Hi, in principle it is correct that beta turns and the neigbouring beta strands get converted into random coil. Urea and the Guanidine cation (the latter is usually paired with chloride anions) both bind preferentially to the  peptide backbone. Therefore both disrupt intramolecular interactions in the protein. Based on that the protein unfolds into random coil - this includes all secondary structure elements including beta turns. You can start reading here http://www.ncbi.nlm.nih.gov/pubmed/18498104 and look for the work of D.W. Bolen and M. Auton and also for Serge Timasheff and the terms "preferential hydration" and "preferential interaction".

Best regards.

Thank you very much for Alexander Tischer's answer! But I wonder if the random coil displays the same vibration absorption as the loop structure or they acturally are the same structure? 

The vibrational signature of a structural unit solely depends on the degree of coupling and delocalization. This might affect by the chemical used in the denaturation process and the "severeness" of denaturation.  Assumed you are trying to reveal structural information by looking at the backbone vibrational modes, I recommend you to read this reference:

http://www.ncbi.nlm.nih.gov/pubmed/12621861

It is a very well written article that explains the molecular origin of all vibration peaks observed. 

It is not easy to clarify the difference between coil and loop, as they both lack of a fixed pattern. The vibrational adsorption may vary from molecules to molecules. You may also seek for simulation studies to verify your assumption.

Hope this help.

Thanks to Wei Liu. It helps me a lot.

NMR can distinguish between a turn a loop based on specific NOE patterns that occur in turns but not  in loops. I think everything else will have difficulty.

Hi

I would say that in equilibrium you sometimes get on-or off-pathway loops after adding denaturants, but mostly, these loops are temporal folding-promoting loops which are transient intermediates that you would not see in equilibrium, but rather in kinetic experiments. Note that sometimes, these transient folding-promoting loops end up as two antiparallel helices or antiparallel beta strand followed by an alpha helix.

See the works from the group of Elisha Haas focusing on Adenylate Kinase protein refolding through utilization of time-resolved ensemble FRET with small dyes:

1. Orevi T, Ben Ishay E, Gershanov SL, Dalak MB, Amir D, Haas E. Fast closure of N-terminal long loops but slow formation of β strands precedes the folding transition state of Escherichia coli adenylate kinase. Biochemistry. 2014 53(19):3169-78

2. Lerner E, Orevi T, Ben Ishay E, Amir D, Haas E. Kinetics of fast changing intramolecular distance distributions obtained by combined analysis of FRET efficiency kinetics and time-resolved FRET equilibrium measurements. Biophys J. 2014 106(3):667-76.

3. Ben Ishay E, Rahamim G, Orevi T, Hazan G, Amir D, Haas E. Fast subdomain folding prior to the global refolding transition of E. coli adenylate kinase: a double kinetics study. J Mol Biol. 2012 423(4):613-23. 

4. Orevi T, Ben Ishay E, Pirchi M, Jacob MH, Amir D, Haas E. Early closure of a long loop in the refolding of adenylate kinase: a possible key role of non-local interactions in the initial folding steps. J Mol Biol. 2009 385(4):1230-42.