Increased Disarray of Extracellular Matrix Collagen‐I Fiber Network and Compromised Biomechanics in Aortae From Marfan‐Syndrome Mice Assessed Through Combined Opto‐Biomechatronics

Figures

• 

  FIGURE 1. Workflow of experimental procedure: (A) Schematics of aorta segment preparation. (1) The aorta was explanted and segmented (2) into A. ascendens (asc.) and A. descendens (desc.). Ring segments of 0.5 mm width were sectioned off from asc. and desc. aortas (3). The rings were linearized by one longitudinal cut (4). (B) Reflected light microscopic image of a linearized aorta segment with a width (w) of 0.5 mm suspended in PBS solution. (C) Reflected light microscopic image of a linearized aorta segment that is mounted in the experimental setup with each end affixed to one clamp. (D) The MechaMorph opto‐biomechatronics system is shown with its major functional components labeled. (E) A side‐view sketch of the experimental setup shows an aorta segment mounted in the MechaMorph in PBS solution and the multiphoton microscope dipping objective optically accessing the segment from above. (F) The force/strain‐time graph shows a typical experimental sequence. Strain on the sample is quickly increased step‐wise by the voice coil actuator, and the passive force response of the sample is recorded by the load cell. (G) A representative 3D reconstruction of an image stack of the aorta sample…

• 

  FIGURE 2. Data fitting and mechanical modelling of passive visco‐elastic aortic tissue biomechanics: (A) The force/strain‐time graph of a single strain step in the experimental sequence is shown with the force traces in black and the strain traces in red. The step‐change in strain Δε results in a force response ΔF(t) with an instantaneous force maximum Fmax that decays over time and reaches the plateau force ΔFeq in equilibrium. The force decay is described by fitting a second‐order exponential function (B(i)) (shown in green) to the experimental data of each step‐change of strain Δεj). The mechanical model (B) of the sample consists of an elasticity component with a Young's Modulus E and two viscous components with dynamic viscosities η1 and η2. The equilibrium stress σeq after each strain change j is determined with (B(ii)) using the sample's cross‐sectional area, obtained from the simultaneous microscopy assessment. The Young's modulus of a sample Es is obtained as the slope over all plotted stress‐strain changes in an experimental set on one sample (B(iii)). Each of the viscosities ηi is determined for each strain step‐change j using the time constants ti,j (B(iv)).

• 

  FIGURE 3. Marfan aortae are at least ten times stiffer than wt aortae. (A) The combined stress‐strain diagram for all samples of the Marfan Syndrome (MFS) A. asc. group is shown (each grey scale representing one individual replicate). A linear fit is performed for each sample to determine the mean slope in the linear range (up to e ∼ 1.0) which describes the respective Young's modulus Es of each sample. (B) The Es values from all MFS samples is roughly an order of magnitude higher than the Es values in the wildtype (wt) samples. All differences between sample groups are statistically significant (ANOVA, Tuckey, **: p < 0.001, (m/n) denotes m replicates from n animals). Individual recordings from a female animal are marked in (A).

• 

  FIGURE 4. Marfan aorta strips have significantly higher circumferential dynamic viscosities: (A) and (C) show exemplary viscosity‐strain diagrams of all samples in the MFS‐asc. group. Pearson correlation tests show that neither the higher viscosity η1 nor the lower viscosity η2 are correlated to strain (Pearson p > 0.05). Dynamic viscosities are similar in Marfan asc. and desc. aortas, while ascending aortas are more viscous over descending ones in the wt. Both high (η1, (B)) and low (η2, (D)) viscosities are significantly higher in Marfan aortas over wt aortas (ANOVA on Ranks with Dunn's, **: p < 0.001). (n/m) in (B) and (D) denotes n individual instantaneous stretch episodes from m technical replicates.

• 

  FIGURE 5. Fiber orientation‐strain correlation: (A) A 3D stack was acquired of each sample after each strain change using simultaneous multiphoton microscopy. Exemplary projections of the stacks at different strain levels of an MFS‐asc sample are shown with the backscattered Second Harmonic Generation (bSHG) signal in green and the Autofluorescence (AF) signal in red. Scale bar: 100 µm. (B) The orientations of the structures in the bSHG signal that predominantly originate from fibrillary collagen were determined using the cubic spline method (σ = 2, Orientation J plugin for Image J). (C) The respective orientation deviation Ψ defined as the standard deviation of a Gaussian fit curve over the orientation distribution histogram is shown. (D) The orientation deviation of the collagen fibers exhibits a negative correlation to strain in all samples of all sample groups. The MFS groups exhibit higher orientation deviations (higher disarray of fibers) on all strain levels. (E) Some of the slopes in the Ψ‐ε diagram are significantly different (ANOVA Tukey ** P<0.001, ° P>0.05). The wt‐asc group shows an especially low slope due to already low Ψ (higher order of fibers) at low strains. MFS aortas…

• 

  FIGURE 6. In vivo functional vascular assessment of wildtype (wt) and Marfan (MFS) mouse aorta. Representative imaging and quantitative analysis in live mice using ultrasound and the CODA tail‐cuff system for blood pressure. The data depict dynamic changes in (A, B) ascending aortic diameter, (C) pulse wave velocity (PWV), indicative of arterial stiffness; (D, E) depict systolic and diastolic blood pressure values measured in female and male mice using the CODA tail‐cuff system. The CODA system utilizes volume pressure recording (VPR) technology to noninvasively assess arterial pressure through a dual‐cuff setup placed on the tail. The values for the (F) pressure gradient and the (G) maximum velocity of blood flow in the ascending aorta were determined from ultrasound Doppler examinations throughout the cardiac cycle. The age of the examined animals is given in weeks “W” followed by the number of weeks. The n‐numbers are shown in brackets within the diagrams. The Shapiro‐Wilk test was used to test for normality. The Mann‐Whitney as a non‐parametric test was used for comparing two samples with non‐normal distributions, and the Welch's t‐test for samples with normal distribution assuming…

• 

  FIGURE 7. Carotid artery perfusion at graded perfusion pressures. Perfusion experiments were performed on the isolated carotid artery from female MFS mice (22.4 ± 2.0 weeks) and female littermate control mice (20.9 ± 2.0 weeks), while intraluminal pressure was step‐wise increased and decreased (20 – 180 – 20 mmHg). (A) Changes in vascular diameter, demonstrating pressure‐dependent vascular responses. Note that in Marfan mgR/mgR mice (MFS), progressive vascular stiffness impairs the ability of the carotid artery to accommodate pressure increases and fails to fully return to their baseline diameter after pressure perfusion due to viscoelastic hysteresis and irreversible structural changes in the vessel wall as shown in (C) (wt, littermate control mice). Data are presented as mean ± SEM from n = 8 – 9 experiments (Welch's t‐test). (B) Carotid artery wall thickness at graded perfusion pressures. Wall thickness was measured at a perfusion pressure of 80 mmHg and at a maximal pressure of 180 mmHg. Medial layer destruction led to progressive weakening and focal thinning of the arterial wall in mgR/mgR (MFS) mice but not in littermate control mice (wt). Data are presented as mean ± SEM from n = 5…