I can see two issues here. One is that your tissues is not perfussed with a fixative and crioprotected afterwards, that will result very often in freezing and cutting artifacts. The other is that by definition ethanol baths will dehydrate your tissue and lost of water plus weak fixation may be the reason you see tissue shrinkage and other artifacts.

I suggest perfussion of your animals plus crioprotection with a sucrose solution. If perfussion is not possible a post-fixation by immersion followed by crioprotection. Also avoid ethanol and xylene baths using an aqueous mounting media and a compatible counter-staining

How do you process the fresh sample? It looks from your protocol that your fresh sample is just covered with OCT and then put at -80deg, is that correct? If so, this will result in freezing damage as the tissue is left to cool slowly and will evetually freeze at different rates throughout the tissue.

When I process a fresh sample for frozen sections, the sample is snap-frozen in liquid nitrogen-cooled hexane and then stored in liquid nitrogen until use.  When required, i then mount the sample on the chuck in OCT, freeze with cryospray then section.  

Persevere if you can, there are (some!) advantages of frozen sections over FFPE!

With regards to the rest of your protocol, it looks OK to me.  However, I use IPA and not ethanol (very similar, but not quite the same!) for dehydration prior to xylene.

I agree the initial processing of the tissue may be causing the artifacts. I use dry ice to rapidly freeze fresh tissue embedded in OCT, as long as the tissue block isn't over 4-5 mm thick, freezes quickly and then i store at -80C until sectioning.

The other place to trouble shoot would be the H2O2 step...how thick are your sections?? look at tissue before then after H2O2 treatment and see if the bubbling action of the H2O2 is damaging the tissue, can use a quick hematoxylin or toluidine blue 0.1% to stain nuclei so you have more contrast to compare. if the H2O2 is causing problems maybe reduce the time OR what are you diluting the H2O2 in...try diluting H2O2in PBS if you aren't already.

it looks from both images that you have some freezing artifact.This occurs from storing your tissue in  freezers that have afreeze/thaw cycle, allowing your tissue to freeze slowly or refreezing. I one that does not have a thaw cycle; most -80C do not have one, though. Freezing tissue slowly allows ice crystals to grow large, which shatter cells, and in turn allows your tissue to shrink. NEVER refreeze your tissue if thawed, and ALWAYS snap freeze your tissue over Liquid nitrogen (best in OCT over isopentane resting in liquid nitrogen). Only if you have a very small piece of tissue can you use dry ice or -80C but it is not really cold enough to 'snap freeze', so artefact of ruptured cells can occur easily. Also, as David Rottland pointed out, the use of ethanol dehydrates tissue, resulting in the loss of water which shrinks tissue. If together with freezing artifact (ruptured cells), the use of ethanol can cause your tissue to shrink inconsistently in every dimension,  leaving your tissue warped in shape from it's original. I suggest fixation (in 4% PFA;10%formalin, as your do; at 4C) then cryoprotecting in 20% sucrose/PBS, then submerging the tissue in your OCT in mold, the snap freezing directly in freezing isopentane over liquid nitrogen. Store this at -80C until you cut it.. Also keep your tissue on dry ice when transporting it to the cryostat. Use a wet mount (aqueous) as also suggested by David above.

I am so sorry that I delayed for one month. 

I will try to use cryoprotection and use liquid nitrogen to freeze the samples next time.

Thank you so much.