Let me give you an example of 'painful environment,' something many people can probably relate to.
There's a kid. They're outgoing, talk to everyone, and adults constantly tell their parents what a smart kid they have.
Kid grows up a bit, and during puberty is constantly ridiculed in social environments. Eventually the kid withdraws into themselves - a shell of their former, outgoing self.
In their twenties, even though they aren't in the negative environment of getting ridiculed for every single last thing anymore, the mere act of talking to someone new induces feer and anxiety.
They know they need to be more social. That they aren't depresses them. They find alcohol a decent enough social lubricant, but it doesn't work as well as they hoped. So they use more of it. They may add a couple more drugs into the mix - a line of cocaine in the john will help liven them up, surely - and end in a flat-out spiral of drug and alcohol abuse coupled with crippling depression and social anxiety.
Their learned response to the stimulus of the social is emotional pain.
They go on prozac; they go into therapy; the increased brain plasticity helps the therapy hold; their response to social interaction is no longer flight-or-fight.
Considered faulty circuits because stimuli was repeatedly presented with pain/an electric shock?
This article also doesn't tell us why it's seemingly plastic.
Also, it was only once the mice were removed the painful environment that ... so how about just remove people from a negative environment?
What happens to the mice if you put them on Prozac, and the pain is still there? Do they learn to fear it even greater/stronger?
Try this with MDMA along with therapy and you'll desensitize the mice/people too; http://www.maps.org/mdma/protocol/