A biodegradable covalent organic framework for synergistic tumor therapy

Stimulus-responsive biodegradable nanocarriers with tumor-selective targeted drug delivery are critical for cancer therapy. Herein, we report for the first time a redox-responsive disulfide-linked porphyrin covalent organic framework (COF) that can be nanocrystallized by glutathione (GSH)-triggered biodegradation. After loading 5-fluorouracil (5-Fu), the generated nanoscale COF-based multifunctional nanoagent can be further effectively dissociated by endogenous GSH in tumor cells, releasing 5-Fu efficiently to achieve selective chemotherapy on tumor cells. Together with the GSH depletion-enhanced photodynamic therapy (PDT), an ideal synergistic tumor therapy for MCF-7 breast cancer via ferroptosis is achieved. In this research, the therapeutic efficacy was significantly improved in terms of enhanced combined anti-tumor efficiency and reduced side effects by responding to significant abnormalities such as high concentrations of GSH in the tumor microenvironment (TME).


S3
FBS-containing culture media, or soybean trypsin inhibitor (2.0 mg/mL) was used to discontinue the dissociation. All animal procedures were reviewed and approved by the Ethics Committee of Shandong Normal University (Jinan, P. R. China), approval number AEECSDNU2022050. All the animal experiments complied with relevant guidelines of the Chinese government and regulations for the care and use of experimental animals. Nude mice (BALB/cJGpt-Foxn1nu/Gpt, aged 4 weeks) were purchased from Hangzhou Ziyuan Laboratory Animal Technology Co., Ltd. The nude mice were housed in a pathogen-free facility and kept in a temperature-controlled room set to light and dark cycle of 12 h each. To establish the MCF-7 xenograft model, MCF-7 cells (10 7 cells) suspended in HBSS (40 μL) were subcutaneously injected into the flanks of each mouse. The length (L) and width (W) of the tumor were determined using digital calipers. The tumor volume (V) was calculated by the formula: V = 1/2 × L × W 2 .

Experimental Instrumentations
Fourier transform infrared (FT-IR) spectra were obtained in the 4000~400 cm -1 range using a Thermo Scientific Nicolet iS50 FT-IR Spectrometer equipped with a diamond attenuated total reflection (ATR) module. Each spectrum was an average of 16 scans. Ultraviolet-visible (UV-vis) absorption spectra were recorded on a Shimadzu UV-2700 Double Beam UV-vis Spectrophotometer using 10 mm quartz cuvettes. Ultraviolet-visible absorption spectra were recorded on a Shimadzu UV-2700 Double Beam UV-Vis Spectrophotometer. Electron paramagnetic resonance (EPR) spectra were recorded on a Bruker A300 EPR Spectroscopy. Liquid chromatography-tandem mass spectrometry (LC-MS/MS) analyses were carried out on SCIEX ExionLC AD and QTRAP 6500+ LCMS/MS Systems. Powder X-ray diffraction (PXRD) patterns were obtained on a Rigaku SmartLab SE X-Ray Powder Diffractometer with Cu Kα line focused radiation (λ = 1.5405 Å) from 2θ = 2.00° up to 30.00° with 0.01° increment. Nitrogen-adsorption isotherms were measured at 77 K with a Micromeritics ASAP2020 HD88 Surface Area and Porosity Analyser. Before measurement, the samples were degassed a in vacuum at 120 °C for 8 h. The Brunauer-Emmett-Teller (BET) equation was used to calculate the specific surface areas. The pore size distribution was derived from the sorption curve using the non-local density functional theory (NLDFT) model. Transmission electron microscopy (TEM) images were recorded on a Hitachi HT7700 120 kV Compact-Digital Transmission Electron Microscope. High-angle annular dark field scanning transmission electron microscopy (HAADF-STEM) images and elemental mapping images were recorded using an FEI Talos F200X High-Resolution Scanning Transmission Electron Microscope. To prepare the TEM samples, the nanomaterial was dispersed in methanol by sonication for 5 min and the dispersion was placed on a carbon-coated copper TEM grid (300 mesh) and dried at room temperature. Hydrodynamic particle size and Zeta potential were measured using Malvern Zetasizer Nano ZS90 System. Microplate assays were carried out on a Molecular Devices SpectraMax i3x Multi-Mode Microplate Detection System. Cell counting was performed on a Thermo Fisher Scientific Invitrogen Countess II Automated Cell S4 Counter equipped with Countess Cell Counting Chamber Slides. Photomicrographs of biological samples were taken with a Leica DMI3000 B Inverted Fluorescence Microscope with an objective lens (10×, 20×, and 40×). Laser scanning confocal fluorescence images of cells were captured with a Leica TCS SP8 Confocal Laser Scanning Microscopy equipped with 405, 458, 488, 514, 561, and 633 nm lasers. Glass bottom dishes and 4/8-well chamber slides (Cellvis, Mountain View, CA, USA) were used for cell culture to provide biological replicates of each experiment. Before live cell imaging, the original culture media or DPBS was replaced with HBSS supplemented with HEPES (15 mM) and GlutaMAX to provide better-buffering capacity under normal CO 2 concentration. For imaging, the scan speed was 400 Hz and transmitted light was used to find the areas of interest to reduce photodamage to the biosample.

