Construction of a model of the Histidine-containing Phosphocarrier protein from Mycoplasma capricolum (HPr) using homology modeling techniques. A model for Histidine-containing Phosphocarrier protein from Mycoplasma capricolum (HPr) has been calculated using homology modeling techniques. The sequence of HPr was used with the Profiles-3D program [1] to search known protein structures as a method for identifying structural motifs compatible with this sequence. This analysis identified the mutant Bacillus subtilis Histidine-containing Phosphocarrier protein (2HPR), Streptococcus faecalis Histidine-containing Phosphocarrier protein (1PTF) and E. coli Histidine-containing Phosphocarrier protein (1POH) as possibly suitable models (in that order) and gave sequence alignments. Multiple sequence alignment of the 4 sequences was then performed. No sequence alignment suggested insertions or deletions so 2HPR was used to obtain the starting backbone coordinates. Identical side-chains to 2HPR in the multiple sequence alignment were also taken from this model. When an identity to one of either 1PTF or 1POH existed in the multiple alignment the model side-chain was positioned by minimizing the rms of the identical atoms and all the backbone atoms, producing a global rigid fit of the 2 proteins to construct the model.
An automated search of a rotamer library was used to find the minimum energy conformations of the other side-chains. This represented the starting point for 3 series of energy minimization.
All software used in this study was provided by Biosym Technologies. The Brookhaven Protein Data Bank was the source of all experimentally determined protein structures.
1. Luthy, R., McLachlan, A.D. & Eisenberg, D., Proteins, 10, 229-239 (1991)
Construction of a model of the Histidine-containing Phosphocarrier protein from Mycoplasma capricolum (HPr) using homology modeling techniques.
Homology modelling methods have been used to construct a model of the Histidine-containing Phosphocarrier protein (Hpr) from Mycoplasma capricolum, based on its homology to known structures of this protein from other organisms. The sequence for the Hpr protein [1] was compared with the sequences for the Histidine-containing Phosphocarrier proteins from B.subtilis [2] (PDB entry 2hpr), S. faecalis [3] (PDB entry 1ptf) and E. coli [4] (PDB entry 1poh). This comparison was made using both the FASTA program [5] and a multiple sequence alignment technique [Biosym Technologies]. The comparison indicated that all four sequences could be aligned without the need for insertions or deletions. The highest sequence identity (46% in a 74 amino acid region) was with the B. subtilis structure and this was chosen as the basis for construction of the M. capricolum structure.
The backbone coordinates from the B. subtilis structure were used. Residue Met-1 (missing from the B. subtilis structure) was built in the conformation found in the S. faecalis structure and Gly-89 was built in an extended conformation. Non-mutated side chains were taken directly from the B. subtilis structure. The catalytically important residues His-15 and Arg-17 exist in a variety of conformations in the three known structures and these were modelled in their conformation from the B. subtilis structure. A sulphate ion that is located between these residues was also used in the modelling process. Side chains for residues that are identical between M. capricolum and S. faecalis or E. coli (but that are different in B. subtilis) were built in the conformation from S. faecalis or E. coli, as appropriate (some were later manually adjusted). Side chains that are similar between M. capricolum and one of the reference proteins (for example, Ile vs. Leu) were built in conformations that were similar to those of the reference proteins. The remaining side chain conformations were adjusted using a combination of a manual scan of a side chain rotamer library for individual residues, and an automated procedure that attempts to minimise the energy of a group of side chains by selecting side chain conformers in a systematic fashion [Biosym Technologies]. Residue Phe-4 was adjusted so that the side chain occupied a "hole" left by the Phe-6 -> Ala-6 replacement.
The structure was then relaxed using energy minimisation in several stages:
The structure was assessed using the Profiles-3D program [6] and geometric checks [Biosym Technologies]. These indicated that the protein was folded in a reasonable conformation. Finally, the protein structure was placed in a box of waters and more thoroughly energy minimised. A short molecular dynamics simulation was then run (all protein atoms, and the sulphate ion, were completely free to move). This was carried out in order to assess the usefulness of minimisation and dynamics for refining model-built protein structures. The results of this study will be presented at the meeting.
1. Zhu et al., J. Biol. Chem. 268, 26531-26540 (1993)
2. Herzberg et al., Proc. Natl. Acad. Sci. USA, 89, 2499-2503 (1992)
3. Jia et al., Nature, 361, 94-97 (1993)
4. Jia et al., J. Biol. Chem. 268, 22490-22501 (1993)
5. Pearson and Lipman, Proc. Natl. Acad. Sci. USA, 85, 2444-2448 (1988)
6. Luthy et al., Nature, 356, 83-85 (1992)
Structural Biology home page Asilomar Conference home page