Folds and motifs

•Using SCOP, CATH, CE to find structural homologs.

•What class/architecture, etc does your molecule fall in?

•Use CE to identify structural homologs for your protein of interest.

•Motifs: classic turn types. Extended turn types.

•Use Rasmol to find the glycines in your structure. That are the phi angles? What type of turn is it?

The SCOP database

• Contains information about classification of protein structures and within that classification, their sequences

• http://scop.berkeley.edu

SCOP classification heirarchy

(1) class

(2) fold

(3) superfamily

(4) family

(5) protein

(6) species

global characteristics (no evolutionary relation)Similar “topology” . Distant evolutionary cousins?Clear structural homologyClear sequence

homologyfunctionally identical

unique sequences

protein classes

1. all 2. all 5. multidomain (21)6. membrane (21)7. small (10)8. coiled coil (4)9. low-resolution (4)10. peptides (61)11. designed proteins (17)

number of sub-categories

possibly not complete, or erroneous

class: proteinsFolds:

TIM-barrel (22)

swivelling beta/beta/alpha domain (5)

spoIIaa-like (2)

flavodoxin-like (10)

restriction endonuclease-like (2)

ribokinase-like (2)

chelatase-like (2)

Mainly parallel beta sheets (beta-alpha-beta units)

Many folds have historical names. “TIM” barrel was first seen in TIM. These classifications are done by eye, mostly.

fold: flavodoxin-like

Superfamilies:

1.Catalase, C-terminal domain (1)

2.CheY-like (1)

3.Succinyl-CoA synthetase domains (1)

4.Flavoproteins (3)

5.Cobalamin (vitamin B12)-binding domain (1)

6.Ornithine decarboxylase N-terminal "wing" domain (1)

7.Cutinase-like (1)

8.Esterase/acetylhydrolase (2)

9.Formate/glycerate dehydrogenase catalytic domain-like (3)

10.Type II 3-dehydroquinate dehydratase (1)

3 layers, ; parallel beta-sheet of 5 strand, order 21345

Note the term: “layers”

These are not domains. No implication of structural independence.

Note how beta sheets are described: number of strands, order (N->C)

fold-level similaritycommon topological features

catalase flavodoxin

At the fold level, a common core of secondary structure is conserved. Outer secondary structure units may not be conserved.

Superfamily:Flavoproteins

Flavodoxin-related (7)

NADPH-cytochrome p450 reductase, N-terminal domain

Quinone reductase

These molecules do not superimpose well, but side-by-side you can easily see the similar topology. Sec struct’s align 1-to-1, mostly.

Family: quinone reductasesbinds FAD

NADPH quinone reductase

quinone reductase type 2

Proteins:

Different members of the same family superimpose well. At this level, a structure may be used as a molecular replacement model.

CATH

• Class• Architecture• Topology• Homology

http://www.biochem.ucl.ac.uk/bsm/cath_new/index.html

Remote homologs are more likely than close

ones

0

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

0.18

0 20 40 60 80 100

percent identity for structural homologs

likelihood

Structural conservation is

stronger than sequence conservation

The existence of large numbers of remote homologs shows us that true structural similarity is hard to see in the amino acid sequence.

beta-catenin versus clathrin (from FSSP database)

Structural homology is sometimes difficult to see

Rasmol practice

Using your protein of interest:

Find all glycines.

select gly

Label each alpha-carbon by number

label %r

Measure the phi-angle by eye (align N and CA. R-handed is positive) Write down the residue number and phi angle.

Rasmol exercises

(1) Color the ribbon drawing of 1qajA.pdb (it’s in your home directory) from the N-terminus (blue) to C-terminus (red)

(2) Divide 1qajA.pdb into domains of different color.

(3) In 1qajA.pdb, display only residues 66-87. What are the first and last residues with helical backbone angles?

(4) Find all of the atoms that H-bond to the sidechain of Ser92.

Secondary structure

Alpha helix

Right-handed Overall dipole N+->C- 3.6 residues/turn i->i+4 H-bonds

3 types of Alpha helix

ae

b

f c

g

dTwo ways to display position of sidechain on a helix.

non-polar

amphipathic

polar

For amphipathic and non-polar, sidechains line up in a favorable way.

