex-based photodynamic molecular beacon that can be selectively triggered by oligonucleotides hybridization ex vivo^28–30 by means of Watson–Crick base pairing. The photosensitizer and the quencher are kept in close proximity in the “off state” based on FRET (fluorescence resonance energy transfer) effect by DNA-programmed assembly. To switch the photosensitizer “on,” a more complementary strand is introduced to hybridize with the quencher moiety and causes photosensitizer moiety release, allowing a restoration of the ¹O₂ production. In addition, the ¹O₂ luminescence recovery efficiency of DNA duplex-based photodynamic molecular beacons is higher than that of traditional hairpin-based photodynamic molecular beacons.³¹ Recently, a zipper molecular beacons based on the similar mechanism has also been proposed.³² These approaches enhance the target selectivity in PDT over its intrinsic targeting mechanism. 

In this study, to selectively destroy retinoblastoma cells, we designed a DNA duplex-based MDMX mRNA triggered photodynamic molecular beacon based on the amplification of MDMX gene expression in retinoblastoma for PDT (Fig. 1). The designed beacon is hybridized with two complementary DNA sequences (Fig. 1A). A formed toehold domain is used to initiate the displacement reaction of the MDMX mRNA and the beacon. The reaction would open the duplex DNA strand to trigger the photosensitizer molecular on (Fig. 1B). As the first step of implementing the beacon as a therapeutic strategy for retinoblastoma treatment, we validate the targeting effect of MDMX mRNA-triggered photodynamic molecular beacon to retinoblastoma cells, and the PDT effect in vitro. 

Figure 1

 

Synthesis of photodynamic molecular beacon and its ¹O₂ generation triggered by MDMX mRNA. (A) Synthesis protocols of PPa-DNA1 (PD₁) and BHQ₃-DNA2 (BD₂). (B) Schematic of the beacon triggered by MDMX mRNA. (1) The formed PD₁-BD₂ with photosensitizer (PS) quenching; (2) MDMX mRNA binds to PD₁-BD₂ through a single-strand toehold domain; (3) PS switched on to yield O₂ after the displacement reaction.

Materials and Methods

Chemicals and Reagents

The following chemicals and reagents were used in our experiments: pyropheophorbide-a (PPa) (J&K Scientific Ltd, Nanjing, Jiangsu, China), BHQ-3 Carboxylic Acid (Biosearch Technologies, Inc., Navato, CA, USA), CCK-8 (Dojindo Laboratories, Kumamoto, Japan), fluoresceinyl cypridina luciferin analog (FCLA) (Tokyo Kasei Kogyo Co., Tokyo, Japan), Lipofectamine 2000 reagent (Invitrogen, Carlsbad, CA, USA), N-Hydroxysuccinimide (NHS), and 1-ethyl-3-(3-(dimethylaminopropyl)carbodiimide (EDC) (Sigma-Aldrich Corp., St. Louis, MO, USA). All the reagents were of analytical grade and used without further purification. Deionized water was used throughout. 

Preparation and Characterization of PD₁-BD₂

Synthesis of PPa-DNA1 (PD₁) and BHQ3-DNA2 (BD₂).

To synthesize PD₁, stock solution of PPa (1 mM) was prepared in DMSO (dimethyl sulfoxide). Labeling buffer (100 mM sodium carbonate, pH 7.5) was prepared by dissolving an appropriate amount of sodium carbonate in water and its pH titrated with HCl and/or NaOH. A 15mer-oligonucleotide (DNA 1) modified with amino groups at the 5′ end was obtained from Invitrogen. The oligonucleotide was dissolved in the labeling buffer to yield a 100-μM solution. The reaction mixture included (in order of addition): labeling buffer (162 μL), PPa (60 μL), EDC (12 μL, 100 mM), NHS (6 μL, 100 mM), and DNA 1 (60 μL, 6 nmol). The reaction mixture was incubated at the room temperature overnight and then purified with dialysis. To detect the intrinsic optical properties of PD₁, ultraviolet-visible (UV-VIS) spectrum was performed. The similar synthesis process was performed for BD₂ by using BHQ3 (42 μL,1.44 mM), 21mer-oligonucleotide (DNA2) (60 μL, 6 nmol), and labeling buffer (180 μL). The sequence of DNA1 and DNA2 were 5′-AAAACTGCCGCTTTT-3′ and 5′-ATCTTCAAAAGCGGCAGTTTT-3′, respectively. 

