Dear Petr, I observed both open and closed protein conformations discretely disordered in the crystals of tyrosine phenol-lyase. In some cases, I could model two ligand conformations in the disordered active site, but the polypeptide chain could be modeled in the prevalent conformation only (for example see: Milic et al., J. Am. Chem. Soc. 2011, 133, 16468-16476, http://dx.doi.org/10.1021/ja203361g). In two crystals soaked with two different ligands, there was approximately 50 % occupancy of each conformation, so I could model both open and closed polypeptide conformations (affecting 129 amino acid residues). Details about these two structures can be found in my doctoral thesis (D. Milic, Doctoral Thesis, University of Zagreb, 2010; https://bib.irb.hr/datoteka/472684.Dalibor_Milic_Doctoral_Thesis.pdf; Section 4.4 "Discrete disorder of the whole protein domain in the TPL crystals"). Some of the interactions between protein subunits observed for the open conformation are disrupted after the closure of one of the protein subunits. To compensate for these broken interactions, new interactions are formed within the protein subunit in the closed conformation. I wrote more about that in my Thesis, page 106.
Another example of such discrete disorder of a polypetide chain affecting 53 residues is described in Lehwess-Litzmann et al, Nat. Chem. Biol. 2011, 7, 678-684 (http://dx.doi.org/10.1038/nchembio.633), PDB ID 3S0C (http://dx.doi.org/10.2210/pdb3s0c/pdb), chain D. I guess there are many more such structures. To find similar examples in PDB, one would need to search for entries where there are alternate location indicators present for an arbitrary long stretch of neighboring residues. I hope my answer is helpful. Best regards, Dalibor -- Dr. sc. Dalibor Milić Heinrich-Heine-Universität Düsseldorf Institut für Biochemische Pflanzenphysiologie Gebäude/Raum: 26.02/01.74 Universitätsstraße 1 D-40225 Düsseldorf Germany Telephone: +49 211 81-13722 E-mail: dalibor.mi...@hhu.de http://www.BioChemPlant.hhu.de