Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2023 Aug 28;63(16):5204-5219.
doi: 10.1021/acs.jcim.3c00558. Epub 2023 Aug 9.

Identifying Metal Binding Sites in Proteins Using Homologous Structures, the MADE Approach

Vid Ravnik  1 , Marko Jukič  1   2   3 , Urban Bren  1   2   3
Affiliations

Identifying Metal Binding Sites in Proteins Using Homologous Structures, the MADE Approach

Vid Ravnik et al. J Chem Inf Model. .

Abstract

In order to identify the locations of metal ions in the binding sites of proteins, we have developed a method named the MADE (MAcromolecular DEnsity and Structure Analysis) approach. The MADE approach represents an evolution of our previous toolset, the ProBiS H2O (MD) methodology, for the identification of conserved water molecules. Our method uses experimental structures of proteins homologous to a query, which are subsequently superimposed upon it. Areas with a particular species present in a similar location among many homologous protein structures are identified using a clustering algorithm. Dense clusters likely represent positions containing species important to the query protein structure or function. We analyze well-characterized apo protein structures and show that the MADE approach can identify clusters corresponding to the expected positions of metal ions in their binding sites. The greatest advantage of our method lies in its generality. It can in principle be applied to any species found in protein records; it is not only limited to metal ions. We additionally demonstrate that the MADE approach can be successfully applied to predict the location of cofactors in computer-modeled structures, e.g., via AlphaFold. We also conduct a careful protein superposition method comparison and find our methodology robust and the results largely independent of the selected protein superposition algorithm. We postulate that with increasing structural data availability, additional applications of the MADE approach will be possible such as non-protein systems, water network identification, protein binding site elaboration, and analysis of binding events, all in a dynamic manner. We have implemented the MADE approach as a plugin for the PyMOL molecular visualization tool. The MADE plugin is available free of charge at https://gitlab.com/Jukic/made_software.

PubMed Disclaimer

Conflict of interest statement

The authors declare no competing financial interest.

