Could you please explain ..how you measure the extinction coefficient?

There are two things.. if you measure it using absorbance at 280 nm, then there is less chances of error.

If you are using colorimetric method, the absorbance value is dependent upon the used buffer ( Value will change with change in buffer pH).

Tough without seeing a spectrum and conditions. How are you determining the protein concentration? Is your protein in the same buffer conditions as the reference value? Does your expressed protein have the expected spectrum? I assume you have optimised spectral conditions so as to avoid obvious artefacts? The difference you mention is about 2.5. You are correctly adjusting for the cuvette length (just had to ask)?

Thanks dears,

I wanted to measure the extinction coefficient at 590 nm. 

i know the protein concentration from absorption at 280 nm. i calculated the protein extiction coefficient by sequence in Expasy program. 

but, i need the  extinction coefficient at 590 nm not 280 nm.

and i am measuring the absorption spectra by Nanodrop and the length is 1 mm. based on different concentration, i have different absorption at 590 nm. so, the slop is the  extinction coefficient value. 

So, I measured in different buffer such as PBS pH.7.5 and 100 mM Tris, 50 mM NaCl pH.7.5; and i did not see any difference in spectra.

@Kurt: I see the shift in maximum emission of mRuby protein. normally, it is 600 nm but the maximum of my mruby spectrum is 570 nm.

Just ran the sequence of mRuby2 through my software and got an epsilon (280nm) of 23020 M-1 cm-1 at 280 using the method of Gill & von Hippel. It has 237 amino acids and a molecular mass of 26478.54. Is this similar to what you got?

Nanodrop is not the tool of choice, but should not be expected to be that off. The shifted fluorescence maximum is concerning. Have you expressed a fusion protein? 

http://www.addgene.org/54771/

here, I ordered mRuby2 in pBAD vector. 

It is only mRhuby protein without any other protein.

I have never worked with mRuby2. Did you express it with the His-tag? Have you cleaved it? Not knowing how sensitive it is, I wonder if the tag or possibly bound Ni2+ (if purified this way) is causing troubles. If so, I would have thought there would have been product info or reports in the literature. Sorry to not be able to be of more help.

When usingthe 280 nm absorbance as a measure of protein concentration, are you taking into account the absorbance of the chromophore at 280 nm? Calculations based on the sequence (Trp, Tyr, Phe, and Cys residues) can be in error if (a) the protein contains another chromophore, as in the case of fluorescent proteins, and (b) if the protein is not completely denatured in guanidine hydrochloride when the absorbance measurement is made.

As for the 590 nm absorbance of the chromophore, if the protein is not pure or if some of it is not properly folded, you will not get an accurate measurement of the extinction coefficient.

Hi there,

The extinction coefficient of 113000M-1.cm-1 is the one observed for the  lambda max of  excitation of the fluorophore which is 559nm for mRuby2 (ie wavelength at which the protein is absorbing the most). What you are actually measuring according to your posts is absorbance at 590nm. So you are calculating absorbance coefficient at this specific wavelength which is less when refering to the mRuby2 absorption spectrum shape. For a quick check just run absorption spectrum between 450nm and 650nm...

Thanks all,

@Dominique Liger: You are right. I am calculating absorption coefficient at this specific wavelength. So, do you know how can i measure the extinction coefficient of 113000M-1.cm-1?

Hi again,

This coefficient is valid for mRuby2 absorbance at 559nm.

Hi,

yes, I measured the absorbance from 240 to 800 nm and calculate the extinction at 559 nm. but the extinction is around 45000. So, I do not understand why it should be like this. I calculated from the slope of the linear fit.

Hi there,

If your protein determination at 280nm is right (ie protein is pure and protein sequence used for coefficient determination,  measure and calculation are right), it means that the protein fluorophore is not totally formed, stable as absorbance at 559nm is mainly due to this structure... Roughly 40% of your prep is not absorbing properly (if one consider the absorbance at 559nm is only the result of fluorophore absorption)...

There are several potential problems involved in measuring the extinction coefficients of fluorescent proteins.  One important issue is incomplete formation of the chromophore.  Another is the choice of protein assay.  Patterson et al (1997) Biophys J. 73: 2782, discuss these issues and describe the bicinchonic acid assay (BCA) that they found worked best.