Precise Detection and Visualization of Cyclooxygenase-2 for Golgi Imaging by a Light-Up Aggregation-Induced Emission-Based Probe

Abstract

Open AccessCCS ChemistryCOMMUNICATION1 Feb 2022Precise Detection and Visualization of Cyclooxygenase-2 for Golgi Imaging by a Light-Up Aggregation-Induced Emission-Based Probe Yuchao Luo†, Song Zhang†, Hui Wang, Quan Luo, Zhigang Xie, Bin Xu and Wenjing Tian Yuchao Luo† State Key Laboratory of Supramolecular Structure and Materials, Department of Chemistry, Jilin University, Changchun 130012 †Y. Luo and S. Zhang contributed equally to this work.Google Scholar More articles by this author , Song Zhang† State Key Laboratory of Supramolecular Structure and Materials, Department of Chemistry, Jilin University, Changchun 130012 †Y. Luo and S. Zhang contributed equally to this work.Google Scholar More articles by this author , Hui Wang State Key Laboratory of Supramolecular Structure and Materials, Department of Chemistry, Jilin University, Changchun 130012 Google Scholar More articles by this author , Quan Luo State Key Laboratory of Supramolecular Structure and Materials, Department of Chemistry, Jilin University, Changchun 130012 Google Scholar More articles by this author , Zhigang Xie Department State Key Laboratory of Polymer Physics and Chemistry Institution, Changchun Institute of Applied Chemistry, Changchun 130022 Google Scholar More articles by this author , Bin Xu *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] State Key Laboratory of Supramolecular Structure and Materials, Department of Chemistry, Jilin University, Changchun 130012 Google Scholar More articles by this author and Wenjing Tian *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] State Key Laboratory of Supramolecular Structure and Materials, Department of Chemistry, Jilin University, Changchun 130012 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.021.202101187 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Precision-targeted detection of cancer in a rapid and sensitive fashion remains a significant challenge in prevention, management, and in particular, treatment. Herein, we have designed and synthesized a unique cyclooxygenase-2 (COX-2) fluorescence probe, dimethylamine-9,10-distyrylanthracene-indomethacin (NDSA-IMC), that could visualize the site of highly expressed COX-2 in the Golgi apparatus of cancer cells, using 9,10-diatyrylanthracene derivative as the luminous unit and a selective inhibitor, indomethacin (IMC), as the recognizing moiety for COX-2. In an aqueous solution, the free state NDSA-IMC showed a weak emission in the absence of COX-2 but enhanced emission in the presence of the enzyme due to the restriction of intramolecular motion of aggregation-induced emission-active NDSA-IMC when bound to COX-2. Cell imaging and flow cytometry experiments indicated that NDSA-IMC could discriminate between cancer and normal cells and visualize the Golgi apparatus of cancer cells via specific targeting of COX-2. Therefore, NDSA-IMC might potentially detect early cancer lesions and ultimately mitigate the population of cancer burden in society. Download figure Download PowerPoint Introduction Precision targeting for pathology and therapeutics is intended to develop therapeutic agents that selectively eliminate cancer cells.1–3 Under the same doctrine of precision targeting, it is also important to visualize the morphology or biological activities of these diseased cells to empower diagnosis, surveil therapeutic response, and facilitate surgical resection.4 As a promising noninvasive, real-time, high-sensitivity, and low-cost imaging technology, molecular fluorescence imaging has become a powerful tool for cancer diagnosis and clinical surgery.5–8 In particular, imaging has offered exciting opportunities to selectively detect specific cell populations and malignant tumors such as those bearing markers of disease. But in general, the specific biomolecules of diseases are expressed with low concentrations in the physiological environment; thus, the detection is susceptible to interference. Inspiringly, the exploitation of targeted fluorescent probes that enable imaging of cancer cells with high selectivity and sensitivity would likely have a significant impact on imaging techniques.9–12 Their selectivity is mainly determined by the affinity between specific ligands and biomarkers, and their sensitivity is usually dependent on the