Yun-Fu Hu,
My first thought is, that you have not mentioned a step to block 
endogenous peroxidases (e.g.3% H2O2 in dH20 for 15 mins.). It is 
possible the melanomas harbor more endogenous peroxidases 
than normal tissue (I'm sure you will hear from others who can say 
for sure).

Second, are there any cross reactivity possibilities among the 
target species, the primary Ab species, and the secondary Ab 
species? 

With an appropriate set of controls (deleting various key steps in 
the procedure) you should be able to isolate the problem.
Hope I was of some assistance. Greg 

> Dear Colleagues,
> 
> A group of us here have been trying to stain melanocytes with an 
IgM
> antibody that we generated.  So far, we have observed beautiful 
staining of
> melanocytes in normal or benign (i.e. nevi) skin samples with a 
fairly
> simple protocol -- dewaxing, dehydration, protein block (5-15 
min), primary
> (1-3 hrs at RT), secondary-HRP (30 min at RT) and detection 
(note: no
> antigen retrieval).  However, whenever we tried the same protocol 
on
> melanoma samples, we noticed considerably higher background 
almost
> everywhere.   Lower incubation temperature or lower antibody 
concentration
> didn't help much, either
> 
> Higher background is probably a consequence of non-specific 
staining of the
> inflammatory or necrotic tissues in melanoma samples by some 
impurities in
> the ascite fluid that we are using.  Therefore, it appears that three
> factors could be influencing our staining unfavorably:
> 1) Malignant skin tissue that tend to have higher background;
> 2) Nonpurified ascite fluid that might contain many other 
impurities;
> 3) An antibody of IgM isotype that is often more troublesome than 
the more
> common IgG's.
> 
> Currently, we are trying to purify the IgM antibody from ascite fluid,
> which is a bit tougher than we anticipated.  As an alternative, we 
are also
> growing the hybridoma hoping to get some supe for testing.  
However, we
> might also run into the same problem with the supe if IgM or 
melanoma
> sample is the problem.
> This leads to my questions:
> Is IgM really a problem for immunohistochemical staining on tissue
> sections?
>      If so, what can we do except to generate another hybridoma 
line?
>      If no, how can we remove the non-specific staining besides 
incubation
> of the tissue section with protein block prior to the primary 
antibody
> (that we have tried)?
> 
> Any suggestions or comments will be greatly appreciated.
> 
> By the way, I just recently signed on to this newsgroup.  If you have
> answered these questions in previous posts, I would certainly 
appreciate it
> if you could please forward a copy of your response to me.
> 
> Thank you very much for your help in advance, and have a great 
day!
> 
> Yun-Fu Hu, D.V.M., Ph.D.
> BDDS
> Sparks, MD
> USA
> 
>