Bleomycin

We used 6-week-old male Sprague Dawley rats. Rats were euthanized with CO₂ asphyxiation, and eyeballs were removed. Next, we enucleated lenses in PBS with using a dissection microscope. Enucleated lens were cultured in 2 mL serum-free M199 medium (Sigma) containing 0.1% BSA [35 (link)]. MMS (Nakalai tesque) or Bleomycin (Tokyo chemical industry) were added to the culture medium at final concentrations of 1 mM and 800 μg/mL, respectively. Culture medium was changed once daily, and fresh drug was added to the medium every time it was changed. All lenses were maintained at 37°C in a humidified incubator with 95% room air and 5% CO₂. On the last day of culture, we photographed lenses with a microscope using a DP58 camera (Olympus) attached to an SZX12 stereomicroscope in a 35 mm dish containing 7 mL PBS. All experiments were approved by the Animal Research Committee of the University of Fukui (Approval number: 28091) and were conducted in accordance with the University of Fukui Animal Care and Use Regulations and Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research. The findings of the study were reported in accordance with ARRIVE guidelines.

Recombinant human TGFβ (100-B-001; R and D Systems, McKinley Place, NE, USA) was dissolved in 0.1% bovine serum albumin (BSA) (A1933, Sigma-Aldrich, St. Louis, MO, USA) with HCl (4 mM) at 10 µg/mL. The following materials and reagents were also used: Lipofectamine™ RNAiMAX transfection reagent (13778150; Invitrogen, Waltham, MA, USA); bleomycin (B3972; Tokyo Chemical Industry Co., Ltd., Tokyo, Japan); mouse anti-col1a antibody (sc-59772; Santa Cruz Biotechnology, Santa Cruz, CA, USA); rabbit anti-Tab1 antibody (ab25878; Abcam, Cambridge, UK); Hoechst 33342 (H3570; Invitrogen); osa-miR172d-5p mimic (Fasmac, Tokyo, Japan); Alexa fluor 488-labeled anti-mouse antibody Fab fragment (A11017) and Alexa fluor 555-labeled anti-rabbit antibody Fab fragment (A21428; Invitrogen); phosphate-buffered saline (PBS) (045-29795; Fuji Firm, Tokyo, Japan); AteloGene (1391; Koken, Tokyo, Japan); 4% paraformaldehyde (163-20145) and Lemosol (128-03993; Fujifilm, Tokyo, Japan); Vectashield (H-1000; Vector Laboratories, Burlingame, CA, USA); and Trans Blot nitrocellulose membranes (Protran BA 85; Sigma Aldrich). miRNA data were obtained from previous reports^{19 (link),48 (link),49 (link)} followed by miRDB (http://www.mirdb.org/).

For the orthotopic syngeneic mouse models, the recipient C57BL/6 wild-type mouse was anesthetized with 4 mg/kg of midazolam (Wako), 0.3 mg/kg of medetomidine (Wako), and 5 mg/kg of butorphanol (Wako), and placed on a tilted platform. A 50-µL volume of bleomycin (Tokyo Chemical Industry, Tokyo, Japan) at 0.5 mg/mL was administered to the mouse via the trachea with the use of a cannula. At 2 weeks after bleomycin treatment, a single-cell suspension of KC or AC cells 1 × 10⁵ cells in 50 µL of PBS was introduced into the lungs via the trachea as for bleomycin administration (Supplementary Figure S3, Supplementary Video S1). At the end of experiments, tumors were removed from mice and fixed in 4% paraformaldehyde (Wako) overnight.
For the subcutaneous injection model, a single-cell suspension of AC cells 1 × 10⁶ cells in 100 µL of 50% growth factor–reduced was injected under the skin on the dorsal flank of a nude mouse. Tumor volume (mm³) was calculated every 2 days according to the formula: short axis² × long axis × 0.5236. When tumor size reached ~75 mm³, mice were randomly assigned to receive vehicle (Methyl cellulose, Wako) or crizotinib (Pfizer, New York, NY, USA) at 150 mg/kg by daily oral administration.