Are paraffin or frozen (fixed) sections are better for IHC?
I’ve had great success in the past with frozen or vibrating
microtome sections, and have been trying paraffin lately,
but haven’t got any good results.
Generally frozen sections are better for IHC because the
antigenic content is well preserved (provided the tissue is
snap frozen rapidly, preferably in isopentane, then stored
at -70C). A “good” frozen section cut at about 5 microns
should provide adequate morphology.
The advantages of paraffin tissue blocks is that larger
pieces of tissue can be used, and morphology is a degree
better, storage is easier, etc.
The disadvantage of paraffin blocks is the fact that the
processing of the tissue (especially when preserved in
common fixatives such as formalin or other formaldehyde-
based solutions) cross-links certain proteins in and on
the cells. Preatreatment to “unmask” cross-linked antigens
is often essential. Antigen retrieval techniques include
microwaving in citrate buffer and pressure cooker techniques.
However, some antigens are destroyed by paraffin processing,
so for these the manufacturer of the antibody should
recommend the use of frozen sections only.
Cambridge Antibody Technology
The Science Park, Melbourn,
Royston, Cambridgeshire SG8 6JJ
In general, immunoreactivity is often better in cryostat
sections than in wax sections, however tissue morphology is
usually not as clear. If you are getting satisfactory results
with cryostat sections, then I would probably recommend sticking
with that technique. However, if need to use wax sections for
whatever reason, there are several ways of tweaking the protocal
to try and improve the staining. Any good IHC text book will
outline most of these.
Off the top of my head, I would suggest playing around with the
fixation conditions or trying some form of antigen unmasking
step (particularly if you are currently seeing no specific
staining at all).
Ian Jones, PhD
School of Biological Sciences,
Queen Mary and Westfield College,
University of London, England.
** Inhibiting endogenous peroxidase
1. What is the best way to inhibit endogenous peroxidase
activity before doing an immunohistochemical method?
2. How long can methanol/H2O2 mixture (for quenching
endogenous peroxidases during IHC) be kept? or should
it be freshly made each time before use?
Different people favour different methods! Here are five
suggestions. All are claimed to work well, so probably
you should start with whatever you think is the easiest
We use a homemade version: PBS with 0.03% hydrogen peroxide,
and 0.1% sodium azide. Very gentle; doesn’t knock sections
off slides (frozens); can make up a one-week supply.
Use it once, then discard (we use dropper bottles).
Our PBS is at pH 7.4. We collect the leftover for chemical
disposal of sodium azide.
OR you can purchase DAKO peroxidase blocker with 0.03% H2O2
This block works best with our mouse antibodies as it does
not interfere with some of the IHC staining/per recommendation
of PharminGen. They use DAKO also, and if there are capillary
gaps involved, this does not produce the crummy bubbles that
drive one crazy.
We prepare 600ml vats of methanol/H2O2 for use on a DRS601
and replace these weekly. It’s left on the machine for 5
working days then dumped. We’re handling about 150 ICC
Just a personal note on the use of methanol in blocking
solutions; I have also found that methonal can be harmful
to some antigens, both hemopoetic and some infectious
disease antigens. We have found that performing our
endogenous peroxidase inactivation prior to any antigen
retreival step (either enzyme digestion or heat induced)
works best. For antigens sensitive to methanol and frozen
sections we use PBS containing 0.1% Na azide and 0.5% H2O2
with excellent results. Just be sure to wash the slides
well after this step because the Na azide is a potent
peroxidase inhibitor which will eliminate any specific
staining quite well. Using poly lysine coated slides will
generally keep frozen sections from lifting off.
Brian J. Chelack
Quenching with the glucose oxidase method works very
well, and is very gentle on sections, particularly frozen
sections. The only drawback is a bit more preparation of
solutions, but in the long run is a very COMPLETE quenching,
better than hydrogen peroxide, according the original
publication and method. I highly recommend it.
