Root-mean-square-deviation folding

Example Crippen and Snow reported their success in developing a simplified potential for protein folding. In their model, single points represent amino acids. For the avian pancreatic polypeptide, the native structure is not at a potential minimum. However, a global search found that the most stable potential minimum had only a 1.8 Angstrom root-mean-square deviation from the native structure. 

In order to examine whether this sequence gave a fold similar to the template, the corresponding peptide was synthesized and its structure experimentally determined by NMR methods. The result is shown in Figure 17.15 and compared to the design target whose main chain conformation is identical to that of the Zif 268 template. The folds are remarkably similar even though there are some differences in the loop region between the two p strands. The core of the molecule, which comprises seven hydrophobic side chains, is well-ordered whereas the termini are disordered. The root mean square deviation of the main chain atoms are 2.0 A for residues 3 to 26 and 1.0 A for residues 8 to 26. 

There is considerable similarity between domains 1, 3, and 5 with excellent matching of the (3-strands 168 a-carbon atoms of domain 3 and 149 atoms of domain 5 can be superimposed on domain 1 with a root mean square fit of only 1.0 A. Major deviations occur in the loops between the first and second, and the fourth and sixth strands. The similarity of the smaller mononuclear copper binding domains, 2,4, and 6, is even more pronounced with 147 atoms from domain 4 and 143 from domain 6 fitting domain 2 with a root mean square deviation of only 0.9 A in each case. However, although all 6 domains are based on a cupre-doxin-type fold, the various loop regions deteriorate the match between an even and an odd domain with a typical fit of 1.8 A for only 91 atoms when domain 2 is superposed onto domain 1. The superposition results are summarized as part of Table 2. 

Structural biology provides a final way to define an E2 enzyme. As expected from the strong sequence conservation, the E2 core domain adopts a conserved fold. At the time this article was being prepared, twelve different E2 structures had been deposited in the Protein Data Bank. The average root-mean-square deviation of... 

In large subunit enzymes (PVC and HPII), a short segment of about 30 residues links the a-helical domain to the C-terminal domain (Fig. 8). The latter segment is a conspicuous addition to the small subunit containing about 150 residues folded into a structure that resembles flavodoxin. For example, there is a root mean square deviation of 3.0 A between flavodoxin and approximately 100 residues of the C-terminal domains of either HPII or PVC. This can be compared to the 1.8 A root mean square deviation for 134 centers between the C-terminal domains of HPII and PVC. Unlike the N-terminal end, the final C-terminal residue Ala753 is visible in the structure of HPII. The C-terminal domain contains extensive secondary structure in the form of four a-helices (al5-18) and eight fi-strands (fi9-16). Despite the obvious structural similarity to flavodoxin, there is no evidence of nucleotide binding in the domain and its function remains a mystery. 

Crane et al. first established the three-dimensional fold of NOS by solving the structure of a monomeric form of the mouse iNOS heme domain (78). This version of iNOS was missing the first 114 residues, which are known to be critical for dimer formation and activity (79). The monomer structure was soon followed by the dimeric heme domain structures of mouse iNOS (80), bovine eNOS (81), and the human isoforms of iNOS (82, 83) and eNOS (82). A comparison of eNOS and iNOS reveals that the structures are essentially the same with an overall root-mean-square deviation in backbone atoms of 1.1 A (S3). The sequence identity between human iNOS and bovine eNOS is 60% for 420 residues compared in the crystal structures (83). 

One of the simplest methods is the comparison of the initial structure of the macromolecule to that throughout the trajectory via a distance measure such as the root mean square deviation (RMSD). This method is most informative for a system like a folded protein under native conditions, where the molecule is expected to spend the vast majority of the time in conformations quite similar to the crystal structure. If one computes the RMSD time series against the crystal structure, one expects to see a rapid rise due to thermal fluctuations, followed by a long plateau or fluctuations about a mean at longer timescales. If the RMSD... 

As noted above, the two lobes have very similar folding. This is only to be expected given their high (—40%) sequence identity. The differences, at the level of polypeptide folding, are confined primarily to small insertions and deletions in the loops that join secondary structure elements. These are almost all located on the molecular surface and do not disturb the basic structure—indeed 90% of the main chain atoms of the N-lobe of human lactoferrin can be superimposed on equivalent atoms in the C-lobe with a root-mean-square deviation of only —1.2 A. The agreement would be even closer were it not for the small difference in the closure of the two domains, described above. 

Table 1. Pairwise comparison of the topology and primary sequence of members of the short spacer family. The alpha carbon atoms defining the zinc protease fold (orange segment. Fig. 3) have been used in the topological superposition [56]. The distances refer to the root mean square deviations of this fold between pairs of structures. The corresponding pairwise primary sequence homology is also shown.

The meaning of the reduced models of protein structures is discussed in more detail later in this chapter however, we stress here that the minimal requirements for a low-resolution structure prediction to be correct are (A). The overall topology (the shape of the main chain trace) of the fold is the same as that seen in the experimental structure (B). The obtained secondary structure is very close to that seen in the native structure and the alpha carbon trace root mean square deviation (RMSD) from the native structure is in the range of 3-7 A depending on protein size. This level of accuracy may be of some use for application to protein function annotation [179]. 

Figure 13 Folding of poly-L-leucine during transfer across the water-hexane interface time evolution of the distance between the center of mass (COM) of the peptide backbone and the interface (solid line), and the distance root mean square deviation (RMSD) with respect to the ideal a-helix (dashed line)...

It was found that the Pf PDO molecule can be almost equally divided into two structural units, the N-terminal one (residues 1 to 117) and the C-terminal one (residues 118 to 226). They are connected by the loop between a4 and o5 in the middle of the molecule. Either unit contains four p strands and four a helices and displays exactly the same fold (Fig. 2), despite the fact that their sequence identity is only about 20%. Their superposition reveal.s a root-mean-square (r.m.s.) deviation of 1.23 A for 66 Cot atoms. Interestingly, each of the ly PDO units is basically a thioredoxin fold motif but with an additional a helix, al or a5, inserted at the N terminus. As a result, the two units share a similar fold with other protein... 

---