Controllable gelation of artificial extracellular matrix for altering mass transport and improving cancer therapies

Global alterations in the metabolic network provide substances and energy to support tumor progression. To fuel these metabolic processes, extracellular matrix (ECM) plays a dominant role in supporting the mass transport and providing essential nutrients. Here, we report a fibrinogen and thrombin based coagulation system to construct an artificial ECM (aECM) for selectively cutting-off the tumor metabolic flux. Once a micro-wound is induced, a cascaded gelation of aECM can be triggered to besiege the tumor. Studies on cell behaviors and metabolomics reveal that aECM cuts off the mass transport and leads to a tumor specific starvation to inhibit tumor growth. In orthotopic and spontaneous murine tumor models, this physical barrier also hinders cancer cells from distant metastasis. The in vivo gelation provides an efficient approach to selectively alter the tumor mass transport. This strategy results in a 77% suppression of tumor growth. Most importantly, the gelation of aECM can be induced by clinical operations such as ultrasonic treatment, surgery or radiotherapy, implying this strategy is potential to be translated into a clinical combination regimen.


S3
Conjugation of Trail to Fb-N3. A method of coupling Trail to Fb-N3 was managed by using EDC/NHS chemistry. To conjugate Trail with Fb-N3, 6 mg of Fb-N3 was suspended in 3 mL PBS buffer. 0.5 mL of PBS containing 0.48 mg of 1-(3dimethylaminopropyl)-3-ethylcarbodiimide hydrochloride (EDC.HCl, 2.5 mM) and 0.29 mg of N-hydroxysuccinimide (NHS, 2.5 mM) were added to the Fb-N3 solution and stirred at room temperature for 4 h. The product was collected via dialysis in phosphate buffered saline (PBS) to remove unreacted EDC.HCl and NHS. Trail was then added and the mixture was stirring at room temperature for another 4 h. The resulting Trail-Fb-N3 was purified with ultrafiltration membranes (MwCO 50 kDa) to remove unbound Trail.

