A COF-based nanoplatform for highly efficient cancer diagnosis, photodynamic therapy and prognosis

A covalent organic framework-based nanoplatform has been developed for cancer imaging, photodynamic therapy, and prognosis.


Introduction
Covalent organic frameworks (COFs), as a class of emerging crystalline porous polymeric materials, possess tremendous potential for application in a variety of elds owing to their facile design and well-dened structure. [1][2][3][4] Over the past dozen years, COFs with different chemical constituents and functionalities have been enthusiastically designed and employed for gas adsorption/separation, catalysis, optoelectronics, energy storage, environmental purposes and so on. [5][6][7][8] COFs are also promising candidates for biomedical applications owing to their unique stability, biocompatibility and functional diversity. [9][10][11][12] But the poor dispersibility and unsatisfactory bioavailability caused by their large size are still the major bottlenecks for COFs to be used for biomedical applications. [13][14][15][16][17] Nanoscale COFs possessing better dispersibility and higher bioavailability are more suitable for biomedical purposes.
Photodynamic therapy (PDT), employing a specic laser to excite a photosensitizer (PS) to generate reactive oxygen species (ROS) and kill malignant cells, is highly spatiotemporally controlled and has been approved for clinical cancer treatment. [18][19][20][21] Novel PSs are the bases for efficient cancer PDT. However, traditional PSs with large aromatic structures oen suffer from poor solubility and tend to aggregate in physiological environments, which can signicantly compromise their therapeutic effects. 22,23 Nanoscale PSs with a regular porous structure could ensure high PS density and prevent unwanted aggregation, and they are ideal tools for highly efficient ROS generation. [24][25][26] Thus, nanoscale COF-based PSs should be promising candidates for application in PDT.
Traditionally, cancer diagnosis and treatment are individual and separate processes. [27][28][29][30][31][32] Because the metabolism pathways and biodistribution of diagnostic and therapeutic agents are different, it is difficult to timely and effectively guide the treatment plan based on the diagnosis result. 33 Recently, theranostic probes that simultaneously integrate diagnostic and therapeutic functions have attracted considerable attention, owing to their potential in correctly adjusting the way of treatment and potentiating disease prognosis. [34][35][36][37][38] Therefore, a COF-based theranostic nanoplatform will be highly desirable for cancer detection, treatment, and prognosis.
Herein, we developed a nanoscale COF-based theranostic nanoplatform. A nanoscale porphyrin-based COF was prepared and a TAMRA-labeled survivin antisense strand (TSAS) was integrated onto the COF NPs to give rise to the nanoplatform (termed COF-survivin). The crystalline reticular structure endowed the COF with better stability and higher ROS generation ability in aqueous solution than those of the porphyrin monomer, while the large planar structure composed of the strong p-electron system makes it easy to absorb DNA single strands and quench the uorophore to form a stable nanoplatform. The TSAS could readily form a duplex structure with survivin mRNA (a cancer biomarker, closely related to the genesis and development of several cancers), 39,40 which signicantly weakens the interaction between the TSAS and COF, resulting in the uorophore being far away from the COF. Correspondingly, the uorescence can be restored by the FRET prohibition, enabling specic cancer imaging. Further irradiation of COF-survivin with red light could generate abundant toxic ROS in cancer cells to induce oxidation stress, decrease the mitochondrial membrane potential (MMP), and trigger cell apoptosis. Based on this, COF-survivin was successfully employed for highly selective cancer cell/tissue imaging and efficient PDT. Interestingly, prognostic evaluation was also demonstrated (Scheme 1).

