5.1. Brief introduction
Theranostics combines the of therapeutics and diagnostics. It has become one of the important idea in many cancer researches (Cole et al., 2011). It is because that the researchers believe cancer growth can be hindered during the diagnostic stage and therefore the following treatment is much more effective. One common used feature of cancer for designing the theranostic agent is the EPR (enhanced permeability and retention) effect. Since the tumour usually grow so fast that subsequently causes the leaky vasculatures around cancer site (Ferrari, 2005). On the basis of this feature, many anti-cancer drug are delivered to the cancer site with higher effectiveness. In addition to this, studies on synthesis of drug-loaded carriers are also promising for specifically delivery to cancer tissues. At the early stage of theranostics in 2005, the imaging function starts to be added into the delivery system by involving the imaging contrast agents for clinical imaging like computed tomography, positron emission tomography, and magnetic resonance imaging.
It is well-known that cancer exhibits the genetic instability (includes mutation) and tumour-associated inflammation (Hilvo et al., 2011). These observations eventually lead the researchers to develop the targeted therapeutic strategies, which attack single or few targets that dominates the survival and proliferation of cancer cells. Generally, these targets are relative to the abnormal gene expression. On the other hand, cells can become cancerous due to the non-mutation effect. This non-mutation effect also contributes to the cellular survival and proliferation (Luo et al., 2009). For example, in the case of non-small cell lung cancer, the epidermal growth factor receptor (EGFR) is
activated abnormally and this kind of mutated EGFR can actually be imaged by the PET/CT (Yeh et al., 2011).
The noninvasive imaging technologies also have the advantages over conventional biopsies of cancer tissue in patient, since the imaging cause no trauma in body.
Moreover, noninvasive imaging technologies are able to assess the oncogenic protein’s activities in the cancer call. Therefore, it allows the evaluation of the result of inhibitors treatment at the end of the therapeutic process. The cancer cell is usually stubborn, even under the high-dose of chemotherapeutic agents. This is because the cancer cell is able to quickly get used to the environment and survive through the mutations and redundancy of signalling pathways. Therefore, the goal of complete eliminating the cancer cell can be approached through the combination of the therapeutic drug and imaging agent. Many kinds of nano-materials have been designed to achieve this goal.
They are loaded with the drugs, targeting ligand that recognises the cancerous cell, and multimodal imaging agents. The assemble of the theranostic nano-material allows the user to monitor the result of treatment and hence increases the drug efficacy and safety (Lee et al., 2012). Those novel approaches carry out the newest concepts of cancer treatment, such as the molecular imaging and personalised cancer therapy (Lim et al., 2011).
In the third part of this study, we propose novel selective decoy oligodeoxynucleotide (dODN)-doxorubicin (DOX) complex is reported for cancer theranostics. It eliminates the use of a ligand or carrier for targeted delivery and disassembles into therapeutic dODN and DOX upon encountering over-activated STAT3 in cancer cells. Hence, in situ STAT3 probing and synergistic anti-cancer effect are attained at the same time.
5.2. Experimental
• Assembly of doxorubicin intercalated hpdODN
The doxorubicin hydrochloride was obtained from the Sigma-Aldrich Taiwan. The
sequences of the hpdODN in this study was:
5’-CGTTTCTCATAAAGCGAAGCTTTA-TGAGAAACG-3’. It is in a structure of hairpin helix. With a simple mixture of doxorubicin and hpdODN (1 µg of each) and the remove of un-bonding reagent by Amicon Ultra centrifugal filter (10KDa), the doxorubicin-intercalated hpdODN, DOXON, was obtained. The doxorubicin used in this study was suitable for fluorescence related experiments.
• Circular dichroism and UV-Vis characterization
The circular dichroism (CD) spectropolarimeter (Aviv Biomedical Inc., Model 202) was used to characterise the secondary structures of the DOXON. A quartz cuvette with the 1 mm optical path was filled with 5 µM of sample solution in DPBS, and the CD spectra were obtained by taking the average of triplicates from 300 nm to 220 nm at 37 ̊C. The UV-Vis absorption spectra were measured on an Agilent UV-Vis spectrophotometer, Perkin Elmer (Lambda 25), with the use of 1.0 cm quartz cells. To determine the release profile, biotinylated DOXON was immobilised on the 24-well streptavidin coated plate.
