BIOC-300 Bioinformatics Mini-Project (BIMP)

Monday April 2nd, 2012

Contact

Coordinator

Dr Silvia Vidal
silvia.vidal@mcgill.ca [url]
514-398-2362

Meta-Instructions

• Write a flow sheet. Not required, but you will find that it is much easier to perform the exercises when you organize yourself - just like in the wet lab.
• Use the references/tutorials to guide you. After attending the lecture, you will want to consolidate what you have seen by looking at the step-by-step tutorials written on each tool used in these exercise. The tutorials are in Chapter 3 of the "Introduction to Bioinformatics", a manual originally written for the Bioinformatics Project in the MicroImm lab course.
• Write the report. A good report should be concise and to the point. Show what you have understood, and present it in an orderly fashion, as suggested below.

Instructions

The ABC transporter Histidine permease
with ATP, seen with Jmol.

This exercise intends to show you how to use some of the most basic tools used in applied bioinformatics. This year, we have chosen to use the ATP-binding cassette (ABC) transporter superfamily to illustrate the use of these tools. The NCBI has a free online book on ABC transporters on their website.

Part One: BLAST

The Basic Local Alignment Search Tool (BLAST) finds regions of local similarity between sequences. It is routinely used by millions of people to find sequences in extremely large databases. Typically, we say that we "BLAST" a sequence to say we "search" a sequence, without specifying where we look for it. The pool of sequences to search from is generally the NCBI sequence database, Genbank (which contained in 2005 about 60 billion bp, for 55 million sequence entries!).

Your task is to use the human MDR1 sequence and perform a BLAST to discover similar sequences stored in public databases. We want you to be able to read the BLAST output and parse information from it.

Download: [Human MDR1 sequence]

• You are asked to search the sequence against two different databases: nr and pdb. What does each database consist of, and which search do you expect to take the most time to run?
• If you look at the results from the nr database search, what is the name of the protein and species of belonging of the top hit you find? Is it consistent with the fact that we have inputted a sequence of MDR1?
• Also from the 'nr' database search results, fetch the "Taxonomy reports" and discuss how the hits are distributed among different species. What does it say about MDR1?
• What is the name, function and the species of belonging of the top hit from the search against the pdb database? Given that this is the structural database, are you surprised that the top hit wasn't from the human?
• The PDB id of the structure we use in Part Three is 1B0U. Considering the score and E-value of the BLAST hit corresponding to the sequence of '1B0U', do you think that a structure of MDR1 would resemble that of structure '1B0U'? (After you do Part Three, you will also know what protein '1B0U' is, and from which organism it comes from.)
• You don't need to print anything.

Part Two: ClustalW

ClustalW is a multiple sequence alignment program. On top of aligning sequences, it may also be used to infer phylogenetic trees (while other programs are usually preferred).

Using ClustalW, we compare different ATB-binding cassette (ABC) transporters and try to find conserved motifs. ABC transporters are known to contain sequences that are involved in the binding of the ATP and the catalysis of the ATP hydrolysis reaction, which are present in all members of the superfamily.

These are the known motif sequences of ABC in Histidine permease (PDB id: 1B0U):

Motif name	Consensus sequence	Function	Sequence in Histidine permease (1B0U)	Position in Histidine permease (1B0U)
Walker A (aka P loop)	GxxGxGKST	ATP binding	GSSGSGKS	39-46
LSGGQ (aka linker peptide or signature motif)	LSGGQxQR	ATP binding	LSGGQQQR	154-161
Walker B	hhhhD	D makes water-bridged contact with Mg++	VLLFD	174-178

x represents any amino acid, and h represents a hydrophobic amino acid. See references for a description of the single-letter amino acid code.

Download: [Sequences of ABC transporters] (Full sequence of Histidine Permease crystal ("1B0U..."), and sequences 300-600 of MDR1,3, CFTR and STE6 - they were truncated to simplify the exercise)

• Using ClustalW, perform a multiple sequence alignment on the list of sequences provided here above. See the tutorial on how to use ClustalW.
• Based on the alignment, find the Walker A, LSGGQ and Walker B motifs for sequence MDR1_HUMAN. Give the start and end position of each motif based on MDR1's numbering, and give the sequence.
• The proteins provided in this exercise are fairly diverse in function (see references, under swiss-prot). Where does it match up perfectly and why do you expect it to be so?
• (Optional: what would the alignment look like if we used full-length sequences of MDR, CFTR and STE6? Each ABC has two nucleotide-binding domains, each with its complement of Walker A, B and LSGGQ...))
• Print the alignment.