Results and discussion
COF NPs were prepared based on a solvothermal strategy. 16 According to the FT-IR spectrum (Fig. 1A), the peak at 3350 nm refers to the N-H stretching band; the intensity of the C]O stretching band at 1660 nm is decreased, while a new peak at 1601 nm appeared corresponding to the C]N stretching band, demonstrating the successful condensation between amino and aldehyde groups. Moreover, multiple intense peaks were observed in the PXRD pattern, indicating the successful preparation of crystal COF structures (Fig. S1 †). The TEM (Fig. 1B) and SEM (Fig. S2A †) results reveal that the obtained COF NPs have a uniform quasi spherical structure and good dispersibility. The UV-Vis spectrum ( Fig. 1C) of COF NPs red shied compared to that of the monomer, which should be ascribed to the formation of the strong p-electron system. The unique UV-Vis absorption effect makes COF NPs promising for constructing FRET-based probes. The uorescence properties of COF NPs and the monomer were further investigated, and the reduced uorescence (Fig. 1D) of COF NPs may enhance the ROS generation effect due to the restricted energy decay. 15 Therefore, the ROS generation effect of COF NPs and the monomer was further evaluated. According to the results shown in Fig. 1E, the ROS generation by COF NPs was signicantly higher than that by the porphyrin monomer in aqueous solution owing to the regular structure and reduced energy decay. Further investigation revealed that singlet oxygen ( 1 O 2 ), one of the most toxic ROS, 41 was the main ROS generated by the COF NPs (Fig. 1F).
Inspired by the unique structural characteristics of the obtained COF NPs, we assume that they could effectively adsorb DNA single strands and quench the uorophore. By adding COF NPs into the TSAS solution, the uorescence of TAMRA was quenched and reached a plateau as the concentration of COF NPs reached 150 mg mL À1 ( Fig. 2A and S3 †). Therefore, this ratio was selected to prepare COF-survivin for further investigations. The morphology of COF-survivin was similar to that of COF NPs ( Fig. 2B and S2B †). The zeta potential of the COF NPs decreased from À17 to À22.5 due to the attachment of more negatively charged oligonucleotides (Fig. 2C). The DLS of COF NPs was slightly enhanced because of the successful DNA loading (Fig. 2D). Based on the uorescence standard curve, the DNA loading amount on the COF was calculated to be 0.572 nmol mg À1 (Fig. S4 †). Further investigation demonstrated that the obtained nanoplatform has a good storage stability (Fig. 2E) and anti-DNase I effect (Fig. 2F), mostly due to the formation of strong hydrogen bonds and p-p stacking interaction between the DNA chain and COF NPs, which prevented non-specic DNA release and reduced the interaction between DNase I and DNA molecules. Further addition of the survivin target into the COF-Scheme 1 Schematic illustration of the preparation of theranostic nanoplatform COF-survivin for tumor imaging, PDT and prognostic evaluation applications.
survivin solution could generate stable duplexes, which could reduce the interaction between the TSAS and COF NPs and restore the uorescence by FRET prohibition (Fig. 2F and S5 †). Therefore, COF-survivin is promising for simultaneous cancer imaging and PDT.
Next, the in vitro detection performance of COF-survivin was further investigated. According to the uorescence spectra, the uorescence of the TSAS could be completely restored by adding the survivin target into the probe solution, and the response time was less than 30 min (Fig. 3A, B, S6 and S7 †). Furthermore, benetting from the rule of complementary base pairing, COFsurvivin has excellent target specicity; therefore its uorescence signal can only be recovered in the presence of Tsurvivin (Fig. 3C). In order to be used for biomedical purposes, the biocompatibility of probes is highly important. We investigated the dark toxicity of COF-survivin with MTT assay (Fig. 3D). The cells maintained high viability even when incubated with COF-survivin at a concentration of 100 mg mL À1 for 24 h, implying that the developed theranostic nanoplatform is nontoxic to living cells under dark conditions. Therefore, we further evaluated the in vitro diagnostic effect of COF-survivin. Interestingly, by incubating COF-survivin with MCF-7 and MCF-10A cells, the cancer cell line was specically lit up because survivin was overexpressed, much brighter than the normal cell  line (Fig. 3E). The above phenomenon was also demonstrated by quantitative ow cytometry analysis (Fig. 3F). The reliability of the probe was further identied on A549/Beas-2b cell lines, which also demonstrated that COF-survivin could be used for survivin overexpressed cancer cell imaging. (Fig. S8 †) Therefore, the COF-based nanoplatform could be employed for cancer cell diagnosis.
Subsequently, the in vitro therapeutic effect of COF-survivin was studied. As shown in Fig. 4A, cells still maintained high viability aer being irradiated with the laser alone, implying  that the selected power and exposure time were non-injurious to living cells. However, obvious concentration-and timedependent toxicity was observed on laser irradiated COFsurvivin treated cells, implying the excellent PDT effect of the nanoplatform. Live/dead cell staining assay further proved the excellent cancer cell inhibition effect of COF-survivin (Fig. 4B). We further carried out a series of experiments to investigate the therapeutic mechanism of COF-survivin. As shown in Fig. 4C, obvious green uorescence was observed in laser irradiated COF-survivin pretreated cells, revealing efficient ROS accumulation in this group; while no ROS accumulation was detected in the other groups, demonstrating that the nanoplatform possesses high biocompatibility and the laser parameters were feasible. This result was also proved by quantitative ow cytometry analysis. (Fig. S9 †) Overexpression of ROS in living cells may cause oxidative stress and induce mitochondrial damage, which could further trigger cell apoptosis. Based on the Rhodamine 123 probe, the mitochondrial membrane potential (DJ m ) of different groups was determined. Prominent DJ m loss was detected in the therapy group as shown in Fig. 4D. The expression of apoptosis biomarker caspase-3 in cancer cells with different treatments was detected by immunouorescence imaging assay. The results revealed obvious capase-3 activation in the cells in the therapy group (Fig. 4E). Flow cytometry analysis of cell apoptosis demonstrated that most cells survived in the control groups, while only 7.27% living cells were le in the therapy group (Fig. 4F). Taken together, it was practicable to employ COF-survivin for highly efficient cancer theranostics.
To further evaluate the theranostic application potential of COF-survivin, a series of in vivo experiments were carried out. We rst investigated whether the nanoplatform could differentiate tumor tissue from normal tissue. The result of in vivo uorescence imaging (Fig. 5A) revealed that COF-survivin could effectively light up tumor tissue, while no uorescence signal was observed from the normal tissue injected with the nanoplatform. Thus, COF-survivin could also efficiently discriminate the survivin expression levels in different tissues of living bodies. Subsequently, in vivo tumor PDT effect of the nanoplatform was further investigated. From the tumor H&E staining assay (Fig. 5B), obvious chromatin condensation could be observed in the tumor tissue injected with COF-survivin and treated with a 633 nm laser, while no pathological injury could be observed in the samples of the other three groups. An excellent tumor inhibition effect was realized and the tumor in the mice of the therapy group was completely eradicated (Fig. 5C and D). In vivo PDT with COF-survivin signicantly prolonged the survival rate of tumor-bearing nude mice. All the mice survived over 40 days, while the mice in the control groups were all dead within 18 days (Fig. 5E). Further investigations revealed that the body weight of all the mice did not change obviously (Fig. 5F), and no damage could be observed in the main organs. (Fig. S10 †) Therefore, COF-survivin was highly biocompatible and suitable for in vivo PDT. Survivin is an antiapoptotic protein, and intracellular survivin levels are closely related to the genesis and development of tumors. Since PDT can signicantly oxidize intracellular nucleic acids, further imaging of intratumoral survivin mRNA levels with the COFbased nanoplatform post treatment may help monitor the therapeutic effect to timely guide therapy, which is also known as the prognostic effect of theranostic probes. Therefore, on the 7th day post treatment, the mice of the therapy group were further treated with COF-survivin for in vivo uorescence imaging. Interestingly, no uorescence signal could be observed, indicating that in vivo PDT with the nanoplatform successfully eliminated the malignant tissue (Fig. 5G). All the in vivo experiments demonstrated that COF-survivin was a reliable theranostic nanoplatform for highly efficient in vivo tumor imaging, PDT and prognosis.