Statistical Analysis
All results are depicted as means ± SD of at least three biological replicates, as indicated in figure legends. And data were compared with the paired or unpaired two-tailed Student's t-test with or without Welch's correction, two-way ANOVA followed by Šídák's post hoc test, as appropriate. ns, no significance (p > 0.05), *p < 0.05, **p < 0.01.

Determination of sulfhydryl groups on nano DSPP-COF by Ellman's method
The Ellman reagent DTNB, i.e. 5,5'-dithiobis(2-nitrobenzoic acid), reacts with sulfhydryl groups to displace benzoic acid, i.e. TNB. DTNB has a characteristic absorption peak at 325 nm, while TNB shows a strong absorption peak at 412 nm under weakly basic conditions, and the sulfhydryl group concentration and absorbance values are in accordance with the Lambert-Beer law. The reaction of the sulfhydryl group with DTNB proceeded quantitatively, so the sulfhydryl group content in the sample could be determined by UV-Vis spectrophotometry.
Ellman's reagent preparation: 2 mg of DTNB was dissolved in 2 mL of a buffer solution with pH 7.8. Establishment of the standard curve: 4-Aminothiophenol was used as the sulfhydryl standard. 200 µL of 4-Aminothiophenol was mixed with 200 µL of Ellman's reagent and 1 mL of pH 7.8 buffer solution and incubated for 10 min at 37 °C. The absorbance at 412 nm was then measured using a multi-mode microplate detection system. Then the standard curve of concentration and absorbance was established. Then the absorbance at 412 nm was measured using a multi-mode microplate detection system.

Drug loading experiments
A mixture of nano DSPP-COF (5 mg) and 5-Fu (5 mg) in methanol (5 mL) was sonicated for 10 min. After being stirred (800 rpm) at 25 C for 24 h, the resulting solids were separated by vacuum filtration and washed with water for 3 times. Finally, the obtained 5-Funano DSPP-COF was resuspended in water and stored at 4 C until use.

ESR-Trapping Test
The acetone dispersion of nano DSPP-COF or 5-Funano DSPP-COF (200 μL, 50 μg/mL) and 10 μL TEMP in a test tube was irradiated with a red LED (50 mW/cm 2 ) for 60 s. Then, the resulting system was characterized using a Bruker EMX plus model spectrometer operating at the X-band frequency (9.4 GHz) at room temperature.

Photodynamic Property
A mixture of nano DSPP-COF or 5-Funano DSPP-COF (2 mL, 50 μg/mL) and DPBF (200 μL, 1 mM) in DMF in a quartz dish was irradiated with a red LED (50 mW/cm 2 ) for 1 min. The absorbance of DPBF at 414 nm in the mixture was recorded at 10 s intervals. The 1 O 2 generation S24 rate was determined from the reduced absorbance over time. To characterize the difference in the rate of 1 O 2 introduced by different lasers, the ratios A/A 0 of absorbance A and the initial absorbance A 0 at 414 nm at different irradiation times were calculated and plotted as the ordinate for the irradiation time. The dispersion of nano DSPP-COF or 5-Funano DSPP-COF (2 mL, 50 μg/mL) was used as the reference for this UV−vis measurement.

