0.20 μm filter

Four sterilization methods were tested to obtain a sterile bioink: filtration, autoclaving, UV exposure, and pasteurization. The first method was filtration, based on the insertion of the SA and GEL components dissolved in sterile PBS in 20 mL syringes and then passed into a 0.20 μm filter (Corning, Corning, NY, USA). The second method was UV exposure, where the two powders, dissolved in sterile PBS, were placed under UV rays for 1 h. The third method tested was pasteurization where the SA and GEL solutions, placed in 50 mL centrifuge tube, were placed in a 72 °C bath for 1 h. The last method tested was autoclaving: the two components dissolved in sterile PBS were collected in 50 mL centrifuge tube and autoclaved at 121 °C for 20 min.

Twenty mg of plant extract was placed into a test tube. D₂O (pH 7.0) containing 0.1% (w/w) TSP was added and vortexed for 1 min at room temperature. After 20 min ultrasonication at room temperature, the mixture was passed through a 0.20 μm filter (Corning, USA). The supernatant was transferred into a 5 mm NMR tube. Proton NMR spectra were acquired using a 400 MHz NMR spectrometer (Bruker, USA). Chemical shift referencing, baseline correction and phasing were performed. Next, the NMR spectral data were processed using MestReNova (Mestrelab Research, USA) software to adjust the peak alignment, normalization and scaling. Multivariate statistical analysis was used to identify differences among samples. PCA and O-PLS analyses were performed using SIMCA-P+ version 12 (Umetrics Inc., Sweden). The data were mean-centered and scaled to Pareto. To confirm the assignment of correlated resonances, statistical total correlation spectroscopy (STOCSY) was employed [31 (link)]. Moreover, the resonances of interest were searched against online metabolite databases such as biological magnetic resonance data bank (BMRD) and human metabolome database (HMDB) [32 (link)]. To further confirm the metabolite assignment, two-dimensional (2-D) NMR experiments, including correlation spectroscopy (COSY) and heteronuclear multiple bond coherence (HMBC), were performed.

Each replicate was weighed out and ground thoroughly in methanol:chloroform:water (1:1:0.7, v/v/v). The volume of the extraction solvent was adjusted according to weight of the sample (1 mL of solvent / 100 mg of sample). The upper aqueous phase was collected after centrifugation at 1000 g at 4 °C for 15 min and further evaporated using a speed vacuum concentrator (Labconco, MO, USA) at 40 °C until dry. The crude extracts were stored at − 80 °C prior to analysis. A total of 600 μl buffer containing 100 mM sodium phosphate, pH 7.4 in D₂O, 0.1 mM 3-trimethysilypropionic acid (TSP) (Cambridge Isotype Laboratories, Tewksbury, MA, USA) as a chemical shift reference (δ¹H = 0 ppm) and optionally 0.2% NaN₃ was added to dissolve the sample. Mixture was sonicated using ultrasonicator (JeKen, China) for 10 min and filtered through 0.20 μm filter (Corning, USA) before centrifugation at 12,000 g at 4 °C for 5 min. An equal amount of 30 μl was aliquoted from all samples and pooled for the QC. Then, a total of 550 μl of supernatant was transferred into NMR tube for metabolic profiling. Proton NMR spectra were acquired using a 400 MHz NMR spectrometer (Bruker, USA) with CryoProbe Prodigy and Carr−Purcell−Meiboom−Gill (CPMG) pulse sequence [RD−90˚−(τ−180˚−τ)n−acquisition] was applied to analyse the samples at 310 K in 64 scans.

Human saliva was collected from healthy volunteers under a Houston Methodist Research Institute Institutional Review Board human subjects protocol and processed as described previously (16 (link), 17 (link)). Briefly, saliva was collected from five healthy donors, pooled, clarified by centrifugation at 45,000 × g for 15 min, and sterilized with a 0.20-μm filter (Corning Inc.). The resulting sterile saliva was used for subsequent transposon mutant library screening and individual strain growth.