Conclusions
In summary, we developed a COF-based theranostic nanoplatform (COF-survivin) by integrating a TAMRA-labeled survivin antisense oligonucleotide onto a nanoscale porphyrin-based COF. The crystalline structure characteristics of the COF endowed the nanoplatform with excellent stability and ROS generation capability. In the presence of cancer biomarker survivin mRNA, the readily formed duplex detached from the COF, which could restore the uorescence by FRET prohibition and realize selective cancer imaging. Under NIR laser irradiation, the COF could generate abundant 1 O 2 species to induce cancer cell apoptosis, while no obvious dark toxicity was detected. The COF-based theranostic nanoplatform was successfully used for cancer diagnosis, therapy and prognostic evaluation. In view of the excellent designability and multifunctionality of the COF, many other targeting, diagnostic and therapeutic strategies could be rationally integrated to construct versatile COF-based theranostic probes. Compared with theranostic probes based on other nanomaterials such as metal-organic frameworks, this COF-based nanoplatform is heavy-metal free, relatively stable, biocompatible and highly integrated and performs well both in vitro and in vivo. This work opens up new avenues for COF-based probes and will inspire the use of more available tools for the biomedical eld.

Ethical statement
All the animal experiments were conducted and agreed with the Principles of Laboratory Animal Care (People's Republic of China) and the Guidelines of the Animal Investigation Committee, Biology Institute of Shandong Academy of Science, China.

Conflicts of interest
The authors declare no competing nancial interest.