Imaging of a siRNA molecular probe targeting MDM2 in breast cancer

Murine double minute 2 (MDM2) is an oncogene that is important for tumorigenesis, tumor metastasis and chemotherapy resistance. We aimed to synthesize a molecular imaging probe, 99m Tc-HYNIC-siRNA 1489, which could specically bind to MDM2. The [ 99m Tc]HYNIC-siRNA 1489 molecular probe provides an effective way to assess MDM2 expression via SPECT. Method: Three siRNAs were designed and their inhibitory eciencies were tested using western blots and qRT-PCR. The selected siRNA was labeled with the radionuclide 99 m-technetium ( 99m Tc) through the chelator HYNIC. The bioactivity and properties of [ 99m Tc]HYNIC-siRNA 1489 were determined before mouse imaging. Imaging and the biodistribution of the probe were used to assess its targeting ability.

MDM2 is an E3 ligase that mediates SIRT6 degradation, and the interaction between MDM2 and SIRT6 is dependent on AKT1-mediated SIRT6 phosphorylation at Ser338 [10]. Moreover, many studies have shown that MDM2 overexpression is associated with breast cancer that have invasive potential and resistance to chemotherapy [10]. So MDM2 should be considered a promising target for breast carcinoma diagnosis and therapy.
Molecular imaging is a noninvasive method for diagnosing and imaging disease. With the development of molecular imaging technology, many imaging agents have been used in medicine. Small interfering RNA (siRNA), composed of a double strand of RNA, is responsible for RNA interference pathway (RNAi)based gene silencing. It is incorporated into RNA-induced silencing complex (RISC) and then Argonaute-2 cleaves the sense strand after siRNA is generated via dicer processing of a long double-stranded RNA or a synthetic siRNA is delivered into the cytoplasm [13]. But siRNAs have some inherent characteristics, such as off-target effects and biological instability, which need to be addressed before they can be incorporated in clinical applications. Chemical modi cations of siRNAs have been shown to improve their inherent properties. For example, representative backbone modi cations of siRNA include phosphorothioate (PS), ribose 2'-OH, 2'-O-methyl (2'-OMe) modi cations. We decided to choose a 2'-OMe modi cation of the purines based on the following reasons. First, this modi cation site is the most attractive as it does not need to be recognized to target the mRNA cleavage process via the RNA induced silencing complex (RISC) [14]. Second, 2'-OMe modi cation increases the resistance of RNA to nucleic acid enzymes and its thermal stability (T m ) increases 0.5-0.7 °C per modi cation [14]. Third, modi cation of purine sites does not affect the silencing e ciency of siRNA.
In the present investigation, we used a chemically modi ed siRNA radiolabeled with 99m Tc to assess Corp. (Shanghai, China). A primary amine was added at the 5' end of the antisense RNA. The amine provided a coupling site for HYNIC, which is important for radionuclide labeling. Quantitative real-time PCR (qRT-PCR) and western blots were used to select the optimal siRNA for inhibiting MDM2.
The siRNA (5 µg/µL) was diluted in 25 mM ammonium acid carbonate buffer (pH = 8.5). HYNIC (GL Biochem, Shanghai, China) was dissolved in dimethylformamide (10 mg/mL). Then siRNA was added in the proportions 20:1 (HYNIC:siRNA). The compounds reacted for 1 h at room temperature. The mixture was puri ed with a Sep-Pak C18 column (Waters, USA). The C18 column was activated by 20 mL of acetonitrile and washed with 20 mL deionized water. The siRNA and HYNIC mixture was diluted to 1 mL with deionized water and then loaded on the Sep-Pak C18 column, which was rinsed once with 10 mL 25 mM ammonium bicarbonate (pH = 8.5), once with 10 mL 25 mM ammonium bicarbonate/5% acetonitrile, and twice with 10 mL 5% acetonitrile, and then siRNA was eluted with 1 mL 30% acetonitrile four times. The HYNIC-siRNA solution was collected in fractions of 8 drops/vial, HYNIC-siRNA was detected at 260 nm, and vials with the highest concentrations of HYNIC-siRNA were combined. Then the product was dried by rotary evaporation for 6 h, and the HYNIC-siRNA was stored at -80 °C.
HYNIC-siRNA solution was prepared and tricine and EDDA were used as co-ligands. Ten µg HYNIC-siRNA were dissolved in 20 µL 25 mM ammonium acid carbonate (pH = 8.5 an incubator (37 °C and 5% CO 2 ) for 6 h. The opti-MEM was removed and complete medium was added.
After 24 h, the MCF-7 cells were collected and used for qRT-PCR and western blotting.

