A
'
/A
A
/
0
/
0
$ig. +" absorption and phase shift
<<
0
" the mass follows the agitator almost instantaneously 0*
(
0
" resonance ( - -#*0
>>
0
" the agitation is so fast that the mass cannot follow0 and the
amplitude A
'
becomes almost 1ero2 ( %.#* There is no
energy transfer between agitator and resonator.
The optical equivalent to the mechanical resonance (
0
is when
the eigenfrequence of the electrons is equal to the frequency of the
incident light0 this is an absorption line in an absorption! spectrum. The
energy
h E
%!
is used to raise the energy of the electronic system correspondingly
e&citation!. There is also a phase shift due to the fact that the oscillating
electrons act as a 3ert1ian oscillator emitting secondary radiation. This
secondary radiation suffers from a delay with respect of the e&citation
since the phase shift is negative. The propagating wave in the matter0
now is a superposition of all secondary waves 3uygens4 principle! and
the velocity of the light in matter
m
is smaller than the velocity of light in
)
resonance curve
phase shift
vacuum c. The quotient is termed refractive index n of the matter and is
correlated with the dielectric constant of the matter"
m
c
n
)!
with
c
respectively
m
( +!
n is a function of the wave length respectively the frequency . This
phenomenon is called dispersion e. g. of light by a prism!. The
dispersion curve decreases usually with but in the vicinity of an
absorption band anormal dispersion is observed and the refractive
inde& increases with the wave length.
Resonance and phase delay result in absorption and
refraction.
$ig. 5" absorption- and dispersion curve
Although it might look like that0 the dispersion curve is not the first
derivative of the absorption curve. 6ven far away from the absorption
ma&imum there is always still a wave-length dependent refractive inde&
see a quart1 prism! even in the completely transparent region.
+
A
n
absorption curve
dispersion curve
In optically anisotropic systems0 e. g. some crystals like calcite0 in
oriented polymer films or oriented liquids flow-birefringence! or solutions
with oriented molecules e. g." lyotropic solutions in a shear 0 electric or
magnetic field! also an anisotropy of physical properties e&pansivity0
heat conductivity7! is observed. Correspondingly0 refractive inde& and
absorption become directional.
The e&tinction coefficient molar or specific! is different whether
the e&tinction E eq. 5! is measured parallel
11
! or orthogonal
! linear
dichroism! to the draw-direction of an unia&ially drawn! film and also n
11
n
0
respectively. If the mica plate has a thickness of d #0#+. mm then the
two beams will show a phase delay of /5 /5-plate!. The superposition
of these two sine-waves phase-shifted against each other by /5 creates
a circular polari1ed wave after the /5-plate0 see fig 8.
$ig. 8" generation of circular polarised light with a /5-plate.
Is the vector of the incident beam E
,
_
0
ln
5!
with I
#
the intensity of the incident beam and I the intensity of the light
after the sample0 ! is the e&tinction coefficient that is either the
specific quantity when the concentration is the mass density Ag/LB the
or the molar quantity mass! concentration c Amol/LB0 and d is the length
of the sample cell. Csually0 the difference"
E() E
L
() - E
'
() ( d A
L
()
'
()B ( d () 8!
is determined and the ellipticity molar or specific! is calculated.
The ellipticity is e&pressed in terms of () and has the following
geometrical e&planation0 see fig. . and fig. -.
c
Also DabsorbanceD0 A. 3owever0 the term e&tinction may be better" attenuation! usually also
considers effects of luminescence and scattering.
,
$
ig. ." linear polari1ed light as the sum of left and right circular polarised
light of the same intensity without phase delay left!. Ehen there is a
phase delay and a difference in intensity0 then elliptically polari1ed light
results as the sum right!. Fiven actual values in C:-spectroscopy0 the
ellipse would still appear to be as thin as a single line given the scaling
above.
?
$ig. -" geometrical e&planation of the ellipticity
Csually the ellipticity is given by"
( )[ ] rad E E
R L r
4
303 . 2
respectively
( ) [ ] deg
180
4
303 . 2
R L d
E E
,!
d
The difference in e&tinction is usually very small %#
-)
G7%#
-+
G! but can
be determined very accurately. And the results are given in mdeg %#
-+
deg ( %#
-+
*!.
