nmr and optical birefringence studies of aligned pf1 …chemgroups.ucdavis.edu › ~augustine ›...
TRANSCRIPT
The measurement of chemical shifts and scalar couplings as spectral
peak positions and splittings yields primary and secondary molecular
structure information, while changes in peak intensities due to the
relaxation-based NOE provide 3D tertiary structure and quaternary
structural details (ie. binding site structure). Recently, many researchers
have shown that higher resolution molecular structures can be obtained in
liquid phase by measurement of residual dipolar couplings which can be
spectrally reintroduced using magnetically aligned phospholipid bicelles,
bacteriorhodopsin purple membranes, or bacteriophages to destroy
isotropic molecular tumbling.
Our lab has been investigating an alternate method of molecular
alignment to recover residual dipolar and quadrupolar couplings, namely
the application of strong DC and AC electric fields. This approach has the
advantages of field strength variability and alignment switching on the time
scale of multi-dimensional NMR experiments. While the techniques work
well for electrically insulating liquids using the hardware and pulse
programming highlighted in this introduction, the highly conducting aqueous
solutions of interesting biological molecules present obvious experimental
challenges. Below we present recent work involving the electrical
orientation of Pf1 bacteriophage including preliminary NMR results and
optical birefringence studies. The designs of the HV DC power supply and
HV power switch used for these studies are included.
Application of an electric field E
will partially align molecules in a
liquid due to electric dipole or
polarizability.
C6D5NO2 2H NMR B=7.04T
The E = 0 kV/cm deuterium NMR
spectrum for nitrobenzene is three
peaks (above-top). Applying a 64
kV/cm DC field partially aligns the
molecules, recovering quadrupolar
splittings and giving the six peak
spectrum shown above-bottom.
D2O2H NMR splittings in magnetically aligned Pf1Electron micrograph of Tobacco Mosaic Virus
It is well known that macromolecules like viruses, DNA fragments, and filamentous
bacteriophages align when subjected to strong magnetic or electric fields. The alignment
of these species can partially align solute molecules allowing measurement of residual
couplings as shown by A. Pardi and co-workers. This is demonstrated in the 2H NMR
spectra above-right in which magnetically aligned Pf1 phage partially align D2O giving rise
to quadrupolar splittings. The measurement of residual couplings in solute molecules in
this fashion provides a practical means by which E field alignment techniques may be
developed and experimentally fine tuned.
As shown below, while orientation in combined E and B fields is calculable using
Kirkwood's rotational diffusion formalism, many experimental conditions need to be
investigated before the NMR studies may procede.
Pulse sequences (see below) were
developed to counteract the spectral
effects of molecular motion using
continuous or, in aqueous samples,
pulsed fields to minimize heating,
electrolysis, and electrophoresis.
The sequences also make use of
multiple dimensions for simplification
of anticipated congested spectra.
Shown here are results of
nitrobenzene at 55 kV/cm.
),(QP),(P)tD(d
),(dP 2
2
ϕθ+ϕθ∇=ϕθ
+θ
ϕ∂∂ϕθ
ϕ∂∂+
θ∂∂θϕθ
θ∂∂
θ=ϕθ
2
12
D
Dcot
W),(P
Wsin),(P
sin
1
kT
1),(QP
on a table top where the condition of the
solution may be monitored, the method of
optical birefringence was employed.
