The glomerular hydraulic conductivity (permeability to water - L_p) and oncotic reflection co-efficient (permeability to macromolecules - s ) can be determined in isolated normal human glomeruli using an adaptation of a glomerular swelling technique the details and validity of which have been previously described in detail [Am J Physiol 251 (Renal fluid Electrolyte Physiol) 1986; 20: F627-F634 and J AM Soc Nephrol 1992; 3(6): 1260].

In summary, consider the volume of an isolated glomerulus allowed to equilibrate in 5% human serum albumin (HSA)(Point A, Figure 1A). Transfer of the glomerulus to a solution of 3% HSA will result in an oncotic gradient across the glomerular filtration barrier (GFB). The resulting movement of water into the glomerulus will cause the glomerulus to swell (Figure 2A & 2B, FOR VIDEO OF SWELLING see Figure 3). Even if water reaches the mesangium rather than the capillary, it must still have crossed the GFB to get there. After 30 seconds expansion ceases, the glomerular volume = V₁ (Figure 1A).

Replacement of the external solution with one of equal osmotic potential but composed of dextran (unable to cross GBF), means that any subsequent change in glomerular volume is due to "leakage" of albumin across the GBF, ie the glomerulus will shrink slightly (V₂). Glomerular volume changes are videotaped and diameters measured using NIH image off-line (volume = 4/3p r³) and s= V₂/V₁. This method requires the use glomeruli with intact afferent and efferent arterioles which can be blocked as in a classical Landis-Michel technique these glomeruli account for approximately 1% of sieved glomeruli. This method has been used primarily to determine the influence of nephrotic plasma on permeability to protein. However, it can be adapted to measure L_p rather than just glomerular filtration co-efficients.

Analysis of glomerular volume change with time produces a volume expansion curve (Figure 1A). The initial rate of volume change depends on the oncotic pressure gradient ie dV/dtµDp . The factors controlling movement of fluid across the GFB are defined by Starlingís equation: J_v/A=L_p[(P_c-P_i)-s (p_c-p_I)], in which J_v/A= the rate of trans-endothelial fluid flux/unit area, P_c and P_i and p_{c and pI} are the capillary and interstitial hydrostatic and oncotic pressures respectively. In this model P_c and P_i are zero, Jv=dV/dt. Therefore dV/dt=L_pA[-s (p_c-p_I)] or dV/dt= L_pAs [-Dp ]. A plot, therefore of dV/dt against -Dp will be a straight line graph with a gradient of L_pAs . L_p can be derived because s is known from previous experiment ie V₂/V₁. A is determined from known glomerular volume occupied by capillary-23% [Diabetes 1992; 41(9): 1106-12] and average capillary width on confocal microscopy analysis at adjacent MRC Imaging Centre after FITC-labelled endothelial specific lectin (B. simplificolia). Glomerular swelling is conducted in a unique "glomerular channel" attached to an in vivo microvascular examination rig constructed and specially designed to maximise the rate of fluid exchange without altering the orientation of the glomerulus. Within the chamber the glomerulus in which the glomerulus is held in a 30-50m m holding pipette in a flow of fluid that can be changed "upstream".
Back to MVRL