Date of Award
8-1988
Document Type
Dissertation - Restricted
Degree Name
Doctor of Philosophy (PhD)
Department
Biomedical Engineering
First Advisor
Joel Myklebust
Second Advisor
William Barber
Third Advisor
James Heinen
Fourth Advisor
Russel Niederjohn
Fifth Advisor
Allen Cowley
Abstract
Starling first described the important role of plasma proteins in determining fluid distribution between vascular and extravascular spaces. Recently evidence has become available to suggest that the distribution of plasma proteins in perivascular spaces is not uniform as has always been assumed. We developed, validated, and applied a new technique of two dimensional microspectrophotometry to obtain the spatial distribution of water, plasma proteins, collagen in the vascular and avascular regions of the extracellular matrix of rat mesentery tissue. Multispectral (280 nm, 320 nm, 705 nm, 800 nm 1500 nm), 360 um x 360 um images of selected tissue regions were digitized into 512 x 512 pixel images of 8 bits resolution. Estimation of the nonspecific absorption and wavelength dependent scattering from 320 nm, 705 nm images allowed calculation of the tissue protein and tissue collagen distribution images from the 280 nm absorbance image. Water distribution images were calculated in a similar manner from 800 nm, 1500 nm absorbance images. Protein concentration averages (1.88 $\pm$ SD 0.8 g/dl) calculated with a scattering correction are not significantly different (p $<$ 0.2) than direct measurement by electrophoresis (1.8 $\pm$ SD 0.2 g/dl). The 320 nm correction of many single beam microspectrophotometry studies yields a value (4.29 g/dl) which is significantly (p $<$ 0.001) too high. The scattering corrected images show clear detail, apparent protein aggregation, straight fiber like collagen absorption. Two way ANOVA analysis (16 x 16 sections) of avascular tissue show significant spatial variation (P $>$.999). Fourier analysis shows more collagen-protein organization near blood vessels. Venule radial protein gradients are significantly longer than arterioles (t = 4.84, p $<$ 0.005) but not exchange vessels (t = 1.32, p $<$ 0.4). Venule perivascular protein (3.73 $\pm$ 1.56 g/dl) concentration is significantly higher than for exchange vessels (2.39 $\pm$ 0.26 g/dl, t = 2.23, p $<$ 0.1) or arterioles 2.17 $\pm$ 0.19 g/dl, t = 2.6, p $<$ 0.05). Diffusion limited aggregation modeling yields protein distributions similar to that found in perivascular tissue. Tunnels (3 to 4 g/dl), with slowly decaying radial gradients of protein suggesting an absence of walls, have been found near some exchange vessels. Injection of saline boluses indicates that the tunnels are also a preferential water path. Thus there appears to be a great deal of organization and structure in tissue regions assumed to be homogeneous.