With the addition of STAT3 or BSA, the doxorubicin in the supernatant solution was measured with the UV-Vis.
• In vitro STAT3-triggered DOX release
Biotinylated DOXON is immobilised onto a streptavidin-coated 24-well plate and followed by adding the DPBS containing 0.5 µM of STAT3 or BSA protein. The fluorescence intensity of released DOX was measured at 596 nm with an excitation wavelength of 480 nm through a micro-plate reader (Infinite M200 PRO, Tecan) with different time interval. The saturated fluorescence intensity is determined by measuring the signal with 50 µM STAT3 protein (30 min incubation) and considered as the maximum intensity (set as 100%). The signal obtained from the as-prepared complex is considered as the background (set as 0%).
• ELONA assays for KD estimation
ELONA assays were used for the estimation of the binding affinity (KD) of DOXON toward phosphorylated STAT3 and the BSA protein was used as the negative control (obtained from Cell Signalling Technology). For protein coating, 100 µL of 10 nM protein in DPBS was added to each well of 96-well polystyrene plate and incubated at 4 ̊C overnight. The wells were then washed four times with 0.05% Tween 20 in DPBS before being blocked with 1% BSA. A 100 µL of biotinylated DOXON in the DPBS was then added and incubated at room temperature for 2 hours. The wells were than washed four times with 0.05% Tween 20 in DPBS. The 100 µL of streptavidin–HRP solution was added and incubated at room temperature for 30 minutes. Afterward, the wells were washed with DPBS supplemented with 0.05% (v/v) Tween 20 and 0.1%
(w/v) BSA. Then the TMB substrate solution was used to quantify the binding complex.
The enzymatic reaction was stopped by adding 100 µL of 2 M H2SO4 and followed by absorbance measurement at 450 nm.
• MTT cell proliferation assays
The cell viability of DOXON-treated cells were assessed with Vybrant® MTT cell proliferation assay. 50 µL of 12 mM MTT solution was added to each well with 1 mL medium. After 2 hours of incubation, 750 µL of the medium was removed and 500 µL of DMSO was added into each well. After another 10 minutes of incubation, the samples were well mixed and measured for the absorbance at 540 nm.
• Gene expression determination
The down-stream mRNA expressions of STAT3 and p53 in response to the ligand transfection were measured using reverse transcription-polymerase chain reaction (RT-PCR) with Qiagen Rotor-Gene Q®. The mRNA extraction and cDNA amplification were carried out with the Cells-to-CT™ Power SYBR® Green Kit according to the protocol from the manufacturer. The primer sequences are Bcl-xL: 5’- GGAAAGCGTAGACAAGGAGATGC3’ and 5’ GGTGGGAGGGTAGA GTGGATGGT3’; Bax: 5’CATGTTTTCTGACGGCAACTTC3’ and 5'AGGGC CTTGAGCACCAGTTT3', and GAPDH: 5’ GACCACAGTCCATGCCATC -ACTGC-3’ and 5’- ATGACCTTGCCCACAGCCTTGG-3’.
5.3. Results and Discussion
Targeted drug delivery and dual pathway intervention strategy have shown promising opportunities for cancer therapy. For example, environment-sensitive carrier and co-delivery of synergistic drugs have both resulted in enhanced anti-cancer efficacy. To further improve selectivity, antibody and aptamer have been investigated for targeted drug delivery systems recently. The targeting ligand carries a drug and releases it while binding to a specific molecule. However, both carrier and ligand become debris after mission complete and may give rise to immunogenicity problems. Here, we propose a novel, proof-of-concept selective decoy-drug complex for effective cancer treatment and in situ oncogenic transcription factor probing. Our dODN-DOX complex (DOXON) is a “drug-in-drug” system and itself alone can achieve highly selective targeted co- delivery and release, oncogenic STAT3 probing and synergistic induction of effective STAT3 over-activated cancer cell apoptosis via two distinct pathways.
