Only 7 days later in dark and +4C storage, the signal reduced 60-70%, it seems unexpected. There may be two reasons related with cover solution including hoescht, or expire date of Alexa 488 dye.
Photobleaching seems unlikely as the explanation because you kept the specimen in the dark because and Alexa Fluor 488 is pretty resistant to photobleaching.
As mentioned above it is unlikely to be photobleaching as you have not exposed it to light. Alexa488 will experience some loss of intensity over time (you should try not to directly compare images taken over different days) but 70% in a week seems a lot. Without more information it is hard to know exactly what is causing this. Do some reading about optimal fixation and mounting techniques for maintaining fluorescence. You should use and antifadant/fluorescent protoctor, And for storage you should seal off the edges of the cover slips. Also check the pH of your other solutions such as the hoescht, Alexa488 has a specific (albeit broad pH range).
Finally consider what the application of Alexa488 is, e.g. are you using it conjugated to a protein/antibody/ester, is it possible that the protein you are binding to is degrading rather than the Alexa488
It's a pity but my only answer is that something wrong with the refrigerator at the lab, there might be electricity shortage during weekend ( temp. increase up to 8-10C for few hours can ruin the sample), following days i can get it, from other sample at that fridge. Thank you very much...
have you used an embedding medium? Does it include an antifade substance? Could you have anything in your embedding which is quenching the Alexa signal?
Yeap i used glycerol in DPBS (1:1 dilution, so it is very dense), and seal the edge of coverslip (nicely). Hundred times i practiced drop, cover, seal at nearly dark room.
Actually i used several protein, the last one i had photobleacing problem, however, i stain them for 2 days and imaging at tha 3rd and 4th day at most. Now i am not sure about if the protein was degraded or Alexa was fading out.
Still, people said about imaging the one month old samples :)