SLIDE 1
Michael Schroeder Joachim Haupt
Biotechnology Center TU Dresden
Protein Structure
Melissa Adasme
Heim et al. Chemical Society Reviews, 2010
Hierarchical structure of proteins
2
Protein Structure Michael Schroeder Joachim Haupt Melissa Adasme - - PowerPoint PPT Presentation
Protein Structure Michael Schroeder Joachim Haupt Melissa Adasme - - PowerPoint PPT Presentation
Protein Structure Michael Schroeder Joachim Haupt Melissa Adasme Biotechnology Center TU Dresden Hierarchical structure of proteins Heim et al. Chemical Society Reviews, 2010 2 wwpdb.org 3 Gorbi.irb.hr 4 Actinidin and Papain 50%
SLIDE 2
SLIDE 3
wwpdb.org 3
SLIDE 4
Gorbi.irb.hr 4
SLIDE 5
Actinidin and Papain
50% sequence ID, same structure
5
SLIDE 6
Hemoglobin and Leghemoglobin
11% sequence ID, same structure
6
SLIDE 7
Same superfamily, but not family Same family Twilight zone
7
SLIDE 8
Structure prediction
Secondary structure prediction Homology modelling Ab initio prediction Rosetta Evfold 8
SLIDE 9
Structure prediction
David Jones, UCL, 1997
9
SLIDE 10
Structure prediction Ramachandran plot
10
SLIDE 11
Rosetta
11
SLIDE 12
Rosetta
12
SLIDE 13
EVfold
Structure Prediction from Multiple Sequence Alignments
Marks et al., PLoS One, 2011
13
13
SLIDE 14
Evfold
Direct Information Statistical Model
Marks et al., PLoS One, 2011
14
14
SLIDE 15
EVfold
Examples
Marks et al., PLoS One, 2011
15
15
SLIDE 16
16
SLIDE 17
SCOP: Structural Classification of Proteins
FOLD CLASS
top
SUPERFAMILY FAMILY C1 set domains (antibody constant) V set domains (antibody variable)
All alpha All Beta Alpha/Beta Alpha+Beta
Trypsin-like serine proteases Immunoglobulin-like Transglutaminase Immunoglobulin
Distant common ancestor Low sequence similarity Closer evolutionary relationship >30 sequence identity
17
SLIDE 18
All beta All alpha
18
SLIDE 19
Alpha and beta Alpha plus beta
19
SLIDE 20
Membrane proteins
Aquaporin
1ih5 20
SLIDE 21
Hou et al. PNAS, 2003 21
SLIDE 22
Structure Alignment
22
SLIDE 23
23
SLIDE 24
G proteins (P-loop)
Conformational change
- n GTP or GDP binding
24
SLIDE 25
Citrate Synthase
1cts, 5cts 25
SLIDE 26
Lactoferrin
- iron-binding protein in secretions such as milk or tears
- Rotation of 54º upon iron-binding
1lfh,1lfg 26
SLIDE 27
Dynamic programming for structure alignment
- Sequence a and b
- Score for ai and bj = 1 / distance d(ai, bj)
- No gap penalties
27
SLIDE 28
Double Dynamic programming for structure alignment
- For all pairs ai and bj:
- Superpose ai and bj
- Dynamic programming of previous slide
- Add best aligned residues to overall alignment
- Double dynamic programming is basis of SSAP
used for CATH
28
SLIDE 29
Scoring a structural alignment
- Sequences: matching residues
- Structure: root mean square deviation (RMSD)
29
SLIDE 30
RMSD: Root Mean Square Deviation
- Distance of points a = (ax,ay) and b=(bx, by)?
30
a b
SLIDE 31
RMSD: Root Mean Square Deviation
- RMSD=average distance of aligned atoms
- Given the distances δi between n aligned atoms,
the RMSD is defined as:
31
SLIDE 32
Quality of alignment
- RMSD is measured in Ångstrom (Å)
- 1 Å = 0.1 nm
- RMSD = 0 Å: Identical structures
- RMSD < 3 Å: Similar structures
32
SLIDE 33
Pitfalls of RMSD
- RMSD is size dependent
- All atoms treated equally
(e.g. core residues less flexible than surface residues)
- Best alignment not always minimal RMSD
33
SLIDE 34
Large complexes
34
SLIDE 35
GroEL/GroES
35
SLIDE 36
Tuukkanen et al. Proteomics, 2010
Modelling large complexes: Set1
36
SLIDE 37
What is a domain?
37
SLIDE 38
What is a domain?
- Functional: independent unit
- Physiochemical: hydrophobic core
- Topological:
- Intra-domain distances of atoms are minimal,
- Inter-domain distances maximal
38
SLIDE 39
Domains are independent units
P-loop (green/orange) Winged helix DNA binding domain P-loop domain (green & orange) DNA polymerase III domain
1in5 1a5t 39
SLIDE 40
By Michael Schroeder, Biotec, 40
Domains have hydrophobic core
Kyte & Doolittle, JMB, 1982
SLIDE 41
By Michael Schroeder, Biotec, 41
Intra-domain distances minimal
- Distances between atoms
within domain are minimal
- Distances between atoms of
two different domains are maximal
SLIDE 42
Motifs
42
SLIDE 43
Interface helices: L-x-S-I-[GP]
Glutamate symporter Serine protease Cytochrome bc1
2nwl 1kb9 2ic8 Marsico et al. BMC Bioinf., 2010 43
SLIDE 44
Protein interactions
44
SLIDE 45
Families and Interfaces
- One family pair = one interface?
- No, 40% of families have alternative binding modes
- One interface = one family?
- 10% of interfaces have more than one interactor
Kim et al., PLoS CB, 2006 45
SLIDE 46
Two domains, many interface
Long chained cytokine (center) and fibronectin type III domains (peripheral)
Kim et al., PLoS CB, 2006 46
SLIDE 47
One interface, many families
Subtilisin/Chymotrypsin. Catalytic triad (Asp, His, Ser) in pink
47
SLIDE 48
Viral mimicry of native interfaces
48
SLIDE 49
By Mic hae l Sch roe der, Biot ec Henschel et al., Bioinformatics 2006
Baculovirus p35 mimics human protein
Human Caspase Human IAP Viral p35 Human Caspase
+ =
49
SLIDE 50
One interface, many families
Caspase 3, Inhibitor of Apoptosis (green), and Baculovirus p35 (yellow)
Henschel, et al. Bioinf., 2006 50
SLIDE 51
HIV Nef mimics human protein
Henschel et al., Bioinformatics 2006 Human SH3 Human HIV NEF Human SH3
+ =
51
SLIDE 52
HIV Nef mimics native PxxP motif
Nef Kinase
Henschel et al., Bioinformatics 2006
Virus Human
52
SLIDE 53
Henschel et al., Bioinformatics 2006 Human Cyclophilin Human Cyclophilin HIV capsid Human Cyclophilin
+ =
HIV capsid mimics human protein
53
SLIDE 54
Drug repositioning
54
SLIDE 55
One interface, many families
55
SLIDE 56
Heinrich et al. J Cancer Res clin Onc, 2011 56
SLIDE 57
57