Discovering structural motifs using a structural alphabet: application to magnesium-binding sites

Abstract

Background: For many metalloproteins, sequence motifs characteristic of metal-binding sites have not been found or are so short that they would not be expected to be metal-specific. Striking examples of such metalloproteins are those containing Mg2+, one of the most versatile metal cofactors in cellular biochemistry. Even when Mg2+-proteins share insufficient sequence homology to identify Mg2+-specific sequence motifs, they may still share similarity in the Mg2+-binding site structure. However, no structural motifs characteristic of Mg2+-binding sites have been reported. Thus, our aims are (i) to develop a general method for discovering structural patterns/motifs characteristic of ligand-binding sites, given the 3D protein structures, and (ii) to apply it to Mg2+-proteins sharing <30% sequence identity. Our motif discovery method employs structural alphabet encoding to convert 3D structures to the corresponding 1D structural letter sequences, where the Mg2+-structural motifs are identified as recurring structural patterns.

Results: The structural alphabet-based motif discovery method has revealed the structural preference of Mg2+-binding sites for certain local/secondary structures: compared to all residues in the Mg2+-proteins, both first and second-shell Mg2+-ligands prefer loops to helices. Even when the Mg2+-proteins share no significant sequence homology, some of them share a similar Mg2+-binding site structure: 4 Mg2+-structural motifs, comprising 21% of the binding sites, were found. In particular, one of the Mg2+-structural motifs found maps to a specific functional group, namely, hydrolases. Furthermore, 2 of the motifs were not found in non metalloproteins or in Ca2+-binding proteins. The structural motifs discovered thus capture some essential biochemical and/or evolutionary properties, and hence may be useful for discovering proteins where Mg2+ plays an important biological role.

Conclusion: The structural motif discovery method presented herein is general and can be applied to any set of proteins with known 3D structures. This new method is timely considering the increasing number of structures for proteins with unknown function that are being solved from structural genomics incentives. For such proteins, which share no significant sequence homology to proteins of known function, the presence of a structural motif that maps to a specific protein function in the structure would suggest likely active/binding sites and a particular biological function.

Figure 1

Zn-binding site structural motifs derived from the structural alphabet representation of 3 Zn-finger proteins. For each protein, the PDB entry and chain is given, followed below by its amino acid sequence (in capital letters), aligned with the corresponding structural alphabet representation (lower-case letters); 'Z', means a letter cannot be assigned to this residue (see Methods). Zn²⁺-binding residues are underlined and in bold. Only the first 30 amino acid residues are shown. The C_α root-mean-square deviation RMSD of 1LAT and 2NLL from 1HCQ are 1.73 and 1.33 Å, respectively, whereas that of 1LAT from 2NLL is 1.25 Å.

Figure 2

The percentage letter frequency distributions of first-shell amino acid residues (gray), second-shell amino acid residues (white), and all amino acid residues (black) in the Mg²⁺-proteins. There is a total of 25,406 amino acid residues in the Mg²⁺-proteins, of which 250 are in the first shell, while 898 are in the second shell

Figure 3

The percentage secondary structure frequency distributions of first-shell amino acid residues (gray), second-shell amino acid residues (white), and all amino acid residues (black) in the Mg²⁺-proteins. The amino acid residues found in α-helices, β-strands, or loops are according to the secondary structure information in the PDB files.

Figure 4

The conserved local structures of the 4 Mg²⁺-structural motifs. (a) e(24–47)h(24)k, (b) f(1)h(109–349)b, (c) f(2)h(126–158)m, and (d) k(26–29)h(1)a.

Figure 5

The conserved binding site of 2 nonhomologous Mg²⁺-proteins. (a) Cartoon diagram of the metal-binding domain in N-acylamino acid racemase (1SJC). (b) Cartoon diagram of the metal-binding domain in gamma enolase (2AKZ). (c) Superposition of the first-shell structural letters of 1SJC (blue) and 2AKZ (yellow).

Figure 6

Conversion of the 3D protein backbone into a 1D structural alphabet representation. The first 2 and the last 2 residues are assigned 'Z', meaning a letter cannot be assigned at these residues. The first valid assignment is 'd', at Val 3 and spanning residues 1 to 5. The next is assigned to Asp 4 and spans residues 2 to 6.