In this work, we prove our idea by tackling STAT3-associated cancer cell lines with our DOXON conjugate, which is composed of a novel, specific anti-STAT3 hairpin dODN (hpdODN) designed by our group and a broad-spectrum anti-cancer drug DOX that is also a structure-sensitive fluorescent reporter. Through drug-DNA intercalation, the DOX-carried hpdODN (DOXON) is prepared. Figure 5.1 illustrates how DOXON works for cancer theranostics. Before internalisation, DOXON remains non-dissociated and does not emit fluorescence. After internalisation and specific recognition of over-activated STAT3, DOXON dissociates into therapeutic dODN for STAT3 binding and releases DOX for taking anti-cancer effect. In addition, free DOX emits fluorescence indicating the presence of over-activated STAT3. Since the disassembly of DOXON is caused by STAT3 binding competition, the decoy-DOX conjugate is selective and will
dODN and DOX exert dual pathway intervention and yield a synergistic anti-cancer effect: dODN down-regulates STAT3-related anti-apoptotic expression, and DOX induces pro-apoptotic p53 process. Thus, an enhanced apoptotic efficacy for cancer cells can be anticipated.
Figure 5.1 The schematic illustration of the DOX-intercalated hpdODN (DOXON) complex for theranostic purpose. The DOXON remains complete and non-fluorescent before entering the cell. Internalised and encountering STAT3, it starts to dissociate into therapeutic dODN and DOX, exerting dual pathway apoptosis (anti-STAT3 and pro-p53). Meanwhile, free DOX becomes an emission beacon for STAT3 probing.
The DOXON was self-assembled through spontaneous dODN-DOX intercalation because of hpdODN’s B-form double helix structure (figure 5.2). Self-assembly of the dODN-DOX conjugate and the STAT3-reponsive DOX release profile were characterized by circular dichroism (CD) spectroscopy and in vitro competitive STAT3 binding assay, respectively. The data are shown in figure 5.3. The CD spectra give clear distinct fingerprints for hpdODN (line 1, figure 5.3) and DOXON (line 6, figure 5.3), respectively. The ellipticity variation around 240 nm indicating DOX intercalation and the corresponding induced structure change of hpdODN. When the ratio of DOX to hpdODN increases (line 1→6), the ellipticity at 260 nm goes from zero to a positive value, which is presumed to be a result of intercalation-induced helix unwinding. This suggests that DOXON features a different structure from hpdODN after carrying DOX.
The possible unwinding actually plays a important role in the following STAT3 binding, since the changed structure may bring a different binding efficiency. Generally, the double stranded DNA binds the the transcription factor with the partial melting.
When the DOX is intercalated into the DNA and enlarging the space between two strands of the sequence, the STAT3 protein will bind to the sequence more easily.
Therefore, the binding affinity is of DOXON is expected to be enhanced due to this effect. However, on there other hand, the intercalated DOX may occupy the binding site of the STAT3 protein, which then results in the decreased binding affinity of the sequence to the protein. The key point is that which factor contributes more. With different ratio of contribution, the binding affinity of the DOXON can be higher, lower, or even equal to the bare hpdODN. In the following experiment result, the measured binding affinity may able to answer this question.
