To put that into perspective - Y. pestis' main chromosome is a little over 4 million basepairs. So it is only broken pieces of code that they can recover, not the whole, intact bacterium. No risk here.
http://www.nature.com/nature/journal/v478/n7370/full/nature1...
Of course, passing through humans for a half century, and then through these rodents, it won't be quite the same, but...
And many can be countered with standard mutations, i.e. something that messes with the surface protein that transports them inside. When I was doing a bit of E. Coli microbiology in the summer of 1977 the conventional wisdom was for a generic antibiotic, like the colicin we were working with, 1 in a million bacterial would spontaneously develop a mutation (which I further did the grunt work to prove the ones we had were of the surface protein type, by and large). Streptomycin, though, would take 1 in a billion.
The real trick is to avoid creating and breathing in an aerosol, which has I'm sure killed more than one researcher before he and his doctor realized what was happening, I heard of one such case in the '70s when I was doing work on E. Coli during a summer program involving salmonella. If you get too much lung tissue---lots of surface area---infected too quickly, well, see pneumonic plague, it's anthrax equivalent, which an Air Force flight surgeon's manual said the first clear diagnostic sign from a patient is his sudden death, etc.
And echoing loxs, it's very unlikely they're working with viable bacteria, if they're just sequencing DNA or the like they're far, far from something viable, and I don't believe it creates spores.