Lecture.5.12 - People.cs.uchicago.edu

Download Report

Transcript Lecture.5.12 - People.cs.uchicago.edu

Structural Bioinformatics
• Basic constraints on the structure of
gene products
• Admissible molecular phenotypes
• Disease and molecular malfunction
• Emergence of disease tied up to
evolution of complexity
Diagnosing disease at a
molecular level:
a bottom-up approach
to medicine
How can we tell malfunction at
the nanoscale?
What constitutes abnormality in
a “molecular phenotype”?
Water is nurturing, it sustains life, but it
also imposes severe constraints on
what life may be like.
These constraints become apparent at
the molecular level but have been
largely overlooked.
Imbalance  disease
Some background
The protein molecule contains polar and nonpolar
groups.
The polar groups interact in very specific ways as
the chain collapses. These interactions only prevail
in water if they are properly “wrapped” by the
nonpolar groups.
A. Fernández and H. A. Scheraga, Proceedings of the National
Academy of Sciences USA 100, 113-118 (2003)
Microenvironment of a hydrogen bond
CHn (n=1,2,3)
Carbonyl O
r
Amide N
HB
Ca
r
Ca
desolvation spheres
r=15
PDB
Code
1aa2
1bz0 (a)
1bz0 ()
1lou
1ris
1aue
256b
1ubi
1gb4
1srl
2ptl
1crc
1hhh
1mim
1ifb
1hhg
1e4j
1e4k
1gff-1
1csk-A
1c3t
C3
Q
r
 (%)
771
1450
1472
726
690
750
1182
465
240
120
222
408
1338
954
645
1404
675
699
1836
333
315
52
95
99
47
45
49
75
31
16
8
16
28
86
64
45
95
44
46
124
23
21









15.00
13.87
14.57
15.56
14.90
14.33
14.77
15.34
15.20
14.81
14.47
15.00
10.18
12.03
12.00
13.05
12.87
11.80
16.05
10.06
10.14
12.83
16.33
9.60
12.68
17.62
8.83
11.09
12.11
11.15
11.58
12.01
10.78
toxins
1fas
1f94
1jwi
171
261
300
23
25
25
7.43
10.44
12.00
17.08
22.80
23.51
prions
1dxo
1dwz
1b10
1qlx
1qm0
1qm1
645
648
699
684
648
639
59
60
58
58
57
56
10.93
10.80
12.05
11.79
11.37
11.25
21.8
24.2
21.3
19.6
20.2
21.4
Hydrogen-bond
desolvation
across the PDB
Worst wrapper (survives
through S-S bridges)
HIV-1 protease
under-wrapped HB (dehydron)
HIV-1 protease
under-wrapped HB (dehydron)
wrapping
Yint
Y
Complex name – PDB Code
We have a
“completedesolvationshell rule”.

int
10-3[Å-2]
10-3[Å-2]
HLA antigen A-2 + 2-microglobulin – 1i4f
6
36
1.58
3.01
Ig-light chain dimer – 1jvk
8
26
1.78
3.34
Transthyretin dimer – 1bm7
5
14
1.01
3.25
Insulin dimer - 6ins
5
22
2.80
4.51
HIV-1 protease dimer + inhibitor -1a30
7
12
1.87
4.71
SIV protease dimer - 1siv
4
14
1.06
2.65
Chey complex - 1fqw
4
10
1.02
5.07
Defensin dimer - 1dfn
8
14
2.72
10.01
Antitrypsin polyms. - 1d5s
14
22
1.01
2.76
Bombyxin - 1bon
4
5
0.60
3.02
FcRIII receptor + Ig - 1e4k, B-C
7
22
0.97
7.08
Colicin + ligand - lbxi
6
12
0.92
3.97
Colicin + ligand - 1emv
5
11
0.86
3.20
Serpin + ligand - 1as4
14
31
1.40
2.02
Troponin heterodimer - 1pon
6
10
1.34
4.54
MHC, antigen+receptor - 1im9, A-D
3
22
0.84
2.22
Are dehydrons relevant to
biology or artifacts resulting
from an in vitro isolation of
folding domains?
malfunction and
wrapping
hemoglobin
-subunit
-FG corner
(90,94)
(90,95)
Sickle-cell anemia mutation
Quaternary a12
interface
(5,8)
Glu6-(Phe85, Leu88) interface
Sickle-cell anemia
health
One mutation
disease
Human prion in cellular form: the most under-wrapped of all chains in PDB
cellular
Whatever stabilizes the -kernel
favors the conversion into the scrapie
form.
scrapie (hypothetical)
W WT
T
Q217V
Mouse Doppel
same fold, but different wrapping
and…
no conversion into scrapie form!
Protein-X epitope is well
wrapped
(unlike in the prion)
Given our average size
genome, where does our
complexity come from?
How is this complexity
linked to disease?
Loner
Team
myoglobin
oxygen carrier in
muscle
Being more underwrapped, our
proteins
are more
interactive.
Their structural
integrity requires
binding partners.
(But then there are
more chances
something might
go wrong)
SH3 domain
a: caenorhabditis elegans
b: homo sapiens
ubiquitin
c: escherichia coli
d: homo sapiens
hemoglobin
e: paramecium (monomer)
f: homo sapiens (tetramer)
scale-free interactome
through
domain-wrapping
analysis
mus
musculus
n= domain connectivity
escherichia coli
homo
sapiens
Disease: a prize we pay
for our complexity.
A rational approach to therapy
requires understanding complexity at
its most basic level. Wrapping might
be a key concept, since it reveals
deficiencies in the relation
with the solvent environment.
Evolution of proteomic complexity
If the protein fold is conserved,
what molecular latitude is
available to evolution?
A: minor alteration of wrapping; B: structure susceptibility is altered;
C: dehydrons conserved, new dehydrons formed concurrently with gene
duplication; D: dehydrons are not conserved; E: structural integrity compromised.
evolution
pea leghaemoglobin
human haemoglobin
disease
Sickle-cell anemia
4
3
2
1
Extent of wrapping of yeast domain folds versus the ancestry of the proteins. r-value
dispersions in an ancestry group are shown as error bars. Selected families are plotted.
Listed in decreasing dehydron density, they are: group 4: P-loop NTP hydrolases (signal
transduction), ARM repeat; group 3: protein kinases (PK), phospholipase C/P1
nucleases, class II aaRS biotin synthetases; group 2: Rossman fold domains, NAD(P)
binding, trypsin-like serine proteases, EF-hand; group 1: nucleotydyl transferases.
Molecular basis for the evolution of
proteomic complexity
Accretion of protein connections is
autocatalytic, since the rate of formation of
dehydrons is proportional to the number of preexisting dehydrons. The latter, in turn, define
the susceptibility of the structure to mutation.