This will no doubt make news when it is published (though I can already picture the absurd headlines):
From this years General Meeting for the American Society for Microbiology, there was a very intriguing interview.
Mycobacterium vaccae, found in soil, previously known to effect serotonin among other things when injected in mice, and improve mood in early human trials when injected, was tested in mice on learning processes by Dorothy Matthews and a colleague.
She states in the video (embedded below) that they found mice who were fed the (live) bacteria navigated a maze about twice as fast as control mice and were less anxious. In a subsequent experiment, the effect after learning was maintained even when the bacteria was removed from the diet. She also found that a trend for a stronger maze memory recall after a 3 week rest.
She summarizes this and other related research in the video (~20 minutes in length).
Importantly, she wisely does not recommend consuming this as a probiotic or from any yogurts that inevitably abuse these scientific findings. Instead, interacting with nature is the best idea at this point. This is good advice for other reasons as well.
And this certainly cannot be generalized to all other genus’, as M. Vaccae may exert its effects through its unique components.
Most likely in my opinion, a lack of microbial exposure in an environment that is too clean may cause you to miss out on microbes such as these that positively effect physiology. I wonder if this effect would be evident in wild (uncaged) mice, for instance. This is no doubt just the tip of the iceberg, and we still need human trials to confirm.
I will update with more details when the paper is published.