Western Blots.
MCF-7 cells were lysed in RIPA buffer with PMSF (100:1) for 30 min on ice, then centrifuged at 12,000 × g at 4 °C for 15 min. The supernatant was collected and mixed with SDS-PAGE buffer then denatured, and the total protein was measured using a BCA Protein Assay Kit (Tiangen, China). Equal amounts of protein were separated by 10% SDS-polyacrylamide gels and transferred onto polyvinylidene di uoride membranes (Millipore, USA). Membranes were blocked using 5% non-fat milk in TBST for 1 h at 37 °C, followed by overnight incubation with anti-MDM2 (ab178938; Abcam, USA) and anti-β-actin (20536-1-AP; Proteintech, China) antibodies at 4 °C. Unbound antibody was removed by three 10-min washes in TBST solution. Membranes were then incubated for 1 h at room temperature with secondary antibody (ZB-2301; Zhong Shan Gold Bridge, China) conjugated with HRP, followed by three 10-min washes with TBST buffer. The blots were visualized with ECL reagent (Advansta, USA) and a Tanon from each group were sacri ced by inhalation of iso urane followed by cervical dislocation. Major internal organs including the brain, lungs, liver, spleen, kidneys, stomach, intestines, muscle, thyroid and blood, together with the tumor, were collected and weighed, and the radioactivity was measured via a gamma counter. The radioactivities of tissues were represented by the percentage of injected dose per g of organ (%ID/g). The mass ratio of the tumor to the contralateral muscle (T/M) was also calculated.

Blocking assay
Six NU/NU nude mice with MCF-7 xenografts were used in blocking assay. Non-labelled siRNA 1489 (1 mg/kg) encapsulated with TurboFect was injected into MCF-7 xenografts for 3 hours before injecting imaging agent. Then repeat the imaging procedure, the radioactivities of tissues were represented by the percentage of injected dose per g of organ (%ID/g).
2.10. Statistical Analysis. Each experiment was performed in triplicate with duplication for each sample tested. Statistical analysis was performed with SPSS 18.0. Data are expressed as the mean ± SD. Statistical differences were analyzed via one-way analysis of variance. P < 0.05 was considered statistically signi cant.

Expression of MDM2 in Breast Cancer Cell Lines.
To choose suitable cell lines with high levels of MDM2 qRT-PCR and western blots were used to evaluate MDM2 expression in MCF-7 and MDA-MB-231 cells. The results showed that the relative expression levels of MCF-7 and MDA-MB-231 mRNAs were 1 ± 0.14 and 0.40 ± 0.06 (p < 0.05, n = 3), respectively (Fig. 1a).
The protein levels of MCF-7 and MDA-MB-231 cells were 1 ± 0.05 and 0.35 ± 0.15 (p < 0.05, n = 3), respectively (Fig. 1b). This meant that MCF-7 had a high level of MDM2 and was suitable for the imaging experiment.
3.2. Inhibitory Effects of siRNA In Vitro.
Three siRNA sequences were designed and their inhibitory effects were investigated to select the best one. The inhibitory effects of each siRNA were estimated according to the protein expression and mRNA content determined by western blots and quantitative real-time PCR (Fig. 1c and d). The results of transfection with the three siRNA sequences showed substantially lower MDM2 mRNA and protein levels than those of the negative control siRNA and blank control groups transfected with Opti-MEM (p < 0.05, n = 3). There were no signi cance differences among the three siRNA sequences.

