Mouse tumour tissue processing
Briefly, the tumour tissues were resected, chopped into small pieces, and digested in a 5% CO2 incubator at 37 °C for 20 min with RPMI-1640 (Biowest, L0498) containing collagenase type IV (Worthington Biochemical, LS004189, 1 mg/mL), hyaluronidase (Sigma, H6254, 1 mg/mL), and DNase I (Sigma, DN25, 1.5 mg/mL). The digested tissues were minced, filtered through a 70 μm cell strainer, and made into single-cell suspensions.
Ultrasound imaging
Ultrasound imaging was performed using a Vevo 2100 System (VisualSonic) with an LZ550 probe (frequency 40 MHz) to measure the size of each tumour twice a week, 7 days after the injection of the cancer cells until the end of the study. Tumour-bearing mice were anaesthetised with an intraperitoneal injection of a mixed solution of ketamine (100 mg/kg) and xylazine (10 mg/kg). The skin was covered with ultrasound gel, and images were acquired using a probe. Tumour tissues in the abdominal cavity were defined as a different echogenicity from the normal pancreas on ultrasound imaging. The long diameter (LD) and short diameter (SD) were measured. Tumour volume was calculated using the following formula: tumour volume = (LD × SD2)/2.
Masson-trichrome staining
For direct visualization of collagen deposition, Masson-trichrome staining was performed. Briefly, paraffin-embedded, formalin-fixed tissue blocks were cut into 4-μm thick slices. The slices were deparaffinized and then subjected to Masson's trichrome staining by using reagents and kits from Sigma-Aldrich following the manufacturer’s protocols. Slides were analyzed using Nikon Eclipse Ts2R microscope and the image quantification was performed by using Image J program.
IHC analysis
Paraffin-embedded, formalin-fixed tissue blocks were cut into 4 μm thick slices. IHC analysis was performed for CD66b, CD3, CD4, and CD8 using the Ventana BenchMark XT Staining system. Briefly, the IHC sections were baked and deparaffinised, and antigen retrieval was performed for 24 min at 100 °C with Cc1 solution (Ventana, #05279801001). Endogenous peroxidase was blocked using 3% hydrogen peroxide for 4 min at 37 °C. The following primary antibodies were then used: anti-CEACAM8/CD66B (G10F5 clone, Novus Biologicals, Cat: NB100-77808, 1:100 dilution) as a PMN-MDSC marker; anti-CD3 (polyclonal, DAKO, Cat: A0452, pre-diluted) as a pan-T cell marker, anti-CD4 (SP35 clone, Ventana, Cat: 790-4423, pre-diluted) for helper T cells, and anti-CD8 (SP57 clone, Ventana, Cat: 790-4460, pre-diluted) for cytotoxic T cells. An OptiView universal DAB kit (Ventana, #760-700) was used for the enzyme-substrate reaction. After immunostaining, the slides were counterstained with haematoxylin, and the haematoxylin shading was changed to a blue colour by applying a bluing agent. IHC assessments were performed using a Nikon Eclipse Ts2R microscope (Nikon, Japan). For each marker, the cells were manually counted in five randomly chosen areas per tissue section using sections from 10 patients.