Besides analysing the secondary structure change of the DOX-intercalated hpdODN, the complex is also investigated with the UV-Vis spectrum to figure out the intercalation
feature. For the free DOX, a characteristic absorption band at 495 nm in the UV-Vis spectrum can be seen (figure 5.4). This feature is very similar with a widely used FITC fluorescent dye. FITC has excitation and emission spectrum peak wavelengths of approximately 495 nm and 519 nm. Therefore the DOX emission is also observable around 520 nm. However, being very different from the most fluorescent dyes, DOX is a one with intercalation-quench emission feature. As we increased the ratio of hpdODN to DOX, the absorption bands decreased due to the interaction between the base and the intercalated DOX. The affinity of DOX towards hpdODN is estimated based on Benesi–
Hildebrand equation. From the insert in figure 5.4, the dissociation constant (KD) is calculated to be 45 ± 5.3 µM. The dissociation constant reported in the previous literature is 31.25 µM, which is quite consistent with our obtained value. In order to estimate the DOX capacity of the hpdODN, the DOXON is melted and measured the absorbance. It is found that approximate every 3.9 base pair of hpdODN would carry one DOX. That is, each DOXON conjugate contains 3.85 DOX on average. This value is close to the theoretical value reported in the previous literature (Bae et al., 2011), however is lower. The lower DOX capacity can be resulted from the lower GC content of the hpdODN (40%), compared to the sequences used in the previous literature (average 50 % or even higher).
Figure 5.2 The 2D and 3D structure prediction of the hpdODN generated by the mfold web server i and RNA composer web server ii. (i http://mfold.rna.albany.edu/ and ii http://rnacomposer.cs.put.poznan.pl/)
Figure 5.3 CD spectra of the DOX-hpdODN mixtures. They are mixed with the molar ratios of 0, 0.2, 0.4, 0.6, 0.8, and 1.0 (lines 1→6). The red dash line was obtained from the 5.0 µM doxorubicin in DPBS.
Figure 5.4 UV-Vis absorption spectra of doxorubicin in the presence of hpdODN at the concentrations = 0, 5.0, 10.0, 15.0, 20.0 and 25.0 µM. The doxorubicin concentration is 10 µM in DPBS.
It should be noted that the ionic strength plays a crucial role in DOXON preparation.
The average of DOX-intercalation capacity for a hpdODN decreases from 3.85 to 3.26 while the concentration of NaCl increases from 50 mM to 750 mM (figure 5.5). This phenomenon of the salt-increasing induced lower drug capacity is also observed in the hpdODN-SYBR green complex and the hpdODN-Quinacrine system. The proton effect has been used to explain this kind of phenomenon. Here we further check the melting temperature of the hpdODN under different concentration of the sodium chloride. It can seen from the figure that the melting temperature is increased alone with the increasing concentration of the sodium chloride. On the basis of this obtained result, it could be deduced that the decreased DOX capacity is resulted from the more stable structure of the sequences. When the sequence’s structure become more stable, the DOX is hard to get the space (between the unwinding strands) to complete the intercalation. The more stable structure of the sequence is also a bad news for the the STAT3 binding. Overall, the high concentration of the salt seems not beneficial to the either hpdODN to the complex system.
To demonstrate STAT3-selective and responsive release of DOX from DOXON, we conducted an in vitro competitive binding assay and characterised the dynamic control release profile as shown in figure 5.6. In this experiment, hpdODN (of DOXON) was biotinylated and was immobilised on a streptavidin-coated plate well surface, and then STAT3 or bovine serum albumin (BSA) protein was added to compete DOX binding.
The time-dependent DOX release profiles in figure 5.6 were recorded by a UV-Vis spectrometer. As a result, STAT3 binding induces 90% release of DOX within 1 hour, whereas non-specific BSA binding only causes approximate 10% DOX release at hour 1 (cf. self-leakage of DOX is around 6%). This demonstrates high selectivity and fast response of our DOXON drug conjugate, especially when compared with some reported
DNA-based control release systems such as 35% drug release after 5 hours (Zhu et al., 2015)(triggered by heat) and 30% drug release after 24 hours (Alexander et al., 2011)(triggered by the cDNA). Presumably, the DOXON selectively releases a large quantity of DOX attributed to the considerable difference in binding affinity between STAT3 and DOX toward hpdODN. The mechanism is similar to a quantum dot FRET aptamer beacon working upon protein binding-induced DNA reporter dye displacement, as reported by our group. When the STAT3 binds to hpdODN, intercalated DOX molecules are displaced away and are then released into the cell. By contrast, a non-specific protein such as BSA cannot compete with DOX for hpdODN binding.