0 0.5 1 1.5time sec
1 mg/mL Pf1
I/Io
arb
. uni
ts
The wave number κ of the electromagnetic wave is given by κ=2πn/λ. With
the polarizer at 45° off vertical, we can write ( )t1E as
( ) ( ) ( ) jtxktxt �cos2
E�cos2
E oo1 ω−κ+ω−κ=E
Transforming into a frame given by k ′� and j′� with k ′� at 45° from k� gives
( ) ( )( ) ( )( )jktxjktxt ′+′ω−κ+′−′ω−κ= ��cos2
E��cos
2
E oo1E
( )( ) ( )[ ]( ) ( )[ ]
′ω−κ+ω−κ
+′ω−κ+ω−κ=
⊥
⊥
jtxtx
ktxtxt
�coscos
�coscos
2
E
||
||o2E
( ) ( ) ( )
( ) ( )j
xtx
kxtx
t
′
κ−κ
ω−κ+κ+
′
κ−κ
ω−κ+κ=
⊥⊥
⊥⊥
�2
sin2
2sinE
�2
cos2
2cosE
||||o
||||o2E
Subtracting and adding the term( ) ( )
kxtx ′
κ−κ
ω−κ+κ ⊥⊥ �
2sin
2
2cos ||||
( ) ( ) ( ) ( )
( ) ( ) ( )′
ω−κ+κ+
′
ω−κ+κ
κ−κ+
′
κ−κ−′
κ−κ
ω−κ+κ=
⊥⊥⊥
⊥⊥⊥
jtx
ktxx
kx
kxtx
t
�2
2sin�
2
2cos
2sinE
�2
sin�2
cos2
2cosE
||||||o
||||||o2E
( ) ( ) ( ) ( )[ ]( ) ( )[ ( ) ]jtxktxx
kxxtxt
′ω−σκ+′ω−σκκ∆+
′κ∆−κ∆ω−σκ=
�2
2sin�2
2cos2sinE
�2sin2cos2
2cosE
o
o2E
( ) ( ) ( ) ( )[ ]( ) ( ) jtxx
kxxtxt
′ω−σκκ∆+
′κ∆−κ∆ω−σκ=
�2
cos2sinE
�2sin2cos2
2cosE
o
o3E
( ) ( ) ( ) jtxxt ′ω−σκκ∆= �2
cos2sinEo4E
( ) ( ) ( )δ−=κ∆−=κ∆∝ cos12
Ecos1
2
E2sinEI
2
o
2
o22
o xx
( )δ−= cos1I
I
o
−=δ
oI
I1arccos
( ) λ−π=κ−κ=κ∆=δ ⊥⊥ xxx )nn(2 ||||
0 400 800 1200 16000
0.5
1
1.5
microseconds
radians
350 us pulse, 1.0 mg/mL Pf1
600 1000 1400 1800 22000
0.4
0.8
1.22.0 mg/mL 1.0 mg/mL
0.5 mg/mL
volts
radians
birefringence at end of 350 µs pulse
0 500 1000 1500
0
0.5
0.97 mg/mL
0.47 mg/mL
microseconds
norm
alizedI/I o
Cooperativity observed
0 400 800 1200 1600
radians
microseconds
0
0.2
0.4
0.6
V=1069+/-14 Volts, 0.5 mg/mL Pf1
Future Work
Acknowledgements
HV Power Supply and Switch
The HV power supply and
switch diagrammed here
were designed to 1) provide
enough power so as not to
current-limit the applied
voltage and 2) switch on
timescales faster than that
cabable of a mechanical
vacuum relay.
The power supply as
shown delivers 2.1 kV DC at
500 mA continuous, 2.5 A
pulsed. The solid-state
switch is constructed of
APT1001 type MOSFETs
each capable of sustaining
1000 V DC at 11 A. As
drawn, the switch will pulse
up to 3000 V DC at 44 A and
is TTL controlled for pulse
lengths of 150 µs to 2 ms.
The TTL source may be a
pulse generator, oscilloscope
or NMR console.
D2O2H NMR splitting vs. buffer concentration
0 1 2 3 4 5
0
1
2
3
seconds
I/I oarb.units
12V DC birefringence of Pf1 in Tris/EDTA buffer
Buffer Concentration
1 mM5 mM10 mM15 mM20 mM
1 kHz birefringence of TMV in phosphate buffer
Buffer Concentration
0.01 mM0.05 mM0.10 mM0.50 mM1.00 mM
time ms
δarb.units
When linearly polarized light passes through an
anisotropic distribution of assymetric molecules like
Pf1 which has a different refractive index along its
different axes, the light becomes elliptically polarized.
This is true when an electric field is used to partially
align the molecules. Using a series of optics a zero
background measurement of this change in the light
polarization may be made. According to the
mathematics describing the electric field vector of the
light (see bottom of poster), the light observed is a
measure of the degree of anisotropy (alignment) of the
sample.
