methanol vs paraformaldehyde fixation

[Deana Leonard]

Hello-

I have a question about paraformaldehyde fixation vs. methanol fixation of
TC cells. I'm trying to subsequently immuno-stain with my affinity purify
antibody. The problem I'm having is with the two fixation protocols I get
entirely different results.

With the paraformaldehyde fixation my protein appears to be excluded from
the nucleus except in early prophase of the cell cycle (when the DNA is
condensed) it is in the nucleus.

BUT

staining after Methanol/acetone fixation shows exclusively nuclear staining.
I'm stumped how to determine which is the correct localization of my
protein. Is one method more reliable than the other? From what I've read
paraformaldehyde is considered to be more reliable. Does every one agree?

I've stained four different cell lines and they all give the above
discrepancy based on fixation method. My protein does not express well in
cells- I've tried making a GFP-tagged construct and a HISMAX tagged
construct with no luck getting noticeable expression.

I'm using 4% paraformaldehyde for 30 minutes followed by 15 minutes of
triton extraction buffer (0.5 % triton X) to permebolize the cells.

For the methanol fixation I'm using 1:1 methanol to acetone for 10 minutes.

All done at room temperature.

Does anyone have any advice?
Thanks,
Deana

[Wolfgang Schechinger]

Dear Deana,

I think, methanol is more likely to precipitate proteins. Formaldehyde will
modify free amino groups (like e-NH2 from lysines).
You could try glutardialdehyde which will crosslink protens by the same
chemistry.

Acetone is likely to kill your GFP fluorescence in the fusion.

-----
this message consists of randomly assembled ascii characters
-----
Dr. Wolfgang Schechinger
Institute of Biomedical Sciences
Academia Sinica, Taipei, Taiwan R.O.C.
e mail wolfsc at ibms dot sinica dot edu dot tw
-----

---

[W. Schul]

Dear Deana,

You have a classical case of the fixation paradox. Fixation does not
lock everything in place instantaneously. Methanol/Acetone will
precipitate proteins and permeabilize the cells at the same time. A
higly soluble protein may actually be extracted by this method,
especially from the cytoplasm, and not detected. The well known PCNA
staining to detect cells in S-phase depends on this trick because PCNA
is present in both the cytoplasm and the nucleus in unbound form and
therefore gives a very low signal in cells after methanol fixation
(extraction) except in S-phase when PCNA is attached to DNA so these
cells give a strong nuclear PCNA labelling. Formaldehyde fixation
crosslinks proteins and DNA with very little extraction (that's why
you have to permeabilize with Triton afterwards to get your antibodies
in). In the case of PCNA it results in a strong cytplasmic and nuclear
labelling in all phases of the cell-cycle.
In your case, is it possible that in formaldehyde fixed cells there is
a very weak nuclear labelling that is hard to detect against the
strong cytoplasmic labelling while the cytoplasmic protein is lost
during methanol/acetone fixation so the nuclear labelling is
detectable? Confocal laser scanning microscopy may show you the
nuclear labelling in the formaldehyde fixed cells.
Formaldehyde crosslinking may also cause problems with permbeability
and mutilation of your epitope. Maybe your antibodies can't reach the
nucleus after formaldehyde fixation; shorter fixation and longer
Triton treatment may answer that.
Good and reliable immunolocalization is more than just a standard
protocol, you have to try and optimize. Also overexpression of a GFP
fusion protein can give you the wrong picture.
Glutaraldehyde is similar to formaldehyde but don't use it with
immunofluorescent labelling because it gives strong autofluorescent
background.

Good luck,

Wouter Schul