Comparison of hydrogen determination with X-ray and neutron crystallography in a human aldose reductase-inhibitor complex.

Protonation states determination by neutron (2.2 A at room temperature) and X-ray (0.66 A at 100 K) crystallographic studies were compared for a medium size enzyme, human aldose reductase (MW=36 kDa), complexed with its NADP+ coenzyme and a selected inhibitor of therapeutic interest. The neutron resolution could be achieved only with the ab initio fully deuterated protein and the subsequent crystallization in D2O of the complex. We used the largest good-quality crystal (1.00x0.67x0.23 mm, i.e. volume of 0.15 mm3) that we were able to grow so far. Both studies enable the determination of protonation states, with a clear advantage for neutrons in the case of less-ordered atoms (B>5 A2). Hydrogen atoms are best determined by a complementary analysis of the Fourier maps obtained from both methods.

High-resolution neutron protein crystallography with radically small crystal volumes: application of perdeuteration to human aldose reductase.

Neutron diffraction data have been collected to 2.2 Angstrom resolution from a small (0.15 mm(3)) crystal of perdeuterated human aldose reductase (h-AR; MW = 36 kDa) in order to help to determine the protonation state of the enzyme. h-AR belongs to the aldo-keto reductase family and is implicated in diabetic complications. Its ternary complexes (h-AR-coenzyme NADPH-selected inhibitor) provide a good model to study both the enzymatic mechanism and inhibition. Here, the successful production of fully deuterated human aldose reductase [h-AR(D)], subsequent crystallization of the ternary complex h-AR(D)-NADPH-IDD594 and neutron Laue data collection at the LADI instrument at ILL using a crystal volume of just 0.15 mm(3) are reported. Neutron data were recorded to 2 Angstrom resolution, with subsequent data analysis using data to 2.2 Angstrom. This is the first fully deuterated enzyme of this size (36 kDa) to be solved by neutron diffraction and represents a milestone in the field, as the crystal volume is at least one order of magnitude smaller than those usually required for other high-resolution neutron structures determined to date. This illustrates the significant increase in the signal-to-noise ratio of data collected from perdeuterated crystals and demonstrates that good-quality neutron data can now be collected from more typical protein crystal volumes. Indeed, the signal-to-noise ratio is then dominated by other sources of instrument background, the nature of which is under investigation. This is important for the design of future instruments, which should take maximum advantage of the reduction in the intrinsic diffraction pattern background from fully deuterated samples.