In the past, researchers experienced this issue, especiallywith flaviviruses. When you freeze your virus you provoke an acidification of the media, this acidification result in a conformation change of the envelop and a drastic loss of infectivity. To solve this problem you have to add HEPES (25mM) to your media before freezing.

Franck Touret Thank you so much for this answer!! Is there any source that shows the effect of acidification on the envelop protein? Also, does the acidification only happens at the moment of freezing, or does it further increase as the medium stays longer in the freezer?

I think it is well documented ( it is a fusion class 2 envelop glycoprotein) you can find answers in this paper. Article Structures and Mechanisms of Viral Membrane Fusion Proteins:...

I do not know in detail the kinetic of this acidification but in our experience ( working with many flaviviruses and biobanking) in -80 they are relitavely stable for few years. If you want more stability then you have to lyophilized them. Good luck

My understanding of the mechanisms of virus degradation with time after freezing are primarily two:

A. Formation of water crystals. This is affected by, and both drives and is driven by development of micro-zones of concentrated salts. I believe that is the source of the micro-local pH change referred to above.

B. Damage from slower than optimum freezing, which exacerbates crystal formation over years. It is common to just stick the vial into a freezer box, put the box in the freezer and leave. But if you think about it, freezer boxes have thermal mass, and warm up when exposed to the air. The cardboard or plastic dividers warm up pretty fast, because they are thin. And the divider of the box, and the air surrounding the vial have thermal mass as well as insulating properties. (Think of what it would be like if a giant robot picked up your building, put it down in a desert where the temperature was 145 C, took off the roof, and set a barrel down next to you that was that temperature, using manipulators that were larger in diameter than the room.) The more slowly your sample freezes, the more micro-zones of salts and the more ice crystals form. (Although, when the rate slows way down, in the presence of DMSO, etc., this results in desiccation with cells, and smaller, more granular crystals that do less damage. Viruses don't do as well with that method, although I have never tried it.) The larger the diameter of your vial, the slower the freeze-zone progresses. It is the rate at which the interface between frozen and liquid water progresses that determines number and size of salt pockets, and how much of the water is glass versus crystals. Crystal seeds can slowly grow, -80 things still happen.

This is why some labs will first flash-freeze in LN2, (or maybe dry ice mixed with EtOH), then transfer that to a -80 freezer-box, sometimes using chilled tongs, forceps, etc.. Experience shows that dessication/lyophilization helps, and that's what physics would predict. I've attached a BEIR poster that presents their data on storage of influenza, which is another enveloped virus. Flaviviruses have a tighter envelope.

You can also get some damage by thawing too slowly. There will be some, because there is always a freeze-thaw boundary that forms when thawing just as during freezing. Sometimes less attention is paid to the thawing process.

You might also consider using siliconized tips and vials to prevent virus present from sticking. If titers are low this might help.

Brian Hanley Thank you really a lot for your amazing answer! My PI was not quite convinced that the titre could drop even at -80C, so I'll measure it with plaque assay before carrying out further investigation on this issue. Also nice tips regarding thawing: I didn't know this before so now I will thaw them in 37 water bath instead of RT!