In a crystal, proteins are fully hydrated. If a protein crystal is dried, it stops diffracting because the native structure of the protein is lost. It is therefore not surprising that in those cases, where structures have been determined in a crystal (X-ray crystallography or electron diffraction) and in solution (NMR), they agree very well (compare, for example, cytochrome B5 https://www.ebi.ac.uk/pdbe/entry/pdb/1ehb and https://www.ebi.ac.uk/pdbe/entry/pdb/1ehb).

Dear Engelbert,

Let me make a few comments on your statements.

Generally, protein samples for X-ray crystallography or electron diffraction are prepared through low-temperature (~0°C) lyophilization in a vacuum of their frozen aqueous solutions. As a result, you get protein material samples that always contain some residual amount of so-called non-frozen water (eutectic water, but not hydration water), and the structure of the samples takes on properties similar to those of crystal hydrates or crystal solvates with a rigid, "frozen" structure. This is what allows us to obtain X-ray and ED diffractograms and spectra with more or less clear reflections and signals, just like with typical crystalline substances. In other words, these methods study rigid, crystal-like, “dead” protein structures. And, if our organisms consisted of such proteins, then, most likely, we would look like DRIED MUMMIES. Isn’t that right?

Next, I give a typical example of Inconsistencies between NMR and X-ray data obtained for the same protein (see link to the article and its abstract below).

About NMR data. There are serious doubts about how closely the conditions (e.g., concentrations, temperature, pH, ionic composition, replacing H2O with D2O, high-speed rotation of samples under high g, which can deform the labile structure, especially for larger proteins, etc.) used for testing proteins match the actual "physiological" conditions in which proteins exist in living matter.

About the difference between water in crystal hydrates and water of hydration. The simplest example of this can be considered using the example of a typical crystal hydrate CuSO4*5H2O. In it, each Cu-“ion” contains 5 water molecules in its environment, but in an aqueous solution of CuSO4, each Cu-“ion” is surrounded by at least three layers of closely bounded water, consisting of 20-30 molecules of hydration water (i.e., hydrated Cu-“ion”).

A similar situation applies to proteins in an aqueous environment. There, they are always in a hydrated state and are surrounded by special "layers" of closely bounded water. Moreover, the spatial structure of this water is cooperatively linked to the labile structure of the protein. In other words, a change in one structure leads to a change in the other, and vice versa. For more details about the special properties of closely bounded, interfacial, the exclusion zone water (EZ-water), you can check the links:

J. Zheng, et al., and G. H. Pollack. Surfaces and interfacial water: Evidence that hydrophilic surfaces have long-range impact. Advances in Colloid and Interface Science 127 (2006) 19. http://dx.doi.org/10.1016/j.cis.2006.07.002. And similar research there https://www.researchgate.net/publication/6838304_Surfaces_and_interfacial_water_Evidence_that_hydrophilic_surfaces_have_long-range_impact. Also

G.V. Andrievsky et al. Peculiarities of the Antioxidant and Radioprotective Effects of Hydrated C60 Fullerene Nanostuctures In Vitro and In Vivo. Free Radical Biology & Medicine, 47 (2009) 786. https://doi.org/10.1016/j.freeradbiomed.2009.06.016 .

----------

Ref about Inconsistencies between NMR and Х-ray data

Walker, F.A. The heme environment of mouse neuroglobin: histidine imidazole plane orientations obtained from solution NMR and EPR spectroscopy as compared with X-ray crystallography. J Biol Inorg Chem 11, 391–397 (2006). https://doi.org/10.1007/s00775-006-0095-8. And similar research there https://www.researchgate.net/publication/6838304_Surfaces_and_interfacial_water_Evidence_that_hydrophilic_surfaces_have_long-range_impact

Abstract

The 1H NMR chemical shifts of the heme methyl groups of the ferriheme complex of metneuroglobin (Du et al. in J. Am. Chem. Soc. 125:8080–8081, 2003) predict orientations of the axial histidine ligands (Shokhirev and Walker in J. Biol. Inorg. Chem. 3:581–594, 1998) that are not consistent with the X-ray data (Vallone et al. in Proteins Struct. Funct. Bioinf. 56:85–94, 2004), and the EPR spectrum (Vinck et al. in J. Am. Chem. Soc. 126:4516–4517, 2004) is only marginally consistent with these data. The reasons for these inconsistencies appear to be rooted in the high degree of aqueous solution exposure of the heme group and the fact that there are no strong hydrogen-bond acceptors for the histidine imidazole N–H protons provided by the protein. Similar inconsistencies may exist for other water-soluble heme proteins, and 1H NMR spectroscopy provides a simple means to verify whether the solution structure of the heme center is the same as or different from that in the crystalline state.

Sebastian Schmitt

Dear Sebastian,

The philosophy of this issue is probably that over the past 50 years of similar research, an "ideal" drug has yet to be developed. The mass gathering for the "games" in computer modeling organized by the winners is unlikely to bring us closer to this goal. The reasons why I gave in my answer to Engelbert Buxbaum on this Q&A forum (Pl., see).

Engelbert Buxbaum

Dear Engelbert,

Let me make a few comments on your statements.

