Miscellany

For want of a crystal, a grant was lost

For more than a decade study groups of the United States National Institutes of Health (NIH) routinely have denied funding for applications dealing with crystallization of membrane proteins, claiming that the outcome of such projects is too unpredictable to justify support. The practice is widely known as the no crystals, no grant rule, applied by the U.S. crystallographic community despite the fact that the NIH does not accept it as an official review criterion. Anthony N. Martonosi

Since 1993, a series of spectacular accomplishments - notably in Europe and Japan, but also including NIH-funded investigators in the U.S. - have convincingly demonstrated that X-ray crystallographic studies of membrane proteins are feasible (Table 1).

The structure of many extremely important and medically relevant membrane proteins remain to be determined. To encourage changes in study section philosophy, the U.S. NIGMS drafted a set of recommended "Criteria for the Evaluation of Structure Determination Proposals". These non-binding suggestions were distributed by the Div. of Research to relevant NIH study sections, some of which are now using these criteria. It appears that progress is being made in considering the importance of the crystallization of membrane proteins in its own right.

Table 1. High-resolution structures that include a membrane embedded domain1

Protein Year of publication
Bacteriorhodopsin1-41990, 1996, 1997
Bacterial photoreaction centers5-131984, 1986, 1993, 1994, 1996
Light harvesting complexes14-16 1984, 1995, 1996
Photosystem I17 1996
Porins18-271991, 1992, 1994, 1995, 1997, 1998
Alpha-hemolysin281996
Prostaglandin synthase-I291993
Prostaglandin synthase-II301996
Cytochrome oxidase31-331995
Cytochrome bc1 complex34-361996, 1997

1Structures solved for substantially intact proteins which included a membrane-embedded domain. The list does not include structure solved for extramembrane domains of membrane proteins, membrane-embedded fragments, small peptides, nor peripheral membrane proteins not tightly associated with the lipid core. NIGMS is interested in tracking progress in this area. If you note omissions to this list, please contact preuschp@nigms.nih.gov

Peter C. Preusch, John C. Norvell, James C. Cassatt and Marvin Cassman