BSA Permeability
Update 8/15/14: Replaced the plots with updated versions including more data for NPN (now n = 3) and cellulose (n = 4,) as well as new high-flux polyethersulfone membrane data (n = 3.) I also dropped one clearly erroneous datapoint from the NPN dataset to dramatically cut down the standard deviation.
—
This is a followup to my previous post, found here.
I’ve tested the permeability of our NPN membranes as well as cellulose triacetate for BSA. BSA is much larger than cytochrome C, so by extrapolation from the previous data I expected NPN to reduce permeability for BSA relative to free diffusion more than it did for the smaller molecules. This is exactly what I observed:
Since this data was gathered in order to get a feel for how our material will behave during hemodialysis, it’s unfortunate that we’re passing BSA because unlike beta-2-microglobulin (same size as cytochrome C) and urea, it’s preferable for BSA to be retained by the blood instead of passed into the dialyzate. However, the dramatic reduction in BSA permeability relative to cytochrome C is promising — smaller pore sizes could restrict BSA considerably better than these 40-48 nm pores (BSA is about 14x4x4 nm in size) while retaining most of the material’s permeability to smaller molecules.
Arguably more interesting than the NPN result, however, are the cellulose triacetate and high-flux results. While the ratio of NPN’s permeability has always decreased relative to free diffusion for larger solutes, the same ratios for cellulose and high-flux membranes here are roughly equal to those for cytochrome C.
I suspect that this is due to the relatively simple, or “ideal” nature of NPN as a filter, compared to the tortuous and thick cellulose and high-flux membranes, although I’d want to look into the theory more closely before making that claim outright. In any event, however, this points towards sharper (or at least more linear) MW cutoffs for NPN as compared to traditional membranes, a feature desirable in most any filter application.