Cell Uptake and Subcellular Localization
To study the cell uptake, intracellular distribution, and subcellular localization, nano DSPP-COF was labelled with the fluorescent dye Bodipy-CHO. Briefly, nano DSPP-COF-Bodipy was prepared as follows: a mixture of nano DSPP-COF (1 mg), Bodipy-CHO (1 mg), and acetic acid (20 μL, 6 M) in ethanol (2 mL) was stirred at 70 °C for 24 h in the dark. After fully washing with ethanol, the resulting solids were re-dispersed into DPBS (1 mL) to afford a stock solution of nano DSPP-COF-Bodipy (1 mg/mL). The structure of Bodipy-CHO and the schematic diagram of postsynthetic modifications are illustrated below: , and 37 °C while pre-treating with DCA (15 mM, inhibiting aerobic glycolysis through inhibiting pyruvate dehydrogenase kinase), CPZ (10 μg/mL, clathrin-dependent endocytosis inhibitor), MβCD (10 mg/mL, caveolin-dependent endocytosis inhibitor), and AMR (37.5 μg/mL, micropinocytosis inhibitor) for 1 h. The cellular uptake was reflected by the MFI of green fluorescence. Data were presented as mean ± SD (n = 4). Scale bar, 50 μm.

Intracellular GSH Measurements
The level of intracellular GSH was measured using a GSH assay kit based on the 5,5'-dithiobis(2nitrobenzoic acid) (DTNB) colorimetric method. Experimentally, cells were seeded and cultured in 60 mm culture dishes for 24 h and treated with nano DSPP-COF (2.0 mL, 10 μg/mL) for 4 h in a CO 2 incubator. After being rinsed with DPBS carefully, the cells were cultured for an additional 24, 48, or 72 h and taken for GSH measurements according to the manufacturer's guidelines of the assay kit. Colorimetric signals were measured by absorbance at 412 nm using a multi-mode microplate detection system. The GSH content was normalized to total protein amount of the cell lysates from a parallel plate and expressed as a percentage value relative to the control group value.

CCK-8 Cell Viability Assays
Standard CCK-8 assay was applied to evaluate the cell cytotoxicity of the nanodrugs. Experimentally, ~5000 cells were cultured in 96-well plates for 24 h and treated with nano DSPP-COF or 5-Funano DSPP-COF (100 μL, 0~20 μg/mL) for 4 h in a CO 2 incubator. Then, the cells were washed with DPBS carefully. For PDT, the cells were exposed to red LED (50 mW/cm 2 , 8 min), and cultured for an additional 24 h. Subsequently, the CCK-8 solution (10 μL) was added to each well and the plate was incubated in a CO 2 incubator for about 2 h. The absorbance at 450 nm was determined using a multi-mode microplate detection system.
For the selectivity of 5-Funano DSPP-COF on normal and cancer cells, cell viabilities of MCF-7 and MCF-10A cells were treated with 5-Funano DSPP-COF for 4 h and cultured for an additional 72 h. Subsequently, the CCK-8 solution (10 μL) was added to each well and the plate was incubated in a CO 2 incubator for about 2 h. The absorbance at 450 nm was determined using a multi-mode microplate detection system. For the toxicity of 5-Fu, we incubated the MCF-7 cells with the same concentration of 5-Fu as in 5-Funano DSPP-COF for 4 h and continued incubation for 24 h after the drug was removed. Subsequently, the CCK-8 solution (10 μL) was added to each well and the plate was incubated in a CO 2 incubator for about 2 h. The absorbance at 450 nm was determined using a multi-mode microplate detection system.

S26
In the CCK-8 cell viability assay, the cells without treatment were used as the control. The wells without cells were used as blanks. The cell viability was expressed as a percentage value relative to the control group value.

Calcein-AM/PI Double Staining
Cells were seeded into 60 mm culture dishes and incubated overnight in a CO 2 incubator. After removal of the culture medium, the cells were incubated with DPBS dispersion of nano DSPP-COF or 5-Funano DSPP-COF (2 mL, 20 μg/mL) for 4 h in a CO 2 incubator. For PDT, the cells were exposed to red LED (50 mW/cm 2 , 8 min). For the Fer-1 addition group, MCF-7 cells were first pretreated with Fer-1 at a concentration of 20 µM for 1 h, followed by the same treatment as above. After additional 24 h incubation, cells were collected by centrifugation after digestion with trypsin solution, washed with DPBS twice carefully, and incubated with calcein-AM (500 μL, 2 μM) and PI (500 μL, 4 μM) for 15 min. Finally, the cells were washed with DPBS twice carefully, and imaged with a laser scanning confocal microscope. The green images of living cells were excited by 488 nm light, and the emission wavelength range was collected at 520 ± 20 nm. The red images of dead cells were excited by 514 nm light, and the emission wavelength range was collected at 640 ± 20 nm.

