Hematoxylin and bluing reagent

After ex vivo imaging, the specimen was formalin-fixed, sectioned and then paraffin-embedded. IHC of the formalin-fixed paraffin-embedded tumor samples was performed on a BenchMark Ultra autostainer (Ventana Medical Systems, Oro Valley, AZ, USA). In brief, 3 µm paraffin sections were deparaffinized with the EZ prep solution (Ventana Medical Systems), and heat-induced antigen retrieval was carried out using Cell Conditioning 1 (Ventana Medical Systems). Hereafter, sections were incubated with an anti-c-MET antibody (clone SP44; Roche Diagnostics, Rotkreuz, Switserland), and an UltraView Universal DAB Detection Kit (Ventana Medical Systems) was used to visualize the c-Met expression. Slides were counterstained with Hematoxylin and Bluing Reagent (Ventana Medical Systems). Due to the lack of a quantitively read-out in routine immunohistochemistry, the c-Met expression levels were visually scored by a dedicated pathologist. Scoring was based on the intensity of the UltraView signal (four classifications were used: no staining, weak, moderate and strong, respectively −, +, ++, +++). Both membrane and cytoplasmic staining were noted. Since c-Met is considered a membrane-bound biomarker [9 (link)], membranous staining was leading in the scoring.

Tissue sections (4 μm thick) were entirely stained with automated stain kit (Table S4) (Ventana Medical Systems Inc., Oro Valley, AZ, USA). Slides were heated to 95 °C, and cell conditioning solutions (Ventana Medical Systems Inc.) #1 (cc1, cat.#950-124) or #2 (cc2, cat.#950-123) were added for set lengths of time for antigen retrieval (refer to Table S4 for cell conditioning solution and antigen retrieval time). Pre-diluted antibodies (Table S5) were added manually. All slides were incubated at 37 °C. Antigen-antibody reaction was revealed using Universal DAB detection kits (Ventana Medical System Inc.) (refer to Table S4). At the end of the experiment, counterstaining was achieved with hematoxylin and bluing reagent (Ventana Medical System Inc.). H&E staining was performed using the Varistain XY model of the Shandon Multi-Program Robotic Slide Stainer (Thermo Fisher Scientific Inc.) and following a standard H&E protocol. Tissue slide sections were scanned with a VS-110 microscope (Olympus, Center Valley, PA, USA) with a 20X objective. The OlyVIA v2.9 software (Olympus) was used for image analysis.

Staining of all antibodies was performed using the Benchmark XT autostainer (Ventana Medical System Inc.). Antigen retrieval was performed with Cell Conditioning 1 (Ventana Medical System Inc., number 950-124) during 30 or 60 minutes for most antibodies although the Cell Conditioning 2 (Ventana Medical System Inc., number 950-124) was used for TP53. Prediluted antibodies were manually added to the slides and incubated at 37°C for 20 to 60 minutes. The following antibodies and dilutions were used: Ki67 (1 : 500; Clone SP6, RM-9106, NeoMarkers), p53 (1 : 200; Clone DO-1, sc-126, Santa Cruz Biotechnologies), ALDH1 (1 : 400, Clone 44/ALDH, 611194, BD transduction lab), CD44 (1 : 100, Clone 2F10, BBA13, RD System), and CD117 (1 : 200, c-kit, Dako). Staining was revealed using the UltraView universal DAB detection kit (Ventana Medical System Inc., 760-500). Counterstaining was achieved with hematoxylin and bluing reagent (Ventana Medical System Inc., number 760-2021 and number 760-2037). Substitution of the primary antibody with phosphate-buffered saline served as a negative control. All sections were scanned using a VS-110 microscope with a 20x 0.75NA objective with a resolution of 0.3225 μm (Olympus). Images were analysed with the OlyVIA software (Olympus).