Helix caps

Please go to http://isites.bio.rpi.edu/

Click on:Glycine alpha (Schellman) C-cap Type 1

Glycine alpha (Schellman) C-cap Type 2

Proline alpha C-cap

Frayed alpha helix

Serine alpha N-cap (capping box)

Exercise: Find a helix cap in 1QAJ. What type is it?

Capping motifs stabilize helices.

Proline helix C-capProline helix C-cap

Sequence pattern=...nppnnpp[HNYF]P[DE]n

structural variability

Locations of non-polar (magenta) and polar (green) sidechains

note: hydrophobic cluster

Pro blocks helixD or E stabilizes tight turn

“n”=non-polar“p”=polar[...]=alternative aa’s

Two Helix motifs

corner(helix-turn-helix)

EF-hand

(binds Ca2+)

corner motif

motif pattern (summary):

nnpp[nH]Gn[PDS][Px]pn

Exercise: Find a corner motif in myoglobin.

“n”=non-polar“p”=polar[...]=alternative aa’s

red=favorableblue = unfavorablegreen=neutral

backbone angles:green=psired=phi

I-sites motif(Also described by A. Efimov)

Find a corner in 1DWT

Download 1dwt from www.rcsb.orgOpen using Rasmol

Rasmol commands:

Display->Cartoons

Color->structure

select alpha

label %m

color labels yellow

Find the corner pattern (sequence and structure).

Can you see the motif in the sequence?

1 GLSDGEWQQV LNVWGKVEAD IAGHGQEVLI RLFTGHPETL EKFDKFKHLK HHHHHHH HHHHHHHHHT HHHHHHHHHH HHHHHTHHHH HTTTTTTT

51 TEAEMKASED LKKHGTVVLT ALGGILKKKG HHEAELKPLA QSHATKHKIP SHHHHHTTHH HHHHHHHHHH HHHHHHHTTT HHHHHHHH HHHHHTS

101 IKYLEFISDA IIHVLHSKHP GDFGADAQGA MTKALELFRN DIAAKYKELG HHHHHHHHHH HHHHHHHHST TSS HHHHHH HHHHHHHHHH HHHHHHHHHT

151 FQG

EF-hand

Exercise: Rasmol 1CLL. Follow instructions on next page

1CLL exercise. How does the EF-hand bind

calcium?restrict hetero and CA (comments in blue...) spacefill (this displays only calciums)click to get residue # (x = residue number)restrict within (10., x)center selected (so you can rotate it easily)Display->ball-and-stick (from menu item. you see atoms and bonds)select protein and within (10., x)Display->sticks (now protein part has think bonds)select hoh and within (10., x) ( selects nearby waters)spacefillcolor cyan (or your favorite water color...)select within (3., x) ( atoms within bonding distance)label ( these are the atoms bonded to calcium x)color labels yellow (..so you can see them)

Calcium causes a conformational change

in the EF hand.red=1CLL

green=1cfd

Create a Rasmol scriptWrite a file using jot or vi (for example: vi script.ras)Put keyboard commands in the script.Start RasmolInvoke the script using the ‘script’ command(for example: script script.ras)

Useful keyboard commands for scripts:

load xxx.pdb (loads a new PDB file)zap (closes PDB file, blanks the display)pause (stops the script until you hit a key)

Create a Rasmol script to survey many

structuresList the pdb files in the xtal directory (ls xtal/*.pdb)

Create a script file (myscript.ras) as follows:

load xtal/file1.pdb (use one of the PDB files listed)backbone 0.6color blueselect hohspacefill 1.0color redset unitcell onpausezapload xtal/file2.pdb

(repeat as before for each file listed)

Run the script: script myscript.ras

Next, change the script and run it again (i.e. change the colors to green and magenta)

Nested scripts

Tired of having to edit the script 10 times over??Write a “nested” script.

Create a file called “mynestedscript.ras”Put in all the commands except the ‘load’ line. Just one copy.The last line is ‘zap’.