fluorescence contrast before and after probe binding to biomolecular targets.13,14 Accordingly, it is crucial to design highly selective and sensitive targeting probes for distinct biomolecules expressed in specific diseases. Golgi apparatus plays an important role in the posttranslational modification, sorting, and packing proteins within cells.15,16 In cancer cells, the Golgi apparatus transports and secrets some essential proteins, including enzymes such as Cyclooxygenase-2 (COX-2). It has been suggested that COX-2 is upregulated in almost all cancer cells but less expressed in normal cells, and the overexpressed COX-2 promotes tumor growth, angiogenesis, and metastasis of cancer cells. Therefore, COX-2 could be regarded as a valuable cancer cell marker; hence, an evaluation of COX-2 activity is of significant importance to cancer diagnosis, therapy, and monitoring. Hence, visualization of COX-2 in the Golgi apparatus of cancer cells has great potential in clinical analysis or medical intervention, as it gives valuable information on cell bioactivities. Several research groups have conducted studies on the detection of the COX-2 enzyme.17–24 In 1996, Stallings and co-workers17 reported three inhibitors for COX-2: flurbiprofen, indomethacin (IMC), and SC-558. Later, Marnett and co-workers18,19 synthesized a series of COX-2 fluorescent inhibitors composed of IMC and organic dyes and proved that the inhibitors with n-butyldiamine linker were the best COX-2 targeted drugs for bioimaging. Recently, Peng’s group20–23 designed photoinduced electron transfer-based COX-2 specific fluorescent probes by combining organic dyes with IMC and realized the visualization of the Golgi apparatus of cancer cells. However, the above-mentioned COX-2-specific fluorescent probes were developed based on traditional organic dyes such as rhodamine, Nile blue, BODIPYs, and so on. The accumulation of these fluorogens in Golgi often leads to high local concentration, which, in turn, results in low fluorescent quantum yield due to the aggregation-caused quenching effect. On the other hand, the small Stokes shift of traditional dyes sometimes makes it difficult to select suitable excitation lasers. In general, the low ability to resist photobleaching of traditional dyes reduces the imaging signal markedly after multiple excitations. These disadvantages of traditional dyes often lead to a decrease in signal-to-noise ratio and sensitivity, and limit their applications in biosensing and bioimaging considerably.24 Therefore, it is exceptionally paramount and imperative to develop COX-2-specific fluorescent probes that could image cancer cells, while minimizing background interference and increasing signal strength. Fluorophores with aggregation-induced emission (AIE) characteristics possess advantages in imaging detection25–27 such as aggregation-induced fluorescence enhancement, large Stokes shift and anti-photobleaching ability.28–30 Recent advances in microenvironment-sensitive AIE fluorescent probes have been achieved, which modulate their fluorescence intensity or color in response to local microenvironmental changes31; they were particularly efficient to light up biomolecular targets, which often enable probes to show nonemissive or weak emission in the initial state, while boosting their emissions following their reaction or binding to specific analytes.32,33 Considering the specificity recognition of IMC to COX-2 and the luminescent characters of AIE molecules, we combined the IMC moiety and one 9,10-diatyrylanthracene (DSA) derivative unit, a typical AIE fluorogen,34,35 to achieve a fluorescent light-up probe (NDSA-IMC) for COX-2. The resultant fluorescence intensity of NDSA-IMC was almost nondetectable in the buffer solution, while the fluorescence signal was significantly enhanced when the probe was bound to COX-2 due to the restriction of intramolecular motion mechanism. Thus, the specific NDSA-IMC responsive mode enabled highly sensitive and selective COX-2 detection and visualization, thereby enabling cancer cells to discriminate against normal cells. Results and Discussion NDSA-IMCs were designed and synthesized using a DSA derivative with an ammonium salt modification as a fluorophore unit, four carbon alkyls as a linker unit, and a compound with an amido bond, IMC, to serve as a connection to the target unit. The synthetic route of the NDSA-IMC and the chemical structure of