In my understanding, MRE should not depend on the protein concentration due to the fact that ellipticity spectrum is converted to normalized value, which is independent of polypeptide length. Moreover, MRE is used to represent the ellipticity for individual protein residues instead of whole protein molecule and, thus MRE used to compare the secondary structure content of different protein irrespective of their molecular weight. 

In your case, the MRE is changing with concentration and is indicative of aggregate formation or concentration dependent interactions. 

MRE should not change with concentration unless there is some intermolecular interaction or solubility issue. Double check you have everything the same, especially path length. Also beware of the difference between Mean Residue Molar Ellipticity and Molar Ellipticity https://www.photophysics.com/resources/7-cd-units-conversions

Dear Anitha Ramu, 

Yours is a very interesting and appropriate question! …the question of whether the signal of my measurement is proportional to the mass of my molecule/protein?

In case of Circular Dichroism (CD) the answer to this question should be easy to find out since it’s an Absorbance-based spectroscopy and the Lambert-Beer law does apply. Basically, this means that the CD signal, when properly corrected, will be proportional to the concentration of the optically active molecules in your sample. Please check the following paper (and references therein) for an excellent protocol on CD measurements for analyzing proteins, How to prepare the samples, and other important aspects to consider such as buffers etc.

Greenfield, N. J. (2006). Using circular dichroism spectra to estimate protein secondary structure. Nature Protocols, 1(6), 2876–90. http://doi.org/10.1038/nprot.2006.202

Now, going to the finer point of your question: Why there’s a difference in the MRE or molar CD of your protein with that value reported in the literature?

Since the CD it’s a fancy absorbance measurement (in principle), the correct estimation of concentration of your sample will have a big impact on the signal recovered and therefore in the corrected molar CD. How can you be sure that the concentration of your sample is as good as the one reported in that reference you are referring to? ... more importantly, what can assure that they made the correct estimation of protein concentration?

There’s also a very important point that needs attention: Is the calibration of your instrument good? The CD machine (spectropolarimeter) relies on amplifiers that help to enhance the CD signal since the CD signal is a very tiny part of the whole absorbance of the sample. Please check the ref listed above for details on how to make sure the calibration of the instrument is decent for this kind of measurements.

So, no need to be disappointed!, at this point your measurement is as good as any other out there!

Mean Residue Ellipticity (MRE) is different from Molar Ellipticity. But if mdeg is converted correctly to MRE or ME, they should be independent of concentration, pathlength etc. Mean Residue Ellipticity is also normalized for number of amino acid residues to enable cross-protein comparisons. If your conversions are done correctly, you should explore many other factors such as CD instrument performance, accuracy of your concn measurement etc.All the best. 

CD can be used with many different applications and depending on what you want to learn you can analyze your data in different ways. For instance, let's say your goal is to compare different samples, or the same sample under different conditions, and you are not sure that your protein concentration are not accurate (or simply you have many samples and don't want to go through the trouble to accurately check concentration for each one); or you want to titrate a ligand, ora  a buffer; instead of looking at CD (or MRE) you can compare the G-factor*, i.e. the ratio between UV CD signal and UV absorbance signal. This way you will normalize for protein concentration and no need to worry if your protein assay was accurate. Clearly, with this approach you don't get secondary structure elements composition, but in most cases that is not very accurate anyways.

*Baker BR, Garrell RL. G-factor analysis of protein secondary structure in

solutions and thin films. Faraday Discuss. 2004;126:209-22; discussion 245-54.

PubMed PMID: 14992408