Transcript Imaging RNA structures and folding intermediates using electron cryo-microscopy Tobin R. Sosnick Dept.

Imaging RNA structures and folding
intermediates using electron cryo-microscopy
Tobin R. Sosnick
Dept. of Biochemistry and Molecular Biology
Institute for Biophysical Dynamics
University of Chicago
RNA can form irregular tertiary structures
Ribosome
Sarcin Ricin loop
P RNA subdomain
Group I Intron
tRNA
How do they obtain these structures?
Kinetics --- Thermodynamics --- Structure
Only the sequence is needed!
“Mother Folding”
RNA
Primary
Sequence
(nucleotides
amino acids)
Native
Unfolded
Metal requirement: Tertiary RNAs typically require divalent cations to fold.
Two classes of cation-RNA interactions
1. non-specific binding (counter-ion condensation, Me+ or Me2+ )
2. High occupancy metal sites are observed in crystal structures of tertiary RNAs (often Me2+).
duplex
tRNA
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
P4-P6 domain
+
+
+
+
+
+
+
+
+
Me2+
+
+
+
+
+
+
+
+
+
Non-specific binding occurs at low [Me], stabilizes secondary structures
Specific binding at higher [Me2+] is involved in tertiary folding transitions.
“Mg2+ core”
Bacillus stearothermophilus Rnase P RNA
Specificity
S-domain
(~150 nt)
UG
ACG
A
200
A
CA
GA
A
UA
G
G
CU
A
160
G
P10.1
G
U
C GG
A
U
AU A
UC
GAG
C
G
U
U G
U
U
A
GA P12
U CU
GC
A
G
A
C
G
CA C
G
UC U A C
U CG
U
CU
A
G
140
A
G
AG
A U
AA A
P11
GA
G
G
A
C
A 220
A
G
A
A G CC A
G
C
A
A A GGCU G C
C
U G GC G
U U C U GA C
C U P10 240
G
U
P15 A A A U
P5
CG
G
P9
A G C GA G A A A C C C
U
120
G CGU
G
U
A
C
C
G G G 260
P8 U U P7 G U U C G C U C
U
A
G
G
G
C A
U
A
U A
60
UG
280
G
G G C
A
U
AG G
A C GA
C
80 C G
A
A P15.1
A
A
A
G
U
C G
C CU U C U U
C
U
U
A
A
G
C
P5.1 C
U
CU GA
G C
G GA A G A G
U
C
G
AC
U
U
300
A
A
AG
G
100
A
G
A
GCA
A
UG
GA
AU
A
A UC
G
U
A
G
P18 U U C
G
A
G
A
U
40
G U
P3
A
A
G
U
GA A U C U GU AG
340 G
G
GA
U
G
A 320
A
C
U C U A GA C GUC
G
CG
U
G
U
C
C G P4
A
U
20 U A
U UG
U
A U
A G A AC G
P2 A U
G
360
C
1
U G
U
G P19
P1
G C
C
G U U C U U AA C GU U C G G C GC G UA
G
A A CA
A GA U U GC AA G A
C
400 C
A
A
C
U U C GGU A C AA
U UCA
Catalytic
C-domain
(~255 nt)
380
Pace and coworkers, PNAS 2005
Mg2+-induced equilibrium folding of C-domain of P RNA
UI
K Mg
K IN
Mg
6.0
CD(260)
5.0
U
N
I
4.0
Abs (260)
2.5
2.4
2.3
2.2
At low [Me2+], non-specific
(and specific) binding,
3.0
CD(287)
1.7
1.6
U1
U2
|
Ui
1.5
1.4
1.3
UI
K Mg
Ieq
K IN
Mg
N
At high [Me2+], energetics of
dominated by specific
binding
fluor.
0.60
+n
0.55
0.50
0.45
0.01
Mg
0.1
2+
1
(m M )
Fang, Pan & Sosnick
Biochemistry, 1999
Mg2+
Summary of C-domain folding
 1msec collapse, 2o +some 3o,
non-specific and some specific Mg2+
Uurea
0% surface buried
Ieq
85%
buried
fast
I2k
I1k
98 %
buried
Rg = 180 Å
Limiting step
Consolidation of specific
metal binding site
98%
buried
fast
N
100%
buried
Rg = 46 Å
Rg = 39 Å
Rg = 38 Å
Fast, local conformational search
Limited by Me2+ binding site formation: Defines tertiary structure
Fang, Pan & Sosnick, PNAS, 2002
Structural characterization of intermediates
U
Mg2+
I
?
Mg2+
N
B. subtilis Specificity domain of RNase P RNA
Krasilnikov et al, Nature 2003
native
Site-resolved information
from chemical and
nuclease cleavage
U to I transition
U
C
U
G
A C
U
C
C
G
A
A
A
140 A
A
P10.1
P7
5’
86
g C G a G
3’ 239c G C U C
A
100
G A
G C G A U C C
U
 