To hybridize the complementary strands, PD₁ (10 μL, 10 μM) and BD₂ (10 μL, 10 μM) in 100 mM sodium carbonate (pH 7.5) were heated in a thermocycler at 95°C for 5 minutes and cooled slowly to 5°C (2°C per minute). To characterize the duplex DNA complex, quenching efficiency of fluorescence and ROS production from PD₁ was measured. PD₁ with a fixed concentration 1 μM was mixed with BD₂ solutions at different concentration (0.5, 1, 2, 4 μM), then hybridized. Reactive oxygen species production was measured by the basic PDT chemiluminescence measurement system, which is based on measurement of chemiluminescence from FCLA when activated by PDT-generated ROS.^33,34 The chemiluminescence (532 nm) from FCLA was measured with 633-nm light excitation. The background was subtracted and the data were normalized relative to PD₁. The fluorescence spectra of PD₁ were obtained using an LS-55 fluorescence spectro-photometer (Perkin-Elmer, Waltham, MA, USA) with an excitation of 488 nm. The data of the fluorescence intensity were normalized relative to PD₁. 

Gel electrophoresis was used to confirm the synthesis of PD₁-BD₂. PD₁ (10 μL, 10 μM) and BD₂ (20 μL, 10 μM) were hybridized. DNA 1 (10 μL, 10 μM) and DNA 2 (20 μL, 10 μM) were hybridized to serve as positive control. Ten microliters of each hybridization production was loaded onto a 10% native polyacrylamide gel (29:1 acryl/bisacryl) in 1× Tris-borate-EDTA (TBE). Gels were run at room temperature for 1 hour at 120 V. The gel was conducted with silver staining and photographed by a Nikon 4500 digital camera. 

The Specificity Activation of PD₁-BD₂ in Solution

A short 21mer-nucleotide sharing the complementary sequence of the DNA2 (DNA target 1) and a short 21mer-nucleotide with single-base mismatch (DNA target 2) were synthesized (5′-AAAACTGCCGCTTTTGAAGAT-3′ and 5′-AAAACTGTCGCTTTTGAAGAT-3′). PD1-BD2 with a fixed concentration at 1 μM was mixed with the perfectly matched DNA (10 μM) and single-base mismatched DNA (10 μM). Reactive oxygen species production was measured by FCLA chemiluminescence with 633-nm excitation wavelength. The fluorescence intensity of PPa was measured with 488-nm excitation wavelength. 

Cell Line and Transfection

The human retinoblastoma cell lines Y79 and WERI-Rb1 were cultured in RPMI-1640 Medium (1640, GIBCO, Grand Island, NY, USA) supplemented with 10% fetal calf serum (FCS) in 5% CO₂, 95% air at 37°C in a humidified incubator. D407 (human retinal pigment epithelium cell line) cells were cultured with DMEM supplemented with 5% fetal bovine serum (FBS) in a humidified atmosphere of 5% CO ₂ at 37°C. 

Gene Silencing.

For gene silencing in Y79 and WERI-Rb1 cells, we used MDMX shRNA and negative control shRNA. pGPU6/GFP/Neo vector carrying MDMX shRNA (5′-GTGATGATACCGATGTAGA-3′) and nonsense sequence shRNA (negative control) (5′-GTTCTCCGAACGTGTCACGT-3′) were synthesized and constructed by Shanghai GenePharma Company (China). 

