Dear Silvana,

CUMS has been considered as a valid animal model of depression based on applying 2 different kind of stressors (one physical and one psychological) to animals daily. Social isolation is used in this paradigm as one of psychological stressors. In fact, Willner and Muscat designed this protocol at early 80's to show that this method is a translation of the condition in which human expose to  daily stressors and challenges during lifetime. Thus, animals in CUMS have a social life but are daily exposed to various stressors for 28 days for example.

Willner, Paul. "Validity, reliability and utility of the chronic mild stress model of depression: a 10-year review and evaluation." Psychopharmacology 134.4 (1997): 319-329.

Hill, Matthew N., et al. "Neurobiology of chronic mild stress: parallels to major depression." Neuroscience & Biobehavioral Reviews 36.9 (2012): 2085-2117.

Good Luck

Shayan

Dear Napoleon,

Thanks a lot for the answer! Indeed, I've never done individual housing and I thought pair-housing is better especially for rats (that get depressed if isolated).

Dear Shayan,

Thank you very much for the references! I am new to the field and currently trying to fix my own protocol. Actually I saw in the Willner et al., 1987 that rats were single-housed and grouped housing was used as a stressor. But I still think pair-housing is better.

Dear Silvana,

Isolatoin housing is a stressor.  When animals are paired with an unfamiliar conspecific it is a stressor at first until dominance hieararchy is established. After that period it is less stressful than isolation.  Some investigaotrs have rotated pair-wise matches to ensure unfamiliar pairings. Another stressor is overcrowding.

That said, after muliiple attemps and significant expense we could not get consistant, reliable results from the model (using sucrose preference as an endpoint).  After digging around at SFN and in the literature we realize that we are not the only ones to be frustrated by the model. Some investigators seem to get an effect- many do not.  My advise would be to determine replicability between two runs (assuming you get an effect on the first run)- if you dont replicate then drop it.  We made the mistake of pursuing it over several attempts and variations and lost a lot of time and resources.

Greg 

Dear Greg,

Thank you a very much for the suggestion. I thought that measuring the serum corticosterone levels before/after the CUS protocol would be more reproducible than the sucrose preference test (but as I said I am completely new to the field, so I could be missing this point). I guess that the best is to do both controls.

Isolation housing and crowing both are stressors. In crowing generally up to 8 rats per cage are housed. 

Hi, Silvana!

Great advice from Greg.

I have used both paired and isolated rats for different models of stress.

And although there is no 'right' answer, I think you should consider the following:

a) isolated animals are stressed. So, are controls real controls?

But, when you keep animals in pairs you will most likely  make your stress less effective. Besides, you can monitor how much animals eat and drink as an easy measurement, and also behavioral changes that will be affected by the  animals the tested animal was paired with. If your CUS model includes  'pairing' or 'crowding' then keeping animals isolated should be a pre-requisite in most cases.

Honestly, find a good reference and stick with it. There a few good reviews on the subject.

b) most stress models are  are variable. You should make sure  your model works in YOUR lab with YOUR animals. Experimental condition, house keeping, strain, provider... All affect the outcome. So, yes, I think you are on the right track measuring cort, for example. Pick the timing wisely. You can also test the animals in different behavior tests like forced swim, tail suspension, anxiety tests like elevated plus maze, and memory tests like Morris water maze. In addition control the animals weight throughout the stress protocol and preferably their consumption of food and water too. 

c) sucrose preference test is certainly useful and will hardly compromise any other test, but keep in mind that it is EXTREMELY variable. And you should keep the animals isolated to get 'clean' data. So, do not use it as the sole endpoint to validate the model in your lab.

d) as a side note, I have discussed the protocol with other people and listend o comments about the fact that the very last stressor of CUS changes the outcome of the protocol... I never tried it myself though.

We have recently published a paper with CUS as stress model in mice. Take a look.

Good luck!

Article Caffeine acts through neuronal adenosine A 2A receptors to p...

Dear Silvana

Between the stress sessions the animals should be housed in pairs or groups, because the isolation is a stressor by itself.

If social stress itself *is* the stressor you want, you can use, for example, the social defeat paradigm.

Good luck

Dear Shira,

Thank you very much for the answer.