all the compounds used are shown in Figure 1a and Supporting Information Scheme S1.28 All compounds were purified by silica gel chromatography; after recrystallization, their structures were verified by NMR spectroscopy, matrix-assisted laser desorption ionization time-of-flight mass spectrometry (MALDI-TOF MS), and elemental analysis ( Supporting Information Figures S1 and S2). Figure 1 | (a) Discrimination of COX-2 enzyme via specific NDSA-IMC probe. (b) The absorbance and emission spectra of NDSA-IMC (5 μM), UV–vis (DMSO): λmax (ε) = 420 nm, fluorescence (DMSO): λex = 435.5 nm, λem = 574 nm. (c) Fluorescence spectra of NDSA-IMC (3 μM) in the absence/presence of COX-2 (2 μg/mL) in 3 μM Tris–HCl buffer with pH 7.4. (d) Fluorescence response of NDSA-IMC at 545 nm on the addition of other biomolecules (5 μg/mL) and COX-2 (2 μg/mL) in 3 μM Tris–HCl buffer with pH 7.4: 1. Control; 2. Glutathione; 3. Adenine; 4. Pepsin; 5. Lysozyme; 6. l-Cysteine; 7. Hypoxanthine; 8. Chymotrypsin; 9. Adenosine monophosphate (AMP); 10. Adenosine triphosphate (ATP); 11. BSA; 12. Papain; 13. trypsin; 14. DNA; 15. COX-2. Download figure Download PowerPoint The absorption and emission spectrum of NDSA-IMC in dimethyl sulfoxide (DMSO) are shown in Figure 1b, demonstrating a maximum absorption at 420 nm and an intense emission at 574 nm. A typical AIE characteristic of NDSA-IMC is displayed in Supporting Information Figure S3a, showing that its fluorescence intensity increased gradually when the glycerol fraction (fw) in DMSO solution increased from 0% to 90%. The fluorescence quantum yield and lifetime of NDSA-IMC in DMSO solution were measured to be 0.005 and 2.12 ns. When COX-2 was added to the solution, it exhibited a remarkable enhancement in quantum yield to 0.45, and the complex lifetime increased to 4.19 ns, with a slight reduction of the extinction coefficient (ε = 20,500 M−1 cm−1), probably due to the formation of NDSA-IMC-COX-2 complex ( Supporting Information Table S1). To confirm the detection ability of NDSA-IMC of COX-2, fluorescent titration experiments were carried out by adding varying concentrations (0 μg/mL–2 μg/mL) of COX-2 in a 5 μM NDSA-IMC buffer solution (buffer = 3 μM Tris–HCl, pH 7.4). As shown in Supporting Information Figure S3b, the fluorescence intensity of NDSA-IMC intensified progressively with an increasing concentration of COX-2. In particular, the fluorescence emission intensity of NDSA-IMC was enhanced an 80-fold, compared with the emission of the control buffer solution, when 2 μg/mL of COX-2 was added (Figure 1c). Furthermore, a plot of the fluorescence intensity against the COX-2 concentration ranging from 0.14 to 0.8 μg/mL provided a perfect straight line, suggesting the plausibility of utilizing the NDSA-IMC probe for COX-2 quantification. Moreover, the detection limit for COX-2 was determined experimentally to be 0.24 μg/mL ( Supporting Information Figure S3c). Besides sensitivity, selectivity is also a crucial parameter for evaluating the ability of the probe to identify the target. In particular, bioimaging probes need to show unique responses to the targeted species beyond other potentially biological coexisting factors in a complex living system. As shown in Figure 1d, prominent fluorescence of NDSA-IMC is observable in the presence of various biomolecules such as DNA, RNA, glutathione, bovine serum albumin (BSA), and other biomolecules, but ∼20-fold increased fluorescence could be obtained in the presence of COX-2 under the same conditions. To clarify the mechanism of the fluorescence enhancement, we used the molecular docking method to investigate the specific binding of NDSA-IMC to COX-2. As shown in Supporting Information Figure S4, in the presence of COX-2, the IMC moiety of NDSA-IMC penetrated farthest into the hydrophobic channel of COX-2; it was stabilized by three amino acids Ser530, Tyr355, and Arg120, consistent with the interaction of IMC with COX-2 ( Supporting Information Figure S4a). The linker between the IMC and DSA moieties (10.9 Å) was a bit longer than the distance from the binding site to the large hydrophobic cavity of COX-2 (8.294–9.294 Å) ( Supporting Information Figure S4b), which enabled the NDSA fluorophore moiety to immobilize on the surface of COX-2 without affecting the specific binding interactions. The IC50 values of NDSA-IMC and IMC, the parameters used to assess the binding capacity at 50% inhibiting