When the cells grew to 70% to 80% confluence, transfection was performed with 0.8 μg shMDMX using the Lipofectamine 2000 according to the manufacturer's instructions in serum-free medium. The serum-free medium was replaced with fresh culture medium 5 hours after and the cells were incubated for an additional 24 to 48 hours for expression. D407 cells, as a control, were also subjected to the same process. 

Total RNA Extraction and Semiquantitative RT-PCR.

To determine the level of MDMX expression in Y79, WERI-Rb1, and D407 cells, as well as their MDMX-knockdown ones, total RNA was extracted from the cells (2 × 10⁶ cells mL⁻¹) according to the manufacturer's instructions using the TRI reagent (Sigma-Aldrich Corp.). The RNA concentrations were determined by measuring the corresponding optical densities (OD) at 260 nm. First-strand cDNA was synthesized with the SuperScript II First-Strand Synthesis System for RT-PCR (Invitrogen). The primers used for semiquantitative RT-PCR of human MDMX mRNA were as follows: human MDMX, forward (FW), 5′-GCCTTGAGGAAGGATTGGTA; reverse (REV), 5′-TCGACAATCAGGGACATCAT; human GAPDH, FW, 5′-ACCACAGTCCATGCCATCAC; REV, 5′-TCCACCACCCTGTTGCTGTA. 

The murine double minute x level in Y79 was also assessed with the similar process to determine the reaction capability of PD₁-BD₂ with MDMX mRNA after the cells were incubated with PD₁-BD₂, MDMX shRNA, and DNA2. 

PD₁-BD₂ Triggered by MDMX mRNA in Living Cell

Cancer cells (1 × 10⁴ per well) growing in 35-mm Petri dishes were incubated with the targeting PD₁-BD₂ probe for 4 hours. After being centrifuged, the cells were rinsed with PBS and replaced with fresh cell medium. The cells were imaged by a commercial laser scanning microscope (LSM 510/ConfoCor 2, combination system Zeiss, Jena, Germany) equipped with a Plan-Neofluar 40×/1.3 NA Oil DIC objective. The targeting probe was excited at 633 nm, and its fluorescence emission was recorded through a 650- to 700-nm IR band-pass filter. 

Quantification of Fluorescence Intensity

Flow cytometry was used to quantitatively assess fluorescence intensity of PD₁-BD₂ in cells. Y79, WERI-Rb1, and their shMDMX cells were incubated with PD₁-BD₂ and PD₁ in a 12-well microplate for 4 hours. The cells were centrifuged and washed, then resuspended in ice-cold PBS. The fluorescence histogram of cells in different treatments was obtained from 10,000 cells with flow cytometry (FACScan, Becton Dickinson, Mountain View, CA, USA). 

Photodynamic Treatment and Cell Viability

To test the phototoxicity of PD₁-BD₂, the cell viability was evaluated with CCK-8 assay (Dojindo Molecular Technologies, Inc., Kumamoto, Japan). Y79 cells were incubated with the PD₁-BD₂ samples at various concentrations (PPa, 0.2∼0.6 μM) in a 96-well microplate (six wells per group) at 37°C under 5% CO₂ /95% air for 4 hours. Then, the mixtures were washed three times with PBS to remove PD₁-BD₂ in the solution. The tumor cells were irradiated with a 633-nm laser (30 J cm⁻² over 5 minutes). Cell cytotoxicity was assessed with CCK-8 24 and 72 hours after the laser irradiation. Y79 cells were also incubated with 0.6 μM PD₁-BD₂ and irradiated with various laser fluence (10, 30, and 50 J cm⁻²), then assessed with CCK-8 24 hours after the laser irradiation. OD: the absorbance value at 450 nm was read with a 96-well plate reader (INFINITE M200, Tecan, Switzerland), to determine the viability of the cells. 

Y79 and shMDMX Y79 cells were incubated with PD₁-BD₂, PD₁, and PPa samples at 0.6 μM PPa concentration in a 96-well microplate for 4 hours. Cell viability was assessed with CCK-8 48 hours after laser irradiation. With the similar protocol, the cell viabilities of Y79, WERI-Rb1, and D407 were also assessed after the PD₁-BD₂ incubation. 