The values of are frequently normali1ed. In protein chemistry this
is the mean molar ellipticity per residue per amino acid!0
mrd
.
d
rad better" radian because the unit DradD is the non-unit of absorbed radiation! is the unit of the plane
angle0 based on the unit circle to give the angle in radian measure. rad ( %.#* %.# deg0
respectively %.#*!0 % rad 8?0)-87* $requently0 DradD is dropped0 so that ( %.#*0 /2 ( -#*0 etc.
.
?!
M ( molar mass0 c ( molar concentration0 d ( length of the sample cell0
N
r
( number of residues amino acids!. The value of
mrd
is in practice
usually of the order of %#
5
.
In general C:-measurements are useful for the following problems"
:etermining if a protein is folded and gain information about the
secondary and tertiary structure0 see fig. %#. This enables to iden-
tify conformational changes during processes0 comparison with
mutants or proteins from different sources
<tudy conformational changes under stress such as p30 heat0
denaturants etc0 see fig. %%0 see also DThe Increment =ethodD.
Interactions with ligands drugs0 other proteins0 lipids7! that
change the conformation
:etermination of the influence of the solvent conditions on the ther-
mal reversible folding/unfolding.
$ig. %#" e&le for a standard curve showing a standard curve of
polyL-lysine! in -helical conformation %##G! at p3 %#..0 -sheet
conformation %##G! at p3 %%.%0 heating to 8?*C and re-cooling0 and in
-
random coil conformation %##G! at p3 ?. $or analysis of a spectrum
that contains all these structures deconvolution is required.
$ig. %%" influence of trifluoro ethanol T$6! content on the conformation
of the protein AHI-.
The protein AHI- consists in aqueous solution of 8-G random coil and
5% G helical conformation per molecule!. Addition of T$6 increases the
helical content0 see fig. %#0 until at 5#G T$6 a saturation is reached with
?%G helical conformation and )-G coil. The content of the different
conformations can be determined by curve fitting.
The secondary structure can be investigated in the far-CJ region
%-# nm7)8# nm! where signals are caused by a regular folding of the
peptide bond which is the chromophore!0 see figs - and %#. Although
the conformational contribution of a certain secondary structure can be
determined0 it is not possible to specify which particular residues are
involved.
%#
The near-CJ region )8# nm7+8# nm! can serve to investigate
certain aspects of the tertiary structure of proteins. In this region the
chromophores are aromatic amino acids phenylalanine )8# nm-)?# nm0
)?# nm - )-# nm tyrosine0 ).# nm - ++# nm tryptophan! and disulphide
bridges broad signals throughout near CJ!. All these can be sensitive
against changes of the tertiary protein structure" the presence of
significant near-CJ signals can indicate a well-defined folded structure
while the absence of near-CJ signals occurs in ill-defined three-
dimensional structures. Although the signal intensity is much smaller
compared with the far-CJ0 even small changes in tertiary structure can
cause changes in the spectrum e. g." protein-protein interactions0
changes of the solvent conditions!. $or primary0 secondary etc.
structures see addendum.
>he has a small e&tinction coefficient because of high symmetry
and it is also the least sensitive to alterations in its environment.
Absorption ma&ima at )850 )8,0 ),) and ),? nm vibronic bands!.
Tyr has lower symmetry then >he and therefore has more intense
absorption band. Tyr has absorption ma&imum at )?, nm and a
shoulder at ).+ nm. 3ydrogen-bonding to the hydro&yl group leads
to a red-shift of up to 5 nm. The dielectric constant affects the
spectrum also.
Trp has the most intense absorption band centered at ).) nm.
3ydrogen-bonding to the K3 can shift the
%
L
a
band by as much as
%) nm.
%%
:isulfide <-<! spectra have a broad band at )8# - +## nm with no
vibronic structure.
$ig. %)" small but significant change of the tertiary structure of a protein
due to different amounts of an antibody.
%)
$ig. %+" folding and unfolding DmeltingD! of a protein in three different
buffers.
Optical Rotation Dispersion L':! occurs because chiral
substances refract L-respectively '-circular polari1ed light in a different
way. Fenerally spoken C: spectroscopy these days has superseded
L':. The measured value is the rotation
is
obtained"
%+
2
3
2
3
2 2
+
n d
M
n d c
%#!
where the term
2
3
2
+ n
accounts for the fact that the system is in solution
and not in vacuum. n is the refractive inde& of the solvent.
