Fiber Matrix Descriptors from Permeability Data Without Requiring Membrane Thickness: Theory, Results, and Optimization

Abstract

Objective: Permeability of basement membrane and all other barriers contains a term for membrane thickness (Δx). This naturally leads to development of methods for measuring Δx that are imprecise, inaccurate, expensive, subject to preparation artifact, and inattentive to variability. Although height and shape of permeability (P) vs. probe radius (α) curves are sensitive to Δx, ln(P) or ln(P/free diffusivity or D_o) curves have shapes independent of Δx. It should, thus, be possible using such characteristics to determine fiber radius (r_f) and void volume ratio (ϵ) without Δx. We developed such a method to derive membrane structure by the standard model of Ogston and present its experimental evaluation.

Methods: Basement membranes were self-assembled using 1: 1 Matrigel: 0.01 M Tris/150 mM NaCl/1.0 mM CaCl₂ buffer on 0.4-μ polycarbonate supports with transport measured in diffusion chambers using FITC-labeled hydroxyethyl starch probes from 25 to 102 Å in radius. Sampling was at 0.5 hr and then for each hour up to 5. Other membranes were measured 7 days after formation.

Results: The best fit of the new technique occurred at 3 hr with R² = 0.949 ± 0.003 SEM, r_f = 36.8 ± 2.4 Å, and ϵ = 0.87 ± 0.02. Membranes studied for 7 days showed more variability but essentially the same characteristics.

Conclusions: Membrane thickness is not necessary to reduce permeability of basement membrane to structure, and optimum sampling time is 3 hr.

Key Words:

• Fiber radius
• void volume
• model fit
• computational methods
• Ogston formulation.