|II.B. β structure
III.C. Parallel α / β Domains
|(1981) The Anatomy and Taxonomy of Protein Structure Adv. in Prot. Chem. 34:167-339.
Parts for this lesson: II.B. β Structure, and III.C. Parallel α/β Domains.
A general introduction to singly-wound parallel α/β barrels:
|(1989) Structural Principles of α/β Barrel Proteins: The Packing of the Interior of the Sheet Proteins: Structure, Function, and Genetics 5:139-148|
Overview of families of α/β loops:
|(1987) Structural and sequence patterns in the loops of βαβ units. Protein Engineering 1: 173-181|
1. Kinemage file abBarrel.kin.gz (20KB)
Run through the animation of the triose P isomerase (TIM) barrel in Kin.1, to get a feeling for why it is described as "singly-wound". This is the first such barrel that was found, but by now there are hundreds of them known.
Compare that TIM barrel with the two alpha / beta barrel domains of IGPS-PRAI in Kin.2.
How many beta strands are in each of the barrels? _________, _________, _________
Are all the beta strands connected by +1x crossovers? _________
Does every crossover connection contain a helix? _________
In file abBarrel.kin, kinemage 3 shows the type of interior packing typical of alpha / beta barrels. There are always at least three layers of hydrophobic sidechains, each layer perpendicular to the axis of the barrel, and sometimes there are additional hydrophilic layers on each end, as here.
What 4 residue types make up the central layer (layer 2) here?
_________, _________, _________, _________
Why are there only 4 sidechains in each layer, rather than 8?
2. Kinemage file baBarrel.kin.gz (16KB)
Main chain, H-bonds, and sequence labels are shown. This is an idealized parallel alpha / beta-barrel modeled on TIM (triose P isomerase), PYK (pyruvate kinase), KGA (KDPG aldolase), TAA (taka-amylase), and GAO (glycolate oxidase). A very large number of parallel α / β-barrel proteins are known, all with 8 strands in the same singly-wound topology (with some minor exceptions). In this model structure all residues have the same phi, psi. (-114°, 124°). Strands were initially fit to a real barrel and then idealized (symmetrized) as much as possible. (Departures from symmetry vary, some barrels being flattened and others round but conical.)
The above diagram shows the H-bonding pattern in babarel, as viewed from the outside and flattened into 2D. The H-bonds are, on average, perpendicular to the strands, so the strand twist gives the H-bond direction a twist, or effective offset from one strand to the next. Turn on & off "row perp." to see balls on the Cbetas for one row of 6 inward-pointing sidechains that are all adjacent in the direction perpendicular to the strands.
What is the total offset or shear (in residues) if one follows the perpendicular average H-bond direction all the way around the barrel back to the starting strand? _________
What is the offset per strand? _________
Turn on the side chains. Why do they look so bristly?
On the H-bond diagram above, the dots represent alpha-carbons.
Circle the ones whose side chains extend toward the inside.
As you saw in Kin.3 of abBarrel.kin, draw lines to connect the Calphas for each of the two central layers of 4 internal sidechains.
Now find the two flanking layers of 4, and connect their Calphas. Do they also look fairly regular? __________
The four layers are made up of sidechain types: _____, _____, _____, and _____.
Considering conformation, H-bonding, and side chain direction, what is the rotational symmetry of this structure around the central axis of the barrel? (That is, is it 2-fold, 4-fold, 8-fold?) ___________________________
3. Kinemage file 1tim_a.kin.gz (32KB)
To start with, turn on just main chain and H-bonds, to compare this real barrel with the idealized babarel. Check that there are really 8 strands, that each connection is righthanded, and that they each move over by a single strand and all in the same direction.
Looking from the side of the barrel (view 2), estimate the angle (twist) between the strands in front and the ones in back. ______________
Turn end-on to the barrel (view 1), so that in projection you see a central ellipse of beta strands surrounded by a ring of helices. The distance from one Calpha to the next is always a constant 3.8 Å, so that can be used as a yardstick to compare other distances.
Measure the flattening of the barrel cross-section by estimating its major and minor axes in Calpha-Calpha units.
______________ : ______________
Turn on side chains. The table below gives the composition of residue types on the inside surfaces of 3 singly-wound parallel beta barrels. What are the five commonest residue types here? (circle them)
Large hydrophobics predominate, of course; however, one curious feature is the large number of glycines, which in general are found in turns and loops rather than regular secondary structures. Turn off "non-gly" to see the glycines. Glycines have 3 unusual roles: 1) Flexibility; 2) Adoption of forbidden (usually positive phi) conformations; and 3) Smallest side chain to fit in tight positions. In this case, however, flexibility is unlikely, the Gly phi,psi values are more or less normal beta, and in many cases there are holes inside the barrel next to the Gly.
Turn "non-gly" back on, and see if you can find one of these holes, next to Gly ____.
From the babarel model building, we believe the function of these glycines is to relieve a bad contact between the internal Cbeta and a CO on the neighboring strand.
Referring to the Thornton paper, locate an alpha-beta connection in TIM that belongs to the ab1 family (at about residue ____ ) and one that belongs to the ab3 family (near ____ ), and find the defining conformational and sequence features in them.
Why are none of the other connections in families? ______________ .
4. Use a browser to visit the SCOP site, http://www.bio.cam.ac.uk/scop/
Enter the classification at the top of the heirarchy, go to class "3. Alpha and beta (a/b)" and then to the "1. TIM a/b barrel" fold. Use the next-to-right top icon to do one level of outline expansion, to view down to the "family" level.
As of the June 2009 release 1.75, SCOP listed 33 superfamilies and 103 homologous families that have this fold. Most are enzymes and some entire metabolic pathways are populated with TIM barrels.
In the original 1981 Anatax, KDPG aldolase had a proposed but unconfirmed retracing as a classic TIM barrel. Find an E. coli KDPG aldolase structure in superfamily 11, under class I aldolases, and download the PDB file. What is its PDB code? _______ What is the resolution? ______Å (look in the PDB file header) Import it into KiNG and make a ribbon kinemage of chain A (refer to the kinemage-construction tutorial if needed). How many β strands are in the barrel? ____ Do they all have +1x righthanded crossovers? _____ Does it match the ribbon schematic in Fig. 75 of Anatax? ______ (If not, in what way?)
Pick one of the specific TIM barrel PDB files in superfamilies 12-17 to download: what is its PDB code? ______ what is the protein's name? ___________________________________.
Use KiNG to create and look at a simple kinemage of it. What aspects of this structure are classic and what is atypical for a TIM barrel? _________________
[If you don't find your initial structure choice interesting or tractable, try another one.]
Imagine that you are making a kinemage to add to this section of Anatax, to illustrate the idea of the more varied examples of TIM barrel folds now seen (or, if your example is not very different, make a kin illustrating TIM as the classic example). Make a fairly simple kinemage of your chosen structure for this purpose (ribbon, ball&stick, or whatever), with a clear overview and at least one view that shows some notable feature, and write a short explanatory caption about the feature(s) into the kinemage text window. If appropriate, provide button control to turn on or off elements of your kinemage to aid the reader in understanding your caption.
Email your kin to the TA and write the submitted filename here ___________________________.