Conjugation of diatrizoic acid (DA) to Fb-N3.
To conjugate DA to Fb-N3, 6 mg of DA was suspended in 4 mL PBS buffer. Equal amounts of EDC and NHS were added to the DA suspension and stirred at room temperature for 2 h to activate DA. Fb-N3 was then added and the mixture was reacted for another 1.5 h. The resulting DA-Fb-N3 was collected and purified with ultrafiltration membranes (MwCO 10 kDa).
The fluorescence labeling of proteins. Rhodamine B isothiocyanate labeled Fb-N3 was managed by using the reaction between rhodamine B isothiocyanate and Fb-N3. In brief, 10 mg of Fb-N3 was suspended in 5 mL PBS buffer. 0.5 mg of the rhodamine B isothiocyanate was added to the Fb-N3 solution and stirred for 4 h. Resulting rhodamine B labeled Fb-N3 was collected and purified with ultrafiltration membranes (MwCO 10 kDa) to remove unbound rhodamine B isothiocyanate.
The labeling of Fb-N3 and Ptb-DBCO was conducted respectively with Cy7-NHS S4 and Cy5-NHS in PBS buffer at pH 7.4. 10 mg of Fb-N3 or Ptb-DBCO was suspended in 5 mL PBS buffer. 0.5 mg of dyes were added to the Fb-N3 solution and stirred for 4 h. The products were then collected and purified with ultrafiltration membranes (MwCO 10 kDa) to remove unbound dyes.
Cell migration assay. In the cell migration experiment, CT26 cells were seeded in the 24-well plates (5 × 10 5 cells per well) and cultured for 24 h. The medium was removed and the cells were washed twice with Versene followed by adding 1 mL of Versene (containing 0.05% trypsin) and keeping for 10 to 15 min at 37 o C. 3 mL of DMEM (10% FBS) was added followed by centrifugation at 150-200 g for 5 min. The medium was removed and the cells were resuspended in 1.5 mL of DMEM for later use. The prepared CT26 cells were then added into the top chambers of Transwell. Then, aECM was prepared by adding thrombin (1 mg mL -1 , 10 μL) and Fb (3 mg mL -1 , 0.1 mL) into the cells for forming gel-like clots. The lower chambers were added with 500 μL of 1640 medium as a chemoattractant. After incubation with aECM or PBS for 48 h, the number of cells migrated to the lower chambers was counting under a light microscope at 40 × magnification (n = 5).
Cell invasion assay. 200 μL of the 1640 medium containing rat tail type I collagen (100 mg mL -1 ) was gelled in the upper chamber of Transwell insert and incubated at 37 o C for 6 h. The gels were washed thrice with PBS and dried at 37 o C. The prepared CT26 cells were then added into the top chambers of collagen-coated Transwell insert. Then, aECM was prepared by adding thrombin (1 mg mL -1 , 10 μL) and Fb (3 mg mL -1 , 0.1 mL) in to the cells for forming gel-like clots. The lower chambers were added with 500 S5 μL of serum-containing 1640 medium as a chemoattractant. After incubation with aECM or culture medium for 48 h, the number of cells migrated to the lower chambers was counted under a light microscope at 40 × magnification (n = 5).
Tracking experiments for cancer cell motility. To achieve long-term observation of cell migration, microscope stage-top incubator system was used. Tracking experiments for cancer cell motility were performed over 11 h. During the observation, 4T1 cells were cultured in a 5% CO2 atmosphere at a constant temperature of 37 o C. The movement routes of CT26 cells within 11 h were analyzed with an Image J plug-in, MTrack2. Both speed and mean squared displacement values were quantitatively analyzed from 100 cells from 4 repeated experiments by using Image J plug-in, MTrack2.
The preparation of multicellular tumor spheroids (MTS). 1% (w/v) of agarose gel was covered on the 96-well plates to prevent the cell adhesion. Then, cancer cells (4T1, CT26, MCF-7 or HT29 cells) were seeded into wells (about 1000 cells per well) and agitated for 5 min followed by incubation at 37 o C for 7 days. The uniform and compact MTS were used for the further studies.
The disassembly of MTS. The disassembly of MTS was carried out by using a shake cultivation. MTS in 96-well plates was treated with aECM or culture medium. After being washed with PBS for three times, MTS was shaken on a shake cultivation at 300 rpm. The disassembly of MTS was observed by fluorescent microscope every 2.5 min.
Cell viability study. The cell viability was firstly studied in 2D cell culture system. CT26, 4T1, MCF-7 and HT29 cells were respectively seeded in 96-well plates (5 × 10 3 S6 cells per well). After 24 h of growth, the cells were treated with aECM for another 24 h. The cell viability was also studied in 3D MTS model. The prepared MTS was treated with aECM for 24 h. After being washed with PBS for three times, the cells were added with 10 μL of CCK-8 solution and cultured for 2 h. The UV-vis absorptions at 450 nm (test wavelength) and 690 nm (reference wavelength) of cells were measured using a microplate reader. Cell viability (%) was calculated, and data was presented as mean ± standard deviation (SD) in triplicate.
In vitro metabolic study. CT26, 4T1, MCF-7 and HT29 cells were respectively seeded in 96-well plates (5 × 10 3 cells per well). After 24 h of growth, the cells were treated with aECM for another 24 h. For the 3D MTS model, 1% (w/v) of agarose gel was covered on the 96-well plate to prevent cell adhesion. Then, cancer cells (4T1, CT26, MCF-7 or HT29 cells) were seeded into wells (about 1000 cells per well) and agitated Enzymatic degradation of aECM in vitro. The degradation of aECM by MMP-2 in vitro was studied using FITC-labeled aECM (FITC-aECM). Briefly, the FITC-aECM was prepared by adding thrombin (1 mg mL -1 , 50 μL) to FITC-labeled fibrinogen (3 mg mL -1 , 0.5 mL) in 24-well plates and standing for 5 min at 37 o C to activate the S7 fibrinogen adequately for gel-like clots formation. Then, the clots were separated and washed by PBS several times until almost all the free FITC-labeled fibrinogen was removed. The prepared FITC-aECM was incubated in PBS supplemented with or without MMP-2 (1.5 μg mL -1 ) at 37 o C. MMP-2 was added every 2 days and the fluorescence of the supernatant was recorded every day. The hydrolysis rates were calculated.
Enzymatic hydrolysis of aECM in vivo. The BALB/c mice with CT26 cell (1 × 10 6 cells per mouse) tumors inoculated on their right flank were used for in vivo aECM enzymatic hydrolysis study. When the tumor volume reached to ~100 mm 3 , the in vivo gelation was induced. Cy7-Fb-N3 was injected intravenously (5 mg mL -1 , 100 μL) to the mice. 1 h later, an ultrasonic treatment (600 W, 5 s) was carried out in the tumor position to trigger the gelation of fibrinogen. 12 h after these treatments, Ptb-DBCO (2 mg mL -1 , 100 μL) was injected through tail vein. A second ultrasonic treatment (600 W, 5 s) was performed to induce the clotting. For the control group, the pan MMPs inhibitor, Batimastat, was intraperitoneal injected at a dose of 30 mg kg -1 . Live animal fluorescent imaging (IVIS) was used to track intratumoral degradation of Cy7 labeled aECM at 1st, 5th and 10th days.
Metabonomics and transcriptomics analysis. Metabonomics analysis was carried out according to the previous work of our group. The BALB/c mice with CT26 cell (1 × 10 6 cells per mouse) tumors and 4T1 cell (1 × 10 6 cells per mouse) tumors inoculated on their right flank were used for metabonomics analysis. When the tumor volume reached to ~100 mm 3 , aECM gelation was induced. CT26 tumors and 4T1 tumors were S8 collected 3 days after the gelation and stored at −80 o C for later use.

