Anything is possible, I guess, but I really doubt that your antibody is binding to itself. Such a phenomenon would be expected to create aggregation (and likely precipitation) of your antibody in solution. I think the signal you're seeing is simply non-specific binding of your detection antibody. This can be reduced by better blocking of your membrane and by reducing the concentration of you gold-labeled antibody.

I would try running an irrelevant antibody or mouse and/or rat serum instead of PBS and seeing if that reduces the signal.

Yes it is possible

For shoot analysis you can use in ELISA one MAB for immolization on plastic and another biotinolatad as second MAB with or without antigen and determine comparative affinity constant interaction between MAB – Ag and MAB - MAB

.  I doubt that either of the suggestions you make above is the explanation.  I agree that this is likely non-specific binding of the labelled antibody. Labeling another antibody or using a different label on your monoclonal will tell you if the membrane is binding the label or the antibody directly, on the other hand, if you just want to make it work then finding an appropriate blocking agent should do the trick. 

It's possible.

2 issues behind: 

1. monoclonals decline to aggregate if they are purified. To avoid this, the use of an innert protein like albumin (1 ... 5%) or innert serum (FCS) could avoid this. Alos pH shifts can help you there.

2. some monoclonals are real "autobodies", recocnizing them by themself. It's a real antibody-antigen-interaction which can be influenced by adding free antigen. normally the affinity between autobodies is much lower than the affinity between antigen - antibody. Thats very difficults to handle this in an assay.

I agree with previous comments that anything is possible, but  if you assume that the background is due to autoreactivity that means that you have to drop that antibody that you most likely have invested a lot of time (and money in). Personally, I would first have another look at unspecific binding. I would either add 0.05-0.1% Tween 20 or a carrier protein to the detection antibody. You probably have used BSA with the detection antibody. I would try 1% dried milk powder (low natural levels of bovine IgG), 1% mouse serum or chicken serum to see if this reduces the signal. I would also try a different batch of BSA. BSA is usually purified by Cohn fractionation (it usually say fraction V on the bottle) and that means that the BSA contains IgG (up to 5%), You do not know from which animals the BSA was purified so there is a chance that the BSA is purified from an animal that has rheumatoid factors. If such BSA is used as a blocker you will get a high background. Addition of mouse serum or mouse IgG to the detection antibody will reduce the signal, but it may not necessarily inhibit it completely. The purified antibodies that you use are either bound to a surface or to a detector. This induces structural changes to the antibodies that resembles aggregated antibodies. If you want to reduce the effect of HAMA/RF mouse serum is usually effective, unspecific mouse monoclonal better and heataggregated mouse monoclonal the most effective inhibitor of HAMA/RF interactions.