"ummary
C: has an important role in the structural determinants
of proteins
3owever0 the effort e&pended in determining secondary
structure elements is usually not worth it because it is
somewhat unreliable.
The real power of C: is in the analysis of structural
changes in a protein upon some perturbation0 or in
comparison of the structure of an engineered protein to
the parent protein.
C: is rapid and can be used to analy1e a number of
candidate proteins from which interesting candidates
can be selected for more detailed structural analysis
like K=' or M-ray crystallography.
#nvestigation of the "econdary "tructure of a $rotein
=onitoring
)))nm
of a protein as a function of temperature or chemical
denaturant yields information about protein stability.
The thermodynamic parameters0 F
u
0 3
u
0 <
u
0 T
m
0 C
p
can be
determined
$ig. %5
,
_
,
_
,
_
m
m p
m
T
T
T T T c
T
T
H G ln 1
%5
( )
1
1
1
1
]
1
,
_
+
+
RT
T
T
c T
T
H
T T T c H
m
p
m
m p
unfolded
ln
exp 1
1
1
$ig. %8 left" C: of hemoglobin0 elastase---0 and lyoso1ym72 $ig.
right"C: of various secondary structures from reference" -heli& %!0
antiparallel -leaf )!0 -leaf +!0 coil 5!0 see also fig. %#
The goal is to determine the fraction of basis set spectra that add up to
give the C: spectrum of the protein or the other way around"
deconvolute the spectrum into its components.
The absorption should be less than %.# usually N #.+! for cell
pathlengths of #.#8 to % cm in order to maintain reasonable signal-to-
noise ratios and accurate C: measurements.
%8
%
)
+
5
>rotein concentration used is typically % mg/mL
;uffer is typically %# m= phosphate with low salt if any.
Solvent Cut-Off (A=1.0) for
Two Different Cell !t"len#t"$
Compound %.# mm #.#8 mm
3
)
L %.) %?,
$
,
i>rL3 %?5.8 %,+
$
+
6tL3 %?-.8 %?#
=eL3 %-8.8 %.5
6tL3 %-, %.,
=eCK %.8 %?8
:io&ane )+% )#).8
Cyclohe&ane %.# %?8
n->entane %?) %,.
The instrument must be calibrated"
typically an aqueous solution of O!-%#-camphorsulfonic acid C<A!" %
mg/mL in % mm cell" ( ).+, at )-#.8 nm0 6 ( %.#)%#
-+
0 ellipticity (
++.8 mdeg. At %-).8 nm ellipticity ( ,-., mdeg. To accurately determine
concentration of C<A solution" 6
).8nm
( #.?5+ in 8 cm cell0
).8nm
( +5.8
=
-%
cm
-%
.
The protein concentration must be known accurately"
%-#nm
( .08## - %%05## =
-%
cm
-%
per residue this is not accurate enough"
as you know the in the far CJ of proteins depends on the secondary
structure!.
%&'nm
(in ( ) *dmCl) + , of Trp residues -.(/' )
0
cm
0
1 , of Tyr residues
0.%&' )
0
cm
0
"ome results sho2ing also results obtained from 3ray4
%,
$ig. %,"cadmodulin and Eco'I endonuc
$ig.%?" C:-spectrum of thymidylate synthetase is ++G 30 )5G A0 )G >0
)%G T0 )#G L. Cpon binding of $dC=> and 80%#-
methylenetetrahydrofolate C: shows -8G A0 -,G T0 O.G L lower
curve!. $or 30 A0 >0 T and L see table above.
5DDE6D7)
>rotein Technique
-
heli&
3
antiparallel-
-sheet
A
parallel-
-sheet
>
-
turn
T
other
L
Eco'I
endonuclease
M-ray ), )# . )8 )%
Eco'I
endonuclease
:econvolution
of C: spectrum
++ )# 8 %? )8
calmodulin M-ray 8- + # - 5%
calmodulin :econvolution
of C: spectrum
,% ) ) - +8
%?
nm
Experimental Conditions
<tandard conditions"
rotein Concentr!tion" #.8 mg/ml
Cell !t" %en#t"" #.8 mm
St!&ili'er$ (Met!l ion$( etc.)) minimum
*uffer Concentr!tion ) 8 m= or as low as possible while maintaining protein
stability
The protein concentration might needs to be ad@usted to produce the best data.