Vascular normalization.
For obtaining the three-dimensional structure of the vessels in and around the tumor, μ-CT scanning was performed on aECM treated tumors (n = 3 for each groups). Mice were sacrificed and perfused intracardiacally with silicone rubber radiopaque compound Microfil (FlowTech, Carver, MA), which polymerized in blood vessels within 20 minutes. After Microfil perfusion, the tumor was excised and preserved in 4% formalin. The blood vessels in the tumor were scanned with μ-CT. The three-dimensional image of the tumor blood vessels was reconstructed by the software that comes with the instrument.

S9
The vascular leakage and perfusion were studied by intravenous injection of FITCdextran and FITC-lectin. Extravascular diffusion was assessed by intravenous injection with FITC-dextran (40 kDa, 0.25 mg per mouse). 10 min after the injection, mice tumors were collected. Cryosections with a thickness of 20 μm were obtained. The blood vessels were stained with anti-CD31 antibody (red fluorescence). The FITCdextran was calculated as the ratio of green fluorescence to red fluorescence.
The vascular perfusion was assessed by intravenous injection with FITC-lectin (10mg kg -1 ). 10 min after the injection, mice tumors were collected. Cryosections with a thickness of 20 μm were obtained. The blood vessels were stained with stained anti-CD31 antibody (red fluorescence). The overlap ratio between lectin+ area and CD31+ area was calculated.
In vivo fluorescence imaging. In order to prove the in vivo gelation of aECM, intravital fluorescence microscopy was performed to visualize this process. To establish the ear tumor model, 50 μL 4T1 cell suspension (containing 1 × 10 6 cells) was injected in the peripheral ear. 7 day after the injection, the mouse was anesthetized by using isoflurane followed by injection with 50 μL FITC-labeled dextran (FITC-dextran; 5%, 150 kDa).
The ear of the mouse was settled on the microslide by coating a piece of coverslip on the microslide to clamp the ear. The animal was then moved to the desk of the fluorescence microscope. The fluorescence microscope was adjusted for FITC-dextran visualization. And then the mouse was injected with RB-labeled fibrinogen followed by ultrasound. The vessel was observed at intervals. Adequate depth of anesthesia should be ensured during all the process.

