Dear Chandni,

I would say the formula you use for normalization is correct. As you use the same peptide (with the same sequence) for all your measurements, the residue number doesn’t change and the observed ellipticity should directly scale with your sample concentration.

Even before transforming the “mdeg” to “MRE”, it is obvious from the spectra that your signal intensity doesn’t scale exactly with the concentration (that is by a factor of 2 going from 100 µM to 50 µM), but rather by a factor of 3. No normalization can adjust this discrepancy, and the reason seems to lie somewhere else.

The peptide concentration difference doesn’t seem high enough to cause major differences in secondary structures as can be observed by e. g. aggregation processes concomitant with signal quenching or structural rearrangements. Therefore it is more likely that the assumed sample concentration is not correct. Maybe you didn’t dry the cuvette completely before measuring the next sample which led to an unwanted dilution of the 50 µM sample. Or you changed the cuvette and for some reason (impurity...) it was not equivalent to the first one with the 100 µM sample. Or you used one common reference buffer, but the composition of the sample solutions were not completely equivalent, e.g. due to incomplete desalting and buffer exchange. We always use small gravity flow size exclusion columns to efficiently exchange the sample buffer composition to the same common CD reference buffer. We take care that the buffer used for this exchange process is the very same as used in the planned CD measurements, so we prepare enough buffer for all these experimental steps. These are just suggestions what can go wrong and cause slight concentration or absorption changes that have a big impact on sensitive CD experiments.

I recommend you to determine the concentration again directly before the CD measurement to rule out such errors. You could take a UV-Vis absorption spectrum of your peptide and should directly see the scaling factor 2 in the measured intensities. The CD-active region close to the peptide bond absorption 190 – 220 nm is of particular interest. If your sample contains impurities there, you can directly see it.

Hope this helps,

Joscha

I've written a tutorial on normalizing CD available here: https://www.photophysics.com/circular-dichroism/chirascan-technology/circular-dichroism-spectroscopy-units-conversions/

Put simply, you are looking at the same peptide at different concentrations. So for your purpose you could just as easily correct to molar epsilon or ellipticity. MRE allows you to compare peptides and proteins of different lengths.

What's happening in the data. If pathlength (you use same cuvette) is the same and the only difference is the concentration, then you'd expect 2 fold change in CD signal going 50uM to 100uM. But you see more like 2.5 to 3 fold. You either have an inaccurate concentration, difference in the pathlength (difference cuvette use the same cuvette), or something physical happening to your peptide like aggregation. I'd check your concentrations and sample prep carefully.

It is an unfortunate situation that CD spectra rely on a quantity like absorbance that is in general not conform with Maxwell's equations (absorbance works only when light does not behave as waves) and dispersion theory (according to which absorbance is in general not linearly related to concentration).

For transmission measurements of strongly diluted solutions and pathlengths of 1 cm you should be on the safe side, otherwise things are unclear... investigations are necessary like in conventional spectroscopy, see e.g. (and references therein):

Article Beer’s law – why integrated absorbance depends linearly on c...

Article Deviations from Beer's law on the microscale - Nonadditivity...

Article Beer's Law – Why Absorbance Depends (Almost) Linearly on Concentration

Article The Electric Field Standing Wave Effect in Infrared Transmis...

Article Employing Theories Far beyond Their Limits-The Case of the (...

VCD spectroscopy is definitely affected due to the need for thin cuvettes and neat or highly concentrated substances...

Thank you all for your valuable insight. I will repeat these experiments to make sure it's not a handling error and if possible I will check the concentration of the sample beforehand.

And Joscha Breibeck, I have a doubt that Jasco CD instrument also measures the absorbance (which could be set up in parallel to mdeg and HT), do you think that would also help in calculating real-time concentration?