To prove selective STAT3-responsive DOX release mechanism, we carried out enzyme-linked oligonucleotide assays (ELONAs) and estimated the binding affinities of phosphorylated STAT3 and BSA toward hpdODN and DOXON, respectively. The data are given in figure 5.7. The affinity constants (KD) for hpdODN, STAT3-DOXON, BSA-hpdODN, and BSA-DOXON binding events are determined to be 16022 nM, 20219 nM, 2.70.2 µM, and 1.10.1 µM, respectively. Notably, the KD of hpdODN toward STAT3 is 281-fold smaller than that toward DOX. This justifies STAT3’s competitiveness to displace intercalated DOX away from hpdODN.
Moreover, the KD of DOXON toward STAT3 is 1.26-fold higher than that of hpdODN, indicating a thermodynamic tendency for DOXON to free DOX and to become naked hpdODN for STAT3 binding. In comparison, the KD of DOXON toward BSA is 2.45-fold smaller than that of hpdODN, suggesting that BSA would bind to DOXON rather than hpdODN and prevent DOX from releasing. According to the above discussion plus the fact that KD of either hpdODN or DOXON toward STAT3 is considerably smaller than that toward BSA, the selective STAT3-responsive DOX release idea in Scheme 1 is secured further.
Figure 5.5 Drug capacity of the DOXON (black line) and melting temperature of the hpdODN (blue line) under different ion strength condition.
Figure 5.6 STAT3-responsive DOX release profile. DOXON was immobilised on 24-well plate 24-wells, and 0.5 µM of protein was added to release DOX.
Figure 5.7 ELONAs for STAT3 binding affinity and specificity assessment for hpdODN and DOXON, carried out at room temperature with BSA as a non-specific control.
The therapeutic effects of DOXON were examined with a MCF-7 cell line, which is known for its STAT3 activation. A CCD-966SK cell without STAT3 over-activation was chosen as a control. The data are shown in figure 5.8A and figure 5.8B.
As it can be seen in column pair 1 of figure 5.8A, the anti-apoptotic gene transcript Bcl-xL (STAT3 pathway) is highly expressed in the STAT3 over-activated MCF-7 cell, whereas the pro-apoptotic gene transcript Bax (p53 pathway) is at a low expression level. Together they prevent such cancer cells from the programmed death. The therapeutic function of hpdODN for down regulating Bcl-xL and that of DOX for up-regulating Bax are proven by the second and third column pairs in figure 5.8A, respectively. It is worthwhile to note that DOXON is capable of regulating both genes toward synergistic apoptotic effect, as evidenced in column pair 4 of figure 5.8A. To determine the MCF-7 proliferation inhibition efficacy, we conducted a direct delivery with serial dilutions of hpdODN, DOX, and DOXON without an additional carrier.
Lipofectamine® 2000 was used as the carrier in the control experiments. All of the experiments led to significant cell death and showed therapeutic effects (at least 10%
cell death with 2 µg/mL) (figure 5.9), but the use of Lipofectamine® only benefits hpdODN. The liposome carrier even decreases the efficacy of DOX and DOXON. This phenomenon can be explained by the DOX-induced aggregation of liposomes, which results in decreased transfection efficiency (figure 5.10). The fact means that DOXON can eliminate the need of an extra carrier for delivery, in addition to its elimination of an extra ligand for STAT3 targeting.
Moreover, DOXON results in the highest cell death whether or not the Lipofectamine® was in presence as compared to hpdODN and DOX. Without liposome, the IC50 of DOXON is 2.28 µg/mL (with 48 hours incubation). This is presumed to the effect of dual pathway intervention (STAT3 and p53) revealed in figure 5.8A, and the inhibition
efficacy is STAT3-specific and synergistic, as disclosed by figure 5.8B. It can be seen
efficacy is STAT3-specific and synergistic, as disclosed by figure 5.8B. It can be seen