Changing this has a profound effect on the data0 so small increments or decrements
are advised. If that does not produce reasonably good data0 a change in buffer
composition may be necessary. It is also a good idea to check the sample for
une&pected aggregation via :ynamic Light <cattering :KA repair en1ymes are an
especially good e&le of this behavior!. If absorption turns out to be a problem0
cells with shorter path #.% mm! and a correspondingly increased protein
concentration and longer scan time might help.
The #ncrement )ethod (to Chec8 for Consistency)
Calibration with standards of well-known secondary structure e. g.
polyamino acids!0 see fig. %#0 allow a splitting of the intensive
e
quantities of the measurement into increments which can be utili1ed for
an appro&imation of structural estimates relative! of secondary
structures0 in particular when the structure changes caused by e&ternal
factors like temperature or p3.
The e&tinction E can be written as a sum of increments in analogy
to thermodynamic quantities like the enthalpy of formation!"
d
c
E
n
i
i i
%%!
This is valid as long as the individual sub-systems i the chromophores!
do not interact with each other. As long as this is the case0 e. g. the
e&tinction doubles when the concentration is doubled.
d
x
c
E
E
i
n
i
i
n
i
i
%)!
+
i
is the molar fraction of the component i.
e
intensive quantity do not depend on the si1e of a system e. g." pressure0 density0 temperature7!0
e&tensive quantities do e. g." volume0 mass0 entropy0 enthalpy7!
%.
1
n
i
i
i
n
i
i
x
x
%+!
$or simplification the number of components or structures! n is assumed
to be n ( )0 e. g. component % heli&!0 component ) coil!. This results for
two different wave lengths
a
and
b
in"
( ) ( )
( ) ( )
( ) ( )
( ) ( )
b b
b b
a a
a a
x and x
2 2
1
2
2 2
1
2
%5!
Is the same molar fraction &
)
found at more than one wave length0 there
is more reliability in the measurements. There should be as many
measurements at different wave lengths as there are components. The
molar fraction can also be determined with different methods. They
should be consistent.
6&le"
The molar ellipticity is an intensive variable. If one assumes0 for e&le0
two different structures P an heli& and a random coil P then the
evaluation of the data at only one wave length is not sufficient. <ame is
true for + structures and evaluation at only two wave lengths.
If there are interactions between the subsystems that influence the
ellipticity0 this can result in misinterpretations. Again0 the only way is to
do calculations at more than one independent methods.
"ome more about $roblems determining the "econdary "tructure
$ar-CJ C: for determining protein structure
n 9 : centered around %%' nm
%-
p 9 : centered around 0/' nm
n 9 : involves nonbonding electrons of O of the carbonyl
9 : involves the electrons of the carbonyl
The intensity and energy of these transitions depends on and
(i;e;. secondary structure)
In a folded protein the amide is in a continuous array.
<or example. the absorption spectrum of poly=lysine in an
helix. sheet. and unordered (random coil) differ due to long
range order in the amide chromophore;
$ar CJ-C: of random coil 'C!
positive at %0% nm (9:)
negative at 0/- nm (n9:)
$ar CJ-C: of -sheet
negative at %0& nm (9:)
positive at 0/( nm (n9:)
$ar CJ-C: of -heli&
exciton coupling of the 9: transitions leads to positive (
9:)
perpendicular
at 0/% nm and negative (9:)
parallel
at %'/ nm
)#
CJ-spectrum
$ar CJ-C:
spectrum
negative at %%% nm is red shifted (n9:)
$igures show deconvolutions of the CJ-spectrum left! and the C:-
spectrum right!.
Cse far-CJ C: to determine amounts of secondary structure in
proteins
generate basis sets by determining spectra of pure helix.
sheet. etc; of synthetic peptides
or deconvoluting CD spectra of proteins 2ith 8no2 structures to
generate basis sets of each of secondary structure
poly-L-lysine QLys!
n
R can adopt + different conformations merely by
varying the p3 and temperature
random coil at p> ?;'
helix at p> 0';&
form at p> 00;0 after heating to -%@C and recooling
)%
C: spectrum of unknown protein ( f
<
! O f
<
! O f
'C
<
'C
!0
where <
!0 <
!0 and <
'C
! are derived from poly-L-lysine basis
spectra.
Recently "
rc
i
i i
S f
, ,