DNA charge transport: correlation with
pathogenic mutations
Stephen A. Wells
(s.a.wells@warwick.ac.uk)
Rudolph A. Roemer, Chi-tin Shih
DNA (Deoxyribonucleicacid)
Linear bio-polymer,
backbone of repeated
sugar-phosphate units,
with paired “bases”
•C ytosine
•A denine
•T hymine
complementary
•G uanine
DNA & electronic transport
Conductor:
[Fink/Schoenenberger,
Nature 398, 407 (1999)]
Semiconductor:
[Porath et al., Nature 403,
635 - 638 (10 Feb 2000)]
Insulator:
[Priyadarshy
et Biochem.
al., J. Phys.71,
B. Giese,
Annu. Rev.
Chem. 100, 17678 51
(1996)]
(2002)
Combining DNA & electronics
Xuefeng Guo et al., Nature Nanotech. 2008
Charge transport and DNA repair
BER (base excision repair) enzyme with [Fe4S4]2+ cluster –
robust to oxidation in the absence of DNA
BER binding to DNA – oxidation activated ([Fe4S4]2+
→
→[Fe4S4]3+)
E. Yavin et al. (JK Barton group), PNAS 103, 3610 (2006).
Tight-binding models
ei-1
ti-1,i
ei
ti,i+1
i-1
A
G
C
T
C
G
ei+1
Slice i
i+1
Tight-binding model parameters
Use ionization
potentials for
onsite energies:
 G  7 .7 5 e V
 C  8 .8 7 e V
 A  8 .2 4 e V
 T  9 .1 4 e V
Phenomenology/guessing for transfer parameters:
Example for 1-D ladder: hopping=0.4 eV
For
ladder model:
“like” hopping 0.35 eV,
“unlike” hopping 0.17 eV
interchain hopping 0.1 eV
Transfer matrix method to extract localisation lengths, Lyapunov
exponents, charge transmission.
Biology
• Mutations in an important oncogene,
p53
• “Guardian of the genome”
• “Point Mutations Effects on Charge
Transport Properties of the TumorSuppressor Gene p53”; Chi-tin Shih et
al., PRL 2008.
Mutation “hotspots” correlate with low charge
transmission.
T(E) in a section of p53 for native and mutant sequences:
T(E)
T
D(E)
E
(DT)2
D2
E
Squared difference in charge transmission for mutant
sequences:
Hotspots
Small CT change
Diagonal model
Ladder model without diagonal
hopping includes an
unphysically large term for
hopping across the hydrogen
bond.
Better to explicitly include
diagonal terms.
We use: 0.1 eV purine-purine
0.01 eV purinepyrimidine
0.001 eV pyr-pyr
Mutation frequency versus change-squared measure:
80
70
60
50
40
30
20
10
0
0
200
400
600
800
1000
1200
T(E)
T
D(E)
E
DT
D
E
Mutation frequency versus linear change measure:
50
40
Change measure B (arb. units)
30
20
10
0
0
-10
-20
-30
-40
-50
200
400
600
800
1000
Mutation frequency in P53
1200
Really?
• The eye is notoriously good at spotting nonexistent patterns.
• Check: variance of distribution with and
without weighting by frequency.
• Unweighted : var 85.9027
• Weighted : var 66.686
Mutation frequency versus linear change measure in 1-D
model:
An interesting way to break the
transfer-matrix method
Set along-strand and diagonal elements all equal:
1
1
1
1
Transfer matrix involves inverse of a singular
matrix:
1
1
1
1
Intepretation 1:
• Carcinogenesis is intimately related to DNA
charge transport and we are modelling it.
• Problems:
– Unrealistic model parameters
– Model of unreal situation!
Intepretation 2:
• Whether a mutation appears in a genetic
disease depends on: the likelihood of DNA
damage, the likelihood of damage detection
and repair, and the effect of the mutation:
coding, regulatory.
• These depend on sequence and our models are
probing the properties of the sequence.
• We derive conductance from localisation
lengths: is charge localisation the important
factor?
Outlook
• Search for more genes with good statistics on
mutation frequency and disease
• Improve model parameters: parameters for
bioinformatics may differ from those for
physical CT
• Medically useful predictions: identify high-risk
mutations for screening.
• Understanding: what is CT model telling us
about sequences?
References
•S.A. Wells, C.T. Shih and R.A. Roemer,
Proceedings of the 32nd International Workshop on
Condensed Matter Theories,
International Journal of Modern Physics B, in
press.
•C.T. Shih, S. Roche, R.A. Roemer, Phys. Rev.
Lett. 100, 018105 (2008)
•S. Roche, D. Bicout, E. Macia, E. Kats, Phys.
Rev. Lett. 92, 109901 (2004)
•S. Roche and E. Macia, Modern Physics Letters
B 18, 847 (2004)
Acknowledgements
• Leverhulme trust for funding
• Your attention.
Modelling rigidity and flexibility in
proteins
Dr. Stephen Wells
Physics/CSC, University of Warwick
s.a.wells@warwick.ac.uk
Overview
•Rigidity in protein structures: hydrogen bonds,
rigid clusters, loss of rigidity
•Protein flexibility: simplified simulations based on
rigid clusters
•Random and directed motions: morphs and
conformational change.
Main-chain rigidity
On hydrogen bonds
“FIRST” software: see “Flexweb.asu.edu”
D.J. Jacobs, A.J. Rader, M.F. Thorpe, and L.A.
Kuhn (2001) Protein Flexibility Predictions using
Graph Theory. Proteins, 44, 150-165.
Hydrogen-bond dilution plots
B.M. Hespenheide, A.J. Rader, M.F. Thorpe and L.A. Kuhn (2002)
Identifying Protein Folding Cores: Observing the Evolution of Rigid
and Flexible Regions During Unfolding. J. Mol. Graph. & Model., 21,
195 207
Proteins and glasses
Protein folding cores
A.J. Rader, B.M. Hespenheide, L.A. Kuhn and M.F. Thorpe
(2002) Protein Unfolding: Rigidity Lost. Proc. Natl. Acad Sci.,
99, 3540-3545.
Extent of mainchain rigidity during
dilution
Horse cytochrome C
Two patterns of
rigidity loss
From rigidity to flexibility
• Rigid clusters provide a basis to simulate
flexible motion
• Motivations:
– Connect static (crystal) and dynamic (NMR)
structures
– Explore large-scale motions
• Cheap and cheerful
Rigidity analysis: barnase
Random motion of barnase
“Framework Rigidity Optimised Dynamic Algorithm” (FRODA)
S A Wells, S Menor, B Hespenheide and M F Thorpe
“Constrained geometric simulation of diffusive motion in proteins.”
Physical Biology 2 S127-S136 (2005).
Comparison to NMR
NMR
X-ray + FRODA
Comparison to NMR
• Comparing globally-aligned RMSD by residue for the FRODA
ensemble and the 1BNR NMR ensemble.
• FRODA: tens of thousands of conformers in tens of minutes
3
NMR
FRODA_long
RMSD(A), ensemble
2.5
2
1.5
1
0.5
0
0
20
40
60
Residue (barnase)
80
100
120
“Morphs”
Special issue (10 April 2009)
illustrated with calmodulin
morph from Yale Morph Server,
PyMOL graphics.
ADK with rigidity
Targeted morph between two known crystal forms
GroEL: cryo-EM fitting
Jolley et al.
Biophysical Journal, 2008
ADK collapse
Inward bias
Ongoing project
• We’ve just begun a collaboration with Mike
Payne to use geometric simulation combined
with ab initio (ONETEP) simulation.
• Generate large-scale motion by biased
geometric simulation (expansion/contraction;
morph; low-frequency eigenmode), then
refine structures and evaluate energy
landscape using ONETEP.
Disulphide isomerase
References

