Detachment and flow adhesion assays using shear stress

Detachment assays were performed using a temperature-controlled parallel-plate flow chamber (FCS2, Bioptechs) at 37°C. Laminar shear stress was determined by using an equation of shear stress, 6Qμ/wh² (dyn/cm²), in which symbols denote the following: Q, flow rate (mL/min); m, viscosity (0.007 P for buffer); w, chamber width (1.2 cm); h, chamber height (0.1 mm).⁶¹ For coating with ligands, capture antibodies were coated on the center of polystyrene disks, then washed and overlaid with recombinant mouse PNAd⁺CD34-Fc, ICAM1-Fc, or MAdCAM1-Fc. For coating with both PNAd⁺CD34-Fc and ICAM1-Fc, PNAd⁺CD34-Fc was coated before ICAM1-Fc. For some experiments, CCL21 (200 nM in PBS) was also coated at room temperature for 30 min after blocking. Purified T cells (2 × 10⁶/mL) were suspended in a warm medium (IMDM containing 1% FBS) and loaded into the chamber, and bright-field images were recorded every 5 s with a 4× objective lens and CCD camera (HAMAMATSU C2741). After cells were settled on the coated substrates for 5 min, shear stress was applied for 1 min at 2 dyn/cm² with a stepwise increment of 1 dyn/cm² every minute to 5 dyn/cm² using a programmed syringe pump (FP-1000, MELQUEST). The attached cells were counted at the end of each shear stress by using an Integrated Morphometry Analysis program of Metamorph (Molecular Devices). Adhesion efficiencies were percentages of attached cell numbers relative to input cell numbers.

For flow adhesion assays, the ligands’ coating was essentially the same as that in detachment assays, with modifications. The upstream part of the disk was masked by a silicon rubber for measurement of cell influx in the flow. Influx and captured cells (tether/rolling/arrest) were counted in the influx area (0.78 mm²) and the capture area (1.05 mm²), respectively, during the same 1000 frames of the imaging period (33.3 s), and normalized for time and area (cell numbers/sec/mm²). Cell accumulation index, a relative degree of cell accumulation in the capture area relative to that in the influx area were calculated by dividing normalized captured cell numbers by normalized influx cell numbers. For tracking of adhesive events of T cells on the capture area, we employed a plugin program TrackMate (ver.5.0.2) for a Fiji program with parameters adjusted for displacement rates of cells below those of flowing cells. Position data (x, y, t) of trajectories were further analyzed with MATLAB (MathWorks). Moving and stopping of cells were defined with threshold displacement rates of 1 μm/frame and 0.5 μm/frame in 100 consecutive frames on PNAd⁺CD34/ICAM1 and MAdCAM1, respectively, to distinguish arrest from rolling. Adhesive events longer than 1 s were analyzed and categorized as follows: tether (1–3.3 s stopping), arrest (>3.3 s stopping), and rolling (moving >1 s). Only the first trajectory of the same cell was analyzed. Capture events (tether, arrest, or rolling) were normalized with time, area, and cell influx (cell numbers/second/mm²). Relative capture frequencies were defined as normalized capture events relative to those of the wild-type control. Average rolling velocities were calculated using displacement and duration longer than 1 s.