concentration of an inhibitor for COX-2, were 0.71 and 0.75 μM ( Supporting Information Figure S5), respectively, indicating that NDSA-IMC and IMC had a similar binding affinity for COX-2. Indeed, the excited state of the isolated NDSA-IMC was annihilated readily after being excited in dilute aqueous solution via a rapid nonradiative decay from the free intramolecular motion of DSA fluorescent moiety. When binding to COX-2, DSA fluorophore moiety could boost emission due to inhibiting the rotation of phenyl ethylene of the DSA moiety ( Supporting Information Figure S4c). In addition, since the hydrophobic cavity of COX-2 homodimer could only hold the IMC moiety due to the small length of flexible linker (10.9 Å), the binding of IMC moiety to COX-2 restricted the intramolecular rotations of the aromatic rotors dramatically ( Supporting Information Figure S4d), resulting in a significant fluorescence light-up of the NDSA-IMC-COX-2 complex. Thus, NDSA-IMC could enter the binding pocket and light up COX-2 in the Golgi apparatus of cancer cells. This pronounced fluorescence enhancement, together with a low background noise of NDSA-IMC, could be beneficial regarding good sensitivity in targeting COX-2. To verify the high selectivity of the NDSA-IMC probe in living biosystems, the receptor-mediated binding ability of the probe to COX-2 was examined in mammalian cells. Since COX-2 was overexpressed in cancer cells but barely in normal cells, NDSA-IMC was incubated for 90 min with three cancer cell lines, Hela (cervical cancer cells), MCF-7 (breast cancer cells), and Hep-G2 (human liver cancer cells), and two normal cell lines, LO-2 (human liver cells) and L929 (mouse fibroblast cells) The confocal fluorescence microscopy imaging excited at 488 nm showed that the two normal cells (LO-2 and L929) afforded very weak fluorescence ( Supporting Information Figure S6), but prominent fluorescence signals were observed from the cancer cells (Hela, MCF-7, and Hep-G2) under the same experimental conditions (Figures 2a–2i). These results demonstrated that NDSA-IMC could distinguish cancer cells from normal cells by specifically labeling overexpressed COX-2 in cancer cells. Furthermore, to verify the specificity of the NDSA-IMC effect, NDSA was also incubated with the three cancer cell types (Hela, MCF-7, and Hep-G2) for 90 min. The confocal fluorescence microscopy imaging after excitation at 488 nm showed very weak fluorescence signals ( Supporting Information Figure S7). In contrast, there were obvious fluorescence signals in the confocal images of cancer cells (Hela, MCF-7, and Hep-G2), after incubation with NDSA-IMC under the same experimental conditions (Figures 2a–2i), confirming that NDSA-IMC specifically exhibited a fluorescent signal due to the covalent introduction of IMC group. Figure 2 | Confocal images of three cancer cells (a–c, Hela cells; d–f, MCF-7 cells; g–i, Hep G2 cells) stained with 4ʹ,6-diamidino-2-phenylindole (DAPI) dye (10 μg/mL) and NDSA-IMC (5.6 μM). (a, d, and g) Channel blue: nucleus imaging stained with DAPI, excitation: 405 nm, scan range: 425–490 nm; (b, e, and h) Channel green: NDSA-IMC stains, excitation: 488 nm, scan range: 530–570 nm; (c, f, and i) Overlay imaging. Scale bar, 20 μm. Download figure Download PowerPoint To further evaluate that NDSA-IMC specifically targeted COX-2, a selective inhibitor, celecoxib, was chosen to specifically bind COX-2 to inhibit the biological activity of the enzyme in cells at the ultratrace level. When MCF-7, cells were incubated with 0 μM (Figures 3a–3c), 6.5 μM (Figures 3d–3f) or 13.0 μM (Figures 3g–3i) celecoxib, the fluorescence intensity decreased gradually with increasing concentrations of celecoxib, indicating that only a few remaining free COX-2 were able to bind with NDSA-IMC. These data confirmed that the probe bound specifically to COX-2 in the cancer cell and exhibited a fluorescent signal as a potent and selective fluorescent probe for COX-2. Figure 3 | Confocal microscopy images of MCF-7 cells pretreated with different amounts of celecoxib for 3 h prior to NDSA-IMC treatment. (a–c) Pretreated with no celecoxib; (d–f) Pretreated with celecoxib of 6.5 μM; (g–i) Pretreated with celecoxib of 13 μM. Scale bar, 20 μm. Download figure Download PowerPoint To determine the subcellular distribution of the probe, BODIPY TR C5-ceramide, a commercial red-fluorescent dye used for staining the Golgi apparatus in live