Flow Cytometry Experiments for the Apoptosis of Cancer Cells

Quantitation of apoptosis and necrosis by Annexin V/PI staining was performed. The Y79 cells and WERI-Rb1 cells with different treatments were collected after 24 hours post PDT. Apoptotic cell death was determined using the BD ApoAlert annexin V-FITC Apoptosis Kit (Becton Dickinson, Biosciences) according to the manufacturer's instructions, with a BD FACS Canto TM II flow cytometer (Becton Dickinson). 

Statistical Analysis

Data are presented as means ± SD from at least three experiments. One-way ANOVA is used to compare the treatment effects. P < 0.05 is considered statistically significant. 

Results and Discussion

PD₁-BD₂ Preparation and Its Characterization Analysis

The two molecular building blocks, PD₁ and BD₂, were synthesized as shown in Figure 1A. The outcomes were then verified by their UV-VIS absorptions (Figs. 2A, 2B). In the PD₁-BD₂ preparation, the freshly synthesized PD₁ and BD₂ would be hybridized to form the complementary strands. To study the quenching efficiency of BD₂ on PPa due to the hybridization, BD₂ was titrated up to 4-fold in Mole concentration excessive that of PD₁. The fluorescence and ROS generation of PPa were then detected and compared. As shown in Figure 2C, when the concentration ratio of BD₂/PD₁ increased to 2, the fluorescence emission from PD₁ was reduced to 41.7% ± 2.3% of the initial value. A similar trend was observed in the quenching of PD₁ ROS generation (Fig. 2D). Reaction oxygen species generation was evaluated with FCLA by using its emission at 532 nm. In the presence of two folds of BD₂, the ROS production from PD₁ was reduced to 33.6% ± 2.8% of the initial value. 

Figure 2

 

Confirmation and characterization of PD₁-BD₂. (A, B) UV-VIS absorption spectra demonstrating the specialized PD₁ and BD₂ each possess the corresponding characterization of their building blocks. (C, D) PPa quenching shown as fluorescence (FL) and ROS production (n = 3). The FL intensity was measured with 488-nm excitation and normalized to that of PD₁ only (C). Reaction oxygen species production was evaluated with FCLA chemiluminescence (CL) (D). (E) Native polyacrylamide gel electrophoresis and silver stain of PD₁ with BD₂, and DNA1 with DNA2.

To further confirm the hybridization of PD₁ and BD₂, polyacrylamide gel electrophoresis (PAGE) was performed (Fig. 2E). The upper band of left lane indicates that the two strands of PD₁ and BD₂ has formed duplex after hybridization, while the lower band presents excess PD₁ and BD₂. The right lane is a positive control in which a corresponding free DNA was used at the same molar ratio. These results suggest that the hybridization of PD₁ and BD₂ forced PPa and BHQ3 into close contact, attenuating the singlet oxygen signal and fluorescence of PPa. 

PD₁-BD₂ was synthesized by PD₁ and BD₂ hybridization. To verify the gene-specific activation of PD₁-BD₂ by MDMX mRNA, the synthesized DNA target 1 and DNA target 2 were titrated to PD₁-BD₂. The resulting fluorescence intensity and ROS production were measured. Initially, the fluorescence intensity of PD₁-BD₂ was significantly less than that of PD1 (P < 0.05), but increased significantly (278% ± 10%) when hybridized with the perfectly matched DNA (P < 0.05). In comparison, only a 46% ± 5% increase was achieved with the single-base mismatched DNA (Fig. 3A). By using a 633-nm laser to irradiate the PD₁ and PD₁-BD₂, the ROS production was measured. A similar result with fluorescence was observed: the MDMX matched DNA strands significantly recovered the ROS producing of PD₁-BD₂ (Fig. 3B). These results suggest that PD₁-BD₂ was specific and sensitive to the matched DNA target. 

