Viewing Molecular Structures

Transcription

1 Viewing Molecular Structures Proteins fulfill a wide range of biological functions which depend upon their three dimensional structures. Therefore, deciphering the structure of proteins has been the quest of scientists for decades. This workshop focuses on handling Cn3D, the NCBI 3D-Viewer for molecular structures. The topics of discussion are the databases MMDB and PDB, the concepts for viewing molecular structures (Sequence- vs. Structure Alignments) and the application, manipulation and display options in Cn3D. 1. The Methods The 3D structures of proteins can be experimentally determined at the atomic level by two different methods. a) X-Ray: Provides the highest resolution. Requires crystallization of protein and usually gives only one model of structure. May be automated in future. Electron crystallography is a method to determine the arrangement of atoms in solids using a transmission electron microscope (TEM). It can complement X-ray crystallography for studies of very small crystals (<0.1 micrometers), both inorganic, organic and proteins, such as membran proteins, that cannot easily form the large 3- dimensional crystals required for that process. b) NMR (Nuclear Magnetic Resonance): Allows structure determination of protein in solution. Variability of solution conditions possible. Provides characterization of intrinsic protein motion in solution. c) Alternatively, computer methods exist for predicting a structure from the linear sequence of the protein, however, the confidence level of the results are low. Regardless of the method used, no protein structure is properly represented by a single conformation. Every protein molecule has intrinsic motion and will shift and change based on molecular properties and environmental conditions. A true protein structure is actually a set of models and we are viewing a single snapshot of the protein in time. Example: - NMR Structure of Calmodulin (1DMO) showing 30 models vs. - X-Ray Crystallography structure (3CLN) of Calmodulin.

2 2. The Database MMDB The MMDB (Molecular Modeling DataBase) contains experimentally-determined 3-dimensional molecular structures, primarily proteins, also including nucleic acids. It is a curated subset of a database called PDB (Protein DataBank). All MMDB records have been checked, annotated, and stored in an ASN.1 format. The structure coordinates in PDB/MMDB records have been obtained experimentally by scientists using X-Ray Crystallography and NMR. The structural data in MMDB has been cross-linked with bibliographic information, the sequence databases, and the NCBI taxonomy. Entrez Structures is the section of the Entrez system. Specialized search fields for Structure include (see Limits) resolution (in angstroms) - [RESO] The experimental method used (X-Ray, NMR, etc.) - [EXPM] See the Structure Help Page for a complete list of fields available ( Note that there is a specific field for searching PDB ID numbers: [PDBACC] Practice question: Find the NMR structure for Mouse Lysozyme M. (mouse[orgn] AND NMR[EXPM] AND lysozyme M ) How do MMDB and PDB differ? PDB is included MMDB that undergoes additional curation. MMDB is integrated (interlinked) with Entrez. MMDB uses the ASN.1 format only. PDB supports several data formats. PDB can export a file in PDB or mmcif formats. MMDB exports ASN.1 as well as PDB or Kinemage. PDB files do not contain complete bond information for biopolymers. Chemical bonds must be reconstructed by the reading/viewing software based on chemistry tables and known bond rules. Since different software can interpret those rules differently, software can be inconsistent the way it draws bonds in PDB-based structures. This is because those bond rules take up a lot of space. MMDB records, however, contain explicit bonding information based on one standard residue dictionary. This means that reading/viewing software doesn't have to interpret the bonds, so the resulting structures are always consistent regardless of viewer software. 3. View and manipulate a structure Cn3D is a helper application for a web browser (ie Netscape or Internet Explorer) that allows viewing of 3-dimensional structures from NCBI's structure database. It is available for free download from the NCBI web site.

3 View a structure with Cn3D - select the appropriate viewing options e.g. Stucture View in Cn3D". (The appropriate software must be already loaded on your computer.) Practice question: View the molecular structure of HIV-1 reverse transcriptase, wild-type (1REV). o Manipulating the Structure Viewer Window Rotating and Moving (clicking and dragging; holding down the 'shift' key while clicking and dragging). Zooming (View - Zoom menu option ; using the "z" key and "x" keys; hold down the 'ctrl' key and click/drag the mouse). Highlighting (structure and sequencewindows are interactive). Restore and Reset (Restore returns the display to the last saved version of the structure. Reset will bring the entire structure back into view). The MENUs of the Structure Window File Menu View Menu Show/Hide Menu Style Menu CDD Menu Window Menu Help Menu The MENUs of the Sequence Window View Menu Edit & Justification Menus (These menus are only used when an alignment has been loaded into Cn3D). Mouse Mode Menu Import Menu (open the Import Window) The Import Window provides a wealth of options for importing, aligning, and manipulating sequences within Cn3D. Multiple sequences can be imported and then aligned using either the standard BLAST algorithm or using a new threading algorithm that combines stored structural information with sequence information to make an alignment. Once the alignments are created, they can then be moved to the sequence and structure windows for viewing and further manipulation. The Import Window Menus View Edit Mouse Mode Unaligned Justification Algorithms Alignments.