Complete inhibition of endogenous peroxidase (including
activity in leukocytes and erythrocytes) can be achieved
by treating formaldehyde- or acetone- fixed smears or
sections with 0.024 M hydrochloric acid in ethanol for
10 minutes. To make this, add 0.02 ml of concentrated
(12 M) hydrochloric acid to 100 ml of ethyl alcohol.
Weir EE + 4 others (1974) Destruction of endogenous peroxidase
activity in order to locate antigens by peroxidase-labeled
antibodies. J Histochem Cytochem 22:51-54.
This simple method doesn’t seem to be much used. I have tried
it, and Yes, it did work.
** Using mouse primary antibodies on mouse tissues
Using a mouse monoclonal on sections of mouse tissue often
makes a strong background staining because the secondary
antiserum binds to mouse immunoglobulin already present
in the tissue. Is there a way to get round this difficulty?
Two published methods seem quite good for this purpose.
They are very briefly summarized below. For practical
details consult the original papers:
Hierck,BP; Iperen,LV; Gittenberger-de Groot,AC; Poelmann,RE
(1994): Modified indirect immunodetection allows study of
murine tissue with mouse monoclonal antibodies.
J. Histochem. Cytochem. 42(11, Nov), 1499-1502.
Mouse monoclonal reacted with HRP-rabbit anti-(mouse serum);
then add excess normal mouse serum & incubate with tissue.
Lu,QL; Partridge,TA (1998): A new blocking method for
application of murine monoclonal antibody to mouse tissue
sections. J. Histochem. Cytochem. 46, 977-983.
Blocking with mixture of Fab and Fc fragments from
rabbit anti-mouse antibody. (Made by papain digestion,
then more Fc added). Stops background staining of
endogenous mouse IgG by the secondary antiserum.
Corazon D. Bucana, Ph.D.
John A. Kiernan
[ This answer does not really explain what to do, but the
advertised product might interest users of mouse monoclonals.]
DAKO just released an immunostaining system for animal tissues.
In particular, it excels with mouse antibodies on mouse tissue.
We engage a novel technology to ensure clean background and high
specificity. Stoichiometric amounts of primary-antibody complex
are preformed before it is exposed to the tissue site. This
eliminates the unwanted reaction between secondary antibody and
Please visit the DAKO Corporation website (www.dakousa.com) to
request literature on the new DAKO ARK (Animal Research Kit). We
presented a poster at the IAP meeting in Boston and this
document is available by mail.
A few highlights: 1. One kit for all animal IHC testing utilizing
mouse monoclonal primary Abs. 2. Use on tissue from any animal
species. 3. Unique process eliminates background staining.
4. Staining results in 45 minutes. 5. Automatable
For more information please contact DAKO Technical Services at
techserv[AT]dakousa.com or call 800-424-0021.
Product Specialist, DAKO Corporation
** Antigen retrieval: A patented or copyright phrase?
I was talking to someone the other day concerning
immunoperoxidase staining and I mentioned the term “antigen
retrieval”. I was told that the term is patented and that it
was not legal to use the phrase. Has anyone else heard that
information. I do know that Biogenex makes and sells
“Antigen Retrieval Solution,” and we use it in our lab.
Is it really true that we cannot talk or write about antigen
retrieval in a general way without the risk of being sued for
some infringement of a copyright or a patent?
This was the subject of some heated discussion in the
HistoNet listserver in 1998. The following remarks are
based on the contributions of people too numerous to
acknowledge individually, and are colored by my own
On the one hand there were the “common sense” viewpoints
making the case that
(a) A combination of two common words could not possibly
amount to an original literary composition (with
copyright assignable to an author or publisher), and
could never be construed as an invention. (A particular
solution could, of course, be invented for the purpose
of retrieving antigens, and patented.)