cells, was employed to co-stain the three cancer cell lines (Hela cells, MCF-7 cells, and Hep-G2 cells) with NDSA-IMC. The fluorescence signal of NDSA-IMC overlapped perfectly with that of BODIPY TR C5-ceramide for two of the cancer cells, revealing satisfactory co-localization (Figures 4a–4f). A similar result was obtained for Hela cells ( Supporting Information Figures S8a and S8b) in green-channel NDSA-IMC fluorescence imaging, overlapping well with the red channel of BODIPY TR C5-ceramide; the merged fluorescence image is shown in Supporting Information Figure S8c. The intensity profiles of the linear regions of interest across Hela cells co-stained with two fluorescent dyes also varied in close synchrony ( Supporting Information Figure S8d), revealing specific imaging of Golgi apparatus in the cancer cells. Figure 4 | Fluorescence images of NDSA-IMC and BODIPY TR C5-ceramide (5.0 μM) in two cancer cells. (a–c) MCF-7; (d–f) Hep G2. (a and d) Stained with NDSA-IMC. (b and e) Stained with BODIPY TR C5-ceramide. (c and f) Merged image. Scale bar, 20 μm. Download figure Download PowerPoint Subsequently, flow cytometry (FCM) was carried out to evaluate the selectivity of NDSA-IMC for COX-2 quantitatively. Various cell lines (Hela, MCF-7, Hep-G2, LO-2, and L929) were incubated with NDSA for 90 min, and the fluorescence (λex: 488 nm, λem: 555 ± 20 nm) in 10,000 individual cells of each population was measured. Histograms demonstrating the strong labeling ability of NDSA-IMC for cancer cells are displayed in Supporting Information Figure S9. Before detection and imaging, an 3-(4,5-Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT) assay (an indicator of cell viability, proliferation, and cytotoxicity) was performed to evaluate the cytotoxicity of NDSA-IMC in living cells: The three cancer cells were treated with 2.5, 5, 7.5, and 10 μM NDSA-IMC solution for 24 h. We found that all the cell survival rates were above 95% ( Supporting Information Figure S10), indicating excellent cytocompatibility of NDSA-IMC. Finally, tumor-bearing mice were established using Hep G2 cells to prove the specific targeting ability of the NDSA-IMC probe in vivo. All tumors in the mice emitted strong fluorescence, examined by a Small Animals Living Imaging System (Xenogen Co., Alameda, CA, USA) 30 min after administration of NDSA-IMC into the tail vein ( Supporting Information Figure S11). Then in the COX-2 expressing tumor region, a strong fluorescence signal was observed. These data indicated that the NDSA-IMC probe accumulated specifically in the ultra-trace COX-2 at the tumor site. Conclusion We coupled an AIE-active fluorescent core to IMC to form the bioprobe, NDSA-IMC, which could specifically light up the Golgi apparatus of cancer cell lines. In the absence of COX-2, NDSA-IMC exhibited weak emission; in contrast, via interaction with accumulated COX-2 in the Golgi apparatus of cancer cells, enhanced fluorescent signal was generated rapidly and selectively. Moreover, the good biocompatibility and low cytotoxicity of NDSA-IMC make it suitable for rapid, highly selective, and sensitive in identifying cancer cells via imaging of the Golgi apparatus. We believe that NDSA-IMC is capable of imaging complex changes of the Golgi apparatus with potential application for early cancer diagnosis. Supporting Information Supporting information is available and includes synthetic route of NDSA-IMC, NMR, and MALDI-TOF data to verify the structure of NDSA-IMC, fluorescence spectra of NDSA-IMC, molecular docking, dose–inhibition curves, confocal images of normal cells treated with NDSA-IMC, confocal images of cancer cells treated with NDSA, NDSA-IMC, and BODIPY TR C5-ceramide, FCM histograms of different cell lines, biological toxicity of NDSA-IMC in three cancer cells, in vivo fluorescence imaging and rebuild with 3D models, comparison of the photophysical properties of NDSA-IMC and bound NDSA-IMC-COX-2. Ethical Statement All animal studies were performed in strict accordance with the NIH guidelines for the Care and Use of Laboratory Animals (NIH Publication No. 85-23 Rev. 1985) and were approved by the Committee of the Animal Use and Care of the Chinese Academy of Sciences. Conflict of Interest There is no conflict of interest to report. Funding Information This work was financially supported by the National Natural Science Foundation of China (grant nos. 21835001, 51773080, and 52073116).