4 4. Viewing alignments Both sequence and structure alignments can be created and viewer in Cn3D. Sequence and structure alignments are different. Sequence alignments are based on BLAST comparisons of sequence identity and similarity. Structure alignments are based on VAST comparisons of structural (spatial) elements. - An individual protein may exist in more than one structural conformation based on environmental conditions or binding to other molecules. - Two proteins may have a similar sequence but different structures. - Two proteins may have a similar structure but different sequences. 4a. Sequence alignments Method 1: If your protein of interest already has a sequence record in the Entrez Protein database: 1. Retrieve your sequence record of interest. 2. Open the Links menu in the upper right corner of the protein sequence record and select Related Structure. This will retrieve a list of protein sequences that are present in 3D structure records and that were found by CBLAST (Cn3D-BLAST) to be similar to your query sequence. Method 2: If your sequence of interest is not yet available in the public database: 1. If you have a protein query sequence, open the Protein BLAST (blastp) page and Choose Search Set: Protein Data Bank proteins (pdb). If you have a nucleotide query sequence and want to compare its translation against protein sequences from resolved 3D structures, open the Translated BLAST (blastx) page and Choose Search Set: Protein Data Bank proteins (pdb) 2. Once the BLAST search results appear, scroll down to the Descriptions portion of the results and press the red "S" button beside any hit of interest to view that hit aligned to your query sequence on the CBLAST display. Method 3: Sequence alignments are created and viewed by importing sequences into the sequence viewing window. Multiple sequences can be imported, but a true multiple sequence alignment is not created. - Display the structure in Cn3D. - Select the "Show Imports" option from the Imports Menu of the Sequence Window. - Select "Import Sequences" option from the Edit Menu of the Import Window - Choose the type of import - from a FASTA file or from Entrez with a gi or accession number. Note that if the molecule has multiple protein chains, you must first select which chain will be used for the alignment.

5 - Upload the sequence either by finding the file or by entering the gi or accession number. - Once you see the sequence imported, go to the Algorithms menu and select a method of alignment. Click on the imported sequence. Once it is aligned, it should show capital letters. - Select "Merge Single" from the Alignments Menu and then click on the imported sequence. The Import Window should become empty and the alignment should be displayed in the Sequence and Structure windows.. Practice Question Align the Aquaporin 1 sequence from dog with the structure of Aquaporin I from human by using the import and align features of Cn3D (gi (Canis familiaris), ) gi (Homo sapiens) 4b. Structure Alignments and VAST Structural alignments can only be constructed and viewed if the structures of both molecules have already been experimentally determined and entered into MMDB. Structural alignments are constructed using a program called VAST (Vector Alignment Search Tool). VAST is a software program that compares three dimensional molecular structures for structural similarity. VAST comparisons are automatically done for every structure in MMDB. Note that VAST can only compare known (experimentally-determined) 3D structures. VAST does NOT do structure predictions for linear protein sequences. Practice Question: Human Aquaporin 1, PDB ID# 1H6I (MMBD ID# ) Chain A with the Crystal Structure Of The E. Coli Glycerol Facilitator (Glpf) With Substrate Glycerol, PDB ID# 1FX8 (MMDB ID# 14737) Chain A. This alignment can be generated in either of the two methods. Method 1: Viewing Structural Alignments in Cn3D 1. From MMDB's structure summary pages, lists of structure neighbors are available by clicking on the chain itself in the graphical portion of the structure summary page. The VAST-result page then presents a graphical representation of the neighboring structures sorted by VAST similarity scores. This is particularly useful when searching for related structures to a structure of interest. Method 2: Importing Structures Within Cn3D Just as the Import Window in Cn3D allows the user to import sequences, it also provides the option of importing structures and VAST structure alignments. In this case, the user chooses the structures and Cn3D queries the stored VAST results to import the specific alignment(s). The MMDB identifier number(s) for the target structure(s) must be known.

6 - Open one of the structures in Cn3D. - From the Sequence Window's Imports Menu select "Show Imports." - From the Import Window's Edit Menu, select "Import Structures /"Via Network." - Enter the MMDB identifier number for the structure to import. o If it asks which chain, select the one(s) to align. - It will create an alignment in the Import Window, but also it will provide a warning that is is necessary to save the alignment to a file and then re-open it in Cn3D. Click on "OK." - From the Import Window's Alignment Menu, select either "Merge Single" or "Merge All" depending on how many structures have been imported. - From the Structure Window's File Menu, select "Save." - From the Structure Window's File Menu, select "Open." - It should open in Cn3D and show the structural alignment. 5. Practice Question. A. View a structural alignment of the Drosophila calmodulin (4CLN) protein with the turkey troponin-c (5TNC) protein. Look at the similarities and differences between these two proteins. C. Sickle Cell vs Wildtype Hemoglobin Compare the molecular sequences and structures of Sickle Cell Hemoglobin (1HbS) and normal Hemoglobin A chain B (1HBB). Can you locate the position of the mutation in the structure? What effect does it have on the structure? Why? D. Protein Conformation Look at two forms of the same protein: the VAST alignment of 1YER with 1YES. E. Ligand Binding Compare the structure of native HIV reverse transcriptase with the same enzyme bound to a non-nucleoside reverse transcriptase inhibitor (NNRTIs), efavirenz. 6. Links - NCBI 3D Macromolecular Structures Resources ( - ExPASy Protein Structure Resources ( - Weblinks from RCSB-PDB - Linklist from the Center for Membrane Proteins in Infectious Diseases (MPID) MPG Bioinformatics Support Service Wiki on bioinformatics tools developed in the Max Planck Society N. Gaedeke Last update October 24 th, 2012