C
A U A A G C U A G G
P8
V1
U
C
U
160
 A
G U
G
C
U
 G
 A
G
 U
 A
U
C
C
U

A
A
U
A
G
A
U
C
A
200 C
U
C
U
A
A
U
G
G
U
G
A
G
A
G
C G U C G G A G
A


U
G C A G U C U
U
120
P11
P9
V1
G
C

A
P10
Ieq contains J11/12
module & fourway junction
T1
KE
P7
5’
86
g C G a G
3’ 239c G C U C
A
100
G A
G C G A U C C
U
 

C
A U A A G C U A G G
P8
DMS
160
 A
G U
G
C
U
 G
 A
G
 U
 A
U
C
C
U
A
TL-receptor
G
G
A
 U
A
G
G
A
Core
G
G220
A
A
C
U
A
C
A
J11/12
module
U
G
C A U G G
A A
P12
G
C A
C 180
G G
G U G C C

P10.1
J12/11
A
A
A

G
G
C
U
G
A C
U
C
C
G
A
A
A
140 A
A
A
G
J11/12
C C
C
U
G
G
A
 U
A
G
G
A
Four-way
junction P10
I to N transition
U
A
G
A
U
C
A
200 C
U
C
U
A
U
G
G
U
G
A
G
A
P12
J12/11
A
G
G
C A
U
G
G220
A
A
C
G
G
U
A
G
C 180
G G
G U G C C

A

G
A
J11/12
C C
A
A
C
A
G
C
C A U G G
A A
C G U C G G A G
A


U
G C A G U C U
U
120
P11
P9
DEPC
P(r) – general shape, from small-angle X-ray scattering
Experimental SAXS
Crystal Structure
0.020
Rg ~ 30 Å
P(r)
0.015
0.010
0.005
0.000
0
40
80
120
r (Å)
Rg – overall size
APS, BioCat beamline
Start from the crystal structure…
Modeling the intermediate with experimental constraints
(SAXS, Nuclease and chemical mapping)
Disrupt TL-receptor
interaction
Rg ~ 2 Å
With Eric Westhof
Baird et al. JMB 2005
Ieq
N
Rotate P10.1 further
Rg ~ 5 Å
Rg ~ 8 Å
Rotate both P10.1
and P12 arm
Intermediate Structure
Ieq
N
Ieq
Role of metal ions in
folding cooperativity
N
Direct mediation of
long-range
contacts
Indirect mediation –
metal binding coupled
to large-scale
conformational change
S-domain
U
Ieq
N
Intermediate Structure (model)
Large molecules light up on EM
But nothing beats a real picture…
• Ribosome – 3 MD
Glaeser group, UC Berkeley,
data from Frank lab, Wadsworth Center
• Proteasome – 750 kD
Hu et al, Mol. Microbiology, 2006
• Generally accepted lower limit ~200 kD
Direct imaging of ‘small’ RNAs
W. Chiu & S. Ludtke, Baylor
Start with molecules with known crystal structures to assess feasibility
82 kD
Catalytic domain RNase P RNA
Side
Top
(blind reconstructions)
Front
Ieq reconstruction
When the intermediate is
stably populated it can be
directly imaged!
Native S-domain
I25 comparison with reconstruction
Model vrs SAXS & CryEM reconstructions
Model
CryoEM
SAXS
Add an extension to enhance image
Native
Ieq with P9ext
CD278
3.4
3.2
3.0
2.8
2.6
Folding behavior unchanged
CD260
5.1
4.8
4.5
4.2
0.01
0.1
1
[MgCl2] (mM)
10
Ieq with P9ext
Front view
Top view
Why doesn’t flexibility blur the image?
P9
P9
P10.1
P10.1
Ieq dimensions: no salt dependence below 1 M NaCl
42
40
Not just electrostatics holding
Ieq in an extended state:
Defined thermodynamic well
Rg (A)
38
36
34
2+
Mg -native
Structure in the core?
32
0.01
0.1
[NaCl] M
1
All-atom
simulations
RNA structure determination using
CryoEM + modeling + all-atom simulations:
A rapid alternative to crystallography?
.
200
UGACGAAG CACUAGA
CGGA A160
UCU UGGUAA
AG 190
UUC U U UGAG
G
C
U170 GA
UG GGC
AGA 210
A
A U AP11
G P12
150CAUCCG
C
C
U
C
C
A
G
U U G
A
220U
G G
G
P10.1AAAAAGGA180
C CAU CGCAAG
G
140 130
P10
230
C
G
AGAGGCUGAC C
P7
UUUCUGACG
G C CGA P5
120 P9
110AGUC90 GCU
GU
UA
C G P8 U
GAU
AAG C
G
U
A
A GA
CU
.
.
.
.
100
modeling
All-atom
simulations
Thoughts:
Include “folding” to help restrict the conformational search in prediction
What about folding cooperativity?
• Tao Pan
Acknowledgments
– Nathan Baird
– Haipeng Gong
– Shahnawaz Zaheer
• Wah Chiu, Steve Ludtke (Baylor, CryoEM)
• Eric Westhof (Strasbourg, modeling)
• Karl Freed (Univ. of Chicago, simulations)
NIH