Taking advantage of the fact that membranes can be designed to recognize, isolate, and transport a desired molecule out of a mixture, researchers at the University of Florida, Gainesville, have carried out chiral separations using an antibody-based nanotube membrane.

Chemistry professor Charles R. Martin and coworkers Sang Bok Lee, David T. Mitchell, and Lacramioara Trofin prepare the antibody-based membranes from alumina films that have cylindrical pores with nanoscopic diameters. They grow silica nanotubes within the pores by sol-gel chemistry. Then they install aldehyde silanes in the inner surfaces of the tubes. Finally, the researchers attach antibody fragments by reacting the aldehyde groups with the free amino sites of the protein.

Using membranes prepared in this way, they successfully separated enantiomers of finrozole [Science, 296, 2198 (2002)]. The compound is being developed by Hormos Medical Corp., Turku, Finland, to treat lower urinary tract symptoms related to hormonal imbalance.

"Exciting" is how Reginald M. Penner, a nanotechnology and materials chemistry expert at the University of California, Irvine, describes the Florida work. "The chiral resolution capability of the membrane system has the potential to change the way drugs are manufactured," he says.

Christopher J. Welch, a process research fellow at Merck who runs a group that develops rapid separations of enantiomers, shares Penner's excitement. He adds that the Florida work "is a nice example of how materials science and nanotechnology together can create something with a lot of promise for actual separations."

Finrozole has two chiral centers, making four stereoisomers possible: RR, SS, RS, and SR. A racemate consisting of the RS and SR enantiomers is in Phase II clinical trials.

	ALL SMILES Martin (right) tests membranes in a U-tube like that held by Trofin. PHOTO BY RAHELA GASPARAC
	Because in principle antibodies can be raised against any molecule, this approach could lead to "a broad class of smart membranes for nearly any type of separation." Charles R. Martin, chemistry professor at the University of Florida, Gainesville

Martin's other coauthors--molecular biologists Tarja K. Nevanen and Hans Söderlund at VTT Biotechnology, in Espoo, Finland--earlier developed Fab antibody fragments that recognize either the RS or the SR enantiomer. They have used those fragments to resolve the racemate by affinity chromatography. Söderlund supplied Martin and coworkers with the RS-selective fragments, noting they have the best properties for the membrane separation studies.

In a U-tube setup, the membrane laced with RS-selective fragments transports the RS enantiomer up to 4.5 times faster than it does the SR enantiomer. Preferential transport increases as membrane pore size decreases.

"As the pore diameter becomes smaller, analytes are forced into close proximity with the antibodies lining the wall," Penner explains. "The nanoscale dimension of the pores is essential to the membrane's operation."

A problem often encountered with using antibodies as recognition elements in separations is that the tight binding between the antibody and the target prevents eventual release of the desired molecule. This problem is easily solved in the antibody-based membrane system through the addition of dimethylsulfoxide to the feed and receiver solutions, the Florida team shows.

At the moment, productivity of the membranes is limited because throughput is low, Martin says. But he points out that, on the basis of his group's work on nanotube membrane separations, various ways to enhance throughput are available.

Chemical & Engineering News
Copyright © 2002 American Chemical Society