Molecular_Modeling
-
Upload
ambra-salamandra -
Category
Documents
-
view
213 -
download
0
Transcript of Molecular_Modeling
-
7/27/2019 Molecular_Modeling
1/48
BasicsMolecular Modeling, lecture 1
-
7/27/2019 Molecular_Modeling
2/48
The course
Biomolecular structure Formation
Interaction
Sequence structure function
Mechanics of biomolecules
Modelling & simulation methods Analyzing computer simulation results
-
7/27/2019 Molecular_Modeling
3/48
Book
Andrew R. LeachMolecular Modelling - Principles andApplications, 2nd editionPrentice Hall 2001, ISBN 0582382106
-
7/27/2019 Molecular_Modeling
4/48
Practicalities
Lecturers:Berk Hess ([email protected])Bjrn Wallner ([email protected])
Lab exercises:Samuel Murail ([email protected])Torben Brmstrup ([email protected])
Lab reports: due one week after eachexercise
mailto:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected] -
7/27/2019 Molecular_Modeling
5/48
Schedule
Tentitative schedule is up at:
http://www.dbb.su.se/Teaching/Courses/Molecular_Modeling
exam: Friday December 17, 9:00-14:00
http://www.dbb.su.se/Teaching/Courses/Molecular_Modelinghttp://www.dbb.su.se/Teaching/Courses/Molecular_Modelinghttp://www.dbb.su.se/Teaching/Courses/Molecular_Modelinghttp://www.dbb.su.se/Teaching/Courses/Molecular_Modelinghttp://www.dbb.su.se/Teaching/Courses/Molecular_Modeling -
7/27/2019 Molecular_Modeling
6/48
Why computer simulations?
Two primary roles:
Numerical experimentsneeds accuracy
Model testing
needs reductionism
Computers are fast enoughfor numerical experiments
Most models are toocomplicated for purelytheoretical reasoning
Allen&Tildesley
-
7/27/2019 Molecular_Modeling
7/48
Molecular modeling
Molecular modeling in biomolecules:mostly numerical experiments
Aim: prediction of macroscopicproperties
Ensemble averages/static properties(binding constants, etc.)
Dynamic properties (rates, mechanisms) Molecular scale: quantum-mechanical
-
7/27/2019 Molecular_Modeling
8/48
Time scales
10-15s 10-12s 10-9s 10-6s 10-3s 100s 103s
Biological Experiments
Molecular dynamics
QM simulations(Atomic detail)
(Electrons)
Coarse-grained models
(Whole proteins)
-
7/27/2019 Molecular_Modeling
9/48
Quantum mechanical simulations
Necessary to describe: electrons & bond formation
hydrogen (sometimes)
Currently at most ~100 atoms
Usually no time dependence
Heavy atoms can be treated classicallyanyway: need to coarse-grain!
-
7/27/2019 Molecular_Modeling
10/48
Atomic-scale simulations
Coarse grained: needs force fields fromquantum mechanical simulations
Good description level for understandingindividual proteins & simple interactions
Can simulate protein dynamics up to
~10s
-
7/27/2019 Molecular_Modeling
11/48
Protein structure
d
-
7/27/2019 Molecular_Modeling
12/48
Protein dynamics
-
7/27/2019 Molecular_Modeling
13/48
Water
Most ubiquitousmolecule in life: water
Forms hydrogen bonds
Strong interaction:~20 kJ/mol, or 8.4 kBT
Completelydetermines protein
structure: responsiblefor hydrophobicity.
d h b
-
7/27/2019 Molecular_Modeling
14/48
Hydrophobicity
Cavities in water disfavored:
small cavities disrupthydrogen bond network
big cavities form surfaces:hydrophobic effect
N l i id
-
7/27/2019 Molecular_Modeling
15/48
Natural amino acids
P id
-
7/27/2019 Molecular_Modeling
16/48
Peptide structure
Backbone degrees of freedom:
Peptide () bond (trans/cis for Pro)
(C-N-CA-C)
(N-CA-C-N) torsions
Side chain degrees of freedom:
1,2,3 torsions
P id
-
7/27/2019 Molecular_Modeling
17/48
Peptide structure
Beta sheet
Alpha helix
Left-handed helix
C f i l
-
7/27/2019 Molecular_Modeling
18/48
Conformational space
How many conformations are there?