Intracellular Total ROS Measurements
Levels of intracellular ROS were measured by the cell-permeable dye DCFH-DA. Experimentally, cells were treated with nano DSPP-COF or 5-Funano DSPP-COF (500 μL, 20 μg/mL) for 4 h in a CO 2 incubator. For the Fer-1 addition group, MCF-7 cells were first pretreated with Fer-1 at a concentration of 20 µM for 1 h, followed by the same treatment as above. And then the cells were washed with DPBS carefully. Afterward, the cells were loaded with DCFH-DA (200 μL, 100 nM) for 15 min in a CO 2 incubator and washed with DPBS twice. For PDT, the cells were exposed to red LED (50 mW/cm 2 , 8 min). Finally, the laser scanning confocal fluorescence images were captured. The green images were excited by a 488 nm light, and the emission wavelength range was collected at 525 ± 20 nm. Cells without any nanodrug treatment were used as a control group.
The cells without treatment were used as the control. The cell viability was expressed as a percentage value relative to the control group value.

Intracellular GPX4 Activity Measurements
Cells were seeded and cultured in 60 mm culture dishes for 24 h and treated with nano DSPP-COF or 5-Funano DSPP-COF (2.0 mL,20 μg/mL) for 4 h in a CO 2 incubator. After carefully rinsed with DPBS. For PDT, the cells were exposed to a red LED (50 mW/cm 2 , 8 min). After that, the cells were cultured for an additional 24 h, and then quickly frozen in liquid nitrogen. Subsequently, the S28 frozen cells were homogenized in lysis buffer on ice and the supernatant was used for GPX4 activity measurements using a glutathione peroxidase assay kit. Colorimetric signals were measured by absorbance at 340 nm using a multi-mode microplate detection system. The GPX4 activity was normalized to the total protein amount of the cell lysates from a parallel plate and expressed as a percentage value relative to the control group value. Similarly, the expression of GPX4 in MCF-7 cells upon formulation treatment was also analyzed by western blotting. The cell lysates containing identical protein (40 µg) were subjected to standard electrophoresis, followed by antibody incubation at 4℃. The dilution ratio for the first antibody was 1:2000 (actin-specific antibody) and 1:2500 (GPX4-specific antibody). Regarding the secondary antibody, the dilution ratio was 1:5000 for both GPX4 and actin. The protein bands were developed via the ECL TM western blotting detection reagents.

Intracellular Lipid Peroxidation Assays
Cells were treated with nano DSPP-COF or 5-Funano DSPP-COF (500 μL, 20 μg/mL) for 4 h in a CO 2 incubator. For PDT, the cells were exposed to red LED (50 mW/cm 2 , 8 min). For the Fer-1 addition group, MCF-7 cells were first pretreated with Fer-1 at a concentration of 20 µM for 1 h, followed by the same treatment as above. And then the cells were washed with DPBS carefully and cultured for an additional 4 h. After that, the cells were incubated with C 11 -BODIPY (200 μL, 2.0 μM) for 30 min in a CO 2 incubator and washed with DPBS twice. Finally, the laser scanning confocal fluorescence images were captured. The green images of the oxidized C 11 -BODIPY dye were excited by a 488 nm light, and the emission wavelength range was collected at 510 ± 20 nm. The red images of the reduced C 11 -BODIPY dye were excited by a 561 nm light, and the emission wavelength range was collected at 591 ± 20 nm. Cells without any nanodrug treatment were used as a control group.

Mitochondrial Membrane Potential Measurements
Mitochondrial membrane potential was measured by a fluorescent lipophilic carbocyanine dye JC-1. Experimentally, cells were treated with nano DSPP-COF or 5-Funano DSPP-COF (500 μL, 20 μg/mL) for 4 h in a CO 2 incubator. And then the cells were washed with DPBS carefully. For the Fer-1 addition group, MCF-7 cells were first pretreated with Fer-1 at a concentration of 20 µM for 1 h, followed by the same treatment as above. For PDT, the cells were exposed to red LED (50 mW/cm 2 , 8 min). After additional 4 h incubation, the cells were incubated with JC-1 (200 μL, 15 μM) for 10 min in a CO 2 incubator and washed with DPBS twice. Finally, the laser scanning confocal fluorescence images were captured. The green images of monomer were excited by 488 nm light, and the emission wavelength range was collected at 530 ± 15 nm. The red images of Jaggregate were excited by 561 nm light, and the emission wavelength range was collected at 590 ± 17 nm. Cells without any nanodrug treatment were used as a control group.

