A1 upright confocal microscope

Frozen sections were fixed in 2% paraformaldehyde/PBS pH 7.4 and blocked with 10% goat serum, 2% BSA, 0.02% fish skin gelatin and 0.05% TritonX100 (Sigma) in PBS for 1 h at room temperature. Paraffin sections of human OSCC were subjected to heat-mediated antigen retrieval (citrate buffer, pH6) prior to blocking. Primary antibodies were incubated overnight at 4 °C, followed by 1 h incubation at room temperature in secondary antibody.
The following primary antibodies were used: Foxp3 (eBioscience, clone FJK-16s, 1/100, and Abcam, clone 236 A/E7, 1/50), anti-Loricrin, anti-Krt76 (Santa Cruz, clone F-12, 1/100, and Sigma, HPA019656, 1/100), Krt14 (Covance, PRB-155P, 1/1000), B220/CD45R (eBioscience, clone RA3-6B2, 1/100), CD3 (BD Pharmingen, clone 17A2, 1/150), CD45 (BD Pharmingen clone 30-F11, 1/150); and secondary antibodies: anti-goat, anti-mouse and anti-rabbit Alexa Fluor 488, 568 and 633 (Life Technologies, 1/300).
EdU staining was performed with a Click-it EdU imaging kit (Life Technologies) according to the manufacturer’s recommendations. DAPI (Life Technologies) was used as a nuclear counterstain. Slides were mounted using ProLong Gold anti-fade reagent (Life Technologies). Images were acquired with a Nikon A1 Upright Confocal microscope. Images were analysed using ICY image analysis software^{55 (link)}.

Pieces of tongue and stomach epithelia were collected and sectioned in a sterilised setup. Sections were fixed in 1:1 acetone-to-methanol and incubated with PNA FISH probes at 55 °C. A universal bacteria probe (BacUni; AdvanDx) and probe for C. dubuliniensis (AdvanDx) were used^{25 (link)}. DAPI was used for nuclear counterstaining. Slides were mounted using ProLong Gold anti-fade reagent (Life Technologies). Images and Z-stacks were acquired with a Nikon A1 Upright Confocal microscope. Evaluation of bacterial location was performed on three-dimensional z-stacks using ICY software.

Tissue was collected at the indicated time points, fixed with 4% paraformaldehyde (PFA) for 20 min at room temperature and frozen embedded in optimal cutting temperature (OCT). Sections of 16 μm thickness or horizontal whole mounts of 60 μm were prepared. Sections of 16 μm thickness were further fixed with 4% PFA for 2 min, followed by PBS washes, blocking buffer and antibody incubation (Supplementary Table 1). Horizontal whole mounts were stained as described previously (Driskell et al., 2013 (link)) with antibodies listed in Supplementary Table 1. Briefly, staining was performed in 1.5 ml Eppendorf tubes with primary incubation overnight at 4°C and secondary incubation with DAPI at room temperature for 1 h. Samples were mounted using a small drop of 100% glycerol (Sigma). All microscopy was performed on a Nikon A1 upright confocal microscope and images were analyzed in Image J. Images displayed from horizontal whole mounts are maximum projection images generated from Z stacks in order to fully capture intact vascular networks. A minimum of 3 sections per mouse were analyzed.

LL2 tumor tissue were cryosectioned (Leica, Buffalo Grove, IL, USA) at 10 µm thickness onto Truebond glass slides VWR, Radnor, PA, USA). Tissue slides were stained with DAPI (1 mg/mL) for 10 min and then imaged on a Nikon A1 upright confocal microscope using a 20×/0.75 ∞/0.17 WD 1.0 air objective. The acquisition parameters were pixel dwell of 1.2 µs, aperture size 1.2 au, at 405 nm laser line power of 5% and gain of 80 was used, at 640 nm laser line power of 30% and gain of 100 was used (values variable depending on power of laser light source). The Nyquist feature was used to obtain images at optimal resolution resulting in an image resolution of 0.2 µm/px for 2 k sized 3 by 3 tile.

Photomicrographs were taken using a Leica DM IL LED Tissue culture microscope (Wetzlar, Germany). Confocal microscopy was performed with a Nikon A1 Upright Confocal microscope (Tokyo, Japan) using 10× or 20× objectives. Imaging of hematoxylin and eosin-stained sections was performed using a Hamamatsu NanoZoomer slide scanner (Hamamatsu City, Japan). Image processing was performed with Nikon Elements, Image J (Fiji) (National Institutes of Health, Bethesda, MD, Photoshop CS6 (Adobe, San Jose, CA), and Icy software (Quantitative Image Analysis Unit, Institut Pasteur, Paris, France).

Cells seeded on PDMS substrates were fixed in 4%PFA for 15 min followed by three washes in PBS. Permeabilisation and blocking (PB) buffer (0.5% skimmed milk powder, 0.25% fish skin gelatin, 0.5% TritonX100, 1× HEPES-buffered saline) was added and incubated at room temperature for 1 h. Substrates were incubated in primary antibodies for 1 h at room temperature. Samples were washed three times in PBS and incubated with secondary antibodies at room temperature for 1 h. Samples were washed three times in PBS and then mounted on glass slides in Prolong antifade mounting reagent (Life technologies) containing DAPI. Samples were imaged using a Nikon A1 upright confocal microscope with a 20× objective. Images are presented as maximum intensity projections.
Primary antibodies used were as follows: anti-β1 integrin mouse monoclonal (clone P5D2; 1:500), anti-involucrin rabbit polyclonal (DH1; 1:200), anti-E-Cadherin mouse monoclonal (clone HECD1; 1:200), anti-Desmoglein 3 mouse monoclonal (Abcam; 1:25), anti-MAL rabbit polyclonal (Abcam; 1:200), anti-Ki67 rabbit polyclonal (Abcam; 1:200), and anti-YAP mouse monoclonal (Santa Cruz; 1:500). Alexa Fluor™ 488 Phalloidin (ThermoFisher Scientific; 1:1000) was also used. Secondary antibodies of the appropriate species were conjugated to Alexa Fluor™ 488, 555 and 647 (ThermoFisher Scientific).