I'm going to try to take a stab at this, but I don't know specifically about
the picrosirius red stain. I'm going to talk in general about stains and
fixation (or lack thereof in the frozen section (FS)), and the relationship
between fixation and staining. I'm going to start out very simple, and then
get a little more complicated, so hang in there. (If you are not interested
in dye theory, delete now, as this is going to get long. But if you like dye
theory (like I do), I hope you enjoy this. And if you do like dye theory,
let me know if you think I'm correct, or if I’m barking up the wrong tree.)
Stains bind to proteins in the tissue, which are made up of amino acids,
which either are positively charged, negatively charged, or non-polar (no
charge). Dyes usually have positive ions and negative ions, but tend to have
more of one type, and thus are considered either positively charged or
negatively charged, and/or tend to bind via one type of bonding over another
(e.g., hydrogen bonds vs. covalent).
Different Proteins in tissue have their own unique shape and density. Dyes
have their own shapes. The two shapes have to be somewhat similar for the
proteins to bind to a dye. A round ball might be able to fit into an open
space in a protein, whereas a stiff long dye (think something the shape of a
ruler) might not be able to fit into that space.
Differently Proteins have their own density (loose to tightly packed). Dyes
have their own sizes (small to very large). The dye has to be able to fit
into the protein, so a large dye may find it difficult to fit into a dense
bunch of protein, whereas a small dye can fit into loose protein and dense
protein.
Therefore, for a dye to bind to a protein, the charged ions on the dyes have
to be the opposite of the charges on the amino acids/proteins (positive
binding to negative). And the charges on the dyes have to line up with the
charges on the proteins, so the dye has to be able to fit into the
protein, and the charges have to line up.
Therefore, if we do something that changes the CHARGES on the protein,
and/or the SHAPE of the proteins, and/or the DENSITY of the proteins, the
dye may bind differently (not at all, very little, or too much).
(Conversely, changing the dye in any way could cause different staining
patterns, but since you said it was the same kit, and since you said later
your boss asked for the stain to be done on a FS, I'm expecting it to have
been the same day or the next, so I don't think the kit went bad, and I'm
assuming you did the stain correctly. So I won't be discussing bad staining
due to bad dyes or performing the stain incorrectly.)
Now, onto fixation vs. frozen section (FS). I'm assuming 10% formalin was
the fixative, or a zinc formalin, or a formalin substitute (glyoxal). It's
the formalin/formaldehyde/glyoxal that is negatively charged. It will bind
with positive amino acids in the protein. Let's assume there were 10 + and
10 - amino acids on the protein, so the overall charge of the protein is a
net zero. Let's bind 4 of the + amino acids with - charged formalin. You now
have 6 + and 10 - amino acids, so you have more negative amino acids than
positive, so your tissue is more negatively charged. You have just changed
the CHARGES on the proteins.
Fixatives cross-link proteins, and pull them in different directions. You
have therefore changed the SHAPE of the protein.
Since the fixative is cross-linking the protein, and pulling the proteins in
different directions, some proteins are going to be pulled further apart,
thus becoming looser in density, while other proteins are being pulled
closer together, or being made denser. You have therefore changed the
DENSITY of the protein.
Most dyes/stains used in histology were designed to be used with formalin
fixed tissue, and are therefore made to work with proteins that have had
their charges changed, their shapes altered, and their density changed,
according to the changes made by formalin.
Frozen sections have NOT have any fixative, and are therefore similar to the
unfixed tissue, without the changes in charged, shape, and density. So it
should make sense that unfixed/FS tissues should/could stain differently
than fixed tissue.
Now, for sirius red specifically. There are several different sirius red dye
molecules. I don't know which one in particular you used, but in this
explanation, it doesn't really matter, because they all belong to the
polyazo dye family. That means they are made up of several benzene rings (5
to 8), held together in a long row (linear - like the ruler I mentioned
earlier) with azo bonds (Nitrogen double bonded to Nitrogen -N=N- which have
hydrogens bonded to them, giving these bonds a positive charge), and several
sulfonic acid groups bound to the benzene rings ( -SO3 ions, which are
negatively charged).
To me, these sirius red dye molecules look very similar to Congo red dye
molecules, and