(b) Methods for enhancing the detection of antigens in
sections have been published in the scientific
literature for several years. All involve treatment
with water, which may be cold or hot, and most
techniques specify other substances to be dissolved in
the water. The solutes include detergents (to damage
cell membranes, helping large antibody molecules to
enter cytoplasm), urea (disturbs protein conformation
and may expose “buried” epitopes), a variety of metal
salts, notably zinc sulfate and lead thiocyanate
(probably work by changing the conformation of the
antigen), and all sorts of buffers, mostly pH 5-6 or pH
8-9. (This probably catalyzes hydrolysis of the
cross-links that formaldehyde makes between nearby
parts of protein molecules. The optimum pH varies with
different antigens. Heat accelerates the reaction, and
can be conveniently delivered in a microwave oven.)
On the other hand (Would it be the Left or the Right?)
were people using these methods daily, in routine
procedures, sometimes with a proprietary solution and
sometimes varying the technique to suit the antigen.
Feeling their freedom of expression (and perhaps also
their livelihoods) threatened, they suggested alternatives
to “antigen retrieval.”
The word “unmasking,” which has a long and honorable
history among histochemists, is a conspicuous improvement
on “retrieval” because it says what happens. The epitopes
of antigens were not retrieved (= brought back), because
they were already there. The hot water and other chemicals
made them accessible to the primary antibody by removing
physical and chemical barriers (“masks”) to the diffusion
of large molecules.
BUT people are human and by nature conservative (= change
can only make things worse), so it’s likely that
“retrieve” will win out over “unmask” despite any logical
arguments. The HistoNet discussions ended when Biogenex
said that the firm did not claim exclusive ownership of
the “antigen retrieval” word pair, and we could say or
write it without being sued.
John A. Kiernan, MB, ChB, PhD, DSc,
Department of Anatomy & Cell Biology,
The University of Western Ontario,
LONDON, Canada N6A 5C1
** p53 protein
What is the significance of immunostaining with
antibody to p53?
First of all, p53 is the antigen in the tissue, with which
the antibody combines (The p is for “protein”). p53 is also
sometimes referred to as a TSG – Tumour Suppressor Gene).
p53 was labelled “Molecule of the Year” by either Science or
Nature about three years ago.
The “wild” type p53 is the normal. It suppresses cell
transformation and/or mutations. It was traditionally
considered to have a very short life and was therefore never
present in concentrations large enough to demonstrate
immunocytochemically. “Mutant” type p53 has a longer “half-life”
and is therefore more easily demonstrated. It used to be that
mutant type p53 was the antigen of interest. Then of course,
things got more complicated.
There are, of course, antibodies to each type of p53 now.
One thing is for sure – p53 is of fundamental importance in
cell transformation. The biggest problem is that many consider
that the expression of p53 is quantitatively related to prognosis
and can therefore, be used to assess treatment outcomes. Whether
quantitation should be by percentage of positive (?tumour) cells
or by intensity of staining in the positive (?tumour) cells is
still open to debate. Whichever it is, it is obviously important
that your results of today can stand statistical comparison with
your results of yesterday or tomorrow. Even more importantly, can
they be used for comparisons with other labs? The patient may
move elswhere for treatment, for example.
One thing I know for certain: it is very easy to make virtually
all cells p53-positive – not just tumour cells – if you tweak your
immunocytochemical method and any heat induced antigen retrieval
you use. A real minefield!
Russ Allison, Wales
** Prevention of fluorescence fading
What is available in the way of chemical additives to aqueous
mounting media, commercial or homemade, to suppress fading of
Jules Elias has a discussion about this in his book
“Immunohistopathology, A practical approach to diagnosis.” ASCP
Press, 1990. He says 1 percent p-phenylenediamine added to the
mounting medium retards fading.
Two references he gives:
Johnson, GD, et al, A Simple Method of Reducing the Fading of
Immunofluorescece During Microscopy. J Immunol Methods
Huff, JC, et.al., Enhancement of Specific Immunofluorescent
Findings with use of para-phenylenediamine mounting buffer.
J Invest Dermatol 78:49, 1982.
Look into Vectashield, it is supposed to a good mounting media
for immunofluorescence. You may not be able to prevent fading
entirely, because the exciting light can cause it. Storage of
the slides, after coverslipping, should be dark, sometimes in
cold, or even in a freezer.