Using this optics setup, the Pf1 alignment due to
pulsed electric fields under a variety of conditions may
be measured. Results suggest DC fields to be more
effective than AC fields of various frequencies (audio
to rf) and low ionic strengths of 1- 5 mM buffer
concentrations to be favorable. (see left)
45o
E1(t)
E2(t)
E3(t)
E4(t)
Polarizer
Kerr Cell
Rhomb
Analyzer45 o
0.0
2.0
4.0
6.0
8.0
0 40 80 120
DC
2 MHz
4 MHz
7.8 MHz
15.3 MHz
19 MHz
lightintensity
Vrms or Vdc
1 mg/mL Pf1
AC versus DC Electric Fields
0 20 40 60 80
degrees from vertical
changeinechomaximum
0
50
100
150
200
250E Field Orientation
100 V applied
20 mg/mL Pf1
0 200 400 600-2000
0
2000
4000
milliseconds
20 mg/mL Pf1
Echo Train Modulation
Hz0 3 6 9 12 150
0.2
0.4
0.6
0.8
1
Echo Train Modulation20 mg/mL Pf1
0 V
100 V
Like the Tobacco Mosaic Virus (left)
Pf1 bacteriophage is rod-like but much
longer: 2000 nm long x 6 nm in diameter.
Its preparation is relatively simple with the
phage and pseudomonas host available
from ATCC. High yields of 400 mg were
obtained...not bad for physical chemists!
One such condition shown to the
right is the ionic strength of the
solution. Others include AC vs.
DC fields and pulse lengths. To
investigate these properties
Using DC power supplies on hand
capable of 1000 V at 20 mA, low buffer
concentrations, and 20 mg/mL Pf1
solutions containing 50% D2O, a series of
preliminary NMR experiments of Pf1
subjected to an E field were conducted.
The results of an attempt to measure a
change in the quadrupolar modulation of
an echo train following a 15 ms DC pulse
are shown to the left. With the E field
oriented parallel to Bo, the observed
decrease in modulation frequency is
suggestive of a lowering of order, possibly
due to turbulent flow due to conduction.
The simple echo sequence to the
lower right was used to check for a
dependence on angle between the E and
Bo fields of the echo intensity. Using the
same concentration of Pf1 and 15 ms 100
V DC pulses, the difference between the
echo maximum with field on and off was
measured for various angles. Little
change in the zero E field echo was
observed for the angles, suggesting the
behavior to be E field induced and not an
artifact of field inhomogeneity due to
sample geometry.
While an E field effect is suggested,
these rather lack-luster results prompted
us to further optimize the experiment
again making use of optical birefringence.
One attempt to further optimize
the experiment was to measure the
birefringence of Pf1 in the bore of
the 7T magnet. Using the setup
shown to the left employing a Kerr
Cell mounted to the top of an NMR
probe chassis fitted with two 90o
prisms, birefringence traces like the
one to the left were observed. Here,
no change in the light intensity is
seen during the E field pulse shown
in the lower trace. Instead, a
change is observed after the pulse
on a long time scale, suggesting the
dominant force to be not the E field
directly, but instead flow alignment
caused by turbulence in the cell
following the pulse. Such turbulence
was observed using Dante encoded
MRI (below) of the PPS sample cell
performed at 7 T using a micro-
gradient coil set.
In order to overcome the strong
orientation due to Bo in Pf1 and
achieve larger E field alignment, a
HV power supply and HV switch
were designed.
1H MRI slice of Pf1 in sample cell with Dante grid.
E=0 E=200 V/cm
Using the HV power supply and switch diagramed
to the left and using the mathematics of birefringence
to calculated the optical retardation δ, the figures to
the left were obtained. The maximum δ achievable is
π/2=1.571 which is almost obtained using this setup.
Because the retardation is also a function of path
length through the sample, it is possible that at 2.1
kV DC and 1-2 mg/mL Pf1 solutions, the alignment is
not yet saturated even though δ≈π/2. In order to
search for saturation, higher voltages should be
tested to see if δ continues to oscillate (see math).
At a Pf1 concentration of 2.0 mg/mL, such an
oscillation may be occurring above applied voltages
of 1400 V as seen on the third graph down.
Having maximized the phage alignment, we will
again turn back to NMR experiments.
We are greatly indebted to the following individuals
and grant agencies
Skipp May
Paul Feldstein
Jeff Walton
Shashi Vyas
April Weekley
NSF #CHE-9984654
The Packard Foundation
Optical Birefringence E Field NMR of Pf1 Optimizing Alignment
Pf1: A Model SystemE Field NMR: Introduction
NMR and Optical Birefringence Studies of Aligned Pf1 Bacteriophage
in AC and DC Electric Fields
Department of Chemistry University of California Davis, CA 95616
Scott A. Riley and Matthew P. Augustine
Optical Birefringence: Math