The effects of strain history on aortic valve interstitial cell activation in a 3D hydrogel environment

FIG. 1.. Multiscale nature of TAV and BAV function. (a) and (b) Renderings of a (a) TAV and (b) BAV during diastole. The Raphe is the area where fusion of two leaflets has occurred in a BAV. Circumferential (C) and radial (R) directions are depicted. (c) and (d) Mean circumferential stretch fields and radial stretch fields for (c) left (L) TAV leaflets and (d) right–left (R–L) fused BAV leaflets, as calculated from human in vivo echocardiogram measurements in Ref. 2, depict that the peak circumferential stretch field is largely reduced and the peak radial stretch field is slightly increased in BAVs. (e) and (f) Overview of data from time vs average stretch plots of (e) porcine TAVs in ex vivo flow loops,1 and (f) computationally simulated BAVs5 derived from data collected in Ref. 1 indicate that in BAVs, there is a large reduction in peak circumferential stretch and root mean squared value of the waveform, slight increase in radial stretch and root mean squared value of the waveform, and a decrease in both circumferential and radial rate of stretch. (g) and (h) Histology of cross sections of bicuspid aortic valve leaflets depicting that (g) highly ordered layered ECM patterning in the…

FIG. 2.. Experimental setup on the 3D high-throughput biaxial oscillatory stretch system (3D HT-BOSS). (a) and (b) The HT-BOSS was implemented to apply oscillatory equi-biaxial stretch up to six 96-well plates simultaneously by utilizing a linear motor that pushes pistons into the bottom of a specialized well plate made with a flexible silicone membrane bottom, which was functionalized to hold PEG gels with AVICs imbedded in them. (c) The 1 Hz waveforms applied by the HT-BOSS, with the physiologic wave moving to the peak stretch of 1.1 as quickly as the motor could do so, the peak matching wave moving to 1.1 within 0.5 s, and the rate matching waveform maintaining the same slope as the physiologic wave and the same RMS value of the peak matching wave, making the peak stretch 1.062.

FIG. 3.. AVIC nuclear evaluation within PEG gels. (a) Overlay of the closed Otsu binarized mask of a representative AVIC nucleus (blue) and its corresponding best fit ellipsoid (red). Overlap of nuclear mask and nuclear best fit ellipsoid depicted in magenta. Yellow point in each frame is the same point in the x-, y-, and z-coordinate system. (b) Schematic of the x-, y-, and z-coordinate system implemented within the gel and the hemi-axial lengths hI, hII, and hIII of the best fit ellipsoid of each nucleus. The ratio of the longest to shortest hemi-axial lengths (hI/hIII) and the mid-length to shortest hemi-axial lengths (hII/hIII) of the best fit ellipsoid were determined to characterize the shape of the AVIC nucleus. Ratios of hI/hIII and hII/hIII at each deformation level were calculated to determine changes in 3D nuclear shape upon deformation. (c) Schematic depicting calculations made to determine the theoretical nuclear ellipsoid reference state hemi-axial lengths, ||Hi||, and the theoretical nuclear ellipsoid hemi-axial lengths at each deformation level, Riλ, from the measured hemi-axial values, hi.

FIG. 4.. 3D VMN and αSMA polymerization in cyclically stretched AVICs. (a)–(c) Maximum intensity flattened projections in the x–y plane and x–z plane for (a) VMN (red), (b) αSMA (green), and (c) the composite for a representative deconvolved image of an AVIC immunostained within a 3D PEG hydrogel. (d) Single slice and orthogonal single slice magnified image of nuclear region of the same cell; the point marked at the center of the crosshatch is the same point in 3D space in all fields of view.

FIG. 5.. Waveform vs AVIC activation. (a) αSMA and (b) VMN intensity per cell volume in gels conditioned for 48 h with designated waveforms and fixed in the reference state. (c) SMAD 2/3 nuclear intensity divided by cell intensity from cells conditioned with indicated waveforms. (D) Primary nuclear aspect ratio measured as the longest hemi-axis divided by the shortest hemi-axis of the nuclear best fit ellipsoid from cells conditioned with indicated waveforms and fixed in the reference state.

FIG. 6.. 3D SMAD 2/3 localization. (a)–(d) Representative deconvolved image of an AVIC that was immunostained within the fixed 3D PEG hydrogel. Orthogonal single slices in the x–y, x–z, and y–z planes are depicted for (a) cell membrane (red), (b) nucleus (blue), (c) SMAD 2/3 (green), and (d) the composite; the point marked at the center of the crosshatch is the same point in 3D space in all fields of view. Nuclear localization of SMAD 2/3 is indicative of its activation.

FIG. 7.. Strong correlations between VMN polymerization, SMAD 2/3 activity, and nuclear shape. (a) Average nuclear eccentricity correlated with average VMN intensity per cell volume (y = −516.2× + 2201, and R2 = 0.8336) and average SMAD 2/3 activity (y = −0.1340x+0.4982, and R2 = 1.000) from AVICs conditioned for 48 h of stretch waveforms. (b) Average SMAD 2/3 nuclear localization correlated with average VMN intensity per cell volume from AVICs conditioned for 48 h of stretch waveforms (y = 3843x + 282.7, and R2 = 0.8294). Dotted lines represent 95% confidence intervals. These trends suggest that a feedback mechanism may exist whereby mechanical stretch inputs affect a system where SMAD 2/3 activity is held in check by VMN polymerization and nuclear shape.