Media.Player

High Thrughput Screen for Drug Toxicity via Nonspecific Effects on Cell Membranes

*Abstract

Inventions:

• Method for high-throughput screening to evaluate whether, and at what dose, test compounds cause nonspecific changes to cell membranes and therefore are likely to cause side effects, by:
• filling vesicles in which gramicidin channel subunits are embedded, with a fluorescent substance that is quenchable with a gramicidin-permeable cation; and
• adding a test drug to a solution with these vesicles; and
• adding cations that can (i) pass through fully formed gramicidin channels and that (ii) quench the fluorophore; and
• measuring the rate of quenching.
• A device for carrying out this method
• An analysis scheme that allows for semi-automatic data analysis.

 

This high-throughput screen for non-specific membrane effects should become widely implemented as a secondary screen for hits from high-throughput single-protein or cell assays. It also will be useful in later iterations of drug discovery, to facilitate pharma's efforts to quickly get a handle on an upper range of dosing for drug candidates, and to stop development of compounds that are likely to cause side effects at relevant doses.

===================

 

As most folks in the biopharma industry know, the last Tufts study of the cost to develop a drug produced a staggering number: ~ $880,000,000. What is perhaps less widely known, is that about $665M of those costs are attributed to candidates that failed, and that fully 90% of that $665M is spent before any compound is tested in humans (presentation by David U'Pritchard; BIO 2009). As the adage goes, one should indeed fail early! In other words, one should eliminate compounds that are unlikely to be effective drugs as soon as possible.

 

The abundance of failure arises from the mind-boggling complexity of trying to identify specific chemical structures (selected from infinite chemical space) that have a desired activity inside the turbulent environment of a living body.

 

In order to fail earlier, medicinal chemists have been trying to pare down chemical space. For instance, Lipinski's widely followed "Rule of 5" lays out qualities that most good drugs have. Additionally, drug companies are increasingly looking for molecules that inhibit very specific targets (e.g. Novartis' Gleevec), by screening against specific proteins, rather than whole cells or whole animals.

 

On a separate track, research led by Weill Cornell's Olaf Andersen has, over the last twenty years, investigated the interactions between cell membranes and membrane proteins. Dr. Andersen's lab and their collaborators have shown that compounds that act directly on the cell membrane are likely to cause changes to membrane proteins and the signals they transmit - the very biochemical definition of "side effect."

 

Ironically, the Rule of 5 may be directing pharma companies directly toward this problem, as one of Lipinski's rules favor lipophilic/amphiphilic compounds -- compounds that adsorb to the membrane/solution interface and therefore are the most likely ones to interact with the membrane itself. And lead compounds emerging from pharma's labs are indeed increasingly lipophilic (Leeson and Springthorpe, 2007). Additionally, the more that companies target specific proteins, the less able they are to weed out -- at an early stage -- compounds that affect the cell membrane.

 

Up to the present, Dr. Andersen has worked in an high-resolution, low-throughput system to examine membrane effects of a variety of compounds. Specifically, he puts subunits of a special kind of ion channel -- the gramicidin channel -- into membranes, adds the test compound, and conducts careful (and relatively slow) single-channel measurements (demonstrated on the website at the link here). Gramicidin channel subunits form membrane-spanning channels through which ions can flow, and the rate at which they break apart -- stopping the ion movement -- is sensitive to membrane properties. A change in ion movement (or current) through the channel is therefore an excellent surrogate marker for changes to the membrane itself. But until now, there was no surrogate marker for changes in the gramicidin channels themselves -- one had to directly measure changes in the flow of current through many single channels.

 

Now, however, Dr. Andersen and his student, Helgi Ingolfsson, have developed fluorescence quench techniques that allow them to summarize changes in gramicidin channel function in a snap. It used to take one-two solid weeks of work to gather meaningful data on a single compound; but now this academic lab, using a system of off-the-shelf parts that is manually operated, can easily test ten compounds in a day. With robotics and with industrial technicians, this number can increase a hundredfold and more.

 

This high-throughput screen for non-specific membrane effects should become widely implemented as a secondary screen for hits from high-throughput single-protein or cell assays. It also will be useful in later iterations of drug discovery, to facilitate pharma's efforts to quickly get a handle on an upper range of dosing for drug candidates, and to stop development of compounds that are likely to cause side effects at relevant doses.

*Licensing

Vibhu Sachdev(212) 746-6187sachdev@cornell.edu

其他

Gramicidin-based fluorescence assay; for determining small molecules potential for modifying lipid bilayer properties. J. Vis. Exp. (44), e2131, DOI:10.3791/2131 (2010)

Thiazolidinedione insulin sensitizers alter lipid bilayer properties and voltage-dependent sodium channel function: implications for drug discovery. J. Gen. Physiol. 138:249-70 (2011)

國家/地區

美國