There is essentially no general difference between autoantibodies and non-autoantibodies except their specificity. Are you looking for autoantibodies to a specific autoantigen or autoantibodies in general?

If you are looking for autoantibodies to a specific autoantigen then it would be best to use a highly purified (preferably homologous) form of that autoantigen (e.g recombinantly expressed) in ELISA or Western blot. Then specific autoantibodies (IgG) will bind while other antibodies will not. Alternatively, if the dominant epitope of that antigen is known it may be possible to use a synthesized peptide containing that specific epitope. However, many antibody epitiopes are conformational so using peptides, which tend to be linear, may not work. An alternative is to use extracts of homologous (or autologous) tissue, cells or cellular compartments in Western blot. If antibody binds to protein of the same size as you protein(s) of interest then it is likely (but not proven) that you have presence of autoantibodies to your protein(s) of interest.

If you are looking for autoantibodies in general to autoantigens primarily expressed in specific tissues, cells or specifific cellular compartments then you can use extracts from homologous (or autologous) tissue, cells or purified cellular compartments (e.g. nuclear extracts) in ELISA or Western blot. However. using such extracts in ELISA may give problems with high unspecific background.

For all such assays it is important to always include proper controls, e.g. non-autoimmune sera/antibodies.

A third alternative, if you want to identify autoantigens (primarily novel autoantigens) is to use cDNA expression cloning (also known as SEREX). This method is based on (auto)antigen display using cDNA expression libraries generated from tissues or cells expressing your autoantigen(s) of interest.

IgM antibodies are in general more polyreactive than IgG (or other class-switched antibodies) so it may be difficult (depending on the autoantigen) to separate autoreactive IgM from non-autoreactive IgM.

The classical way to do this is by immunohistology. SOme years ago when I worked in a diagnostic immunology lab we screened for autoantibodies in patient's serum samples using composite 'blocks' of mouse tissue (liver, ileum, heart. muscle) and looked for binding by immunofluoresence. We had known postive and negative controls and binding to a tissue would allow us to focus on the most likely antigen. Such a screen is very useful if you do not suspect a specific antigen target and allows identification of autoantibodies where the binding would be likely to occur in vivo.

a simple method is also by the use of mercaptoethanol which is usually used for half antibodies. several papers were published on this

In all of this, the most important point remains to identify the reliable autoantigen. one of the modified ways would be to make two aliquots of polyclonal antibody fractions of patient sera; one having only IgG fraction purified by using protein G column and other aliquot should essentially contain all the antibody types accept IgG ( left-out of IgG minus sera). This step increases the chances of specific hit while performing SERPA or MApping. 2nd step should be to equilize the dilution factor while before probing the tissue/cell lysares with (above) purified polyclonal ab fractions. This can be done by checking the conc. of IgG and IgM ( two of the most relevant auto-ab) and accordingly set up the dilution factor, that ideally varies from patient to patient.  All the Best

Hi Marina,

If you find the answer to this , you will deserve a Nobel prize. Many of the natural autoantibodies have a protetcive role , but it is completely unclear how/why they may develop into pathogenic Ab

best, H

My experience is that autoantibodies mainly interact with conformational epitops.Thus, you can run an immunoassay preserving the native conformation of the autoantigen and compare your results with immunoblot. SDS-PAGE destroys the conformation of biomolecules and theres is usually little refolding during the transfer to a blot membrane. All the proteinous autoantigens I have worked with behaved accordingly.

Here is the possible approach:

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4164660/pdf/pone.0107513.pdf