Sample , torsion in 10 degree units
36 states for each torsion
For a 100-residue chain we get:
362 states per residue
(362
)100
=36200
10308
states for the chain Only one is the native structure
L i h l d
-
7/27/2019 Molecular_Modeling
19/48
Levinthals paradox
But proteins obey the laws ofthermodynamics!
Structure must be that with the lowest
free energy
Levinthal: how can a protein do that?
We just saw that there are too manystates! Levinthals paradox
-
7/27/2019 Molecular_Modeling
20/48
B.Robson 1999
A f ldi d i
-
7/27/2019 Molecular_Modeling
21/48
Answer: folding dynamics
The answer lies in the dynamics of how
proteins fold We need to know more about protein
structure
CD S t
-
7/27/2019 Molecular_Modeling
22/48
CD Spectroscopy
Circular dichroism - chirality of amino acids will
rotate polarized light
Amount depends on the environment
Cheap, fast, simple, no sequence resolution
N l M ti R
-
7/27/2019 Molecular_Modeling
23/48
Nuclear Magnetic Resonance Environment will shift frequency of
nuclearspin resonance - chemical shifts
More complex than CD, but sequence
resolved
P t i t t
-
7/27/2019 Molecular_Modeling
24/48
Protein structure
Max Perutz & Hemoglobin
First X-raystructureTook 22 years...
-
7/27/2019 Molecular_Modeling
25/48
Kv1.2 ion channel
Large protein, 25k atoms(Rod MacKinnon)
Hierarchical structure:
Amino acid sequence
Secondary structure(sheets, helices)
Tertiary structure (1 chain)
Quaternary structure(more chains)
P t i t t
-
7/27/2019 Molecular_Modeling
26/48
Protein structure
FABP(Fatty acid binding protein)
NMR structure:
Multipleconformations
S d t t
-
7/27/2019 Molecular_Modeling
27/48
Secondary structure
Local structure is very ordered
Helices
Sheets
Turns
Stable building blocks
Paired hydrogen bonds Good local packing
No interference of side chains
H li
-
7/27/2019 Molecular_Modeling
28/48
Helices
Naturally occurring amino acid
helices are right-handed
Nomenclature: NM-helix
Residue i h-bonds to i+N M atoms per helical turn
310 helix
413 () helix - most common!
Other (very rare) forms: 27 and 516 ()helix
Heli e amples
-
7/27/2019 Molecular_Modeling
29/48
Helix examples
310 helix helix
helix
The -helix is the mostrelaxed of the helical
structures
Helices on the Ramachandran plot
-
7/27/2019 Molecular_Modeling
30/48
Helices on the Ramachandran plot
-helices occupy favorable part of diagram3.6 residues per turn (100 degrees per residue)
27
310
left
Helix dipole
-
7/27/2019 Molecular_Modeling
31/48
Helix dipole
Peptide dipoles parallel, from N to Cterminus
Strong dipole - important in some ionchannels!
Partial + charge at N, partial - charge at C
+-
Partial charges
-
7/27/2019 Molecular_Modeling
32/48
Partial charges
-0.82-0.82
-0.82
+0.41
+0.41 +0.41
+0.41
+0.41
+0.41
Effective average charge + location can
be different from unit charge:
Helix dipoles
-
7/27/2019 Molecular_Modeling
33/48
Helix dipoles
In helix: effective dipole = sum of amino
acid dipole contributions.
Beta sheets
-
7/27/2019 Molecular_Modeling
34/48
Beta sheets
Antiparallel sheets Parallel sheets
Sheet properties
-
7/27/2019 Molecular_Modeling
35/48
Sheet properties
Extended chains
H-bonds between, not
inside individual chains
Pleated sheets
Slightly twisted
-
7/27/2019 Molecular_Modeling
36/48
Pauling, Corey (and partly Branson) - 1951
The protein papers (8 papers in PNAS vol 37)
http://www.pnas.org/misc/classics1.shtml
Tight turns in sheets
http://www.pnas.org/misc/classics1.shtmlhttp://www.pnas.org/misc/classics1.shtml -
7/27/2019 Molecular_Modeling
37/48
Tight turns in sheets
Venkatachalam, 1968 (models)Simple steric repulsion
Type (i+1) (i+1) (i+2) (i+2)
I -60 -30 -90 0
I 60 30 90 0
II -60 120 80 0
II 60 -120 -80 0
IV -61 10 -53 17
VIa1 -60 120 -90 0
VIa2 -120 120 -60 0
VIb -135 135 -75 160
VIII -60 -30 -120 120
Type I Type II
Helices vs Sheets
-
7/27/2019 Molecular_Modeling
38/48
Helices vs. Sheets
Helices:
Local h-bonds
Gradual (but fast) growth
Low initiation barrier Sheets:
Non-local h-bonds
Collective interactions; all-or-nothing
High initiation barrier - very slow
formation
Amino acid properties
-
7/27/2019 Molecular_Modeling
39/48
Amino acid properties
All amino acids are not equal Proline is very rare in alpha helices
Glycine is common in tight turns
Some residues common at helix ends
Differences inside/surface of proteins
What is the cause of these differences,and can it be useful?