Figure 3

 

Recovery efficiency of PD₁-BD₂ after DNA sequence targeting. The recovered fluorescence emission (A) and ROS production (B) of PD₁-BD₂ by mixing with matched or single-base mismatched DNA targets. PD₁ was used as a positive control. *P < 0.05 (n = 3).

PD₁-BD₂ Triggered by MDMX mRNA in Living Cell

The MDMX mRNA expression and the mRNA knockdown efficiency of Y79, WERI-Rb1, and D407 cells were evaluated with the RT-PCR method (Fig. 4A). It shows that the Y79 and WERI-Rb1 cells had significantly higher MDMX mRNA expression, compared to that in shMDMX and D407 cells. The latter were then used as the negative controls. The knockdown efficiencies of the Y79 and WERI-Rb1 cells were 86% ± 2% and 77% ± 1.4%, respectively. The reaction capability of PD₁-BD₂ with MDMX mRNA was also assessed by comparing with DNA2 and MDMX shRNA in Y79 cells. The result (Fig. 4B) indicates that, through the hybridization reaction with MDMX mRNA, the PD₁-BD₂ reduced the mRNA level. 

Figure 4

 

Expression of MDMX mRNA and its specificity of PD₁-BD₂. (A) The expressions of MDMX mRNA in Y79, WERI-Rb1, and D407 cells were measured with RT-PCR after the cells were conducted with null, negative (Control shRNA), and positive (shMDMX) MDMX knockdown. The * indicates the significant difference of shMDMX group with its control. *P < 0.05 (n = 3). (B) The murine double minute x mRNA knockdown effect of PD₁-BD₂ in Y79 cells by comparing with the control, DNA2, and shMDMX Y79. (C, D) MDMX mRNA specificity of PD₁-BD₂ in Y79 and WERI-Rb1 cells. Left, confocal imaging of MDMX positive and negative cells. Right, fluorescence histogram of cells incubated with PD₁-BD₂ and PD₁ detected by flow cytometry. Scale bars in C and D: 10 μm.

To confirm whether the photodynamic beacon can be specifically activated by MDMX mRNA in cancer cells, Y79 and WERI-Rb1 cells were incubated with PD₁-BD₂. The cells with high MDMX expression (Figs. 4C, 4D) showed a more prominant fluorescence signal compared to that in the shMDMX cells. To further confirm the target specificity, flow cytometry was conducted (Figs. 4C, 4D, right). After cells were incubated with PD₁-BD₂, the fluorescence intensities were higher than that of the control and the shMDMX groups, and were comparable with that of the cells incubated with PD₁ (positive control). This indicates that the PD₁-BD₂ had been activated by the MDMX mRNA of cells. It suggests that the gene-specific drug may be used as an efficient candidate for the MDMX high-expressed cell imaging and killing. 

Cellular Cytotoxicity Analysis of PD₁-BD₂

To verify the cytotoxic ROS release of PD₁-BD₂ in cells based on the specific activation by MDMX mRNA, cytotoxicity analysis with CCK8 was implemented. The viabilities of Y79 cells were evaluated by preincubating with an escalating PD₁-BD₂ concentration. The result indicates no significant decrease in cell viability in the cells with only laser and only PD₁-BD₂. Furthermore, 72% ± 1.2% cell lethality was reached with PD₁-BD₂ concentration at 0.6 μM 72 hours incubation after the laser irradiation (Fig. 5A). The medium drug concentration was then used to demonstrate the flexibility of using light dosage to control the PDT cytotoxicity. By using 0.6 μM drug concentration with an increased laser fluence, cell viabilities decreased to 7.7% ± 1% at 50 J cm⁻² (Fig. 5B). 

Figure 5

 

The viability of cells under different treatments. (A) Y79 cells were incubated with different doses of PD₁-BD₂ (0, 0.2, 0.4, and 0.6 μM), and assessed at 24 and 72 hours after 30 J cm⁻² laser irradiation. (B) Y79 cells were incubated with 0.6 μM PD₁-BD₂ and irradiated with different laser fluence. (C) Y79 and shMDMX Y79 cells were incubated with PD₁-BD₂, PD₁, and PPa (0.6 μM), followed by laser irradiation (30 J cm⁻²). (D) Y79, WERI-Rb1, and D407 cells were incubated with PD₁-BD₂, and followed by laser irradiation (30 J cm⁻²). *P < 0.05.