Lysosomal Membrane Permeabilization Detections
Lysosomal membrane permeabilization was measured by a lysosomotropic metachromatic fluorochrome AO. Experimentally, cells were treated with various nano DSPP-COF or 5-Funano DSPP-COF (500 μL, 20 μg/mL) for 4 h in a CO 2 incubator. And then the cells were washed with DPBS carefully. For the Fer-1 addition group, MCF-7 cells were first pretreated with Fer-1 at a concentration of 20 µM for 1 h, followed by the same treatment as above. For PDT, the cells were exposed to red LED (50 mW/cm 2 , 8 min). After additional 4 h incubation, the cells were incubated with AO (200 μL, 15 μM) for 10 min in a CO 2 incubator and washed with DPBS twice. Finally, the laser scanning confocal fluorescence images were captured. The green images of deprotonated AO were excited by 488 nm light, and the emission wavelength range was collected at 530 ± 20 nm. The red images of protonated AO were excited by 488 nm light, and the emission wavelength range was collected at 640 ± 20 nm. Cells without any nanodrug treatment were used as a control group.

Hemolysis Analysis
First, fresh nude mouse blood samples (2 mL) were added to PBS solution (4 mL), and red blood cells (RBC) were separated by centrifugation at 3000 rpm for 10 minutes. After washing 5 times with PBS, dilute purified red blood cells to 20 mL with PBS. For hemolysis assay, 0.2 mL of diluted RBCs suspension was added to 1.0 mL of PBS as a negative control and 1.0 mL of deionized water as a positive control. And 1.0 mL 5-Funano DSPP-COF suspension at a concentration range of 1 to 200 μg/mL. All mixtures were then allowed to stand at 37 °C for 5 h and then centrifuged at 13300 rpm for 10 minutes. Due to the small size of 5-Funano DSPP-COF, it was difficult to separate 5-Funano DSPP-COF completely even by centrifugation at 13300 rmp for 10 minutes. Therefore, we chose the supernatant of the corresponding concentration as a control. The absorbance of 541 nm supernatant was measured by synergy SpectraMax i3x multi-mode microplate reader. The hemolytic percentage of red blood cells was calculated by the following formula: Hemolysis Rate = [ (Dt -Dcc)/(Dpc -Dnc)] ×100%.

In vivo biodistribution
To study in vivo biodistribution, nano DSPP-COF was labelled with the fluorescent dye Cy5-NH 2 . Briefly, nano DSPP-COF-Cy5 was prepared as follows: a mixture of nano DSPP-COF (1.0 mg), Cy5-NH 2 (1.0 mg), and acetic acid (20 μL, 6.0 M) in ethanol (2.0 mL) was stirred at 70 °C for 24 h in the dark. After fully washing with ethanol, the resulting solids were re-dispersed into DPBS (1.0 mL) to afford a stock solution of nano DSPP-COF-Cy5 (1.0 mg/mL). 5-Funano DSPP-COF-Cy5 was prepared by the same method as that used to prepare 5-Funano DSPP-COF, and the resulting solids were re-dispersed into DPBS (1.0 mL) to afford a stock solution of 5-Funano DSPP-COF-Cy5 (1.0 mg/mL).

Fig. S23
The structure of Cy5-NH 2 and schematic diagram of post-synthetic modifications.
To evaluate the biodistribution of 5-Funano DSPP-COF-Cy5 (50 µL,1mg /mL) was intratumorally injected. Then, nude mice with MCF-7 tumors were anesthetized and imaged at different times (4 h, 8 h, 12 h, 24 h, and 0-12 days) using an in vivo imaging system with an excitation wavelength of 640 nm and an emission wavelength of 680 nm. Then, the mice were killed and organs and tumors were separated for ex vivo imaging to determine the biodistribution pattern and retention of 5-Funano DSPP-COF-Cy5 in the tumors. It is important to note that nude mice have to be fed with non-fluorescent chow for 24 hours before imaging.