Two further points to those already made. Firstly, if there are a number
of IgG binding proteins in your extract (not an uncommon occurence) you
can remove the offending bands by pre-incubating the blot with an
irrelevant IgG or serum, i.e. one from a species other than those from
which your primary and secondary antibodies are obtained. Secondly, it
always pays not only to titrate your primary antibody and use it at the
lowest concentration, but also the detecting antibody.