Vectashield is from Vector and it is pricey: $40 for 10 ml.
I think that the anti-fade agents that have already been
mentioned are all good, I must admit I have never used
Vectashield so will not comment on this. However, no mention
has been made of the possible variability in results with these
materials. Most of the anti-fade agents I have tried vary
considerably in their effectiveness. This appears to depend on
the specific antibody used, the fluorescent marker, the
fluorescence ratio of dye to marker molecule, whether the IHC
is direct or indirect and if you remembered to feed your cat
before going to work. As an example using lectin labelling of
cells with direct or indirect techniques, I found that the FITC
label was usually retained for UEA-1 but not for WGA. I would
therefore urge anyone who is going to use anti-fade agents to
try them first on some extra slides to test their
** Background in immunostained cartilage
I have tried to immunostain sections of whole mouse embryos with
several primary antibodies to a nuclear epitope. I am getting
nonspecific antibody staining in cytoplasm and in the connective
tissue around the cartilage.
I have blocked with embryo powder, normal goat serum, normal
horse serum, beat blocking solution from Zymed, and Fab
fragments. What could be reacting with secondary alone?
I do a lot of cartilage and bone IHC markers, mostly on rat, but
have done some mouse tissue. Is your primary made in a mouse?
Even with rat tissue, anti-mouse secondaries can combine
non-specifically with the rat tissue, I put rat serum in my
detection and it helps tremendously with the background.
The different blocking steps you have tested all block
hydrophobic areas (“sticky sites”) in your specimen. Hydrophobic
areas are blocked before the immunoincubation with e.g. normal
serum or BSA. Once blocked these sites generally will not give
rise to background anymore.
Cartilage and perichondrium are composed of collagen fibers with
a positive charge (still present after aldehyde fixation)
embedded in proteoglycans which have a negative charge. Most
antibodies (primaries and secondaries) are negatively charged at
pH 7-8.2. I therfore think that the collagen fibers present in
the cartilage tissue are causing your background problem. This
charge-determined background can be circumvented by
adding negatively charged molecules (e.g. aurion BSA-c) to the wash
and incubation buffers. Another possible cause for background
(a specific binding to proteoglycans) can be prevented by adding
gelatin to your buffers. Do not put both BSA-c and gelatin in
the same buffer, because they have charge-determined affinity
for each other as well.
I invite you to visit our web-site for detailed info on the
topic above. http://www.aurion.nl
Peter van de Plas
** Endogenous biotin in mast cells?
Do mast cells contain any endogenous biotin? They are often
falsely positive in immunostaining methods that use avidin.
Mast cells bind avidin nonspecifically because of ionic attraction
between avidin (a basic protein) and heparin (acid polysaccharide
in MC granules). This results in false positive staining by ABC.
The cure is to use the ABC reagent at pH 9.4. For more information
see Bussolati, G & Gugliotta, P 1983. Nonspecific staining of mast
cells by avidin-biotin-peroxidase complexes (ABC). J. Histochem.
Cytochem. 31: 1419-1421.
John A. Kiernan, Department of Anatomy & Cell Biology,
The University of Western Ontario,
LONDON, Canada N6A 5C1
Bussolati and Gugliotta (J. Histochem. Cytochem., 31(12):
1419-1421, 1983) described binding of ABC to mast cells.
They believed this to be due to both the binding of
avidin basic residues as well as peroxidase to the sulphate
groups of heparin. They showed that binding could be prevented
by using the ABC solution at a pH of 9.4. This high pH does
not affect either previous binding or localisation of antibody
or the affinity of biotin for avidin.
They also showed that the nonspecific binding of avidin
could be blocked by a 30 minute pretreatment of sections with
a synthetic basic polypeptide such as poly-L-lysine (0.01%
in PBS, pH 7.6).
Tony Henwood, Senior Scientist
Royal Prince Alfred Hospital