Part Three: PDB

The Protein Data Bank is a repository of structural data, consisting overwhelmingly of protein structures. Much of the data can be appreciated best using standard molecular visualization tools. Classically, RasMol was used, but several new tools have been developed recently, many of which are web-based and just require Java to run.

We will be using the structure of the first ABC transporter to have been crystallized, which is a Histidine Permease from the bacterium Salmonella typhimurium.

• Go to the PDB site and search the structure id 1B0U - it's a zero, not an O (please see the tutorial in Introduction to Bioinformatics for guidance)
• Gather the following information:
  • Authors of the structure
  • Year of publication
  • Name of the experimental method
  • Functional class of the protein
• View the structure using Jmol, and with the console, highlight the regions corresponding to the Walker A, Walker B and LSGGQ consensus regions of ABC transporters (as seen on the table in Part Two).
• Comment on the spatial location of these regions with respect to the ATP ligand (the tiny small balls and sticks molecule).
• The Walker B motif constitutes what sort of secondary structure? (Alpha helix or beta sheet) Given that the motif is four hydrophobic amino acids and one hydrophillic one (D - aspartate), how do you expect it to be arranged within the protein?
• Take the outermost atom of the aspartate (your choice, indicate which), and double-click to use the 'ruler' function. Find the distance between it and the closest amino acid of Motif A. Even if both motifs are more than a hundred amino acids apart, are you surprised by how close they actually are spatially?
• Print an image of the structure, with at least the motifs highlighted and customizations of your fancy.

Bonus: InterPro

If you thought the rest was too easy, you may venture on this bonus question. InterPro is an integration of the most important databases of protein families, domains and functional sites.

One of the first analyses one would do upon discovering a new protein would be a similarity test against databases like InterPro. Shapes in the world of proteins are finite and natural selection has produced proteins that reuse those shapes in perhaps novel permutations.

InterPro is fairly new (first version was released during Winter 1999) and integrates the most important protein databases such as Pfam, PROSITE and PIR.

InterProScan is the sequence search tool to use with InterPro. If you pass the MDR1 human sequence you got in previous sections to InterProScan, what are the conserved domains you can find? Why do you see a certain overlap between them?

Report

At least two pages to answer the questions asked here. You are encouraged to present other conclusions, ideas of further analyses, etc. The important thing is that you use the tools yourselves, and get a feel for each of them. No BLAST printout is required, but we would still like you to print the ClustalW alignment and the 3-d structure of histidine permease.

References

These reference documents were first written for the "Bioinformatics Project" in MIMM-386 course (equivalent of 300D in Micro). We adapted some of them for BIOC-300.

Introduction to Bioinformatics

You will want to use the tutorials in Section 3 primarily for performing the exercises. Sections 1 and 2 are used as background information on the databases and tools you are using, and the institutes that maintain them. A glossary is found at the end of the manual.

• Introduction to Bioinformatics for Biological Sciences Students [PDF - 2.11 Mb]

• BLAST and CLUSTALW2 tutorials [PDF - 4.5 Mb]

Presentation

Dr Vidal - April 2nd, 2012

• PDF [4.2 MB]
• PPT [46 MB]

External tutorials

• PDB tutorial in Flash

Review articles

These are a few articles proposed to us by Christian Gauthier, a PhD student at UdM. You may skip reading them from one end to the other, but going over the abstract can still give you some insights on the protein under study.

• ATP-Binding Cassette Transporter in Bacteria (Davidson and Chen, 2004 - Quite a long complete review on ABC transporters)
• Structure and mechanism of ABC transporters (Schmitt and Tampe, 2002 - Shorter than the previous)
• The ATP switch model for ABC transporters (Higgins and Linton, 2004 - more on the catalytic reaction)
• Molecular insights into the mechanism of ATP-hydrolysis... (Hanekop et al., 2006 - even more details)
• The structures of MsbA (Reyes et al., 2006 - a similar review to the previous)

Formats and conventions

• About sequence formats (including FASTA)
• Amino acid abbreviations
• The origins of the single-letter amino acid code

Swiss-Prot entries of proteins used

• MDR1_HUMAN
• MDR3_HUMAN
• STE6_YEAST
• CFTR_HUMAN