Amino acid properties
-
7/27/2019 Molecular_Modeling
40/48
Name 3-letter code 1-letter code Abundance G solvationGl cine GLY G 6.89%
Alanine ALA A 7.34% 1.94
Proline PRO P 5.00%Glutamic acid GLU E 6.22% -79.12
Glutamine GLN Q 3.96% -9.38
As artic acid ASP D 5.12% -80.65
As ara ine ASN N 4.57% -9.70
Serine SER S 7.38% -5.06
Histidine HIS H 2.26% -10.27/-64.13
L sine LYS K 5.81% -69.24
Ar inine ARG R 5.20% ~ -60
Threonine THR T 5.85% -4.88
Valine VAL V 6.48% 1.99
Isoleucine ILE I 5.76% 2.15
Leucine LEU L 9.36% 2.28
Metionine MET M 2.32% -1.48Phen lalanine PHE F 4.12% -0.76
T rosine TYR Y 3.25% -6.11
C steine CYS C 1.76% -1.24
Tr to han TRP W 1.34% -5.88
GLU or GLN GLX Z = E OR Q
ASP or ASN ASX B = D OR NAn amino acid XXX X kcal/mol
Amino acid properties
Amino acid properties
-
7/27/2019 Molecular_Modeling
41/48
Amino acid properties
Proline
-
7/27/2019 Molecular_Modeling
42/48
Proline
Proline: Cannot form hydrogen bonds, bulky
side-chain with two carbons connectedtothe backbone nitrogen atom
N-terminus of alpha helices
Turns Normally not inside
helices/sheets
Glycine + Alanine
-
7/27/2019 Molecular_Modeling
43/48
Glycine + Alanine
Glycine
No side chain means no clashes
Flexible ramachandran map
No entropic stabilization
Common in turns (flexible)
Alanine
Methyl side chain
Slight helix preference, but sheet ok
Hydrophobic residues
-
7/27/2019 Molecular_Modeling
44/48
Hydrophobic residues
Normally prefer beta sheets
Side chains protrude onalternating sides
More room for bulkyside chains (often h-phobic)
In particular residueswith two carbons
Polar + charged residues
-
7/27/2019 Molecular_Modeling
45/48
Polar + charged residues Polar:
Prefers turn/loop regions
H-bonds to both water andthe polypeptide chain
Charged:
Occurs on surface, in active sites
Negative charges stabilize helix N-terminus
Positive charges stabilize helix C-
terminus
Helix capping
-
7/27/2019 Molecular_Modeling
46/48
Helix capping
+ - ARGLYSHIS
ASPGLU
Charged residues act as caps for the helixdipole, which stabilizes both the helix andthe charged residue in that position
-
7/27/2019 Molecular_Modeling
47/48
The protonation state ofcharged/polar amino acidsdepends on the current pH
AA pH 7 pKa
GLU -1 4.3
ASP -1 3.9
HIS 0 or +1 6.5
LYS +1 10.5
ARG +1 12.5TYR 0 10.1
CYS 0 9.2Tricky;ver
yclosetoneutralpH
Depends
onenvironm
enttoo
Summary
-
7/27/2019 Molecular_Modeling
48/48
Summary
Amino acid properties Protein structure - aa backbone +
sidechains
More about proteins next lecture!
If needed, read up on protein structure inIntroduction to Protein Structure
Lab on proteins & molecular graphics intwo weeks (Nov 17)
Next Lecture: Wednesday afternoon