FIRST/FRODA: http://flexweb.asu.edu (Thorpe et al.)

ElNeMo: http://www.igs.cnrs-mrs.fr/elnemo/ (Tirion, Tama, Sanejouand et al.)

CC Jolley, SA Wells, P Fromme and MF Thorpe "Fitting low-resolution cryo-EM maps of proteins
using constrained geometric simulations." Biophysical Journal 94, 1613-1621, 2008

CC Jolley, SA Wells, BM Hespenheide, MF Thorpe and P Fromme "Docking of Photosystem I subunit
C using a constrained geometric simulation." JACS 128, 8803-8812 (2006).

SA Wells, S Menor, BM Hespenheide and MF Thorpe "Constrained geometric simulation of diffusive
motion in proteins.“ Physical Biology 2 S127-S136 (2005).

“Molecular docking and spatial coarse graining simulations as tools to investigate substrate
recognition, enhancer binding and conformational transitions in indoleamine-2,3-dioxygenase
(IDO)”, A Macchiarulo, R Nuti, D Bellochi, E Camaioni, R Pellicciari, Biochimica et Biophysica Acta
1774 1058-1068 (2007).

SA Wells, JE Jimenez and RA Roemer "Comparative analysis of rigidity across protein families“
Physical Biology (accepted August 2009)
Acknowledgements
•
•
•
•
•
•
Mike Thorpe, ASU
Rudo Roemer and Emilio Jimenez, CSC
Robert Freedman & David Roper, Warwick bio.
Mike Payne, William Bellfield, Cambridge
Leverhulme, for funding
Thank you for your attention