• α-helix: CD spectrum of right-handed α-helix is characterized by two negative absorption bands of nearly same intensity centered around 222 nm (arises due to n → π* transition) and 208 nm (a fraction of the π → π* transition) and a relatively more intense positive band around 192 nm (a fraction of the π → π* transition).
  
  
• β-sheet: β-sheets display a negative band centered around 216-218 nm (arises due to n → π* transition) and a positive band of comparable intensity at ~ 195 nm (arises due to π → π* transition).
  
  
• β-turn: β-turn is a four residue protein motif that causes the polypeptide backbone to take a turn of approximately 180°. β-turns do not have a well defined spectral signature. A typical β-turn, however, shows a weak negative band around 225 nm (arises due to n → π* transition), a strong positive band between 200 – 205 nm (arises due to π → π* transition), and a strong negative band (arises due to π → π* transition) between 180 – 190 nm.
  
  
• Random coil: Random coil or the unordered conformation displays a weak positive band around 218 nm (arises due to n → π* transition) and a strong negative band (arises due to π → π* transition) below 200 nm.

It is due to these characteristic signatures for the secondary structures that the secondary structural components in the proteins can be identified and estimated.

Many different methods are available for analyzing the circular dichroism spectra of proteins. All these methods work on the assumption that the CD spectrum of the protein is a linear combination of the spectra of its secondary structural elements, plus noise. The ellipticity of the protein at any wavelength can therefore be represented by equation 10.1

(10.1)

where θλ represents the ellipticity of the protein at wavelength, λ; fi, represents the fraction of the ith secondary structural element; and Siλ represents the ellipticity of the secondary structural element, Si at wavelength, λ .

 

 

The methods that are in practice utilize the CD spectra derived from the proteins whose crystal structures have been determined as the reference. A number of algorithms have been developed that utilize the reference spectra database to evaluate the secondary structural components in a protein from its CD spectrum. We shall not be discussing these algorithms but details can be found elsewhere [1- 4].

The initial attempts in deconvoluting the protein CD spectra utilized poly-L-lysine CD spectra as reference. Poly-L-lysine can adopt, depending on the conditions, three different conformations in dilute aqueous solutions. It adopts random coil in aqueous solutions at acidic and neutral pH. At pH 11.2, it adopts a predominantly α-helical conformation. Heating the poly-L-lysine solution at pH 11.2, for 20 minutes at 51°C, results in antiparallel β-sheets. The CD spectra of poly-L-lysine can therefore be, to a very good approximation, be treated as those arising for pure conformations and used for analyzing the structures of unknown proteins. Unlike poly-L-lyine, however, proteins are heteropolymers and the CD spectra of a homopolymer may not represent a good basis for estimating their secondary structural components. Modern methods therefore utilize the CD spectra of the proteins whose structures have been determined by X-ray crystallography as the reference database.