Characterizing nuclear morphology and expression of eNOS in vascular endothelial cells subjected to a continuous range of wall shear stress magnitudes and directionality

Abstract

Complex patterns of hemodynamic wall shear stress occur in regions of arterial branching and curvature. Areas within these regions can be highly susceptible to atherosclerosis. Although many studies have characterized the response of vascular endothelial cells to shear stress in a categorical manner, our study herein addresses the need of characterizing endothelial behaviors over a continuous range of shear stress conditions that reflect the extensive variations seen in the vasculature. We evaluated the response of human umbilical vein endothelial cell monolayers to orbital flow at 120, 250, and 350 revolutions per minute (RPM) for 24 and 72 h. The orbital shaker model uniquely provides a continuous range of shear stress conditions from low and multidirectional at the center of each well of a culture plate to high and unidirectional at the periphery. We found distinct patterns of endothelial nuclear area, nuclear major and minor diameters, nuclear aspect ratio, and expression of endothelial nitric oxide synthase over this range of shear conditions and relationships were fit with linear and, where appropriate, power functions. Nuclear area was particularly sensitive with increases in the low and multidirectional WSS region that incrementally decreased as WSS became higher in magnitude and more unidirectional over the radius of the cell layers. The patterns of all endothelial behaviors exhibited high correlations (positive and negative) with metrics of shear stress magnitude and directionality that have been shown to strongly associate with atherosclerosis. Our findings demonstrate the exquisite sensitivity of these endothelial behaviors to incremental changes in shear stress magnitude and directionality, and provide critical quantitation of these relationships for predicting the susceptibility of an arterial segment to diseases such as atherosclerosis, particularly within complex flow environments in the vasculature such as around bifurcations.

Keywords: Endothelial nitric oxide synthase; Endothelium; Hemodynamics; Kruppel-like factor 2; Mechanical dose response; Mechanobiology; Nuclear area; Nuclear aspect ratio.

Figure 1.

Custom 3-D printed mask used to separate the center and periphery of each well of a culture plate for extraction of RNA. (A) The entire mask attached (via screws at the midpoint) to a baseplate that secured it to a 6-well culture plate. (B) Top view of an example well with dye-colored water to illustrate the efficacy of the mask in separating the solution in the central portion (orange) from the outer portion (green). (C) Underside of one well (and partial of another) showing the separation of the two regions via a silicon gasket (arrows).

Figure 2.

CFD simulations of swirling medium and associated shear stress metrics in a culture plate well under orbital flow. (A) Simulations at 120, 250, and 350 RPM where color map and scale bar indicate fluid height. (B-E) Shear metrics calculated from the center to periphery of the well in each CFD model at each angular velocity, including (B) TAWSS, (C) MagMeanWSS, (D) OSI, and (E) TransWSS.

Figure 3.

Endothelial nuclear morphology under orbital flow as a function of radial position in the well. (A) Confocal tile scans of DAPI from the center to the periphery of a representative cell layer after orbital flow at 250 and 0 (static control) RPM for 72 h with representative high magnification images at selected locations. (B-C) Plots of nuclear area relative to the mean of static controls over the radius of each well from center to periphery at (B) 24 h and (C) 72 h. (D-E) Plots of nuclear aspect ratio over the radius of each well from center to periphery at (D) 24 h and (E) 72 h. Data from controls (0 RPM) are shown at the exact well position, whereas other data are slightly staggered to improve clarity. The n refers to the number of cell layers assessed for each orbital velocity (in RPM). Data are mean ± SD. *Indicates statistically significant difference (p<0.05) compared to static controls.

Figure 4.

Endothelial nuclear major and minor diameters under orbital flow as a function of radial position in the well. (A-B) Plots of nuclear major diameter relative to the mean of static controls over the radius of each well from center to periphery at (A) 24 h and (B) 72 h. (C-D) Plots of nuclear minor diameter over the radius of each well from center to periphery at (C) 24 h and (D) 72 h. Data from controls (0 RPM) are shown at the exact well position, whereas other data are slightly staggered to improve clarity. The n refers to the number of cell layers assessed for each orbital velocity (in RPM). Data are mean ± SD. *Indicates a statistically significant difference (p<0.05) compared to static controls.