In Vitro 99m
Tc-HYNIC-siRNA Stability. The labeling e ciencies of [ 99m Tc]HYNIC-siRNA1489 and [ 99m Tc]HYNIC-control siRNA were 62.17 ± 3.16% and 61.56 ± 4.08% (p > 0.05, n = 4), respectively. The radiochemical purities of the probes were over 94% after puri cation. When the probes were tested at normal temperature (37 °C), the radiochemical purity did not signi cantly change (n = 3). With increasing incubation time, the probes were stable within 6 h and did not show signi cant degradation. Similarly, there was no difference between cultures in phosphate-buffered saline (PBS) and fetal bovine serum (FBS) (Fig. 2a). This indicated that the probes had relatively stable physicochemical properties.

In Vitro Inhibitory Effect of siRNA1489 after Labeling with 99m Tc.
We tested the effect of [ 99m Tc]HYNIC-siRNA1489 via qRT-PCR and western blots in MCF-7 cell lines. There was an obvious inhibiting effect of siRNA1489 unlabeled with 99m Tc. We compared the results for siRNA1489 labeled with 99m Tc to unlabeled siRNA1489 and blank control. From our experiments, radiolabeling of siRNA1489 did not affect the RNAi effect ( Fig. 2b and c). [ 99m Tc]HYNIC-siRNA 1489 still inhibited the expression of MDM2. The relative expressions of MDM2 mRNA and protein were 0.34 ± 0.06 and 0.46 ± 0.09 (p < 0.05, n = 3), respectively ( Fig. 2b and c). This indicated that siRNA conjugated with 99m Tc could be applied to in vitro experiments.  Fig. 3a). In the NC group, the imaging of mouse tumors was not observed at every time point. The blocking experiment was used to assess the speci city of [ 99m Tc]HYNIC-siRNA 1489 (Fig. 3c). The accumulation of [ 99m Tc]HYNIC-siRNA 1489 pre-treated with siRNA 1489 was lower than without siRNA 1489 in MCF-7 xenografts. Our results showed that the MDM2 molecular probe could reveal MDM2 transcription levels in vivo.

Biodistribution Studies.
After the imaging experiments, we measured the biodistributions of the probes in the nude mice with a gamma counter (Fig. 3b) was a high accumulation in the liver (14.92 ± 1.46% ID/g in the NC group and 14.48 ± 0.99% ID/g in the siRNA 1489 group) and kidney (13.34 ± 1.22% ID/g in the NC group and 13.65 ± 1.05% ID/g in the siRNA 1489 group) in both groups at 1 h. Then the radioactivities in the liver and kidney decreased over time. This signi ed that the molecular probe was cleared via the urinary system and the liver.