S10
The BALB/c mice with CT26 (1 × 10 6 cells per mouse) tumors inoculated on their right flank were also used to study the in vivo fluorescence imaging. When the tumor volume reached to ~200 mm 3 , mice were intravenously injected with Cy7-Fb-N3 and Cy5-Ptb-DBCO followed by treatment with ultrasonic in the tumors. Then, the fluorescence imaging of mice was performed by an IVIS Spectrum (PerkinElmer) at different time points.
Additional glucose for impairing the therapeutic effect of aECM. CT26 cells suspended in PBS were subcutaneously injected (1 × 10 6 ) into both left and right flanks of female BALB/c mouse (5-6 weeks). With the US treatment, the gelation of aECM was initiated in both mice tumors. In the left tumor, the intratumoral injection of glucose solution with a dose of 1g kg -1 was performed every day. Meanwhile, the same volume of saline was intratumoral injected into the right tumor. The tumor volume was recorded every two days.
Tissue clearing and 3D tumor fluorescence imaging. Tissue clearing was performed on CLARITY technique. CT26 tumor-bearing mice were sacrificed after aECM treatment for 24 h followed by perfusion with 0.5% NaNO2 and 10 U mL −1 heparin contained ice-cold PBS. And then the tumors of mice were collected and mixed with 4% paraformaldehyde at 37 o C for 2 h. The tumors were then soaked in acrylamide monomer (4%) and 2,2'-azobis [2-(imidazolin-2-yl)propane] dihydrochloride thermoinitiator (0.25%) contained ice-cold PBS at 4 o C for 1 day to ensure enough penetration of the monomer and initiator. Then, the polyacrylamide hydrogel was kept in 37 o C water with N2 protection to initiate the polymerization. After 3 h of S11 polymerization, the system was then washed with PBS contained 0.01% NaN3 and 10% SDS by keeping in 37 o C water for 4 days until became optically transparent hydrogel.
The CLARITY method could only remove lipid molecules without removing biomacromolecules such as proteins or nucleic acids among tissues. Thus, further immunofluorescence staining could be performed. The transparent hydrogel obtained by CLARITY method was then stained with DAPI for 15 min and washed with PBS for several times. The 3D tumor fluorescence imaging was observed on a C1-Si (Nikon) confocal laser scanning microscope.
In vivo CT imaging of fibrinogen. The BALB/c mice with CT26 cell (1 × 10 6 cells per mouse) tumors inoculated on their right flank were used for in vivo CT imaging (n = 3 for each groups). The in vivo CT imaging was performed on a Quantum FX micro-CT imaging system (Perkin-Elmer) by using diatrizoic acid conjugated Fb-N3. When the CT26 tumor volume reached to ~200 mm 3 , the in vivo gelation of aECM was induced. A similar dose of DA-Fb-N3, instead of Fb-N3, was used. The CT imaging was recorded before and after treatment with ultrasonic.
MRI imaging for the permeability of the tumor. The BALB/c mice with CT26 cell (1 × 10 6 cells per mouse) tumors inoculated on their both side of flank were used for in vivo permeability study. The in vivo MRI imaging was performed on a 7.0 T clinical MRI instrument and the T1-weighted MRI scanning was performed on the bilateral tumor-bearing mice (n = 3 for each groups). Briefly, when the tumor volume reached to ~100 mm 3 , the right tumor of each mouse was treated with aECM through subcutaneous injection and the left one was injected subcutaneously with S12 equivoluminal PBS as the control. After 12 h post-administration, the contrast agent gadopentetate dimeglumine (0.1 M, 150 µL) was injected intravenously to the mice.
Hemolysis assay. Blood samples were obtained from heart of mice by using blood collection tube containing Na-heparin as an anti-coagulant. Then, 100 μL Blood samples were respectively added into 900 μL of PBS (as negative control), deionized water (as positive control), Fb and aECM. After co-incubation in shaking incubator for 4 h at 37 o C, the mixtures were centrifuged (300 g, 5 min) to remove unbroken red blood cell and the clots. The supernates were then added into a 96-well plate followed by measuring their absorbance at 540 nm by using a microplate reader. The data are presented as mean ± standard deviation (SD) in triplicate.

Blood routine examination and blood biochemistry analysis.
To evaluate the effects in physiology caused by aECM, the blood biochemical indexes and blood routine indexes were tested. Female BALB/c mice were divided into 3 groups (n = 3 for each groups) and treated respectively with PBS, aECM and aECM-Trail via i.v. administration. After 3 days of post-administration, blood samples were collected from heart (100 μL each mouse). The blood routine examination was carried out by Auto Hematology Analyzer (MC-6200VET), and blood biochemistry analysis was performed by biochemical auto analyzer (MNCHIP, Tianjin, China).