Generally, protein samples for X-ray crystallography or electron diffraction are prepared through low-temperature (~0°C) lyophilization in a vacuum of their frozen aqueous solutions. As a result, you get protein material samples that always contain some residual amount of so-called non-frozen water (eutectic water, but not hydration water), and the structure of the samples takes on properties similar to those of crystal hydrates or crystal solvates with a rigid, "frozen" structure. This is what allows us to obtain X-ray and ED diffractograms and spectra with more or less clear reflections and signals, just like with typical crystalline substances. In other words, these methods study rigid, crystal-like, “dead” protein structures. And, if our organisms consisted of such proteins, then, most likely, we would look like DRIED MUMMIES. Isn’t that right?

Next, I give a typical example of Inconsistencies between NMR and X-ray data obtained for the same protein (see link to the article and its abstract below).

About NMR data. There are serious doubts about how closely the conditions (e.g., concentrations, temperature, pH, ionic composition, replacing H2O with D2O, high-speed rotation of samples under high g, which can deform the labile structure, especially for larger proteins, etc.) used for testing proteins match the actual "physiological" conditions in which proteins exist in living matter.

About the difference between water in crystal hydrates and water of hydration. The simplest example of this can be considered using the example of a typical crystal hydrate CuSO4*5H2O. In it, each Cu-“ion” contains 5 water molecules in its environment, but in an aqueous solution of CuSO4, each Cu-“ion” is surrounded by at least three layers of closely bounded water, consisting of 20-30 molecules of hydration water (i.e., hydrated Cu-“ion”).

A similar situation applies to proteins in an aqueous environment. There, they are always in a hydrated state and are surrounded by special "layers" of closely bounded water. Moreover, the spatial structure of this water is cooperatively linked to the labile structure of the protein. In other words, a change in one structure leads to a change in the other, and vice versa. For more details about the special properties of closely bounded, interfacial, the exclusion zone water (EZ-water), you can check the links:

J. Zheng, et al., and G. H. Pollack. Surfaces and interfacial water: Evidence that hydrophilic surfaces have long-range impact. Advances in Colloid and Interface Science 127 (2006) 19. http://dx.doi.org/10.1016/j.cis.2006.07.002. And similar research there https://www.researchgate.net/publication/6838304_Surfaces_and_interfacial_water_Evidence_that_hydrophilic_surfaces_have_long-range_impact. Also

G.V. Andrievsky et al. Peculiarities of the Antioxidant and Radioprotective Effects of Hydrated C60 Fullerene Nanostuctures In Vitro and In Vivo. Free Radical Biology & Medicine, 47 (2009) 786. https://doi.org/10.1016/j.freeradbiomed.2009.06.016 .

----------

Ref about Inconsistencies between NMR and Х-ray data

Walker, F.A. The heme environment of mouse neuroglobin: histidine imidazole plane orientations obtained from solution NMR and EPR spectroscopy as compared with X-ray crystallography. J Biol Inorg Chem 11, 391–397 (2006). https://doi.org/10.1007/s00775-006-0095-8. And similar research there https://www.researchgate.net/publication/6838304_Surfaces_and_interfacial_water_Evidence_that_hydrophilic_surfaces_have_long-range_impact

Abstract

The 1H NMR chemical shifts of the heme methyl groups of the ferriheme complex of metneuroglobin (Du et al. in J. Am. Chem. Soc. 125:8080–8081, 2003) predict orientations of the axial histidine ligands (Shokhirev and Walker in J. Biol. Inorg. Chem. 3:581–594, 1998) that are not consistent with the X-ray data (Vallone et al. in Proteins Struct. Funct. Bioinf. 56:85–94, 2004), and the EPR spectrum (Vinck et al. in J. Am. Chem. Soc. 126:4516–4517, 2004) is only marginally consistent with these data. The reasons for these inconsistencies appear to be rooted in the high degree of aqueous solution exposure of the heme group and the fact that there are no strong hydrogen-bond acceptors for the histidine imidazole N–H protons provided by the protein. Similar inconsistencies may exist for other water-soluble heme proteins, and 1H NMR spectroscopy provides a simple means to verify whether the solution structure of the heme center is the same as or different from that in the crystalline state.

G. V. Andrievsky : It would be highly unusual to use lyophilised samples in X-ray or electron crystallography. The normal procedure is to drop the sample into propane cooled by liquid nitrogen to vitrify the water (snap-freeze it w/o crystal formation). The propane is near its freezing point of -188 °C and will therefore not form bubbles around the sample, avoiding the Leidenfrost effect.

High spin rates in NMR (magic angle spinning NMR) are not used for soluble proteins, but for immobile ones, e.g., membrane proteins. It avoids the lowering of spectral resolution as the rotational correlation times become longer than the heteronuclear dipole-dipole and spin interaction times in “solid” (in the sense that NMR spectroscopist use this word) samples.

Comparing the crystal water of inorganic salts and the water in a protein crystal makes no sense at all, they are completely different.

In the PDB, I could only find neuroglobin structures obtained by X-ray diffraction, not by NMR. I was therefore unable to verify whether the two would be significantly different. EPR looks only at the immediate environment of the metal centre and can therefore not measure differences in protein structure caused by (de)hydration. The PDB therefore does not contain structures obtained by this method. Small changes of the position of the haem group with respect to the protein are normal, for example during ligand binding. They are not proof of major changes in protein structure.