Create a file called “mymainscript.ras” containing:

load file1.pdbscript mynestedscript.rasload file2.pdbscript mynestedscript.rasetc. etc. etc

Further exercises: just a few examples!

Select atoms by temperature factor: select temperature > 5000 (Rasmol uses B*100, so this selects all atoms with B > 50.00)

Find crystal contacts: select :B & within (8., :A)(This gets all atoms of chain B that are close to chain A. Your chain IDs may vary.)

Can you determine the crystal form (triclinic, orthorhombic, etc) just by looking at the unit cell? (set unitcell on)

Survey the B-factors of waters in several structures using a nested script. (Use: color temperature)

Use logic operators! Find alpha carbons within 10.A or residue x but not within 6.A. (select within (10.,x) and not within (6., x) )

Practice stereo viewing. (stereo -7 for walleyed viewing)

Proline helix C-capProline helix C-cap

Sequence pattern=...nppnnpp[HNYF]P[DE]n

structural variability

Locations of non-polar (magenta) and polar (green) sidechains

note: hydrophobic cluster

Pro blocks helixD or E stabilizes tight turn

“n”=non-polar“p”=polar[...]=alternative aa’s

Two Helix motifs

corner(helix-turn-helix)

EF-hand

(binds Ca2+)

beta-strandAntiparallel:

Parallel:

middle strand

edge strand

Sequence motifs for beta strands

Amphipathic

Found at the edges of a sheet, or when one side of the sheet is exposed to solvent (i.e. 2-layer proteins).

Found in the buried middle strands of sheets in 3-layer proteins.

Hydrophobic

Note preference for beta-branched aa’s: I,V,T

beta hairpins

Two adjacent antiparallel beta strands = a beta hairpin

Shown are “tight turns”, 2 residues in the loop region (shaded). Hairpins can have as many as 20 residues in the loop region.

hairpin motifs

“Serine beta-hairpin” (my naming). A specific pattern (DPESG) forms an incomplete alpha-helical turn 4-residues long.

“Extended Type-1 hairpin”. A type-1 “tight turn” has only 2 residues in the turn. This motif, more common than the tight turn, has an additional Pro or polar sidechain. Pattern: PDG.

diverging turn

“Diverging turns” have a Type-2 beta turn and two strands that do not pair. The consensus sequence pattern is PDG. The residue before G can be anything polar, but not a D or an N.

Exercise: Find the diverging turn in 1CSK.

Greek key motif

2 3 4 1

beta meander

One of the most common arrangements of four strands.

Two permutations. 2341 and 3214

Exercise: Download 2PLT. Find the Greek key motifs.

motifWhen a helix occurs between two strands, they are often paired in a parallel sheet.

The cross-over from one strand to the next is almost always right-handed. It is not known why this is true.

(NOTE: the cross-over is always right-handed even when the two strands are not paired!!)

TOPS diagramsbeta strand pointing up

beta strand pointing down

alpha helix

NC motif

Note R-handed crossover!

motif1 2 3

12 3

31 2

Only 3 are possible.

(with R-handed crossovers)

Efimov’s theory

A. Efimov showed that almost all protein structures can be classified as being one of 7 trees, each starting with a motif and “growing” by one secondary structure unit at time.

Does this recapitulate evolution? the folding pathway?

Domains

Proteins fold heirarchically.

secondary structure --> motifs --> subdomains-->domains --> multidomain proteins --> complexes

Most domains fold independently of the rest of the chain.

All alpha folds

3-helix bundle 1CUK 156-203

4-helix bundle 1NFN

Globin 1DWT

...many many other

alpha/beta folds

“TIM barrel” 8TIM

“Rossman fold” 1HET 176-318

many others...

Exercise: in 1HET, find the “crevice” between two motifs where NAD binding commonly occurs.

Exercise: Draw TOPS diagram for 1HET 176-318.

all beta folds

“Jelly roll” 1JMU 371-500:D, 2PLT

barrels 1EMC

propellers

beta-helixExercise: Draw TOPS diagrams for the two domains above.

Term projects

Look at the description online.

Choose a molecule to study. Check with me before you start.

Look at “Molecule of the Month” at RCSB.org for inspiration. But don’t pick one of these molecules.