Figures

Figure 1
Figure 1
MADE approach: first, we identify a set of protein chains homologous to our query, these are then superimposed onto the query protein, and next dense areas of a particular species are located with a clustering algorithm. Clusters with many members likely represent species important to the query protein structure.
Figure 2
Figure 2
MADE plugin. (A) Define System tab with the PDB ID input field (or Custom Structure for inputting a custom .pdb file). The Use Alphafold structure check allows the use of a protein structure modeled by AlphaFold2. Underneath we specify, if we want to analyze a whole protein chain or a specific binding site, and if we want at most one chain per PDB entry. Below one finds a dropdown list that allows them to choose a sequence identity cutoff for the MMseqs2 clustering data (Or a custom cluster of complexes). There is also a dropdown list to select between different superposition methods. The Find Homologous button finds the set of homologous structures based on the query protein PDB ID and on the chosen sequence identity cutoff. If the query consists of multiple chains with different clusters of homologous structures, select the cluster to use in the Select Cluster field. The Download Complexes button downloads the .pdb files from the PDB database if they are not already present. When the query protein structure is downloaded, the user selects a protein chain or a binding site from the Select BINDING SITE or PROTEIN CHAIN list and runs the calculation with the Launch Calculation button. (B) After the calculation finishes, the user is redirected to the Cluster Analysis tab. Here, one finds the Info panel with the information about the parameters of the calculation, as well as about the identities and chemical formulae of heteroatoms located in the analyzed system (taken from the HETNAM and FORMUL entries of the .pdf files). Below on the left is the Types of HETATMs list, which displays the total number and maximum conservation of clusters of all types of heteroatoms present. To the right lies the Calculated Clusters field, which lists all the different clusters of heteroatoms identified. Selecting an entry from the Types of HETATMs list will filter the Calculated Clusters field to display only clusters matching the selected species. Below the Calculated clusters field are buttons used for displaying the results. The Fetch/Reset button will clear the PyMOL display and fetch the query protein structure. To the left, the Display Selected Clusters button will draw the selected clusters as spheres in the PyMOL viewer. The Clus. B-Site button will show the residues around the already displayed clusters, the Clus. Contacts button will show the distances between clusters and the nearby amino acid residues. Last but not least, the Analysed Area button will draw a box around the area in which the plugin searched for clusters.
Figure 3
Figure 3
Zinc Finger with PDB ID 1RIK analyzed with the MADE plugin. (A) All the homologous protein structures are superimposed (using DeepAlign) upon the original 1RIK structure (cyan). An isotropic displacement of zinc ions (light purple spheres) in the cluster identified by DBSCAN is displayed with red dots. (B) Cys2His2 biding site of 1RIK with the calculated average position of the Zn cluster (red sphere). Numbers represent coordinate bond distances in Å.
Figure 4
Figure 4
MnSOD structure with PDB ID 3OT7 analyzed with MADE plugin. (A) All homologous structures superimposed (using DeepAlign) upon the query 3OT7 structure (cyan). We see two clusters of Mn ions (light purple spheres) in the binding sites of the A and B chains. (B) Predicted Mn binding site in chain A of 3OT7: position of the cluster of Mn ions predicted by MADE (red sphere) as well as clusters of H2O and OH– species (red and blue cross markers, respectively). The coordination of the predicted Mn ion is trigonal bipyramidal (gray lines), and the clusters of H2O or OH– can form hydrogen bonds with nearby residues (yellow dotted lines). Numbers display bond distances in Å.
Figure 5
Figure 5
BCII MβL structure with the PDB accession code 3I0V analyzed with the MADE plugin. The positions of the clusters of Zn ions predicted by the MADE approach (red spheres) are shown as well as a cluster of H2O molecules (red cross marker). Numbers represent bond distances in Å. The coordination of Zn-1 in the 3H binding site is tetrahedral as expected (three His residues and the conserved bridging water cluster). The expected coordination of Zn-2 is trigonal bipyramidal (DHC, the conserved bridging water molecule and another conserved water molecule), our approach, however, does not find a second water cluster that should coordinate Zn-2.
Figure 6
Figure 6
Chain A of AGCO structure with the PDB accession code 3X42 analyzed using the MADE plugin. (A) From the whole protein chain, we can observe many solvent ions. (B) Cu binding site of 3X42, the Cu ion (brown sphere) from 3X42, and the Cu cluster from the MADE approach (red sphere). Three conserved water molecules, Wa, We, and W1 (red cross markers in 3X42, blue cross markers from MADE), are located in the binding site. Hydrogen bonds are indicated with yellow dotted lines and numbers represent bond distances in Å. The Cu ion is bound with a distorted square-planar geometry with three His residues, We and Wa.
Figure 7
Figure 7
Cytochrome c’ structure predicted by AlphaFold2 with UniProt ID P00138 (RCSB PDB CMS ID AF_AFP00138F1). The full Heme C structure was predicted by the MADE approach. Heme C is covalently bound via thioester bonds with Cys-116 and Cys-119, the Fe ion in Heme C has five ligands, four from nitrogen atoms of the porphyrin ring, and one from the polypeptide, where it is coordinated by the nitrogen of His-120. Numbers represent coordinate bond distances in Å.
Figure 8
Figure 8
Example of the results of different superposition algorithms: chain A of 2DOO IMP-1 metallo-β-lactamase superposed onto 6JED (not shown) using various superposition algorithms. ProBiS (green), TM-align, DeepAlign, and GANGSTA+ (shades of yellow and orange) as well as PyMOL align and super (shades of blue).
Figure 9
Figure 9
Diagram of our recommendations for superposition method selection. Use ProBiS for local superposition. DeepAlign and PyMOL align are sequence dependent; we recommend using them with high sequence identity proteins. GANGSTA+ can align proteins non-sequentially and focuses on secondary structure alignment; we recommend using it on systems where it is advantageous (poorly defined loops between SSEs etc.). We recommend using TM-align and PyMOL super on lower sequence identify proteins. The PyMOL algorithms work fastest in the MADE plugin, apply them if speed is important (many chains to superimpose), and the remaining methods are, however, better documented in the scientific literature.
See this image and copyright information in PMC

Similar articles

See all similar articles

References

    1. Tainer J. A.; Roberts V. A.; Getzoff E. D. Metal-binding sites in proteins. Curr. Opin. Biotechnol. 1991, 2, 582–591. 10.1016/0958-1669(91)90084-i. - DOI - PubMed
    1. Harding M. M.; Nowicki M. W.; Walkinshaw M. D. Metals in protein structures: a review of their principal features. Crystallogr. Rev. 2010, 16, 247–302. 10.1080/0889311x.2010.485616. - DOI
    1. Chen A. Y.; Adamek R. N.; Dick B. L.; Credille C. V.; Morrison C. N.; Cohen S. M. Targeting metalloenzymes for therapeutic intervention. Chem. Rev. 2018, 119, 1323–1455. 10.1021/acs.chemrev.8b00201. - DOI - PMC - PubMed
    1. Sobolev V.; Edelman M. Web tools for predicting metal binding sites in proteins. Isr. J. Chem. 2013, 53, 166–172. 10.1002/ijch.201200084. - DOI
    1. Sobolev V.; Levy R.; Babor M.; Edelman M. Metal binding sites in proteins. Experiment. Stand. Cond. Enzyme Charact. 2012, 15, 149–159.

Publication types

-