Too strong PDT dosage would trigger cell necrosis but apoptosis, resulting in many side effects. The crucial factors in determining the type of cell death following PDT include cell type, light dose, photosensitizer concentration, and its subcellular localization.^35,36 Therefore, a moderate dosage in PDT, 0.6 μM drug and 30 J cm⁻² laser, was preferred in our experiment. 

Y79 cells were irradiated with laser in the presence of PD₁-BD₂, PD₁, and PPa (0.6 μM), then the photodynamic toxicity was assessed after a 48-hour incubation with consideration of the previous MDMX knockdown efficiency. The results show that the viability of Y79 cells treated with PD₁-BD₂ decreased by 68% ± 1.3%, being comparable to both the PD₁ and PPa groups. This indicates that the PPa molecules on PD₁-BD₂ had been activated by MDMX mRNA in cells and retained its efficiency in generating cytotoxicity during laser irradiation. In comparison, the treatment had a significantly lower cytotoxicity on the shMDMX Y79 cells. This clearly demonstrates the effectiveness of the modality on the MDMX high expression Y79 cells (Fig. 5C). A comparison of cell viabilities among Y79, WERI-Rb1, and D407 cells was further implemented. In the results, PD₁-BD₂ exhibited significant killing effect on either Y79 and WERI-Rb1 cells (Fig. 5D). The cells with high expressed MDMX mRNA were more sensitive to PD₁-BD₂-based PDT due to its idiotype. 

In order to confirm whether the apoptosis was the prominent cell death mode in the PDT, quantification of apoptosis and necrosis with Annexin V/PI staining and FACS analysis was performed. Figure 6 shows Y79 and WERI-Rb1 cell death mainly through apoptosis in the PD₁-BD₂+laser groups. This observation confirms that with the investigated treatment protocol, apoptosis was a major contribution to the cell death. 

Figure 6

 

Cell apoptosis analysis. Cell apoptosis of Y79 and WERI-Rb1 cells under different treatment protocols were analyzed with FACS. The averaged percentage of cell apoptosis is marked. The inserted histograms show the percentage of cell necrosis and apoptosis in PDT. Y79 and WERI-Rb1 cells were incubated with PD₁-BD₂ (0.6 μM), followed by 30 J cm⁻² laser irradiation (n = 3).

Therefore, the PD₁-BD₂ beacons can be triggered by the endogenetic MDMX mRNA and generated cytotoxicity to induce cell apoptosis. The specific mRNA trigging would improve the killing efficiency to retinoblastoma cells and greatly decrease the toxicity to normal cells. 

Conclusions

In summary, this work demonstrates the synthesis of a sensitive and effective MDMX mRNA-triggered photodynamic molecular beacon for retinoblastoma therapy. The beacon showed an efficient imaging capability on retinoblastoma cells and resulted in an increased cell apoptosis after PDT in vitro. The properties make it a potential tumor-specific drug for imaging and therapy in the future. The protocol is prospective to be used as a versatile method to design other tumor mRNA-based PDT agent assemblies. 

Acknowledgments

The authors thank Junyang Zhao Beijing (Tongren Eye Centre, Beijing Ophthalmology & Visual Sciences Key Lab, Tongren Hospital, Capital Medical University) for providing data analysis assistance. 

Supported by the National Natural Science Foundation of China (91539127; 61361160414; 81371646), the Guangdong Natural Science Foundation (2015A030313394), Science and Technology Planning Project (2014A020212738) of Guangdong, and the Science Technology Program Guangzhou (201607010371). 

Disclosure: Y. Wei, None; C. Lu, None; Q. Chen, None; D. Xing, None