Figure 5.

Endothelial nuclear area as a function of the WSS metrics. (A-H) Plots of nuclear area (relative to the mean of static controls) versus each WSS metric, including (A-B) TAWSS, (C-D) MagMeanWSS, (E-F) OSI, and (G-H) TransWSS at (column 1) 24 h and (column 2) 72 h. Data points were averaged over all cell layers (indicated by n) for each orbital velocity (120, 250, and 350 RPM) at each duration and radial position in the well (multiple data points for an orbital velocity represent the multiple radial locations assessed for the cell layers; TransWSS was only assessed at 250 RPM due to inconsistencies in the data between orbital velocities). The black line represents a best-fit line (solid) and, in some cases, power (dashed) function across all data points to visualize the trend. The Spearman’s correlation coefficient (ρ) and associated p-value are given for each plot. *p<0.05 is considered statistically significant.

Figure 6.

Endothelial nuclear aspect ratio as a function of the WSS metrics. (A-H) Plots of nuclear aspect ratio versus each WSS metric, including (A-B) TAWSS, (C-D) MagMeanWSS, (E-F) OSI, and (G-H) TransWSS at (column 1) 24 h and (column 2) 72 h. Data points were averaged over all cell layers (indicated by n) for each orbital velocity (250 and 350 RPM) at each duration and radial position in the well (multiple data points for an orbital velocity represent the multiple radial locations assessed for the cell layers; TransWSS was only assessed at 250 RPM due to inconsistencies in the data between orbital velocities). The black line represents a linear regression of all data points to visualize the trend. The Spearman’s correlation coefficient (ρ) and associated p-value are given for each plot. *p<0.05 is considered statistically significant.

Figure 7.

Endothelial expression of eNOS under orbital flow as a function of radial position in the well. (A) Confocal tile scans of eNOS from the center to the periphery of a representative cell layer after orbital flow at 250 and 0 (static control) RPM for 72 h with representative high magnification images at selected locations. (B-C) Plots of eNOS relative to the mean of static controls over the radius of each well from center to periphery at (B) 24 h and (C) 72 h. Data from controls (0 RPM) are shown at the exact well position, whereas other data are slightly staggered to improve clarity. The n refers to the number of cell layers assessed for each orbital velocity (in RPM). Data are mean ± SD. *Indicates statistically significant difference (p<0.05) compared to static controls.

Figure 8.

Endothelial eNOS expression as a function of the WSS metrics. (A-H) Plots of eNOS (relative to the mean of static controls) versus each WSS metric, including (A-B) TAWSS, (C-D) MagMeanWSS, (E-F) OSI, and (G-H) TransWSS at (column 1) 24 h and (column 2) 72 h. Data points were averaged over all cell layers (indicated by n) for each orbital velocity (120, 250, and 350 RPM) at each duration and radial position in the well (multiple data points for an orbital velocity represent the multiple radial locations assessed for the cell layers; TransWSS was only assessed at 250 RPM due to inconsistencies in the data between orbital velocities). The black line represents a best-fit line (solid) and, in some cases, power (dashed) function across all data points to visualize the trend. The Spearman’s correlation coefficient (ρ) and associated p-value are given for each plot. *p<0.05 is considered statistically significant.

Figure 9.

Expression of atheroprotective and atherogenic genes at the center and periphery of endothelial cell layers under orbital flow as assessed by RT-qPCR. (A) Shear metric patterns at 250 (solid line) and 120 (dotted line) RPM in the center and periphery regions of each well where RNA was extracted. Gene expression at the center (grey bars) and periphery (white bars) of wells under orbital flow for (B) 24 h at 250 RPM, (C) 72 h at 250 RPM, (D) 24 h at 120 RPM, and (E) 72 h at 120 RPM. Data are mean ± SD and averaged over n=5 cell layers per condition. *p<0.05 is considered statistically significant. Note, the y-axis is scaled differently for the atheroprotective and atherogenic genes to better show differences between center and periphery for all genes.