Shaun D. Black
SHAUN at JASON.UTHCT.EDU
Tue May 14 15:08:29 EST 1996
L. Kempster wrote:
>Can anyone give me a basic discription of how CD works and particularly with
>respect to protein secondary structure determination?
>Please note that I am not a biophysicist!
Asymmetric molecules have the ability to interact with polarized light in a
characteristic fashion. The asymmetry often arises due to chirality;
for proteins, the asymmetry comes from periodic folding of polypeptides
built from chiral alpha carbon atoms in the amino acyl residues.
CD, or circular dichroism, is a technique in which light is polarized into
two circularly-polarized components (one left handed, one right handed).
Asymmetric structures can absorb circularly-polarized light differentially
at a given wavelength and characteristically as a function of wavelength.
When one component of the circularly polarized pair is selectively absorbed,
the resultant "vector addition" yields light that is eliptical. Thus, you
will see CD results expressed as "elipticity"; the more differential the
absorption, the more eliptical the resultant light will be.
Model protein secondary structures show characteristic CD spectra in the
far uv (between about 170 nm and 300 nm). Alpha helix has a strong elipticity
peak near 192 nm with a broad trough between about 200-240 nm; beta sheet
looks somewhat similar but the maximum is at about 198 nm and the max and min
are not as intense as those of alpha helices; and "random" coil structure
has apeak at about 209 nm with a trough around 190 nm. These descriptions
are crude, but get the general idea across. In fact, it is useful to
note that antiparallel beta sheets have somewhat different spectra than
paralell beta structure, and that beta turns also have characteristic
Most methods call for recording of an accurate CD spectrum. This means that
your sample cannot be too concentrated (about .1-.5 OD at 280nm is usually
fine), and that you need to scan relatively slowly and purge effectively
with oxygen-free nitrogen (to minimize noise). Once you have your spectrum,
it can be deconvolved with the assumption that it is the simple sum of the
spectra of each secondary structure scaled by the mole fraction. There are
about a dozen methods published for doing this. Search for Curtis Johnson
or Gerald Fasman; each of their labs has made excellent recent contributions
to CD spectra decompositon. After the calculations are finished, you have
the mole fractions (and thus the percent) of each type of secondary structure
that the given method analyzes for (usually at least alpha, beta, random).
That's it for my CD "primer". I hope it has been somewhat useful! Best
= Shaun D. Black, PhD | Internet address: shaun at jason.uthct.edu =
= Dept. of Biochemistry | University of Texas Health Center, at Tyler =
= World Wide Web: http://pegasus.uthct.edu/UTHCT-Home/Welcome.html =
More information about the Proteins