That's a tricky question. How pure is your isolate? Did you add a His-tag or some other way of easy purification?
I would say your best bet is to make a sizeable batch and purify it as well as you can, use column purification or if need be a SDS-PAGE gel. Then just measure total protein concentration by BCA or Bradford and correct for purity. You can use that as a standard for future reference.
Oh yeah, if you have an antibody (since you mentioned ELISA) you can of course use that to purify.
Total amino acid analysis is the 'gold standard' method, but you'd need to get this done at a suitable facility.
If your protein is pure, then it can be possible to determine an extinction coefficient for absorbance at e.g. 280 nm. This would provide a very easy way to determine the concentration of a pure samples once established.
If you're dealing with a crude sample and want to determine the concentration of just one protein amongst many, you have to have an assay specific for that protein and pure protein to compare it to. Methods such as western blotting will give crude estimates of abundance - none, little, lots - but are not reliable beyond this. See:
A Defined Methodology for Reliable Quantification of Western Blot Data
Sean C. Taylor • Thomas Berkelman • Geetha Yadav • Matt Hammond
Mol Biotechnol (2013) 55:217–226 DOI 10.1007/s12033-013-9672-6
for detailled anaylsis of quantificatin by western blotting.
If your protein is pure, then determine the total protein content by measuring absorbance (extinction coefficient) at 280 nm. The disadvantage of this method is your protein should be in high concentrations.
You can also use amino acid sequencing of your protein and then use appropriate wet chemical methods like lowry and BCA assays which will detect mainly aromatic amino acids (Tyr. Trypt. & PA). Make sure that your protein is pure and is free from interferring substances during the wet chemistry. BCA is more sensitive (micro version upto 1 µg) and resistant to interferrring substances than Lowry assay. Choose a standard protein which is more similar to your protein during these assays.
If you want to measure the specific activity (antigenic portion) out of total protein then go for ELISA or western blot assays.
Agree with Pete Holt above - - For a purified prep - quantitative amino acid analysis (AAA) is the best method if you have access to the analyzer or service provider. Best to perform hydrolysis under HCL and NaOH conditions since tryptophan is destroyed during acid hydrolysis. Good news is you only need to do this once - since (as mentioned above) you can standardize the A280 (or even A230nm) concentration from the AAA values as long as buffer do not interfere.
A very low tech alternative we used years ago (when the earth was still warm) was a technique called dry weight analysis - - very accurate and easy to do but you need access to very sensitive micro balances and very good controls for buffer. Lots of good literature on this goggle - Nozaki et al dry weight analysis
As mentioned above there are several methods for protein quantification. Please find the following suggestions.
1. UV 280 nm (simple and easy if your protein solution is clear)
2. Excitation at 280 and emission spectra from 330-380 nm ( this is good for both soluble and insoluble protein but your protein should contain tryptophan residues)
3. Dye binding assay (BCA or Bradford or other commercial kits) (good for total protein quantification but make sure that your protein has positively charged amino acids)
4. Fluorescent dye binding assay (very sensitive)
5. Amino acid analysis is good but laborious
6. SDS-PAGE ( if your protein got coomassie dye binding regions) you could use calibration standards and calculate your protein content
7. Western blot may be good idea if your got his-tag (you can use commercial standards for the quantification