Discussion
In recent years, molecular biology technology has developed rapidly. Imaging used for the diagnosis of disease is no longer limited to anatomical imaging. Researchers have aimed to use functional imaging and metabolic imaging, and gene imaging has become the focus. Gene imaging reveals changes in cellular physiological processes at the genetic level and it has a high potential to aid early and accurate diagnosis. In gene imaging, RNA interference (RNAi) technology has become a hot research topic because siRNA can effectively combine with target gene mRNA. The siRNA has a high speci city and e ciency. In the past few decades, the therapeutic potential of siRNA has been con rmed in diseases such as viral infections and cancers [16][17][18].
The key to molecular imaging is to synthesize molecular probes that target the tumor, which means these molecular probes must have high speci city. It also requires the target tissue to overexpress certain molecules such as proteins or nucleic acids. In this study, MDM2 was selected as the target gene because it plays an important role in different tumor cells [16][17][18]. MDM2 is the major downstream regulator gene of p53. It can form a negative feedback loop with p53 leading to its degradation via ubiquitination and proteasomes. The mutation of p53 in different cancers will cause immortalization of cells, which is the rst step in tumorigenesis [22]. However, the expression of p53 in cancer cells is too low to be detected in vivo by noninvasive methods. In contrast, the high expression levels of MDM2 in cancer cells provide a new approach to diagnosis and treatment. As a negative regulator, the content of MDM2 could re ect the content of p53 indirectly. Park found that different types of breast cancers that express high levels of MDM2 in both the nucleus and cytoplasm were more malignant [23]. Therefore, the early detection of MDM2 is vitally important for the diagnosis and treatment of tumors. In our research we selected two kinds of breast cancer cells to determine which expresses higher amounts of MDM2. The PCR results proved that the MCF-7 cells expressed MDM2 at higher levels. We used MCF-7 cells to show which siRNA combined with the MDM2 gene. Finally, the results of PCR and western blots revealed that siRNA1489 was the best choice, so it was used in further experiments.
We used TurboFect as a carrier in the in vivo experiments. It is lipophilic and is often used for nonviral transfection. TurboFect could combine with siRNA through electrostatic interactions, and TurboFect-siRNA was adsorbed to the electronegative cytomembrane and transferred into the cells. Then the siRNA and DNA could bind and play a role. The siRNA probe could be carried into the cells by lipo2000 in vitro, but the way it penetrated the cytomembrane in vivo and its intake rate were the keys to the experiment. And these factors are related to the transcription and translation of the target gene. In general, cells can take in large molecules via endocytosis, and small molecules can enter cells through speci c channels or transporters. But there is no speci c channel for siRNA and it is easily excreted by the reticuloendothelial system. Hence, the targeting e ciency may decrease in vivo, so we attempted to increase the probe uptake by using TurboFect for in vivo experiments. In order to solve the problem of rapid degradation and excretion of siRNA in vivo, double strand siRNA was selected in these experiments. Because the half-life of double-chain siRNA is longer than that of single-chain siRNA, the inhibiting e ciency of double-chains siRNA is signi cantly higher than that of single-chain siRNA [24]. Furthermore, we used a 2'OMe modi cation for the sense strand of siRNA because the 2'OMe modi cation did not affect RNA silencing and splicing. It is one of the compatible modi cation methods, as it not only increased the resistance of siRNA to nucleases but also enhanced the interference effect of siRNA and its thermal stability [25]. Another method that enhances resistance to nucleases is the modi cation of U and C bases with uorine. In order to enhance the binding of the siRNA to the target gene, we added two free dTT bases on the 3'end [26]. The amino modi cation on the 5'end was used to couple with HYNIC. Finally, HYNIC was chelated with 99m TcO 4 − . At the same time, the co-ligands, tricine and EDDA, were added to occupy the void position of 99m TcO 4 − . The PCR and western blot results showed that [ 99m Tc]HYNIC-siRNA1489 signi cantly inhibited the expression of mRNA and MDM2 protein after transfection of the MCF-7 cells. The stability of the siRNA-MDM2 probe was validated in PBS and serum, which provided a basis for imaging of animal tumors.
In addition to the characteristics of the molecular probe, radiochemical purity and labeling e ciency were also important. If the radiochemical purity of the probe was low, excessive free 99m TcO 4 − will be absorbed by the thyroid, which would interfere with tumor imaging. Similarly, if the labeling e ciency was low, the molecular probe that was injected into the nude mice would be taken up at lower levels. All of these factors will affect the imaging. So we used chromatography to determine the radiochemical purity and labeling e ciency after purifying the molecular probe, and the results were satisfactory.

Declarations
Ethics approval and consent to participate This project was approved by the Medical Ethics Expert Committee of Harbin Medical University. All procedures performed in studies involving animals were in accordance with the ethical standards of the institutional research committee for animals.

Consent for publication
Not applicable.

Availability of data and materials
The full datas used and analyzed during the current study are available on reasonable request from the corresponding authors.   experiments. *P < 0.05; **P < 0.01 versus the NC group. Figure 3