1067 Acta Crvst. (1996). D52, 1067-1074 An X-ray Analysis of Native Monoclinic Lysozyme. A Case Study on the Reliability of Refined Protein Structures and a Comparison with the Low-Humidity Form in Relation to Mobility and Enzyme Action H. G. NAGENDRA, C. SUDARSANAKUMAR AND M. VIJAYAN Molecular Biophysics Unit, Indian Institute of Science, Bangalore-560 012, India. E-mail: mv@mbu.iisc, ernet, in (Received 8 April 1995: accepted 12 Februao" 1996) Abstract The atomic models of native monoclinic lysozyme obtained by refinement at Bangalore and elsewhere [Young, Dewan, Nave & Tilton (1993). J. Appl. Co'st. 26, 309-319] differed significantly in the flexible regions of the protein molecule. The two models were reconciled starting from regions where they were in reasonable agreement to produce an improved model which yielded an R value of 0.169 for 12 816 observed reflections in the 10-2,~, resolution range. The recon- ciled model was compared with the structure of the 88 % relative humidity form obtained through a water- mediated transformation [Madhusudan, Kodandapani & Vijayan (1993). Acta Co, st. D49, 234-245]. Parts of the flexible regions of the molecule register significant movements during the transformation. The changes resulting from the transformation from the native to the low-humidity forms are pronounced in many of the side chains in the active-site region, thus indicating the relationship between hydration, mobility and enzyme action. The fact that the overall changes in molecular geometry resulting from water-mediated transformation are similar to those which occur during enzyme action, further emphasizes this relationship. 1. Introduction We have been exploring the variability in protein hydration and its structural consequences, including its implication to enzyme action, using an approach involving water-mediated transformations induced by changes in the solvent content in protein crystals resulting from systematic changes in environmental humidity (Salunke, Veerapandian & Vijayan, 1984; Salunke, Veerapandian, Kodandapani & Vijayan, 1985; Kodandapani, Suresh & Vijayan, 1990; Madhusudan & Vijayan, 1991; Madhusudan, Kodandapani & Vijayan, 1993; Radhakishan, Chandra, Sudarsanakumar, Suguna & Vijayan, 1995; Nagendra, Sudarsanakumar & Vijayan, 1995). This exploration has also led to the elucidation of the nature of mobility in hen egg-white lysozyme and bovine ribonuclease A and the identifica- tion of the relatively invariant features in their hydration © 1996 International Union of Crystallography Printed in Great Britain - all rights reserved shell (Madhusudan & Vijayan, 1991; Radhakishan et al., 1995). Among the water-mediated transformations, the one exhibited by monoclinic lysozyme was the most impressive. During the transformation which occurs around a relative humidity of 90% and is accompanied by substantial loss of water, the two crystallographically independent protein molecules in the native form become equivalent. The solvent content of the low- humidity 88% form is as low as 22% which is perhaps the lowest observed in any protein crystal until now. This form diffracts substantially better than the native form and its X-ray analysis has also led to useful information on the water structure associated with proteins (Madhusudan et al., 1993). A detailed comparison of this structure with that of the native form could not be carried out as the available structure of native monoclinic lysozyme was only incompletely refined (Rao, Hogle & Sundaralingam, 1983). There- fore, we undertook a thorough refinement of the structure of the native crystals using data collected on an area detector with the partially refined structure available in the Protein Data Bank (Bernstein et al., 1977) as the starting model. By the time this refinement was completed, another independent refinement of the structure using diffractometer data with the same starting model became available (Young, Dewan, Nave & Triton, 1993). A comparison of these two refined structures and the procedures adopted for the reconciliation of the differences between them form the subject matter of the first part of this paper. Problems associated with the refinement of protein structures especially the possibility of arriving at a crystallographically acceptable, but erroneous, refined structure, have been a matter of concern from the early days of protein structure refinement (Chambers & Stroud, 1977; Vijayan, 1980). Several investiga- tions have been carried out to explore these problems through comparisons of the parameters of the same structure obtained through refinement by different groups, different data sets or both. The results of three of the most recent of these investigations (Daopin, Davies, Schlunegger & Grutter, 1994; Harata, 1994; Fields et al., 1994) underline the reliability of refined structures. The results of a fourth Acta Crystallographica Section D ISSN 0907-4449 © 1996