Microlever

Post-glycated 3D collagen matrices were elaborated as explained above using the same conditions of collagen fibers configuration (confocal reflectance imaging). The ribose-induced mechanical alterations in the collagen gels were examined on the 7^th, 14^th and 21^st days by assessing their Young’s elastic modulus, E, which is indicative of gel’s resistance to deformation. E was measured using a stand-alone AFM (Bioscope, Veeco, USA) coupled to an inverted optical microscope as previously described [33 (link)]. Briefly, the AFM measurements were taken using low spring constant cantilevers with pyramidal tips (nominal k = 0.01 N/m) (Microlever, Veeco), which were calibrated using the thermal noise method [34 (link)]. For each gel location, three force vs. piezo displacement curves (F-z curves) were acquired using a moderate loading force (~1 nN). For each gel location, E and the extent of gel indentation (d) were computed as the averages of the values obtained by least-squares fitting of a contact elastic model to each set of three F-z curves, as reported elsewhere [35 (link)]. The same protocol was applied on at least 9 random gel locations. For each treatment, the E data were normalized using the corresponding value obtained from 0 mM ribose gel at the 7^th day.

The Young’s elastic modulus E of 3D collagen-I gels and 2D PAA gels was assessed with a home-made stand-alone AFM adapted to an inverted optical microscope as described in detail elsewhere [27 (link),36 (link)]. In brief, AFM nanoindentation measurements were performed with low spring constant cantilevers with pyramidal tips (nominal k = 0.03 N/m) (Microlever, Veeco, Santa Barbara, CA, USA), which were calibrated using the thermal noise method [36 (link)]. For each gel location, three force-displacement curves (F-z curves) were acquired at a moderate loading force (~1 nN); E of each F-z curve was computed by least-squares fitting of a contact elastic model, and averaged over the three F-z curves [27 (link)]. The same protocol was applied on at least 9 random gel locations to elicit the final E estimate of the hydrogel.

The cell puncturing experiments were conducted in a JPK NanoWizard^® II AFM (JPK Instrumental). A probe with a flat ended silicon nitride tip (Microlevers, Veeco) of diameter of 1, 2 or 3 μm, and a cantilever spring constant of 0.035 N/m, was used. The AFM tip was first made to indent into the cell at a speed of 0.2 μm/s until an indentation depth of 0.4 μm was reached. This was followed by holding for 30 s, and then retraction at a speed of 0.1 μm/s. Size evolution of the hole in the plasma membrane was captured by time-lapse images, taken at a rate of 10 frames/sec, and then quantitatively analyzed by a MATLAB program (see Supplementary Information A and Supplementary Figure 1 for details). All tests were conducted at 25 °C and within 1 hour after the cells were removed from the incubator.